Minki/linux
1
0
Fork 1
mirror of https://github.com/torvalds/linux.git synced 2024-12-30 06:41:43 +00:00
Code Issues Packages Projects Releases Wiki Activity
2c042e8e54
linux/fs/ceph/dir.c

2065 lines
53 KiB
C
Raw Normal View History

License cleanup: add SPDX GPL-2.0 license identifier to files with no license Many source files in the tree are missing licensing information, which makes it harder for compliance tools to determine the correct license. By default all files without license information are under the default license of the kernel, which is GPL version 2. Update the files which contain no license information with the 'GPL-2.0' SPDX license identifier. The SPDX identifier is a legally binding shorthand, which can be used instead of the full boiler plate text. This patch is based on work done by Thomas Gleixner and Kate Stewart and Philippe Ombredanne. How this work was done: Patches were generated and checked against linux-4.14-rc6 for a subset of the use cases: - file had no licensing information it it. - file was a */uapi/* one with no licensing information in it, - file was a */uapi/* one with existing licensing information, Further patches will be generated in subsequent months to fix up cases where non-standard license headers were used, and references to license had to be inferred by heuristics based on keywords. The analysis to determine which SPDX License Identifier to be applied to a file was done in a spreadsheet of side by side results from of the output of two independent scanners (ScanCode & Windriver) producing SPDX tag:value files created by Philippe Ombredanne. Philippe prepared the base worksheet, and did an initial spot review of a few 1000 files. The 4.13 kernel was the starting point of the analysis with 60,537 files assessed. Kate Stewart did a file by file comparison of the scanner results in the spreadsheet to determine which SPDX license identifier(s) to be applied to the file. She confirmed any determination that was not immediately clear with lawyers working with the Linux Foundation. Criteria used to select files for SPDX license identifier tagging was: - Files considered eligible had to be source code files. - Make and config files were included as candidates if they contained >5 lines of source - File already had some variant of a license header in it (even if <5 lines). All documentation files were explicitly excluded. The following heuristics were used to determine which SPDX license identifiers to apply. - when both scanners couldn't find any license traces, file was considered to have no license information in it, and the top level COPYING file license applied. For non */uapi/* files that summary was: SPDX license identifier # files ---------------------------------------------------|------- GPL-2.0 11139 and resulted in the first patch in this series. If that file was a */uapi/* path one, it was "GPL-2.0 WITH Linux-syscall-note" otherwise it was "GPL-2.0". Results of that was: SPDX license identifier # files ---------------------------------------------------|------- GPL-2.0 WITH Linux-syscall-note 930 and resulted in the second patch in this series. - if a file had some form of licensing information in it, and was one of the */uapi/* ones, it was denoted with the Linux-syscall-note if any GPL family license was found in the file or had no licensing in it (per prior point). Results summary: SPDX license identifier # files ---------------------------------------------------|------ GPL-2.0 WITH Linux-syscall-note 270 GPL-2.0+ WITH Linux-syscall-note 169 ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) 21 ((GPL-2.0 WITH Linux-syscall-note) OR BSD-3-Clause) 17 LGPL-2.1+ WITH Linux-syscall-note 15 GPL-1.0+ WITH Linux-syscall-note 14 ((GPL-2.0+ WITH Linux-syscall-note) OR BSD-3-Clause) 5 LGPL-2.0+ WITH Linux-syscall-note 4 LGPL-2.1 WITH Linux-syscall-note 3 ((GPL-2.0 WITH Linux-syscall-note) OR MIT) 3 ((GPL-2.0 WITH Linux-syscall-note) AND MIT) 1 and that resulted in the third patch in this series. - when the two scanners agreed on the detected license(s), that became the concluded license(s). - when there was disagreement between the two scanners (one detected a license but the other didn't, or they both detected different licenses) a manual inspection of the file occurred. - In most cases a manual inspection of the information in the file resulted in a clear resolution of the license that should apply (and which scanner probably needed to revisit its heuristics). - When it was not immediately clear, the license identifier was confirmed with lawyers working with the Linux Foundation. - If there was any question as to the appropriate license identifier, the file was flagged for further research and to be revisited later in time. In total, over 70 hours of logged manual review was done on the spreadsheet to determine the SPDX license identifiers to apply to the source files by Kate, Philippe, Thomas and, in some cases, confirmation by lawyers working with the Linux Foundation. Kate also obtained a third independent scan of the 4.13 code base from FOSSology, and compared selected files where the other two scanners disagreed against that SPDX file, to see if there was new insights. The Windriver scanner is based on an older version of FOSSology in part, so they are related. Thomas did random spot checks in about 500 files from the spreadsheets for the uapi headers and agreed with SPDX license identifier in the files he inspected. For the non-uapi files Thomas did random spot checks in about 15000 files. In initial set of patches against 4.14-rc6, 3 files were found to have copy/paste license identifier errors, and have been fixed to reflect the correct identifier. Additionally Philippe spent 10 hours this week doing a detailed manual inspection and review of the 12,461 patched files from the initial patch version early this week with: - a full scancode scan run, collecting the matched texts, detected license ids and scores - reviewing anything where there was a license detected (about 500+ files) to ensure that the applied SPDX license was correct - reviewing anything where there was no detection but the patch license was not GPL-2.0 WITH Linux-syscall-note to ensure that the applied SPDX license was correct This produced a worksheet with 20 files needing minor correction. This worksheet was then exported into 3 different .csv files for the different types of files to be modified. These .csv files were then reviewed by Greg. Thomas wrote a script to parse the csv files and add the proper SPDX tag to the file, in the format that the file expected. This script was further refined by Greg based on the output to detect more types of files automatically and to distinguish between header and source .c files (which need different comment types.) Finally Greg ran the script using the .csv files to generate the patches. Reviewed-by: Kate Stewart <kstewart@linuxfoundation.org> Reviewed-by: Philippe Ombredanne <pombredanne@nexb.com> Reviewed-by: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2017-11-01 14:07:57 +00:00
// SPDX-License-Identifier: GPL-2.0
ceph: factor out libceph from Ceph file system This factors out protocol and low-level storage parts of ceph into a separate libceph module living in net/ceph and include/linux/ceph. This is mostly a matter of moving files around. However, a few key pieces of the interface change as well: - ceph_client becomes ceph_fs_client and ceph_client, where the latter captures the mon and osd clients, and the fs_client gets the mds client and file system specific pieces. - Mount option parsing and debugfs setup is correspondingly broken into two pieces. - The mon client gets a generic handler callback for otherwise unknown messages (mds map, in this case). - The basic supported/required feature bits can be expanded (and are by ceph_fs_client). No functional change, aside from some subtle error handling cases that got cleaned up in the refactoring process. Signed-off-by: Sage Weil <sage@newdream.net>
2010-04-06 22:14:15 +00:00
#include <linux/ceph/ceph_debug.h>
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
#include <linux/spinlock.h>
#include <linux/namei.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 08:04:11 +00:00
#include <linux/slab.h>
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
#include <linux/sched.h>
ceph: Switch to generic xattr handlers Add a catch-all xattr handler at the end of ceph_xattr_handlers. Check for valid attribute names there, and remove those checks from __ceph_{get,set,remove}xattr instead. No "system.*" xattrs need to be handled by the catch-all handler anymore. The set xattr handler is called with a NULL value to indicate that the attribute should be removed; __ceph_setxattr already handles that case correctly (ceph_set_acl could already calling __ceph_setxattr with a NULL value). Move the check for snapshots from ceph_{set,remove}xattr into __ceph_{set,remove}xattr. With that, ceph_{get,set,remove}xattr can be replaced with the generic iops. Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com> Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2016-04-13 22:30:17 +00:00
#include <linux/xattr.h>
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
#include "super.h"
ceph: factor out libceph from Ceph file system This factors out protocol and low-level storage parts of ceph into a separate libceph module living in net/ceph and include/linux/ceph. This is mostly a matter of moving files around. However, a few key pieces of the interface change as well: - ceph_client becomes ceph_fs_client and ceph_client, where the latter captures the mon and osd clients, and the fs_client gets the mds client and file system specific pieces. - Mount option parsing and debugfs setup is correspondingly broken into two pieces. - The mon client gets a generic handler callback for otherwise unknown messages (mds map, in this case). - The basic supported/required feature bits can be expanded (and are by ceph_fs_client). No functional change, aside from some subtle error handling cases that got cleaned up in the refactoring process. Signed-off-by: Sage Weil <sage@newdream.net>
2010-04-06 22:14:15 +00:00
#include "mds_client.h"
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
/*
* Directory operations: readdir, lookup, create, link, unlink,
* rename, etc.
*/
/*
* Ceph MDS operations are specified in terms of a base ino and
* relative path. Thus, the client can specify an operation on a
* specific inode (e.g., a getattr due to fstat(2)), or as a path
* relative to, say, the root directory.
*
* Normally, we limit ourselves to strict inode ops (no path component)
* or dentry operations (a single path component relative to an ino). The
* exception to this is open_root_dentry(), which will open the mount
* point by name.
*/
ceph: constify dentry_operations This makes checkpatch happy. Signed-off-by: Sage Weil <sage@newdream.net>
2010-08-03 17:25:30 +00:00
const struct dentry_operations ceph_dentry_ops;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
ceph: periodically trim stale dentries Previous commit make VFS delete stale dentry when last reference is dropped. Lease also can become invalid when corresponding dentry has no reference. This patch make cephfs periodically scan lease list, delete corresponding dentry if lease is invalid. There are two types of lease, dentry lease and dir lease. dentry lease has life time and applies to singe dentry. Dentry lease is added to tail of a list when it's updated, leases at front of the list will expire first. Dir lease is CEPH_CAP_FILE_SHARED on directory inode, it applies to all dentries in the directory. Dentries have dir leases are added to another list. Dentries in the list are periodically checked in a round robin manner. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-01-31 08:55:51 +00:00
static bool __dentry_lease_is_valid(struct ceph_dentry_info *di);
static int __dir_lease_try_check(const struct dentry *dentry);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
/*
* Initialize ceph dentry state.
*/
ceph: switch to use of ->d_init() Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2016-10-29 02:05:13 +00:00
static int ceph_d_init(struct dentry *dentry)
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
{
struct ceph_dentry_info *di;
ceph: add ceph_sb_to_mdsc helper support to parse the mdsc This will help simplify the code. [ jlayton: fix minor merge conflict in quota.c ] Signed-off-by: Xiubo Li <xiubli@redhat.com> Signed-off-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-09-03 13:01:39 +00:00
struct ceph_mds_client *mdsc = ceph_sb_to_mdsc(dentry->d_sb);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
ceph: use kmem_cache_zalloc Use kmem_cache_zalloc() instead of kmem_cache_alloc() with flag GFP_ZERO. Signed-off-by: Geliang Tang <geliangtang@163.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2016-03-13 07:26:29 +00:00
di = kmem_cache_zalloc(ceph_dentry_cachep, GFP_KERNEL);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
if (!di)
return -ENOMEM; /* oh well */
di->dentry = dentry;
di->lease_session = NULL;
ceph: don't use ->d_time Pretty simple: just use ceph_dentry_info.time instead (which was already there, unused). Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
2016-06-22 14:35:04 +00:00
di->time = jiffies;
ceph: handle racing calls to ceph_init_dentry The ->lookup() and prepopulate_readdir() callers are working with unhashed dentries, so we don't have to worry. The export.c callers, though, need to initialize something they got back from d_obtain_alias() and are potentially racing with other callers. Make sure we don't return unless the dentry is properly initialized (by us or someone else). Reported-by: Al Viro <viro@ZenIV.linux.org.uk> Reviewed-by: Yehuda Sadeh <yehuda@hq.newdream.net> Signed-off-by: Sage Weil <sage@newdream.net>
2011-07-26 18:30:15 +00:00
dentry->d_fsdata = di;
ceph: periodically trim stale dentries Previous commit make VFS delete stale dentry when last reference is dropped. Lease also can become invalid when corresponding dentry has no reference. This patch make cephfs periodically scan lease list, delete corresponding dentry if lease is invalid. There are two types of lease, dentry lease and dir lease. dentry lease has life time and applies to singe dentry. Dentry lease is added to tail of a list when it's updated, leases at front of the list will expire first. Dir lease is CEPH_CAP_FILE_SHARED on directory inode, it applies to all dentries in the directory. Dentries have dir leases are added to another list. Dentries in the list are periodically checked in a round robin manner. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-01-31 08:55:51 +00:00
INIT_LIST_HEAD(&di->lease_list);
ceph: add dentry lease metric support For dentry leases, only count the hit/miss info triggered from the vfs calls. For the cases like request reply handling and ceph_trim_dentries, ignore them. For now, these are only viewable using debugfs. Future patches will allow the client to send the stats to the MDS. The output looks like: item total miss hit ------------------------------------------------- d_lease 11 7 141 URL: https://tracker.ceph.com/issues/43215 Signed-off-by: Xiubo Li <xiubli@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-03-20 03:44:59 +00:00
atomic64_inc(&mdsc->metric.total_dentries);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
return 0;
}
/*
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
* for f_pos for readdir:
* - hash order:
* (0xff << 52) | ((24 bits hash) << 28) |
* (the nth entry has hash collision);
* - frag+name order;
* ((frag value) << 28) | (the nth entry in frag);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
*/
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
#define OFFSET_BITS 28
#define OFFSET_MASK ((1 << OFFSET_BITS) - 1)
#define HASH_ORDER (0xffull << (OFFSET_BITS + 24))
loff_t ceph_make_fpos(unsigned high, unsigned off, bool hash_order)
{
loff_t fpos = ((loff_t)high << 28) | (loff_t)off;
if (hash_order)
fpos |= HASH_ORDER;
return fpos;
}
static bool is_hash_order(loff_t p)
{
return (p & HASH_ORDER) == HASH_ORDER;
}
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
static unsigned fpos_frag(loff_t p)
{
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
return p >> OFFSET_BITS;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
}
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
static unsigned fpos_hash(loff_t p)
{
return ceph_frag_value(fpos_frag(p));
}
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
static unsigned fpos_off(loff_t p)
{
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
return p & OFFSET_MASK;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
}
ceph: fix __dcache_readdir() If directory is fragmented, readdir() read its dirfrags one by one. After reading all dirfrags, the corresponding dentries are sorted in (frag_t, off) order in the dcache. If dentries of a directory are all cached, __dcache_readdir() can use the cached dentries to satisfy readdir syscall. But when checking if a given dentry is after the position of readdir, __dcache_readdir() compares numerical value of frag_t directly. This is wrong, it should use ceph_frag_compare(). Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
2014-02-13 11:40:26 +00:00
static int fpos_cmp(loff_t l, loff_t r)
{
int v = ceph_frag_compare(fpos_frag(l), fpos_frag(r));
if (v)
return v;
return (int)(fpos_off(l) - fpos_off(r));
}
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
/*
* make note of the last dentry we read, so we can
* continue at the same lexicographical point,
* regardless of what dir changes take place on the
* server.
*/
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
static int note_last_dentry(struct ceph_dir_file_info *dfi, const char *name,
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
int len, unsigned next_offset)
{
char *buf = kmalloc(len+1, GFP_KERNEL);
if (!buf)
return -ENOMEM;
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
kfree(dfi->last_name);
dfi->last_name = buf;
memcpy(dfi->last_name, name, len);
dfi->last_name[len] = 0;
dfi->next_offset = next_offset;
dout("note_last_dentry '%s'\n", dfi->last_name);
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
return 0;
}
ceph: search cache postion for dcache readdir use binary search to find cache index that corresponds to readdir postion. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-28 09:43:35 +00:00
static struct dentry *
__dcache_find_get_entry(struct dentry *parent, u64 idx,
struct ceph_readdir_cache_control *cache_ctl)
{
struct inode *dir = d_inode(parent);
struct dentry *dentry;
unsigned idx_mask = (PAGE_SIZE / sizeof(struct dentry *)) - 1;
loff_t ptr_pos = idx * sizeof(struct dentry *);
pgoff_t ptr_pgoff = ptr_pos >> PAGE_SHIFT;
if (ptr_pos >= i_size_read(dir))
return NULL;
if (!cache_ctl->page || ptr_pgoff != page_index(cache_ctl->page)) {
ceph_readdir_cache_release(cache_ctl);
cache_ctl->page = find_lock_page(&dir->i_data, ptr_pgoff);
if (!cache_ctl->page) {
dout(" page %lu not found\n", ptr_pgoff);
return ERR_PTR(-EAGAIN);
}
/* reading/filling the cache are serialized by
ceph: fix comments mentioning i_mutex inode->i_mutex has been replaced with inode->i_rwsem long ago. Fix comments still mentioning i_mutex. Signed-off-by: hongnanli <hongnan.li@linux.alibaba.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2022-02-11 03:33:25 +00:00
i_rwsem, no need to use page lock */
ceph: search cache postion for dcache readdir use binary search to find cache index that corresponds to readdir postion. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-28 09:43:35 +00:00
unlock_page(cache_ctl->page);
cache_ctl->dentries = kmap(cache_ctl->page);
}
cache_ctl->index = idx & idx_mask;
rcu_read_lock();
spin_lock(&parent->d_lock);
/* check i_size again here, because empty directory can be
ceph: fix comments mentioning i_mutex inode->i_mutex has been replaced with inode->i_rwsem long ago. Fix comments still mentioning i_mutex. Signed-off-by: hongnanli <hongnan.li@linux.alibaba.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2022-02-11 03:33:25 +00:00
* marked as complete while not holding the i_rwsem. */
ceph: search cache postion for dcache readdir use binary search to find cache index that corresponds to readdir postion. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-28 09:43:35 +00:00
if (ceph_dir_is_complete_ordered(dir) && ptr_pos < i_size_read(dir))
dentry = cache_ctl->dentries[cache_ctl->index];
else
dentry = NULL;
spin_unlock(&parent->d_lock);
if (dentry && !lockref_get_not_dead(&dentry->d_lockref))
dentry = NULL;
rcu_read_unlock();
return dentry ? : ERR_PTR(-EAGAIN);
}
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
/*
* When possible, we try to satisfy a readdir by peeking at the
* dcache. We make this work by carefully ordering dentries on
move d_rcu from overlapping d_child to overlapping d_alias Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2014-10-26 23:19:16 +00:00
* d_child when we initially get results back from the MDS, and
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
* falling back to a "normal" sync readdir if any dentries in the dir
* are dropped.
*
ceph: use i_release_count to indicate dir's completeness Current ceph code tracks directory's completeness in two places. ceph_readdir() checks i_release_count to decide if it can set the I_COMPLETE flag in i_ceph_flags. All other places check the I_COMPLETE flag. This indirection introduces locking complexity. This patch adds a new variable i_complete_count to ceph_inode_info. Set i_release_count's value to it when marking a directory complete. By comparing the two variables, we know if a directory is complete Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
2013-03-13 11:44:32 +00:00
* Complete dir indicates that we have all dentries in the dir. It is
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
* defined IFF we hold CEPH_CAP_FILE_SHARED (which will be revoked by
* the MDS if/when the directory is modified).
*/
ceph: skip invalid dentry during dcache readdir skip dentries that were added before MDS issued FILE_SHARED to client. Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com> Reviewed-by: Sage Weil <sage@inktank.com>
2014-04-06 06:10:04 +00:00
static int __dcache_readdir(struct file *file, struct dir_context *ctx,
ceph: use atomic_t for ceph_inode_info::i_shared_gen It allows accessing i_shared_gen without holding i_ceph_lock. It is preparation for later patch. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Acked-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-11-27 02:47:46 +00:00
int shared_gen)
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
{
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
struct ceph_dir_file_info *dfi = file->private_data;
kill f_dentry uses Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2014-10-31 05:22:04 +00:00
struct dentry *parent = file->f_path.dentry;
VFS: normal filesystems (and lustre): d_inode() annotations that's the bulk of filesystem drivers dealing with inodes of their own Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2015-03-17 22:25:59 +00:00
struct inode *dir = d_inode(parent);
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
struct dentry *dentry, *last = NULL;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
struct ceph_dentry_info *di;
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
struct ceph_readdir_cache_control cache_ctl = {};
ceph: search cache postion for dcache readdir use binary search to find cache index that corresponds to readdir postion. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-28 09:43:35 +00:00
u64 idx = 0;
int err = 0;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
ceph: use atomic_t for ceph_inode_info::i_shared_gen It allows accessing i_shared_gen without holding i_ceph_lock. It is preparation for later patch. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Acked-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-11-27 02:47:46 +00:00
dout("__dcache_readdir %p v%u at %llx\n", dir, (unsigned)shared_gen, ctx->pos);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
ceph: search cache postion for dcache readdir use binary search to find cache index that corresponds to readdir postion. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-28 09:43:35 +00:00
/* search start position */
if (ctx->pos > 2) {
u64 count = div_u64(i_size_read(dir), sizeof(struct dentry *));
while (count > 0) {
u64 step = count >> 1;
dentry = __dcache_find_get_entry(parent, idx + step,
&cache_ctl);
if (!dentry) {
/* use linar search */
idx = 0;
break;
}
if (IS_ERR(dentry)) {
err = PTR_ERR(dentry);
goto out;
}
di = ceph_dentry(dentry);
spin_lock(&dentry->d_lock);
if (fpos_cmp(di->offset, ctx->pos) < 0) {
idx += step + 1;
count -= step + 1;
} else {
count = step;
}
spin_unlock(&dentry->d_lock);
dput(dentry);
}
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
ceph: search cache postion for dcache readdir use binary search to find cache index that corresponds to readdir postion. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-28 09:43:35 +00:00
dout("__dcache_readdir %p cache idx %llu\n", dir, idx);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
}
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
ceph: search cache postion for dcache readdir use binary search to find cache index that corresponds to readdir postion. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-28 09:43:35 +00:00
for (;;) {
bool emit_dentry = false;
dentry = __dcache_find_get_entry(parent, idx++, &cache_ctl);
if (!dentry) {
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
dfi->file_info.flags |= CEPH_F_ATEND;
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
err = 0;
break;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
}
ceph: search cache postion for dcache readdir use binary search to find cache index that corresponds to readdir postion. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-28 09:43:35 +00:00
if (IS_ERR(dentry)) {
err = PTR_ERR(dentry);
goto out;
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
}
spin_lock(&dentry->d_lock);
ceph: avoid dereferencing invalid pointer during cached readdir Readdir cache keeps array of dentry pointers in page cache. If any dentry in readdir cache gets pruned, ceph_d_prune() disables readdir cache for later readdir syscall. The problem is that ceph_d_prune() ignores unhashed dentry. Ideally MDS should have already revoked CEPH_CAP_FILE_SHARED (which also disables readdir cache) when dentry gets unhashed. But if it is somehow MDS does not properly revoke CEPH_CAP_FILE_SHARED and the unhashed dentry gets pruned later, ceph_d_prune() will not disable readdir cache, later readdir may reference invalid dentry pointer. The fix is make ceph_d_prune() do extra check for unhashed dentry. Disable readdir cache if the unhashed dentry is still referenced by readdir cache. Another fix in this patch is handle d_splice_alias(). If a dentry gets spliced into new parent dentry, treat it as if it was pruned (call ceph_d_prune() for it). Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Acked-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-11-27 03:23:48 +00:00
di = ceph_dentry(dentry);
if (d_unhashed(dentry) ||
d_really_is_negative(dentry) ||
di->lease_shared_gen != shared_gen) {
spin_unlock(&dentry->d_lock);
dput(dentry);
err = -EAGAIN;
goto out;
}
if (fpos_cmp(ctx->pos, di->offset) <= 0) {
ceph: periodically trim stale dentries Previous commit make VFS delete stale dentry when last reference is dropped. Lease also can become invalid when corresponding dentry has no reference. This patch make cephfs periodically scan lease list, delete corresponding dentry if lease is invalid. There are two types of lease, dentry lease and dir lease. dentry lease has life time and applies to singe dentry. Dentry lease is added to tail of a list when it's updated, leases at front of the list will expire first. Dir lease is CEPH_CAP_FILE_SHARED on directory inode, it applies to all dentries in the directory. Dentries have dir leases are added to another list. Dentries in the list are periodically checked in a round robin manner. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-01-31 08:55:51 +00:00
__ceph_dentry_dir_lease_touch(di);
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
emit_dentry = true;
}
fs: dcache scale d_unhashed Protect d_unhashed(dentry) condition with d_lock. This means keeping DCACHE_UNHASHED bit in synch with hash manipulations. Signed-off-by: Nick Piggin <npiggin@kernel.dk>
2011-01-07 06:49:33 +00:00
spin_unlock(&dentry->d_lock);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
if (emit_dentry) {
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
dout(" %llx dentry %p %pd %p\n", di->offset,
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
dentry, dentry, d_inode(dentry));
ctx->pos = di->offset;
if (!dir_emit(ctx, dentry->d_name.name,
ceph: fix inode number handling on arches with 32-bit ino_t Tuan and Ulrich mentioned that they were hitting a problem on s390x, which has a 32-bit ino_t value, even though it's a 64-bit arch (for historical reasons). I think the current handling of inode numbers in the ceph driver is wrong. It tries to use 32-bit inode numbers on 32-bit arches, but that's actually not a problem. 32-bit arches can deal with 64-bit inode numbers just fine when userland code is compiled with LFS support (the common case these days). What we really want to do is just use 64-bit numbers everywhere, unless someone has mounted with the ino32 mount option. In that case, we want to ensure that we hash the inode number down to something that will fit in 32 bits before presenting the value to userland. Add new helper functions that do this, and only do the conversion before presenting these values to userland in getattr and readdir. The inode table hashvalue is changed to just cast the inode number to unsigned long, as low-order bits are the most likely to vary anyway. While it's not strictly required, we do want to put something in inode->i_ino. Instead of basing it on BITS_PER_LONG, however, base it on the size of the ino_t type. NOTE: This is a user-visible change on 32-bit arches: 1/ inode numbers will be seen to have changed between kernel versions. 32-bit arches will see large inode numbers now instead of the hashed ones they saw before. 2/ any really old software not built with LFS support may start failing stat() calls with -EOVERFLOW on inode numbers >2^32. Nothing much we can do about these, but hopefully the intersection of people running such code on ceph will be very small. The workaround for both problems is to mount with "-o ino32". [ idryomov: changelog tweak ] URL: https://tracker.ceph.com/issues/46828 Reported-by: Ulrich Weigand <Ulrich.Weigand@de.ibm.com> Reported-and-Tested-by: Tuan Hoang1 <Tuan.Hoang1@ibm.com> Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-08-18 12:03:48 +00:00
dentry->d_name.len, ceph_present_inode(d_inode(dentry)),
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
d_inode(dentry)->i_mode >> 12)) {
dput(dentry);
err = 0;
break;
}
ctx->pos++;
ceph: check directory's completeness before emitting directory entry Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com> Reviewed-by: Sage Weil <sage@inktank.com>
2014-04-08 13:42:59 +00:00
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
if (last)
dput(last);
last = dentry;
} else {
dput(dentry);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
}
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
}
ceph: search cache postion for dcache readdir use binary search to find cache index that corresponds to readdir postion. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-28 09:43:35 +00:00
out:
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
ceph_readdir_cache_release(&cache_ctl);
if (last) {
int ret;
di = ceph_dentry(last);
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
ret = note_last_dentry(dfi, last->d_name.name, last->d_name.len,
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
fpos_off(di->offset) + 1);
if (ret < 0)
err = ret;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
dput(last);
ceph: fix race in concurrent readdir For a large directory, program needs to issue multiple readdir syscalls to get all dentries. When there are multiple programs read the directory concurrently. Following sequence of events can happen. - program calls readdir with pos = 2. ceph sends readdir request to mds. The reply contains N1 entries. ceph adds these N1 entries to readdir cache. - program calls readdir with pos = N1+2. The readdir is satisfied by the readdir cache, N2 entries are returned. (Other program calls readdir in the middle, which fills the cache) - program calls readdir with pos = N1+N2+2. ceph sends readdir request to mds. The reply contains N3 entries and it reaches directory end. ceph adds these N3 entries to the readdir cache and marks directory complete. The second readdir call does not update fi->readdir_cache_idx. ceph add the last N3 entries to wrong places. Cc: stable@vger.kernel.org # v4.3+ Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-07-06 03:12:21 +00:00
/* last_name no longer match cache index */
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
if (dfi->readdir_cache_idx >= 0) {
dfi->readdir_cache_idx = -1;
dfi->dir_release_count = 0;
ceph: fix race in concurrent readdir For a large directory, program needs to issue multiple readdir syscalls to get all dentries. When there are multiple programs read the directory concurrently. Following sequence of events can happen. - program calls readdir with pos = 2. ceph sends readdir request to mds. The reply contains N1 entries. ceph adds these N1 entries to readdir cache. - program calls readdir with pos = N1+2. The readdir is satisfied by the readdir cache, N2 entries are returned. (Other program calls readdir in the middle, which fills the cache) - program calls readdir with pos = N1+N2+2. ceph sends readdir request to mds. The reply contains N3 entries and it reaches directory end. ceph adds these N3 entries to the readdir cache and marks directory complete. The second readdir call does not update fi->readdir_cache_idx. ceph add the last N3 entries to wrong places. Cc: stable@vger.kernel.org # v4.3+ Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-07-06 03:12:21 +00:00
}
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
}
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
return err;
}
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
static bool need_send_readdir(struct ceph_dir_file_info *dfi, loff_t pos)
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
{
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
if (!dfi->last_readdir)
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
return true;
if (is_hash_order(pos))
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
return !ceph_frag_contains_value(dfi->frag, fpos_hash(pos));
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
else
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
return dfi->frag != fpos_frag(pos);
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
}
[readdir] convert ceph Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2013-05-17 20:52:26 +00:00
static int ceph_readdir(struct file *file, struct dir_context *ctx)
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
{
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
struct ceph_dir_file_info *dfi = file->private_data;
[readdir] convert ceph Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2013-05-17 20:52:26 +00:00
struct inode *inode = file_inode(file);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
struct ceph_inode_info *ci = ceph_inode(inode);
ceph: factor out libceph from Ceph file system This factors out protocol and low-level storage parts of ceph into a separate libceph module living in net/ceph and include/linux/ceph. This is mostly a matter of moving files around. However, a few key pieces of the interface change as well: - ceph_client becomes ceph_fs_client and ceph_client, where the latter captures the mon and osd clients, and the fs_client gets the mds client and file system specific pieces. - Mount option parsing and debugfs setup is correspondingly broken into two pieces. - The mon client gets a generic handler callback for otherwise unknown messages (mds map, in this case). - The basic supported/required feature bits can be expanded (and are by ceph_fs_client). No functional change, aside from some subtle error handling cases that got cleaned up in the refactoring process. Signed-off-by: Sage Weil <sage@newdream.net>
2010-04-06 22:14:15 +00:00
struct ceph_fs_client *fsc = ceph_inode_to_client(inode);
struct ceph_mds_client *mdsc = fsc->mdsc;
ceph: record 'offset' for each entry of readdir result This is preparation for using hash value as dentry 'offset' Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-28 07:17:40 +00:00
int i;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
int err;
ceph: choose readdir frag based on previous readdir reply The dirfragtree is lazily updated, it's not always accurate. Infinite loops happens in following circumstance. - client send request to read frag A - frag A has been fragmented into frag B and C. So mds fills the reply with contents of frag B - client wants to read next frag C. ceph_choose_frag(frag value of C) return frag A. The fix is using previous readdir reply to calculate next readdir frag when possible. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-04-24 03:56:50 +00:00
unsigned frag = -1;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
struct ceph_mds_reply_info_parsed *rinfo;
ceph: record 'offset' for each entry of readdir result This is preparation for using hash value as dentry 'offset' Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-28 07:17:40 +00:00
dout("readdir %p file %p pos %llx\n", inode, file, ctx->pos);
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
if (dfi->file_info.flags & CEPH_F_ATEND)
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
return 0;
/* always start with . and .. */
[readdir] convert ceph Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2013-05-17 20:52:26 +00:00
if (ctx->pos == 0) {
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
dout("readdir off 0 -> '.'\n");
ceph: fix inode number handling on arches with 32-bit ino_t Tuan and Ulrich mentioned that they were hitting a problem on s390x, which has a 32-bit ino_t value, even though it's a 64-bit arch (for historical reasons). I think the current handling of inode numbers in the ceph driver is wrong. It tries to use 32-bit inode numbers on 32-bit arches, but that's actually not a problem. 32-bit arches can deal with 64-bit inode numbers just fine when userland code is compiled with LFS support (the common case these days). What we really want to do is just use 64-bit numbers everywhere, unless someone has mounted with the ino32 mount option. In that case, we want to ensure that we hash the inode number down to something that will fit in 32 bits before presenting the value to userland. Add new helper functions that do this, and only do the conversion before presenting these values to userland in getattr and readdir. The inode table hashvalue is changed to just cast the inode number to unsigned long, as low-order bits are the most likely to vary anyway. While it's not strictly required, we do want to put something in inode->i_ino. Instead of basing it on BITS_PER_LONG, however, base it on the size of the ino_t type. NOTE: This is a user-visible change on 32-bit arches: 1/ inode numbers will be seen to have changed between kernel versions. 32-bit arches will see large inode numbers now instead of the hashed ones they saw before. 2/ any really old software not built with LFS support may start failing stat() calls with -EOVERFLOW on inode numbers >2^32. Nothing much we can do about these, but hopefully the intersection of people running such code on ceph will be very small. The workaround for both problems is to mount with "-o ino32". [ idryomov: changelog tweak ] URL: https://tracker.ceph.com/issues/46828 Reported-by: Ulrich Weigand <Ulrich.Weigand@de.ibm.com> Reported-and-Tested-by: Tuan Hoang1 <Tuan.Hoang1@ibm.com> Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-08-18 12:03:48 +00:00
if (!dir_emit(ctx, ".", 1, ceph_present_inode(inode),
[readdir] convert ceph Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2013-05-17 20:52:26 +00:00
inode->i_mode >> 12))
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
return 0;
[readdir] convert ceph Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2013-05-17 20:52:26 +00:00
ctx->pos = 1;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
}
[readdir] convert ceph Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2013-05-17 20:52:26 +00:00
if (ctx->pos == 1) {
ceph: fix inode number handling on arches with 32-bit ino_t Tuan and Ulrich mentioned that they were hitting a problem on s390x, which has a 32-bit ino_t value, even though it's a 64-bit arch (for historical reasons). I think the current handling of inode numbers in the ceph driver is wrong. It tries to use 32-bit inode numbers on 32-bit arches, but that's actually not a problem. 32-bit arches can deal with 64-bit inode numbers just fine when userland code is compiled with LFS support (the common case these days). What we really want to do is just use 64-bit numbers everywhere, unless someone has mounted with the ino32 mount option. In that case, we want to ensure that we hash the inode number down to something that will fit in 32 bits before presenting the value to userland. Add new helper functions that do this, and only do the conversion before presenting these values to userland in getattr and readdir. The inode table hashvalue is changed to just cast the inode number to unsigned long, as low-order bits are the most likely to vary anyway. While it's not strictly required, we do want to put something in inode->i_ino. Instead of basing it on BITS_PER_LONG, however, base it on the size of the ino_t type. NOTE: This is a user-visible change on 32-bit arches: 1/ inode numbers will be seen to have changed between kernel versions. 32-bit arches will see large inode numbers now instead of the hashed ones they saw before. 2/ any really old software not built with LFS support may start failing stat() calls with -EOVERFLOW on inode numbers >2^32. Nothing much we can do about these, but hopefully the intersection of people running such code on ceph will be very small. The workaround for both problems is to mount with "-o ino32". [ idryomov: changelog tweak ] URL: https://tracker.ceph.com/issues/46828 Reported-by: Ulrich Weigand <Ulrich.Weigand@de.ibm.com> Reported-and-Tested-by: Tuan Hoang1 <Tuan.Hoang1@ibm.com> Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-08-18 12:03:48 +00:00
u64 ino;
struct dentry *dentry = file->f_path.dentry;
spin_lock(&dentry->d_lock);
ino = ceph_present_inode(dentry->d_parent->d_inode);
spin_unlock(&dentry->d_lock);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
dout("readdir off 1 -> '..'\n");
ceph: fix inode number handling on arches with 32-bit ino_t Tuan and Ulrich mentioned that they were hitting a problem on s390x, which has a 32-bit ino_t value, even though it's a 64-bit arch (for historical reasons). I think the current handling of inode numbers in the ceph driver is wrong. It tries to use 32-bit inode numbers on 32-bit arches, but that's actually not a problem. 32-bit arches can deal with 64-bit inode numbers just fine when userland code is compiled with LFS support (the common case these days). What we really want to do is just use 64-bit numbers everywhere, unless someone has mounted with the ino32 mount option. In that case, we want to ensure that we hash the inode number down to something that will fit in 32 bits before presenting the value to userland. Add new helper functions that do this, and only do the conversion before presenting these values to userland in getattr and readdir. The inode table hashvalue is changed to just cast the inode number to unsigned long, as low-order bits are the most likely to vary anyway. While it's not strictly required, we do want to put something in inode->i_ino. Instead of basing it on BITS_PER_LONG, however, base it on the size of the ino_t type. NOTE: This is a user-visible change on 32-bit arches: 1/ inode numbers will be seen to have changed between kernel versions. 32-bit arches will see large inode numbers now instead of the hashed ones they saw before. 2/ any really old software not built with LFS support may start failing stat() calls with -EOVERFLOW on inode numbers >2^32. Nothing much we can do about these, but hopefully the intersection of people running such code on ceph will be very small. The workaround for both problems is to mount with "-o ino32". [ idryomov: changelog tweak ] URL: https://tracker.ceph.com/issues/46828 Reported-by: Ulrich Weigand <Ulrich.Weigand@de.ibm.com> Reported-and-Tested-by: Tuan Hoang1 <Tuan.Hoang1@ibm.com> Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-08-18 12:03:48 +00:00
if (!dir_emit(ctx, "..", 2, ino, inode->i_mode >> 12))
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
return 0;
[readdir] convert ceph Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2013-05-17 20:52:26 +00:00
ctx->pos = 2;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
}
ceph: use i_ceph_lock instead of i_lock We have been using i_lock to protect all kinds of data structures in the ceph_inode_info struct, including lists of inodes that we need to iterate over while avoiding races with inode destruction. That requires grabbing a reference to the inode with the list lock protected, but igrab() now takes i_lock to check the inode flags. Changing the list lock ordering would be a painful process. However, using a ceph-specific i_ceph_lock in the ceph inode instead of i_lock is a simple mechanical change and avoids the ordering constraints imposed by igrab(). Reported-by: Amon Ott <a.ott@m-privacy.de> Signed-off-by: Sage Weil <sage@newdream.net>
2011-11-30 17:47:09 +00:00
spin_lock(&ci->i_ceph_lock);
ceph: consider inode's last read/write when calculating wanted caps Add i_last_rd and i_last_wr to ceph_inode_info. These fields are used to track the last time the client acquired read/write caps for the inode. If there is no read/write on an inode for 'caps_wanted_delay_max' seconds, __ceph_caps_file_wanted() does not request caps for read/write even there are open files. Call __ceph_touch_fmode() for dir operations. __ceph_caps_file_wanted() calculates dir's wanted caps according to last dir read/modification. If there is recent dir read, dir inode wants CEPH_CAP_ANY_SHARED caps. If there is recent dir modification, also wants CEPH_CAP_FILE_EXCL. Readdir is a special case. Dir inode wants CEPH_CAP_FILE_EXCL after readdir, as with that, modifications do not need to release CEPH_CAP_FILE_SHARED or invalidate all dentry leases issued by readdir. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-03-05 12:21:00 +00:00
/* request Fx cap. if have Fx, we don't need to release Fs cap
* for later create/unlink. */
__ceph_touch_fmode(ci, mdsc, CEPH_FILE_MODE_WR);
/* can we use the dcache? */
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
if (ceph_test_mount_opt(fsc, DCACHE) &&
ceph: factor out libceph from Ceph file system This factors out protocol and low-level storage parts of ceph into a separate libceph module living in net/ceph and include/linux/ceph. This is mostly a matter of moving files around. However, a few key pieces of the interface change as well: - ceph_client becomes ceph_fs_client and ceph_client, where the latter captures the mon and osd clients, and the fs_client gets the mds client and file system specific pieces. - Mount option parsing and debugfs setup is correspondingly broken into two pieces. - The mon client gets a generic handler callback for otherwise unknown messages (mds map, in this case). - The basic supported/required feature bits can be expanded (and are by ceph_fs_client). No functional change, aside from some subtle error handling cases that got cleaned up in the refactoring process. Signed-off-by: Sage Weil <sage@newdream.net>
2010-04-06 22:14:15 +00:00
!ceph_test_mount_opt(fsc, NOASYNCREADDIR) &&
ceph: avoid dcache readdir for snapdir We should always go to the MDS for readdir on the hidden snapdir. The set of snapshots can change at any time; the client can't trust its cache for that. Signed-off-by: Sage Weil <sage@newdream.net>
2010-07-22 20:47:21 +00:00
ceph_snap(inode) != CEPH_SNAPDIR &&
ceph: introduce a new inode flag indicating if cached dentries are ordered After creating/deleting/renaming file, offsets of sibling dentries may change. So we can not use cached dentries to satisfy readdir. But we can still use the cached dentries to conclude -ENOENT for lookup. This patch introduces a new inode flag indicating if child dentries are ordered. The flag is set at the same time marking a directory complete. After creating/deleting/renaming file, we clear the flag on directory inode. This prevents ceph_readdir() from using cached dentries to satisfy readdir syscall. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2014-10-22 01:09:56 +00:00
__ceph_dir_is_complete_ordered(ci) &&
ceph: add caps perf metric for each superblock Count hits and misses in the caps cache. If the client has all of the necessary caps when a task needs references, then it's counted as a hit. Any other situation is a miss. URL: https://tracker.ceph.com/issues/43215 Signed-off-by: Xiubo Li <xiubli@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-03-20 03:45:00 +00:00
__ceph_caps_issued_mask_metric(ci, CEPH_CAP_FILE_SHARED, 1)) {
ceph: use atomic_t for ceph_inode_info::i_shared_gen It allows accessing i_shared_gen without holding i_ceph_lock. It is preparation for later patch. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Acked-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-11-27 02:47:46 +00:00
int shared_gen = atomic_read(&ci->i_shared_gen);
ceph: add caps perf metric for each superblock Count hits and misses in the caps cache. If the client has all of the necessary caps when a task needs references, then it's counted as a hit. Any other situation is a miss. URL: https://tracker.ceph.com/issues/43215 Signed-off-by: Xiubo Li <xiubli@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-03-20 03:45:00 +00:00
ceph: use i_ceph_lock instead of i_lock We have been using i_lock to protect all kinds of data structures in the ceph_inode_info struct, including lists of inodes that we need to iterate over while avoiding races with inode destruction. That requires grabbing a reference to the inode with the list lock protected, but igrab() now takes i_lock to check the inode flags. Changing the list lock ordering would be a painful process. However, using a ceph-specific i_ceph_lock in the ceph inode instead of i_lock is a simple mechanical change and avoids the ordering constraints imposed by igrab(). Reported-by: Amon Ott <a.ott@m-privacy.de> Signed-off-by: Sage Weil <sage@newdream.net>
2011-11-30 17:47:09 +00:00
spin_unlock(&ci->i_ceph_lock);
ceph: skip invalid dentry during dcache readdir skip dentries that were added before MDS issued FILE_SHARED to client. Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com> Reviewed-by: Sage Weil <sage@inktank.com>
2014-04-06 06:10:04 +00:00
err = __dcache_readdir(file, ctx, shared_gen);
ceph: do not carry i_lock for readdir from dcache We were taking dcache_lock inside of i_lock, which introduces a dependency not found elsewhere in the kernel, complicationg the vfs locking scalability work. Since we don't actually need it here anyway, remove it. We only need i_lock to test for the I_COMPLETE flag, so be careful to do so without dcache_lock held. Signed-off-by: Sage Weil <sage@newdream.net>
2010-10-18 21:04:31 +00:00
if (err != -EAGAIN)
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
return err;
ceph: do not carry i_lock for readdir from dcache We were taking dcache_lock inside of i_lock, which introduces a dependency not found elsewhere in the kernel, complicationg the vfs locking scalability work. Since we don't actually need it here anyway, remove it. We only need i_lock to test for the I_COMPLETE flag, so be careful to do so without dcache_lock held. Signed-off-by: Sage Weil <sage@newdream.net>
2010-10-18 21:04:31 +00:00
} else {
ceph: use i_ceph_lock instead of i_lock We have been using i_lock to protect all kinds of data structures in the ceph_inode_info struct, including lists of inodes that we need to iterate over while avoiding races with inode destruction. That requires grabbing a reference to the inode with the list lock protected, but igrab() now takes i_lock to check the inode flags. Changing the list lock ordering would be a painful process. However, using a ceph-specific i_ceph_lock in the ceph inode instead of i_lock is a simple mechanical change and avoids the ordering constraints imposed by igrab(). Reported-by: Amon Ott <a.ott@m-privacy.de> Signed-off-by: Sage Weil <sage@newdream.net>
2011-11-30 17:47:09 +00:00
spin_unlock(&ci->i_ceph_lock);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
}
/* proceed with a normal readdir */
more:
/* do we have the correct frag content buffered? */
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
if (need_send_readdir(dfi, ctx->pos)) {
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
struct ceph_mds_request *req;
int op = ceph_snap(inode) == CEPH_SNAPDIR ?
CEPH_MDS_OP_LSSNAP : CEPH_MDS_OP_READDIR;
/* discard old result, if any */
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
if (dfi->last_readdir) {
ceph_mdsc_put_request(dfi->last_readdir);
dfi->last_readdir = NULL;
ceph: fix possible double-free of mds request reference Clear pointer to mds request after dropping the reference to ensure we don't drop it again, as there is at least one error path through this function that does not reset fi->last_readdir to a new value. Signed-off-by: Sage Weil <sage@newdream.net>
2010-03-10 20:03:32 +00:00
}
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
if (is_hash_order(ctx->pos)) {
ceph: choose readdir frag based on previous readdir reply The dirfragtree is lazily updated, it's not always accurate. Infinite loops happens in following circumstance. - client send request to read frag A - frag A has been fragmented into frag B and C. So mds fills the reply with contents of frag B - client wants to read next frag C. ceph_choose_frag(frag value of C) return frag A. The fix is using previous readdir reply to calculate next readdir frag when possible. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-04-24 03:56:50 +00:00
/* fragtree isn't always accurate. choose frag
* based on previous reply when possible. */
if (frag == (unsigned)-1)
frag = ceph_choose_frag(ci, fpos_hash(ctx->pos),
NULL, NULL);
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
} else {
frag = fpos_frag(ctx->pos);
}
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
dout("readdir fetching %llx.%llx frag %x offset '%s'\n",
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
ceph_vinop(inode), frag, dfi->last_name);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
req = ceph_mdsc_create_request(mdsc, op, USE_AUTH_MDS);
if (IS_ERR(req))
return PTR_ERR(req);
ceph: preallocate buffer for readdir reply Preallocate buffer for readdir reply. Limit number of entries in readdir reply according to the buffer size. Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
2014-03-29 05:41:15 +00:00
err = ceph_alloc_readdir_reply_buffer(req, inode);
if (err) {
ceph_mdsc_put_request(req);
return err;
}
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
/* hints to request -> mds selection code */
req->r_direct_mode = USE_AUTH_MDS;
ceph: send LSSNAP request to auth mds of directory inode Snapdir inode has no capability. __choose_mds() should choose mds base on capabilities of snapdir's parent inode. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-07-26 04:48:08 +00:00
if (op == CEPH_MDS_OP_READDIR) {
req->r_direct_hash = ceph_frag_value(frag);
__set_bit(CEPH_MDS_R_DIRECT_IS_HASH, &req->r_req_flags);
ceph: voluntarily drop Fx cap for readdir request MDS need to rdlock directory inode's filelock when handling readdir request. Voluntarily dropping CEPH_CAP_AUTH_EXCL avoids a cap revoke message. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-11-23 10:28:16 +00:00
req->r_inode_drop = CEPH_CAP_FILE_EXCL;
ceph: send LSSNAP request to auth mds of directory inode Snapdir inode has no capability. __choose_mds() should choose mds base on capabilities of snapdir's parent inode. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-07-26 04:48:08 +00:00
}
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
if (dfi->last_name) {
req->r_path2 = kstrdup(dfi->last_name, GFP_KERNEL);
ceph: kstrdup() memory handling Currently, there is no check for the kstrdup() for r_path2, r_path1 and snapdir_name as various locations as there is a possibility of failure during memory pressure. Therefore, returning ENOMEM where the checks have been missed. Signed-off-by: Sanidhya Kashyap <sanidhya.gatech@gmail.com> Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-03-21 16:54:58 +00:00
if (!req->r_path2) {
ceph_mdsc_put_request(req);
return -ENOMEM;
}
ceph: make seeky readdir more efficient Current cephfs client uses string to indicate start position of readdir. The string is last entry of previous readdir reply. This approach does not work for seeky readdir because we can not easily convert the new postion to a string. For seeky readdir, mds needs to return dentries from the beginning. Client keeps retrying if the reply does not contain the dentry it wants. In current version of ceph, mds sorts CDentry in its cache in hash order. Client also uses dentry hash to compose dir postion. For seeky readdir, if client passes the hash part of dir postion to mds. mds can avoid replying useless dentries. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-04-05 16:54:05 +00:00
} else if (is_hash_order(ctx->pos)) {
req->r_args.readdir.offset_hash =
cpu_to_le32(fpos_hash(ctx->pos));
ceph: kstrdup() memory handling Currently, there is no check for the kstrdup() for r_path2, r_path1 and snapdir_name as various locations as there is a possibility of failure during memory pressure. Therefore, returning ENOMEM where the checks have been missed. Signed-off-by: Sanidhya Kashyap <sanidhya.gatech@gmail.com> Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-03-21 16:54:58 +00:00
}
ceph: make seeky readdir more efficient Current cephfs client uses string to indicate start position of readdir. The string is last entry of previous readdir reply. This approach does not work for seeky readdir because we can not easily convert the new postion to a string. For seeky readdir, mds needs to return dentries from the beginning. Client keeps retrying if the reply does not contain the dentry it wants. In current version of ceph, mds sorts CDentry in its cache in hash order. Client also uses dentry hash to compose dir postion. For seeky readdir, if client passes the hash part of dir postion to mds. mds can avoid replying useless dentries. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-04-05 16:54:05 +00:00
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
req->r_dir_release_cnt = dfi->dir_release_count;
req->r_dir_ordered_cnt = dfi->dir_ordered_count;
req->r_readdir_cache_idx = dfi->readdir_cache_idx;
req->r_readdir_offset = dfi->next_offset;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
req->r_args.readdir.frag = cpu_to_le32(frag);
ceph: define 'end/complete' in readdir reply as bit flags Set a flag in readdir request, which indicates that client interprets 'end/complete' as bit flags. So that mds can reply additional flags in readdir reply. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-27 09:48:30 +00:00
req->r_args.readdir.flags =
cpu_to_le16(CEPH_READDIR_REPLY_BITFLAGS);
ceph: kstrdup() memory handling Currently, there is no check for the kstrdup() for r_path2, r_path1 and snapdir_name as various locations as there is a possibility of failure during memory pressure. Therefore, returning ENOMEM where the checks have been missed. Signed-off-by: Sanidhya Kashyap <sanidhya.gatech@gmail.com> Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-03-21 16:54:58 +00:00
req->r_inode = inode;
ihold(inode);
req->r_dentry = dget(file->f_path.dentry);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
err = ceph_mdsc_do_request(mdsc, NULL, req);
if (err < 0) {
ceph_mdsc_put_request(req);
return err;
}
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
dout("readdir got and parsed readdir result=%d on "
"frag %x, end=%d, complete=%d, hash_order=%d\n",
err, frag,
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
(int)req->r_reply_info.dir_end,
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
(int)req->r_reply_info.dir_complete,
(int)req->r_reply_info.hash_order);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
ceph: handle frag mismatch between readdir request and reply If client has outdated directory fragments information, it may request readdir an non-existent directory fragment. In this case, the MDS finds an approximate directory fragment and sends its contents back to the client. When receiving a reply with fragment that is different than the requested one, the client need to reset the 'readdir offset'. Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com> Reviewed-by: Sage Weil <sage@inktank.com>
2013-09-18 01:44:13 +00:00
rinfo = &req->r_reply_info;
if (le32_to_cpu(rinfo->dir_dir->frag) != frag) {
frag = le32_to_cpu(rinfo->dir_dir->frag);
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
if (!rinfo->hash_order) {
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
dfi->next_offset = req->r_readdir_offset;
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
/* adjust ctx->pos to beginning of frag */
ctx->pos = ceph_make_fpos(frag,
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
dfi->next_offset,
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
false);
}
ceph: handle frag mismatch between readdir request and reply If client has outdated directory fragments information, it may request readdir an non-existent directory fragment. In this case, the MDS finds an approximate directory fragment and sends its contents back to the client. When receiving a reply with fragment that is different than the requested one, the client need to reset the 'readdir offset'. Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com> Reviewed-by: Sage Weil <sage@inktank.com>
2013-09-18 01:44:13 +00:00
}
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
dfi->frag = frag;
dfi->last_readdir = req;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
ceph: convert bools in ceph_mds_request to a new r_req_flags field Currently, we have a bunch of bool flags in struct ceph_mds_request. We need more flags though, but each bool takes (at least) a byte. Those add up over time. Merge all of the existing bools in this struct into a single unsigned long, and use the set/test/clear_bit macros to manipulate them. These are atomic operations, but that is required here to prevent load/modify/store races. The existing flags are protected by different locks, so we can't rely on them for that purpose. Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Yan, Zheng <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-02-01 18:49:09 +00:00
if (test_bit(CEPH_MDS_R_DID_PREPOPULATE, &req->r_req_flags)) {
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
dfi->readdir_cache_idx = req->r_readdir_cache_idx;
if (dfi->readdir_cache_idx < 0) {
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
/* preclude from marking dir ordered */
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
dfi->dir_ordered_count = 0;
ceph: record 'offset' for each entry of readdir result This is preparation for using hash value as dentry 'offset' Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-28 07:17:40 +00:00
} else if (ceph_frag_is_leftmost(frag) &&
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
dfi->next_offset == 2) {
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
/* note dir version at start of readdir so
* we can tell if any dentries get dropped */
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
dfi->dir_release_count = req->r_dir_release_cnt;
dfi->dir_ordered_count = req->r_dir_ordered_cnt;
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
}
} else {
ceph: add newline to end of debug message format Some of dout format do not include newline in the end, fix for the files which are in fs/ceph and net/ceph directories, and changing printk to dout for printing debug info in super.c Signed-off-by: Chengguang Xu <cgxu519@icloud.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-01-30 08:29:17 +00:00
dout("readdir !did_prepopulate\n");
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
/* disable readdir cache */
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
dfi->readdir_cache_idx = -1;
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
/* preclude from marking dir complete */
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
dfi->dir_release_count = 0;
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
}
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
/* note next offset and last dentry name */
if (rinfo->dir_nr > 0) {
ceph: define struct for dir entry in readdir reply This avoids defining multiple arrays for entries in readdir reply Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-28 01:37:39 +00:00
struct ceph_mds_reply_dir_entry *rde =
rinfo->dir_entries + (rinfo->dir_nr-1);
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
unsigned next_offset = req->r_reply_info.dir_end ?
2 : (fpos_off(rde->offset) + 1);
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
err = note_last_dentry(dfi, rde->name, rde->name_len,
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
next_offset);
ceph: fix memory leak in ceph_readdir when note_last_dentry returns error Reset the last_readdir at the same time, and add a comment explaining why we don't free last_readdir when dir_emit returns false. Signed-off-by: Xiubo Li <xiubli@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2022-03-05 11:52:59 +00:00
if (err) {
ceph_mdsc_put_request(dfi->last_readdir);
dfi->last_readdir = NULL;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
return err;
ceph: fix memory leak in ceph_readdir when note_last_dentry returns error Reset the last_readdir at the same time, and add a comment explaining why we don't free last_readdir when dir_emit returns false. Signed-off-by: Xiubo Li <xiubli@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2022-03-05 11:52:59 +00:00
}
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
} else if (req->r_reply_info.dir_end) {
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
dfi->next_offset = 2;
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
/* keep last name */
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
}
}
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
rinfo = &dfi->last_readdir->r_reply_info;
ceph: record 'offset' for each entry of readdir result This is preparation for using hash value as dentry 'offset' Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-28 07:17:40 +00:00
dout("readdir frag %x num %d pos %llx chunk first %llx\n",
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
dfi->frag, rinfo->dir_nr, ctx->pos,
ceph: record 'offset' for each entry of readdir result This is preparation for using hash value as dentry 'offset' Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-28 07:17:40 +00:00
rinfo->dir_nr ? rinfo->dir_entries[0].offset : 0LL);
[readdir] convert ceph Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2013-05-17 20:52:26 +00:00
ceph: record 'offset' for each entry of readdir result This is preparation for using hash value as dentry 'offset' Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-28 07:17:40 +00:00
i = 0;
/* search start position */
if (rinfo->dir_nr > 0) {
int step, nr = rinfo->dir_nr;
while (nr > 0) {
step = nr >> 1;
if (rinfo->dir_entries[i + step].offset < ctx->pos) {
i += step + 1;
nr -= step + 1;
} else {
nr = step;
}
}
}
for (; i < rinfo->dir_nr; i++) {
struct ceph_mds_reply_dir_entry *rde = rinfo->dir_entries + i;
ceph: fix readdir EOVERFLOW on 32-bit archs One of the readdir filldir_t callers was passing the raw ceph 64-bit ino instead of the hashed 32-bit one, producing an EOVERFLOW in the filler callback. Fix this by calling the ceph_vino_to_ino() helper to do the conversion. Reported-by: Jan Smets <jan.smets@alcatel-lucent.com> Tested-by: Jan Smets <jan.smets@alcatel-lucent.com> Signed-off-by: Sage Weil <sage@newdream.net>
2010-11-18 17:15:07 +00:00
ceph: record 'offset' for each entry of readdir result This is preparation for using hash value as dentry 'offset' Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-28 07:17:40 +00:00
BUG_ON(rde->offset < ctx->pos);
ctx->pos = rde->offset;
dout("readdir (%d/%d) -> %llx '%.*s' %p\n",
i, rinfo->dir_nr, ctx->pos,
ceph: define struct for dir entry in readdir reply This avoids defining multiple arrays for entries in readdir reply Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-28 01:37:39 +00:00
rde->name_len, rde->name, &rde->inode.in);
ceph: record 'offset' for each entry of readdir result This is preparation for using hash value as dentry 'offset' Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-28 07:17:40 +00:00
ceph: define struct for dir entry in readdir reply This avoids defining multiple arrays for entries in readdir reply Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-28 01:37:39 +00:00
BUG_ON(!rde->inode.in);
ceph: record 'offset' for each entry of readdir result This is preparation for using hash value as dentry 'offset' Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-28 07:17:40 +00:00
ceph: define struct for dir entry in readdir reply This avoids defining multiple arrays for entries in readdir reply Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-28 01:37:39 +00:00
if (!dir_emit(ctx, rde->name, rde->name_len,
ceph: fix inode number handling on arches with 32-bit ino_t Tuan and Ulrich mentioned that they were hitting a problem on s390x, which has a 32-bit ino_t value, even though it's a 64-bit arch (for historical reasons). I think the current handling of inode numbers in the ceph driver is wrong. It tries to use 32-bit inode numbers on 32-bit arches, but that's actually not a problem. 32-bit arches can deal with 64-bit inode numbers just fine when userland code is compiled with LFS support (the common case these days). What we really want to do is just use 64-bit numbers everywhere, unless someone has mounted with the ino32 mount option. In that case, we want to ensure that we hash the inode number down to something that will fit in 32 bits before presenting the value to userland. Add new helper functions that do this, and only do the conversion before presenting these values to userland in getattr and readdir. The inode table hashvalue is changed to just cast the inode number to unsigned long, as low-order bits are the most likely to vary anyway. While it's not strictly required, we do want to put something in inode->i_ino. Instead of basing it on BITS_PER_LONG, however, base it on the size of the ino_t type. NOTE: This is a user-visible change on 32-bit arches: 1/ inode numbers will be seen to have changed between kernel versions. 32-bit arches will see large inode numbers now instead of the hashed ones they saw before. 2/ any really old software not built with LFS support may start failing stat() calls with -EOVERFLOW on inode numbers >2^32. Nothing much we can do about these, but hopefully the intersection of people running such code on ceph will be very small. The workaround for both problems is to mount with "-o ino32". [ idryomov: changelog tweak ] URL: https://tracker.ceph.com/issues/46828 Reported-by: Ulrich Weigand <Ulrich.Weigand@de.ibm.com> Reported-and-Tested-by: Tuan Hoang1 <Tuan.Hoang1@ibm.com> Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-08-18 12:03:48 +00:00
ceph_present_ino(inode->i_sb, le64_to_cpu(rde->inode.in->ino)),
le32_to_cpu(rde->inode.in->mode) >> 12)) {
ceph: fix memory leak in ceph_readdir when note_last_dentry returns error Reset the last_readdir at the same time, and add a comment explaining why we don't free last_readdir when dir_emit returns false. Signed-off-by: Xiubo Li <xiubli@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2022-03-05 11:52:59 +00:00
/*
* NOTE: Here no need to put the 'dfi->last_readdir',
* because when dir_emit stops us it's most likely
* doesn't have enough memory, etc. So for next readdir
* it will continue.
*/
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
dout("filldir stopping us...\n");
return 0;
}
[readdir] convert ceph Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2013-05-17 20:52:26 +00:00
ctx->pos++;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
}
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
ceph_mdsc_put_request(dfi->last_readdir);
dfi->last_readdir = NULL;
ceph: choose readdir frag based on previous readdir reply The dirfragtree is lazily updated, it's not always accurate. Infinite loops happens in following circumstance. - client send request to read frag A - frag A has been fragmented into frag B and C. So mds fills the reply with contents of frag B - client wants to read next frag C. ceph_choose_frag(frag value of C) return frag A. The fix is using previous readdir reply to calculate next readdir frag when possible. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-04-24 03:56:50 +00:00
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
if (dfi->next_offset > 2) {
frag = dfi->frag;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
goto more;
}
/* more frags? */
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
if (!ceph_frag_is_rightmost(dfi->frag)) {
frag = ceph_frag_next(dfi->frag);
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
if (is_hash_order(ctx->pos)) {
loff_t new_pos = ceph_make_fpos(ceph_frag_value(frag),
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
dfi->next_offset, true);
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
if (new_pos > ctx->pos)
ctx->pos = new_pos;
/* keep last_name */
} else {
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
ctx->pos = ceph_make_fpos(frag, dfi->next_offset,
false);
kfree(dfi->last_name);
dfi->last_name = NULL;
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
}
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
dout("readdir next frag is %x\n", frag);
goto more;
}
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
dfi->file_info.flags |= CEPH_F_ATEND;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
/*
* if dir_release_count still matches the dir, no dentries
* were released during the whole readdir, and we should have
* the complete dir contents in our cache.
*/
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
if (atomic64_read(&ci->i_release_count) ==
dfi->dir_release_count) {
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
spin_lock(&ci->i_ceph_lock);
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
if (dfi->dir_ordered_count ==
atomic64_read(&ci->i_ordered_count)) {
ceph: introduce a new inode flag indicating if cached dentries are ordered After creating/deleting/renaming file, offsets of sibling dentries may change. So we can not use cached dentries to satisfy readdir. But we can still use the cached dentries to conclude -ENOENT for lookup. This patch introduces a new inode flag indicating if child dentries are ordered. The flag is set at the same time marking a directory complete. After creating/deleting/renaming file, we clear the flag on directory inode. This prevents ceph_readdir() from using cached dentries to satisfy readdir syscall. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2014-10-22 01:09:56 +00:00
dout(" marking %p complete and ordered\n", inode);
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
/* use i_size to track number of entries in
* readdir cache */
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
BUG_ON(dfi->readdir_cache_idx < 0);
i_size_write(inode, dfi->readdir_cache_idx *
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
sizeof(struct dentry*));
} else {
ceph: introduce a new inode flag indicating if cached dentries are ordered After creating/deleting/renaming file, offsets of sibling dentries may change. So we can not use cached dentries to satisfy readdir. But we can still use the cached dentries to conclude -ENOENT for lookup. This patch introduces a new inode flag indicating if child dentries are ordered. The flag is set at the same time marking a directory complete. After creating/deleting/renaming file, we clear the flag on directory inode. This prevents ceph_readdir() from using cached dentries to satisfy readdir syscall. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2014-10-22 01:09:56 +00:00
dout(" marking %p complete\n", inode);
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
}
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
__ceph_dir_set_complete(ci, dfi->dir_release_count,
dfi->dir_ordered_count);
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
spin_unlock(&ci->i_ceph_lock);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
}
[readdir] convert ceph Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2013-05-17 20:52:26 +00:00
dout("readdir %p file %p done.\n", inode, file);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
return 0;
}
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
static void reset_readdir(struct ceph_dir_file_info *dfi)
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
{
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
if (dfi->last_readdir) {
ceph_mdsc_put_request(dfi->last_readdir);
dfi->last_readdir = NULL;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
}
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
kfree(dfi->last_name);
dfi->last_name = NULL;
dfi->dir_release_count = 0;
dfi->readdir_cache_idx = -1;
dfi->next_offset = 2; /* compensate for . and .. */
dfi->file_info.flags &= ~CEPH_F_ATEND;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
}
ceph: record 'offset' for each entry of readdir result This is preparation for using hash value as dentry 'offset' Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-28 07:17:40 +00:00
/*
* discard buffered readdir content on seekdir(0), or seek to new frag,
* or seek prior to current chunk
*/
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
static bool need_reset_readdir(struct ceph_dir_file_info *dfi, loff_t new_pos)
ceph: record 'offset' for each entry of readdir result This is preparation for using hash value as dentry 'offset' Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-28 07:17:40 +00:00
{
struct ceph_mds_reply_info_parsed *rinfo;
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
loff_t chunk_offset;
ceph: record 'offset' for each entry of readdir result This is preparation for using hash value as dentry 'offset' Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-28 07:17:40 +00:00
if (new_pos == 0)
return true;
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
if (is_hash_order(new_pos)) {
/* no need to reset last_name for a forward seek when
* dentries are sotred in hash order */
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
} else if (dfi->frag != fpos_frag(new_pos)) {
ceph: record 'offset' for each entry of readdir result This is preparation for using hash value as dentry 'offset' Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-28 07:17:40 +00:00
return true;
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
}
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
rinfo = dfi->last_readdir ? &dfi->last_readdir->r_reply_info : NULL;
ceph: record 'offset' for each entry of readdir result This is preparation for using hash value as dentry 'offset' Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-28 07:17:40 +00:00
if (!rinfo || !rinfo->dir_nr)
return true;
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
chunk_offset = rinfo->dir_entries[0].offset;
return new_pos < chunk_offset ||
is_hash_order(new_pos) != is_hash_order(chunk_offset);
ceph: record 'offset' for each entry of readdir result This is preparation for using hash value as dentry 'offset' Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-28 07:17:40 +00:00
}
lseek: the "whence" argument is called "whence" But the kernel decided to call it "origin" instead. Fix most of the sites. Acked-by: Hugh Dickins <hughd@google.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-12-17 23:59:39 +00:00
static loff_t ceph_dir_llseek(struct file *file, loff_t offset, int whence)
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
{
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
struct ceph_dir_file_info *dfi = file->private_data;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
struct inode *inode = file->f_mapping->host;
loff_t retval;
wrappers for ->i_mutex access parallel to mutex_{lock,unlock,trylock,is_locked,lock_nested}, inode_foo(inode) being mutex_foo(&inode->i_mutex). Please, use those for access to ->i_mutex; over the coming cycle ->i_mutex will become rwsem, with ->lookup() done with it held only shared. Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2016-01-22 20:40:57 +00:00
inode_lock(inode);
fs: handle SEEK_HOLE/SEEK_DATA properly in all fs's that define their own llseek This converts everybody to handle SEEK_HOLE/SEEK_DATA properly. In some cases we just return -EINVAL, in others we do the normal generic thing, and in others we're simply making sure that the properly due-dilligence is done. For example in NFS/CIFS we need to make sure the file size is update properly for the SEEK_HOLE and SEEK_DATA case, but since it calls the generic llseek stuff itself that is all we have to do. Thanks, Signed-off-by: Josef Bacik <josef@redhat.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2011-07-18 17:21:38 +00:00
retval = -EINVAL;
lseek: the "whence" argument is called "whence" But the kernel decided to call it "origin" instead. Fix most of the sites. Acked-by: Hugh Dickins <hughd@google.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-12-17 23:59:39 +00:00
switch (whence) {
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
case SEEK_CUR:
offset += file->f_pos;
ceph: fix fall-through warnings for Clang In preparation to enable -Wimplicit-fallthrough for Clang, fix a couple of warnings by explicitly adding a break and a goto statements instead of just letting the code fall through to the next case. URL: https://github.com/KSPP/linux/issues/115 Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org> Reviewed-by: Ilya Dryomov <idryomov@gmail.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2021-03-05 09:59:23 +00:00
break;
fs: handle SEEK_HOLE/SEEK_DATA properly in all fs's that define their own llseek This converts everybody to handle SEEK_HOLE/SEEK_DATA properly. In some cases we just return -EINVAL, in others we do the normal generic thing, and in others we're simply making sure that the properly due-dilligence is done. For example in NFS/CIFS we need to make sure the file size is update properly for the SEEK_HOLE and SEEK_DATA case, but since it calls the generic llseek stuff itself that is all we have to do. Thanks, Signed-off-by: Josef Bacik <josef@redhat.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2011-07-18 17:21:38 +00:00
case SEEK_SET:
break;
ceph: rework dcache readdir Previously our dcache readdir code relies on that child dentries in directory dentry's d_subdir list are sorted by dentry's offset in descending order. When adding dentries to the dcache, if a dentry already exists, our readdir code moves it to head of directory dentry's d_subdir list. This design relies on dcache internals. Al Viro suggests using ncpfs's approach: keeping array of pointers to dentries in page cache of directory inode. the validity of those pointers are presented by directory inode's complete and ordered flags. When a dentry gets pruned, we clear directory inode's complete flag in the d_prune() callback. Before moving a dentry to other directory, we clear the ordered flag for both old and new directory. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-16 12:48:56 +00:00
case SEEK_END:
retval = -EOPNOTSUPP;
ceph: fix fall-through warnings for Clang In preparation to enable -Wimplicit-fallthrough for Clang, fix a couple of warnings by explicitly adding a break and a goto statements instead of just letting the code fall through to the next case. URL: https://github.com/KSPP/linux/issues/115 Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org> Reviewed-by: Ilya Dryomov <idryomov@gmail.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2021-03-05 09:59:23 +00:00
goto out;
fs: handle SEEK_HOLE/SEEK_DATA properly in all fs's that define their own llseek This converts everybody to handle SEEK_HOLE/SEEK_DATA properly. In some cases we just return -EINVAL, in others we do the normal generic thing, and in others we're simply making sure that the properly due-dilligence is done. For example in NFS/CIFS we need to make sure the file size is update properly for the SEEK_HOLE and SEEK_DATA case, but since it calls the generic llseek stuff itself that is all we have to do. Thanks, Signed-off-by: Josef Bacik <josef@redhat.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2011-07-18 17:21:38 +00:00
default:
goto out;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
}
fs: handle SEEK_HOLE/SEEK_DATA properly in all fs's that define their own llseek This converts everybody to handle SEEK_HOLE/SEEK_DATA properly. In some cases we just return -EINVAL, in others we do the normal generic thing, and in others we're simply making sure that the properly due-dilligence is done. For example in NFS/CIFS we need to make sure the file size is update properly for the SEEK_HOLE and SEEK_DATA case, but since it calls the generic llseek stuff itself that is all we have to do. Thanks, Signed-off-by: Josef Bacik <josef@redhat.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2011-07-18 17:21:38 +00:00
ceph: fix ceph_dir_llseek() Comparing offset with inode->i_sb->s_maxbytes doesn't make sense for directory. For a fragmented directory, offset (frag_t, off) can be larger than inode->i_sb->s_maxbytes. At the very beginning of ceph_dir_llseek(), local variable old_offset is initialized to parameter offset. This doesn't make sense neither. Old_offset should be ceph_make_fpos(fi->frag, fi->next_offset). Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com> Reviewed-by: Alex Elder <elder@linaro.org>
2014-02-27 08:26:24 +00:00
if (offset >= 0) {
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
if (need_reset_readdir(dfi, offset)) {
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
dout("dir_llseek dropping %p content\n", file);
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
reset_readdir(dfi);
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
} else if (is_hash_order(offset) && offset > file->f_pos) {
/* for hash offset, we don't know if a forward seek
* is within same frag */
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
dfi->dir_release_count = 0;
dfi->readdir_cache_idx = -1;
ceph: using hash value to compose dentry offset If MDS sorts dentries in dirfrag in hash order, we use hash value to compose dentry offset. dentry offset is: (0xff << 52) | ((24 bits hash) << 28) | (the nth entry hash hash collision) This offset is stable across directory fragmentation. This alos means there is no need to reset readdir offset if directory get fragmented in the middle of readdir. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-04-29 03:27:30 +00:00
}
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
if (offset != file->f_pos) {
file->f_pos = offset;
file->f_version = 0;
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
dfi->file_info.flags &= ~CEPH_F_ATEND;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
}
retval = offset;
}
fs: handle SEEK_HOLE/SEEK_DATA properly in all fs's that define their own llseek This converts everybody to handle SEEK_HOLE/SEEK_DATA properly. In some cases we just return -EINVAL, in others we do the normal generic thing, and in others we're simply making sure that the properly due-dilligence is done. For example in NFS/CIFS we need to make sure the file size is update properly for the SEEK_HOLE and SEEK_DATA case, but since it calls the generic llseek stuff itself that is all we have to do. Thanks, Signed-off-by: Josef Bacik <josef@redhat.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2011-07-18 17:21:38 +00:00
out:
wrappers for ->i_mutex access parallel to mutex_{lock,unlock,trylock,is_locked,lock_nested}, inode_foo(inode) being mutex_foo(&inode->i_mutex). Please, use those for access to ->i_mutex; over the coming cycle ->i_mutex will become rwsem, with ->lookup() done with it held only shared. Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2016-01-22 20:40:57 +00:00
inode_unlock(inode);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
return retval;
}
/*
ceph: fix ceph_lookup_open intent usage We weren't properly calling lookup_instantiate_filp when setting up the lookup intent, which could lead to file leakage on errors. So: - use separate helper for the hidden snapdir translation, immediately following the mds request - use ceph_finish_lookup for the final dentry/return value dance in the exit path - lookup_instantiate_filp on success Reported-by: Al Viro <viro@ZenIV.linux.org.uk> Reviewed-by: Yehuda Sadeh <yehuda@hq.newdream.net> Signed-off-by: Sage Weil <sage@newdream.net>
2011-07-26 18:28:11 +00:00
* Handle lookups for the hidden .snap directory.
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
*/
ceph: don't use d_add in ceph_handle_snapdir It's possible ceph_get_snapdir could end up finding a (disconnected) inode that already exists in the cache. Change the prototype for ceph_handle_snapdir to return a dentry pointer and have it use d_splice_alias so we don't end up with an aliased dentry in the cache. Reported-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2021-03-01 13:01:54 +00:00
struct dentry *ceph_handle_snapdir(struct ceph_mds_request *req,
ceph: fix error handling in ceph_atomic_open and ceph_lookup Commit aa60cfc3f7ee broke the error handling in these functions such that they don't handle non-ENOENT errors from ceph_mdsc_do_request properly. Move the checking of -ENOENT out of ceph_handle_snapdir and into the callers, and if we get a different error, return it immediately. Fixes: aa60cfc3f7ee ("ceph: don't use d_add in ceph_handle_snapdir") Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: Ilya Dryomov <idryomov@gmail.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2021-06-02 16:46:07 +00:00
struct dentry *dentry)
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
{
ceph: factor out libceph from Ceph file system This factors out protocol and low-level storage parts of ceph into a separate libceph module living in net/ceph and include/linux/ceph. This is mostly a matter of moving files around. However, a few key pieces of the interface change as well: - ceph_client becomes ceph_fs_client and ceph_client, where the latter captures the mon and osd clients, and the fs_client gets the mds client and file system specific pieces. - Mount option parsing and debugfs setup is correspondingly broken into two pieces. - The mon client gets a generic handler callback for otherwise unknown messages (mds map, in this case). - The basic supported/required feature bits can be expanded (and are by ceph_fs_client). No functional change, aside from some subtle error handling cases that got cleaned up in the refactoring process. Signed-off-by: Sage Weil <sage@newdream.net>
2010-04-06 22:14:15 +00:00
struct ceph_fs_client *fsc = ceph_sb_to_client(dentry->d_sb);
ceph: fix comments mentioning i_mutex inode->i_mutex has been replaced with inode->i_rwsem long ago. Fix comments still mentioning i_mutex. Signed-off-by: hongnanli <hongnan.li@linux.alibaba.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2022-02-11 03:33:25 +00:00
struct inode *parent = d_inode(dentry->d_parent); /* we hold i_rwsem */
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
/* .snap dir? */
ceph: fix error handling in ceph_atomic_open and ceph_lookup Commit aa60cfc3f7ee broke the error handling in these functions such that they don't handle non-ENOENT errors from ceph_mdsc_do_request properly. Move the checking of -ENOENT out of ceph_handle_snapdir and into the callers, and if we get a different error, return it immediately. Fixes: aa60cfc3f7ee ("ceph: don't use d_add in ceph_handle_snapdir") Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: Ilya Dryomov <idryomov@gmail.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2021-06-02 16:46:07 +00:00
if (ceph_snap(parent) == CEPH_NOSNAP &&
ceph: don't use d_add in ceph_handle_snapdir It's possible ceph_get_snapdir could end up finding a (disconnected) inode that already exists in the cache. Change the prototype for ceph_handle_snapdir to return a dentry pointer and have it use d_splice_alias so we don't end up with an aliased dentry in the cache. Reported-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2021-03-01 13:01:54 +00:00
strcmp(dentry->d_name.name, fsc->mount_options->snapdir_name) == 0) {
struct dentry *res;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
struct inode *inode = ceph_get_snapdir(parent);
ceph: don't use d_add in ceph_handle_snapdir It's possible ceph_get_snapdir could end up finding a (disconnected) inode that already exists in the cache. Change the prototype for ceph_handle_snapdir to return a dentry pointer and have it use d_splice_alias so we don't end up with an aliased dentry in the cache. Reported-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2021-03-01 13:01:54 +00:00
res = d_splice_alias(inode, dentry);
dout("ENOENT on snapdir %p '%pd', linking to snapdir %p. Spliced dentry %p\n",
dentry, dentry, inode, res);
if (res)
dentry = res;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
}
ceph: don't use d_add in ceph_handle_snapdir It's possible ceph_get_snapdir could end up finding a (disconnected) inode that already exists in the cache. Change the prototype for ceph_handle_snapdir to return a dentry pointer and have it use d_splice_alias so we don't end up with an aliased dentry in the cache. Reported-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2021-03-01 13:01:54 +00:00
return dentry;
ceph: fix ceph_lookup_open intent usage We weren't properly calling lookup_instantiate_filp when setting up the lookup intent, which could lead to file leakage on errors. So: - use separate helper for the hidden snapdir translation, immediately following the mds request - use ceph_finish_lookup for the final dentry/return value dance in the exit path - lookup_instantiate_filp on success Reported-by: Al Viro <viro@ZenIV.linux.org.uk> Reviewed-by: Yehuda Sadeh <yehuda@hq.newdream.net> Signed-off-by: Sage Weil <sage@newdream.net>
2011-07-26 18:28:11 +00:00
}
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
ceph: fix ceph_lookup_open intent usage We weren't properly calling lookup_instantiate_filp when setting up the lookup intent, which could lead to file leakage on errors. So: - use separate helper for the hidden snapdir translation, immediately following the mds request - use ceph_finish_lookup for the final dentry/return value dance in the exit path - lookup_instantiate_filp on success Reported-by: Al Viro <viro@ZenIV.linux.org.uk> Reviewed-by: Yehuda Sadeh <yehuda@hq.newdream.net> Signed-off-by: Sage Weil <sage@newdream.net>
2011-07-26 18:28:11 +00:00
/*
* Figure out final result of a lookup/open request.
*
* Mainly, make sure we return the final req->r_dentry (if it already
* existed) in place of the original VFS-provided dentry when they
* differ.
*
* Gracefully handle the case where the MDS replies with -ENOENT and
* no trace (which it may do, at its discretion, e.g., if it doesn't
* care to issue a lease on the negative dentry).
*/
struct dentry *ceph_finish_lookup(struct ceph_mds_request *req,
struct dentry *dentry, int err)
{
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
if (err == -ENOENT) {
/* no trace? */
err = 0;
if (!req->r_reply_info.head->is_dentry) {
dout("ENOENT and no trace, dentry %p inode %p\n",
VFS: normal filesystems (and lustre): d_inode() annotations that's the bulk of filesystem drivers dealing with inodes of their own Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2015-03-17 22:25:59 +00:00
dentry, d_inode(dentry));
if (d_really_is_positive(dentry)) {
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
d_drop(dentry);
err = -ENOENT;
} else {
d_add(dentry, NULL);
}
}
}
if (err)
dentry = ERR_PTR(err);
else if (dentry != req->r_dentry)
dentry = dget(req->r_dentry); /* we got spliced */
else
dentry = NULL;
return dentry;
}
ceph: make logical calculation functions return bool This patch makes serverl logical caculation functions return bool to improve readability due to these particular functions only using 0/1 as their return value. No functional change. Signed-off-by: Zhang Zhuoyu <zhangzhuoyu@cmss.chinamobile.com>
2016-03-25 09:18:39 +00:00
static bool is_root_ceph_dentry(struct inode *inode, struct dentry *dentry)
ceph: hide /.ceph from readdir results We need to skip /.ceph in (cached) readdir results, and exclude "/.ceph" from the cached ENOENT lookup check. Signed-off-by: Sage Weil <sage@newdream.net>
2009-12-02 19:54:25 +00:00
{
return ceph_ino(inode) == CEPH_INO_ROOT &&
strncmp(dentry->d_name.name, ".ceph", 5) == 0;
}
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
/*
* Look up a single dir entry. If there is a lookup intent, inform
* the MDS so that it gets our 'caps wanted' value in a single op.
*/
static struct dentry *ceph_lookup(struct inode *dir, struct dentry *dentry,
stop passing nameidata to ->lookup() Just the flags; only NFS cares even about that, but there are legitimate uses for such argument. And getting rid of that completely would require splitting ->lookup() into a couple of methods (at least), so let's leave that alone for now... Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2012-06-10 21:13:09 +00:00
unsigned int flags)
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
{
ceph: factor out libceph from Ceph file system This factors out protocol and low-level storage parts of ceph into a separate libceph module living in net/ceph and include/linux/ceph. This is mostly a matter of moving files around. However, a few key pieces of the interface change as well: - ceph_client becomes ceph_fs_client and ceph_client, where the latter captures the mon and osd clients, and the fs_client gets the mds client and file system specific pieces. - Mount option parsing and debugfs setup is correspondingly broken into two pieces. - The mon client gets a generic handler callback for otherwise unknown messages (mds map, in this case). - The basic supported/required feature bits can be expanded (and are by ceph_fs_client). No functional change, aside from some subtle error handling cases that got cleaned up in the refactoring process. Signed-off-by: Sage Weil <sage@newdream.net>
2010-04-06 22:14:15 +00:00
struct ceph_fs_client *fsc = ceph_sb_to_client(dir->i_sb);
ceph: add ceph_sb_to_mdsc helper support to parse the mdsc This will help simplify the code. [ jlayton: fix minor merge conflict in quota.c ] Signed-off-by: Xiubo Li <xiubli@redhat.com> Signed-off-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-09-03 13:01:39 +00:00
struct ceph_mds_client *mdsc = ceph_sb_to_mdsc(dir->i_sb);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
struct ceph_mds_request *req;
int op;
ceph: fix security xattr deadlock When security is enabled, security module can call filesystem's getxattr/setxattr callbacks during d_instantiate(). For cephfs, d_instantiate() is usually called by MDS' dispatch thread, while handling MDS reply. If the MDS reply does not include xattrs and corresponding caps, getxattr/setxattr need to send a new request to MDS and waits for the reply. This makes MDS' dispatch sleep, nobody handles later MDS replies. The fix is make sure lookup/atomic_open reply include xattrs and corresponding caps. So getxattr can be handled by cached xattrs. This requires some modification to both MDS and request message. (Client tells MDS what caps it wants; MDS encodes proper caps in the reply) Smack security module may call setxattr during d_instantiate(). Unlike getxattr, we can't force MDS to issue CEPH_CAP_XATTR_EXCL to us. So just make setxattr return error when called by MDS' dispatch thread. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-03-07 02:34:50 +00:00
int mask;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
int err;
assorted conversions to %p[dD] Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2014-10-22 00:11:25 +00:00
dout("lookup %p dentry %p '%pd'\n",
dir, dentry, dentry);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
if (dentry->d_name.len > NAME_MAX)
return ERR_PTR(-ENAMETOOLONG);
/* can we conclude ENOENT locally? */
VFS: normal filesystems (and lustre): d_inode() annotations that's the bulk of filesystem drivers dealing with inodes of their own Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2015-03-17 22:25:59 +00:00
if (d_really_is_negative(dentry)) {
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
struct ceph_inode_info *ci = ceph_inode(dir);
struct ceph_dentry_info *di = ceph_dentry(dentry);
ceph: use i_ceph_lock instead of i_lock We have been using i_lock to protect all kinds of data structures in the ceph_inode_info struct, including lists of inodes that we need to iterate over while avoiding races with inode destruction. That requires grabbing a reference to the inode with the list lock protected, but igrab() now takes i_lock to check the inode flags. Changing the list lock ordering would be a painful process. However, using a ceph-specific i_ceph_lock in the ceph inode instead of i_lock is a simple mechanical change and avoids the ordering constraints imposed by igrab(). Reported-by: Amon Ott <a.ott@m-privacy.de> Signed-off-by: Sage Weil <sage@newdream.net>
2011-11-30 17:47:09 +00:00
spin_lock(&ci->i_ceph_lock);
ceph: add infrastructure for waiting for async create to complete When we issue an async create, we must ensure that any later on-the-wire requests involving it wait for the create reply. Expand i_ceph_flags to be an unsigned long, and add a new bit that MDS requests can wait on. If the bit is set in the inode when sending caps, then don't send it and just return that it has been delayed. Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-01-14 20:06:40 +00:00
dout(" dir %p flags are 0x%lx\n", dir, ci->i_ceph_flags);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
if (strncmp(dentry->d_name.name,
ceph: factor out libceph from Ceph file system This factors out protocol and low-level storage parts of ceph into a separate libceph module living in net/ceph and include/linux/ceph. This is mostly a matter of moving files around. However, a few key pieces of the interface change as well: - ceph_client becomes ceph_fs_client and ceph_client, where the latter captures the mon and osd clients, and the fs_client gets the mds client and file system specific pieces. - Mount option parsing and debugfs setup is correspondingly broken into two pieces. - The mon client gets a generic handler callback for otherwise unknown messages (mds map, in this case). - The basic supported/required feature bits can be expanded (and are by ceph_fs_client). No functional change, aside from some subtle error handling cases that got cleaned up in the refactoring process. Signed-off-by: Sage Weil <sage@newdream.net>
2010-04-06 22:14:15 +00:00
fsc->mount_options->snapdir_name,
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
dentry->d_name.len) &&
ceph: hide /.ceph from readdir results We need to skip /.ceph in (cached) readdir results, and exclude "/.ceph" from the cached ENOENT lookup check. Signed-off-by: Sage Weil <sage@newdream.net>
2009-12-02 19:54:25 +00:00
!is_root_ceph_dentry(dir, dentry) &&
ceph: fix dcache/nocache mount option Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-03-04 08:05:04 +00:00
ceph_test_mount_opt(fsc, DCACHE) &&
ceph: use i_release_count to indicate dir's completeness Current ceph code tracks directory's completeness in two places. ceph_readdir() checks i_release_count to decide if it can set the I_COMPLETE flag in i_ceph_flags. All other places check the I_COMPLETE flag. This indirection introduces locking complexity. This patch adds a new variable i_complete_count to ceph_inode_info. Set i_release_count's value to it when marking a directory complete. By comparing the two variables, we know if a directory is complete Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
2013-03-13 11:44:32 +00:00
__ceph_dir_is_complete(ci) &&
ceph: add caps perf metric for each superblock Count hits and misses in the caps cache. If the client has all of the necessary caps when a task needs references, then it's counted as a hit. Any other situation is a miss. URL: https://tracker.ceph.com/issues/43215 Signed-off-by: Xiubo Li <xiubli@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-03-20 03:45:00 +00:00
__ceph_caps_issued_mask_metric(ci, CEPH_CAP_FILE_SHARED, 1)) {
ceph: consider inode's last read/write when calculating wanted caps Add i_last_rd and i_last_wr to ceph_inode_info. These fields are used to track the last time the client acquired read/write caps for the inode. If there is no read/write on an inode for 'caps_wanted_delay_max' seconds, __ceph_caps_file_wanted() does not request caps for read/write even there are open files. Call __ceph_touch_fmode() for dir operations. __ceph_caps_file_wanted() calculates dir's wanted caps according to last dir read/modification. If there is recent dir read, dir inode wants CEPH_CAP_ANY_SHARED caps. If there is recent dir modification, also wants CEPH_CAP_FILE_EXCL. Readdir is a special case. Dir inode wants CEPH_CAP_FILE_EXCL after readdir, as with that, modifications do not need to release CEPH_CAP_FILE_SHARED or invalidate all dentry leases issued by readdir. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-03-05 12:21:00 +00:00
__ceph_touch_fmode(ci, mdsc, CEPH_FILE_MODE_RD);
ceph: use i_ceph_lock instead of i_lock We have been using i_lock to protect all kinds of data structures in the ceph_inode_info struct, including lists of inodes that we need to iterate over while avoiding races with inode destruction. That requires grabbing a reference to the inode with the list lock protected, but igrab() now takes i_lock to check the inode flags. Changing the list lock ordering would be a painful process. However, using a ceph-specific i_ceph_lock in the ceph inode instead of i_lock is a simple mechanical change and avoids the ordering constraints imposed by igrab(). Reported-by: Amon Ott <a.ott@m-privacy.de> Signed-off-by: Sage Weil <sage@newdream.net>
2011-11-30 17:47:09 +00:00
spin_unlock(&ci->i_ceph_lock);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
dout(" dir %p complete, -ENOENT\n", dir);
d_add(dentry, NULL);
ceph: use atomic_t for ceph_inode_info::i_shared_gen It allows accessing i_shared_gen without holding i_ceph_lock. It is preparation for later patch. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Acked-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-11-27 02:47:46 +00:00
di->lease_shared_gen = atomic_read(&ci->i_shared_gen);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
return NULL;
}
ceph: use i_ceph_lock instead of i_lock We have been using i_lock to protect all kinds of data structures in the ceph_inode_info struct, including lists of inodes that we need to iterate over while avoiding races with inode destruction. That requires grabbing a reference to the inode with the list lock protected, but igrab() now takes i_lock to check the inode flags. Changing the list lock ordering would be a painful process. However, using a ceph-specific i_ceph_lock in the ceph inode instead of i_lock is a simple mechanical change and avoids the ordering constraints imposed by igrab(). Reported-by: Amon Ott <a.ott@m-privacy.de> Signed-off-by: Sage Weil <sage@newdream.net>
2011-11-30 17:47:09 +00:00
spin_unlock(&ci->i_ceph_lock);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
}
op = ceph_snap(dir) == CEPH_SNAPDIR ?
CEPH_MDS_OP_LOOKUPSNAP : CEPH_MDS_OP_LOOKUP;
req = ceph_mdsc_create_request(mdsc, op, USE_ANY_MDS);
if (IS_ERR(req))
fs/ceph: Use ERR_CAST Use ERR_CAST(x) rather than ERR_PTR(PTR_ERR(x)). The former makes more clear what is the purpose of the operation, which otherwise looks like a no-op. In the case of fs/ceph/inode.c, ERR_CAST is not needed, because the type of the returned value is the same as the type of the enclosing function. The semantic patch that makes this change is as follows: (http://coccinelle.lip6.fr/) // <smpl> @@ type T; T x; identifier f; @@ T f (...) { <+... - ERR_PTR(PTR_ERR(x)) + x ...+> } @@ expression x; @@ - ERR_PTR(PTR_ERR(x)) + ERR_CAST(x) // </smpl> Signed-off-by: Julia Lawall <julia@diku.dk> Signed-off-by: Sage Weil <sage@newdream.net>
2010-05-22 10:01:14 +00:00
return ERR_CAST(req);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
req->r_dentry = dget(dentry);
req->r_num_caps = 2;
ceph: fix security xattr deadlock When security is enabled, security module can call filesystem's getxattr/setxattr callbacks during d_instantiate(). For cephfs, d_instantiate() is usually called by MDS' dispatch thread, while handling MDS reply. If the MDS reply does not include xattrs and corresponding caps, getxattr/setxattr need to send a new request to MDS and waits for the reply. This makes MDS' dispatch sleep, nobody handles later MDS replies. The fix is make sure lookup/atomic_open reply include xattrs and corresponding caps. So getxattr can be handled by cached xattrs. This requires some modification to both MDS and request message. (Client tells MDS what caps it wants; MDS encodes proper caps in the reply) Smack security module may call setxattr during d_instantiate(). Unlike getxattr, we can't force MDS to issue CEPH_CAP_XATTR_EXCL to us. So just make setxattr return error when called by MDS' dispatch thread. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-03-07 02:34:50 +00:00
mask = CEPH_STAT_CAP_INODE | CEPH_CAP_AUTH_SHARED;
if (ceph_security_xattr_wanted(dir))
mask |= CEPH_CAP_XATTR_SHARED;
req->r_args.getattr.mask = cpu_to_le32(mask);
ceph: take reference to req->r_parent at point of assignment Currently, we set the r_parent pointer but then don't take a reference to it until we submit the request. If we end up freeing the req before that point, then we'll do a iput when we shouldn't. Instead, take the inode reference in the callers, so that it's always safe to call ceph_mdsc_put_request on the req, even before submission. Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: Luis Henriques <lhenriques@suse.de> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2021-06-18 17:05:06 +00:00
ihold(dir);
ceph: add a new flag to indicate whether parent is locked struct ceph_mds_request has an r_locked_dir pointer, which is set to indicate the parent inode and that its i_rwsem is locked. In some critical places, we need to be able to indicate the parent inode to the request handling code, even when its i_rwsem may not be locked. Most of the code that operates on r_locked_dir doesn't require that the i_rwsem be locked. We only really need it to handle manipulation of the dcache. The rest (filling of the inode, updating dentry leases, etc.) already has its own locking. Add a new r_req_flags bit that indicates whether the parent is locked when doing the request, and rename the pointer to "r_parent". For now, all the places that set r_parent also set this flag, but that will change in a later patch. Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Yan, Zheng <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-01-31 15:28:26 +00:00
req->r_parent = dir;
set_bit(CEPH_MDS_R_PARENT_LOCKED, &req->r_req_flags);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
err = ceph_mdsc_do_request(mdsc, NULL, req);
ceph: fix error handling in ceph_atomic_open and ceph_lookup Commit aa60cfc3f7ee broke the error handling in these functions such that they don't handle non-ENOENT errors from ceph_mdsc_do_request properly. Move the checking of -ENOENT out of ceph_handle_snapdir and into the callers, and if we get a different error, return it immediately. Fixes: aa60cfc3f7ee ("ceph: don't use d_add in ceph_handle_snapdir") Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: Ilya Dryomov <idryomov@gmail.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2021-06-02 16:46:07 +00:00
if (err == -ENOENT) {
struct dentry *res;
res = ceph_handle_snapdir(req, dentry);
if (IS_ERR(res)) {
err = PTR_ERR(res);
} else {
dentry = res;
err = 0;
}
ceph: don't use d_add in ceph_handle_snapdir It's possible ceph_get_snapdir could end up finding a (disconnected) inode that already exists in the cache. Change the prototype for ceph_handle_snapdir to return a dentry pointer and have it use d_splice_alias so we don't end up with an aliased dentry in the cache. Reported-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2021-03-01 13:01:54 +00:00
}
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
dentry = ceph_finish_lookup(req, dentry, err);
ceph_mdsc_put_request(req); /* will dput(dentry) */
dout("lookup result=%p\n", dentry);
return dentry;
}
/*
* If we do a create but get no trace back from the MDS, follow up with
* a lookup (the VFS expects us to link up the provided dentry).
*/
int ceph_handle_notrace_create(struct inode *dir, struct dentry *dentry)
{
stop passing nameidata to ->lookup() Just the flags; only NFS cares even about that, but there are legitimate uses for such argument. And getting rid of that completely would require splitting ->lookup() into a couple of methods (at least), so let's leave that alone for now... Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2012-06-10 21:13:09 +00:00
struct dentry *result = ceph_lookup(dir, dentry, 0);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
if (result && !IS_ERR(result)) {
/*
* We created the item, then did a lookup, and found
* it was already linked to another inode we already
ceph: return error for traceless reply race When we receives traceless reply for request that created new inode, we re-send a lookup request to MDS get information of the newly created inode. (VFS expects FS' callback return an inode in create case) This breaks one request into two requests. Other client may modify or move to the new inode in the middle. When the race happens, ceph_handle_notrace_create() unconditionally links the dentry for 'create' operation to the inode returned by lookup. This may confuse VFS when the inode is a directory (VFS does not allow multiple linkages for directory inode). This patch makes ceph_handle_notrace_create() when it detect a race. This event should be rare and it happens only when we talk to old MDS. Recent MDS does not send traceless reply for request that creates new inode. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-02-04 07:10:48 +00:00
* had in our cache (and thus got spliced). To not
* confuse VFS (especially when inode is a directory),
* we don't link our dentry to that inode, return an
* error instead.
*
* This event should be rare and it happens only when
* we talk to old MDS. Recent MDS does not send traceless
* reply for request that creates new inode.
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
*/
ceph: fix dentry leaks Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-02-02 03:27:56 +00:00
d_drop(result);
ceph: return error for traceless reply race When we receives traceless reply for request that created new inode, we re-send a lookup request to MDS get information of the newly created inode. (VFS expects FS' callback return an inode in create case) This breaks one request into two requests. Other client may modify or move to the new inode in the middle. When the race happens, ceph_handle_notrace_create() unconditionally links the dentry for 'create' operation to the inode returned by lookup. This may confuse VFS when the inode is a directory (VFS does not allow multiple linkages for directory inode). This patch makes ceph_handle_notrace_create() when it detect a race. This event should be rare and it happens only when we talk to old MDS. Recent MDS does not send traceless reply for request that creates new inode. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-02-04 07:10:48 +00:00
return -ESTALE;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
}
return PTR_ERR(result);
}
fs: make helpers idmap mount aware Extend some inode methods with an additional user namespace argument. A filesystem that is aware of idmapped mounts will receive the user namespace the mount has been marked with. This can be used for additional permission checking and also to enable filesystems to translate between uids and gids if they need to. We have implemented all relevant helpers in earlier patches. As requested we simply extend the exisiting inode method instead of introducing new ones. This is a little more code churn but it's mostly mechanical and doesnt't leave us with additional inode methods. Link: https://lore.kernel.org/r/20210121131959.646623-25-christian.brauner@ubuntu.com Cc: Christoph Hellwig <hch@lst.de> Cc: David Howells <dhowells@redhat.com> Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: linux-fsdevel@vger.kernel.org Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
2021-01-21 13:19:43 +00:00
static int ceph_mknod(struct user_namespace *mnt_userns, struct inode *dir,
struct dentry *dentry, umode_t mode, dev_t rdev)
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
{
ceph: add ceph_sb_to_mdsc helper support to parse the mdsc This will help simplify the code. [ jlayton: fix minor merge conflict in quota.c ] Signed-off-by: Xiubo Li <xiubli@redhat.com> Signed-off-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-09-03 13:01:39 +00:00
struct ceph_mds_client *mdsc = ceph_sb_to_mdsc(dir->i_sb);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
struct ceph_mds_request *req;
ceph: rename struct ceph_acls_info to ceph_acl_sec_ctx Also rename ceph_release_acls_info() to ceph_release_acl_sec_ctx(). And move their definitions to different files. This is preparation for security label support. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-05-26 07:35:39 +00:00
struct ceph_acl_sec_ctx as_ctx = {};
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
int err;
if (ceph_snap(dir) != CEPH_NOSNAP)
return -EROFS;
ceph: wait for the first reply of inflight async unlink In async unlink case the kclient won't wait for the first reply from MDS and just drop all the links and unhash the dentry and then succeeds immediately. For any new create/link/rename,etc requests followed by using the same file names we must wait for the first reply of the inflight unlink request, or the MDS possibly will fail these following requests with -EEXIST if the inflight async unlink request was delayed for some reasons. And the worst case is that for the none async openc request it will successfully open the file if the CDentry hasn't been unlinked yet, but later the previous delayed async unlink request will remove the CDenty. That means the just created file is possiblly deleted later by accident. We need to wait for the inflight async unlink requests to finish when creating new files/directories by using the same file names. Link: https://tracker.ceph.com/issues/55332 Signed-off-by: Xiubo Li <xiubli@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2022-05-10 01:47:01 +00:00
err = ceph_wait_on_conflict_unlink(dentry);
if (err)
return err;
ceph: add d_drop for some error cases in ceph_mknod() When file num exceeds quota limit or fails from ceph_per_init_acls() should call d_drop to drop dentry from cache as well. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-07-09 14:17:30 +00:00
if (ceph_quota_is_max_files_exceeded(dir)) {
err = -EDQUOT;
goto out;
}
ceph: quota: support for ceph.quota.max_files This patch adds support for the max_files quota. It hooks into all the ceph functions that add new filesystem objects that need to be checked against the quota limits. When these limits are hit, -EDQUOT is returned. Note that we're not checking quotas on ceph_link(). ceph_link doesn't really create a new inode, and since the MDS doesn't update the directory statistics when a new (hard) link is created (only with symlinks), they are not accounted as a new file. Signed-off-by: Luis Henriques <lhenriques@suse.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-01-05 10:47:19 +00:00
ceph: rename struct ceph_acls_info to ceph_acl_sec_ctx Also rename ceph_release_acls_info() to ceph_release_acl_sec_ctx(). And move their definitions to different files. This is preparation for security label support. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-05-26 07:35:39 +00:00
err = ceph_pre_init_acls(dir, &mode, &as_ctx);
ceph: add selinux support When creating new file/directory, use security_dentry_init_security() to prepare selinux context for the new inode, then send openc/mkdir request to MDS, together with selinux xattr. security_dentry_init_security() only supports single security module and only selinux has dentry_init_security hook. So only selinux is supported for now. We can add support for other security modules once kernel has a generic version of dentry_init_security() Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-05-26 08:27:56 +00:00
if (err < 0)
goto out;
err = ceph_security_init_secctx(dentry, mode, &as_ctx);
ceph: include the initial ACL in create/mkdir/mknod MDS requests Current code set new file/directory's initial ACL in a non-atomic manner. Client first sends request to MDS to create new file/directory, then set the initial ACL after the new file/directory is successfully created. The fix is include the initial ACL in create/mkdir/mknod MDS requests. So MDS can handle creating file/directory and setting the initial ACL in one request. Signed-off-by: Yan, Zheng <zyan@redhat.com> Reviewed-by: Sage Weil <sage@redhat.com>
2014-09-16 12:35:17 +00:00
if (err < 0)
ceph: add d_drop for some error cases in ceph_mknod() When file num exceeds quota limit or fails from ceph_per_init_acls() should call d_drop to drop dentry from cache as well. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-07-09 14:17:30 +00:00
goto out;
ceph: include the initial ACL in create/mkdir/mknod MDS requests Current code set new file/directory's initial ACL in a non-atomic manner. Client first sends request to MDS to create new file/directory, then set the initial ACL after the new file/directory is successfully created. The fix is include the initial ACL in create/mkdir/mknod MDS requests. So MDS can handle creating file/directory and setting the initial ACL in one request. Signed-off-by: Yan, Zheng <zyan@redhat.com> Reviewed-by: Sage Weil <sage@redhat.com>
2014-09-16 12:35:17 +00:00
switch ->mknod() to umode_t Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2011-07-26 05:52:52 +00:00
dout("mknod in dir %p dentry %p mode 0%ho rdev %d\n",
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
dir, dentry, mode, rdev);
req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_MKNOD, USE_AUTH_MDS);
if (IS_ERR(req)) {
ceph: include the initial ACL in create/mkdir/mknod MDS requests Current code set new file/directory's initial ACL in a non-atomic manner. Client first sends request to MDS to create new file/directory, then set the initial ACL after the new file/directory is successfully created. The fix is include the initial ACL in create/mkdir/mknod MDS requests. So MDS can handle creating file/directory and setting the initial ACL in one request. Signed-off-by: Yan, Zheng <zyan@redhat.com> Reviewed-by: Sage Weil <sage@redhat.com>
2014-09-16 12:35:17 +00:00
err = PTR_ERR(req);
goto out;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
}
req->r_dentry = dget(dentry);
req->r_num_caps = 2;
ceph: add a new flag to indicate whether parent is locked struct ceph_mds_request has an r_locked_dir pointer, which is set to indicate the parent inode and that its i_rwsem is locked. In some critical places, we need to be able to indicate the parent inode to the request handling code, even when its i_rwsem may not be locked. Most of the code that operates on r_locked_dir doesn't require that the i_rwsem be locked. We only really need it to handle manipulation of the dcache. The rest (filling of the inode, updating dentry leases, etc.) already has its own locking. Add a new r_req_flags bit that indicates whether the parent is locked when doing the request, and rename the pointer to "r_parent". For now, all the places that set r_parent also set this flag, but that will change in a later patch. Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Yan, Zheng <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-01-31 15:28:26 +00:00
req->r_parent = dir;
ceph: take reference to req->r_parent at point of assignment Currently, we set the r_parent pointer but then don't take a reference to it until we submit the request. If we end up freeing the req before that point, then we'll do a iput when we shouldn't. Instead, take the inode reference in the callers, so that it's always safe to call ceph_mdsc_put_request on the req, even before submission. Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: Luis Henriques <lhenriques@suse.de> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2021-06-18 17:05:06 +00:00
ihold(dir);
ceph: add a new flag to indicate whether parent is locked struct ceph_mds_request has an r_locked_dir pointer, which is set to indicate the parent inode and that its i_rwsem is locked. In some critical places, we need to be able to indicate the parent inode to the request handling code, even when its i_rwsem may not be locked. Most of the code that operates on r_locked_dir doesn't require that the i_rwsem be locked. We only really need it to handle manipulation of the dcache. The rest (filling of the inode, updating dentry leases, etc.) already has its own locking. Add a new r_req_flags bit that indicates whether the parent is locked when doing the request, and rename the pointer to "r_parent". For now, all the places that set r_parent also set this flag, but that will change in a later patch. Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Yan, Zheng <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-01-31 15:28:26 +00:00
set_bit(CEPH_MDS_R_PARENT_LOCKED, &req->r_req_flags);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
req->r_args.mknod.mode = cpu_to_le32(mode);
req->r_args.mknod.rdev = cpu_to_le32(rdev);
ceph: voluntarily drop Ax cap for requests that create new inode MDS need to rdlock directory inode's authlock when handling these requests. Voluntarily dropping CEPH_CAP_AUTH_EXCL avoids a cap revoke message. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-11-23 09:47:15 +00:00
req->r_dentry_drop = CEPH_CAP_FILE_SHARED | CEPH_CAP_AUTH_EXCL;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
req->r_dentry_unless = CEPH_CAP_FILE_EXCL;
ceph: rename struct ceph_acls_info to ceph_acl_sec_ctx Also rename ceph_release_acls_info() to ceph_release_acl_sec_ctx(). And move their definitions to different files. This is preparation for security label support. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-05-26 07:35:39 +00:00
if (as_ctx.pagelist) {
req->r_pagelist = as_ctx.pagelist;
as_ctx.pagelist = NULL;
ceph: include the initial ACL in create/mkdir/mknod MDS requests Current code set new file/directory's initial ACL in a non-atomic manner. Client first sends request to MDS to create new file/directory, then set the initial ACL after the new file/directory is successfully created. The fix is include the initial ACL in create/mkdir/mknod MDS requests. So MDS can handle creating file/directory and setting the initial ACL in one request. Signed-off-by: Yan, Zheng <zyan@redhat.com> Reviewed-by: Sage Weil <sage@redhat.com>
2014-09-16 12:35:17 +00:00
}
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
err = ceph_mdsc_do_request(mdsc, dir, req);
if (!err && !req->r_reply_info.head->is_dentry)
err = ceph_handle_notrace_create(dir, dentry);
ceph_mdsc_put_request(req);
ceph: include the initial ACL in create/mkdir/mknod MDS requests Current code set new file/directory's initial ACL in a non-atomic manner. Client first sends request to MDS to create new file/directory, then set the initial ACL after the new file/directory is successfully created. The fix is include the initial ACL in create/mkdir/mknod MDS requests. So MDS can handle creating file/directory and setting the initial ACL in one request. Signed-off-by: Yan, Zheng <zyan@redhat.com> Reviewed-by: Sage Weil <sage@redhat.com>
2014-09-16 12:35:17 +00:00
out:
ceph: add acl for cephfs Signed-off-by: Guangliang Zhao <lucienchao@gmail.com> Reviewed-by: Li Wang <li.wang@ubuntykylin.com> Reviewed-by: Zheng Yan <zheng.z.yan@intel.com>
2013-11-11 07:18:03 +00:00
if (!err)
ceph: rename struct ceph_acls_info to ceph_acl_sec_ctx Also rename ceph_release_acls_info() to ceph_release_acl_sec_ctx(). And move their definitions to different files. This is preparation for security label support. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-05-26 07:35:39 +00:00
ceph_init_inode_acls(d_inode(dentry), &as_ctx);
ceph: add missing init_acl() for mkdir() and atomic_open() Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
2014-02-11 04:55:05 +00:00
else
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
d_drop(dentry);
ceph: rename struct ceph_acls_info to ceph_acl_sec_ctx Also rename ceph_release_acls_info() to ceph_release_acl_sec_ctx(). And move their definitions to different files. This is preparation for security label support. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-05-26 07:35:39 +00:00
ceph_release_acl_sec_ctx(&as_ctx);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
return err;
}
fs: make helpers idmap mount aware Extend some inode methods with an additional user namespace argument. A filesystem that is aware of idmapped mounts will receive the user namespace the mount has been marked with. This can be used for additional permission checking and also to enable filesystems to translate between uids and gids if they need to. We have implemented all relevant helpers in earlier patches. As requested we simply extend the exisiting inode method instead of introducing new ones. This is a little more code churn but it's mostly mechanical and doesnt't leave us with additional inode methods. Link: https://lore.kernel.org/r/20210121131959.646623-25-christian.brauner@ubuntu.com Cc: Christoph Hellwig <hch@lst.de> Cc: David Howells <dhowells@redhat.com> Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: linux-fsdevel@vger.kernel.org Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
2021-01-21 13:19:43 +00:00
static int ceph_create(struct user_namespace *mnt_userns, struct inode *dir,
struct dentry *dentry, umode_t mode, bool excl)
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
{
fs: make helpers idmap mount aware Extend some inode methods with an additional user namespace argument. A filesystem that is aware of idmapped mounts will receive the user namespace the mount has been marked with. This can be used for additional permission checking and also to enable filesystems to translate between uids and gids if they need to. We have implemented all relevant helpers in earlier patches. As requested we simply extend the exisiting inode method instead of introducing new ones. This is a little more code churn but it's mostly mechanical and doesnt't leave us with additional inode methods. Link: https://lore.kernel.org/r/20210121131959.646623-25-christian.brauner@ubuntu.com Cc: Christoph Hellwig <hch@lst.de> Cc: David Howells <dhowells@redhat.com> Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: linux-fsdevel@vger.kernel.org Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
2021-01-21 13:19:43 +00:00
return ceph_mknod(mnt_userns, dir, dentry, mode, 0);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
}
fs: make helpers idmap mount aware Extend some inode methods with an additional user namespace argument. A filesystem that is aware of idmapped mounts will receive the user namespace the mount has been marked with. This can be used for additional permission checking and also to enable filesystems to translate between uids and gids if they need to. We have implemented all relevant helpers in earlier patches. As requested we simply extend the exisiting inode method instead of introducing new ones. This is a little more code churn but it's mostly mechanical and doesnt't leave us with additional inode methods. Link: https://lore.kernel.org/r/20210121131959.646623-25-christian.brauner@ubuntu.com Cc: Christoph Hellwig <hch@lst.de> Cc: David Howells <dhowells@redhat.com> Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: linux-fsdevel@vger.kernel.org Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
2021-01-21 13:19:43 +00:00
static int ceph_symlink(struct user_namespace *mnt_userns, struct inode *dir,
struct dentry *dentry, const char *dest)
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
{
ceph: add ceph_sb_to_mdsc helper support to parse the mdsc This will help simplify the code. [ jlayton: fix minor merge conflict in quota.c ] Signed-off-by: Xiubo Li <xiubli@redhat.com> Signed-off-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-09-03 13:01:39 +00:00
struct ceph_mds_client *mdsc = ceph_sb_to_mdsc(dir->i_sb);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
struct ceph_mds_request *req;
ceph: add selinux support When creating new file/directory, use security_dentry_init_security() to prepare selinux context for the new inode, then send openc/mkdir request to MDS, together with selinux xattr. security_dentry_init_security() only supports single security module and only selinux has dentry_init_security hook. So only selinux is supported for now. We can add support for other security modules once kernel has a generic version of dentry_init_security() Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-05-26 08:27:56 +00:00
struct ceph_acl_sec_ctx as_ctx = {};
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
int err;
if (ceph_snap(dir) != CEPH_NOSNAP)
return -EROFS;
ceph: wait for the first reply of inflight async unlink In async unlink case the kclient won't wait for the first reply from MDS and just drop all the links and unhash the dentry and then succeeds immediately. For any new create/link/rename,etc requests followed by using the same file names we must wait for the first reply of the inflight unlink request, or the MDS possibly will fail these following requests with -EEXIST if the inflight async unlink request was delayed for some reasons. And the worst case is that for the none async openc request it will successfully open the file if the CDentry hasn't been unlinked yet, but later the previous delayed async unlink request will remove the CDenty. That means the just created file is possiblly deleted later by accident. We need to wait for the inflight async unlink requests to finish when creating new files/directories by using the same file names. Link: https://tracker.ceph.com/issues/55332 Signed-off-by: Xiubo Li <xiubli@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2022-05-10 01:47:01 +00:00
err = ceph_wait_on_conflict_unlink(dentry);
if (err)
return err;
ceph: add d_drop for some error cases in ceph_symlink() When file num exceeds quota limit, should call d_drop to drop dentry from cache as well. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-07-09 14:17:31 +00:00
if (ceph_quota_is_max_files_exceeded(dir)) {
err = -EDQUOT;
goto out;
}
ceph: quota: support for ceph.quota.max_files This patch adds support for the max_files quota. It hooks into all the ceph functions that add new filesystem objects that need to be checked against the quota limits. When these limits are hit, -EDQUOT is returned. Note that we're not checking quotas on ceph_link(). ceph_link doesn't really create a new inode, and since the MDS doesn't update the directory statistics when a new (hard) link is created (only with symlinks), they are not accounted as a new file. Signed-off-by: Luis Henriques <lhenriques@suse.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-01-05 10:47:19 +00:00
ceph: add selinux support When creating new file/directory, use security_dentry_init_security() to prepare selinux context for the new inode, then send openc/mkdir request to MDS, together with selinux xattr. security_dentry_init_security() only supports single security module and only selinux has dentry_init_security hook. So only selinux is supported for now. We can add support for other security modules once kernel has a generic version of dentry_init_security() Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-05-26 08:27:56 +00:00
err = ceph_security_init_secctx(dentry, S_IFLNK | 0777, &as_ctx);
if (err < 0)
goto out;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
dout("symlink in dir %p dentry %p to '%s'\n", dir, dentry, dest);
req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_SYMLINK, USE_AUTH_MDS);
if (IS_ERR(req)) {
ceph: include the initial ACL in create/mkdir/mknod MDS requests Current code set new file/directory's initial ACL in a non-atomic manner. Client first sends request to MDS to create new file/directory, then set the initial ACL after the new file/directory is successfully created. The fix is include the initial ACL in create/mkdir/mknod MDS requests. So MDS can handle creating file/directory and setting the initial ACL in one request. Signed-off-by: Yan, Zheng <zyan@redhat.com> Reviewed-by: Sage Weil <sage@redhat.com>
2014-09-16 12:35:17 +00:00
err = PTR_ERR(req);
goto out;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
}
ceph: switch some GFP_NOFS memory allocation to GFP_KERNEL GFP_NOFS memory allocation is required for page writeback path. But there is no need to use GFP_NOFS in syscall path and readpage path Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-06-13 09:27:05 +00:00
req->r_path2 = kstrdup(dest, GFP_KERNEL);
ceph: kstrdup() memory handling Currently, there is no check for the kstrdup() for r_path2, r_path1 and snapdir_name as various locations as there is a possibility of failure during memory pressure. Therefore, returning ENOMEM where the checks have been missed. Signed-off-by: Sanidhya Kashyap <sanidhya.gatech@gmail.com> Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-03-21 16:54:58 +00:00
if (!req->r_path2) {
err = -ENOMEM;
ceph_mdsc_put_request(req);
goto out;
}
ceph: add a new flag to indicate whether parent is locked struct ceph_mds_request has an r_locked_dir pointer, which is set to indicate the parent inode and that its i_rwsem is locked. In some critical places, we need to be able to indicate the parent inode to the request handling code, even when its i_rwsem may not be locked. Most of the code that operates on r_locked_dir doesn't require that the i_rwsem be locked. We only really need it to handle manipulation of the dcache. The rest (filling of the inode, updating dentry leases, etc.) already has its own locking. Add a new r_req_flags bit that indicates whether the parent is locked when doing the request, and rename the pointer to "r_parent". For now, all the places that set r_parent also set this flag, but that will change in a later patch. Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Yan, Zheng <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-01-31 15:28:26 +00:00
req->r_parent = dir;
ceph: take reference to req->r_parent at point of assignment Currently, we set the r_parent pointer but then don't take a reference to it until we submit the request. If we end up freeing the req before that point, then we'll do a iput when we shouldn't. Instead, take the inode reference in the callers, so that it's always safe to call ceph_mdsc_put_request on the req, even before submission. Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: Luis Henriques <lhenriques@suse.de> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2021-06-18 17:05:06 +00:00
ihold(dir);
ceph: add a new flag to indicate whether parent is locked struct ceph_mds_request has an r_locked_dir pointer, which is set to indicate the parent inode and that its i_rwsem is locked. In some critical places, we need to be able to indicate the parent inode to the request handling code, even when its i_rwsem may not be locked. Most of the code that operates on r_locked_dir doesn't require that the i_rwsem be locked. We only really need it to handle manipulation of the dcache. The rest (filling of the inode, updating dentry leases, etc.) already has its own locking. Add a new r_req_flags bit that indicates whether the parent is locked when doing the request, and rename the pointer to "r_parent". For now, all the places that set r_parent also set this flag, but that will change in a later patch. Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Yan, Zheng <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-01-31 15:28:26 +00:00
set_bit(CEPH_MDS_R_PARENT_LOCKED, &req->r_req_flags);
ceph: kstrdup() memory handling Currently, there is no check for the kstrdup() for r_path2, r_path1 and snapdir_name as various locations as there is a possibility of failure during memory pressure. Therefore, returning ENOMEM where the checks have been missed. Signed-off-by: Sanidhya Kashyap <sanidhya.gatech@gmail.com> Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-03-21 16:54:58 +00:00
req->r_dentry = dget(dentry);
req->r_num_caps = 2;
ceph: voluntarily drop Ax cap for requests that create new inode MDS need to rdlock directory inode's authlock when handling these requests. Voluntarily dropping CEPH_CAP_AUTH_EXCL avoids a cap revoke message. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-11-23 09:47:15 +00:00
req->r_dentry_drop = CEPH_CAP_FILE_SHARED | CEPH_CAP_AUTH_EXCL;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
req->r_dentry_unless = CEPH_CAP_FILE_EXCL;
ceph: set sec_context xattr on symlink creation Symlink inodes should have the security context set in their xattrs on creation. We already set the context on creation, but we don't attach the pagelist. The effect is that symlink inodes don't get an SELinux context set on them at creation, so they end up unlabeled instead of inheriting the proper context. Make it do so. Cc: stable@vger.kernel.org Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: Ilya Dryomov <idryomov@gmail.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-07-28 14:34:20 +00:00
if (as_ctx.pagelist) {
req->r_pagelist = as_ctx.pagelist;
as_ctx.pagelist = NULL;
}
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
err = ceph_mdsc_do_request(mdsc, dir, req);
if (!err && !req->r_reply_info.head->is_dentry)
err = ceph_handle_notrace_create(dir, dentry);
ceph_mdsc_put_request(req);
ceph: include the initial ACL in create/mkdir/mknod MDS requests Current code set new file/directory's initial ACL in a non-atomic manner. Client first sends request to MDS to create new file/directory, then set the initial ACL after the new file/directory is successfully created. The fix is include the initial ACL in create/mkdir/mknod MDS requests. So MDS can handle creating file/directory and setting the initial ACL in one request. Signed-off-by: Yan, Zheng <zyan@redhat.com> Reviewed-by: Sage Weil <sage@redhat.com>
2014-09-16 12:35:17 +00:00
out:
if (err)
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
d_drop(dentry);
ceph: add selinux support When creating new file/directory, use security_dentry_init_security() to prepare selinux context for the new inode, then send openc/mkdir request to MDS, together with selinux xattr. security_dentry_init_security() only supports single security module and only selinux has dentry_init_security hook. So only selinux is supported for now. We can add support for other security modules once kernel has a generic version of dentry_init_security() Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-05-26 08:27:56 +00:00
ceph_release_acl_sec_ctx(&as_ctx);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
return err;
}
fs: make helpers idmap mount aware Extend some inode methods with an additional user namespace argument. A filesystem that is aware of idmapped mounts will receive the user namespace the mount has been marked with. This can be used for additional permission checking and also to enable filesystems to translate between uids and gids if they need to. We have implemented all relevant helpers in earlier patches. As requested we simply extend the exisiting inode method instead of introducing new ones. This is a little more code churn but it's mostly mechanical and doesnt't leave us with additional inode methods. Link: https://lore.kernel.org/r/20210121131959.646623-25-christian.brauner@ubuntu.com Cc: Christoph Hellwig <hch@lst.de> Cc: David Howells <dhowells@redhat.com> Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: linux-fsdevel@vger.kernel.org Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
2021-01-21 13:19:43 +00:00
static int ceph_mkdir(struct user_namespace *mnt_userns, struct inode *dir,
struct dentry *dentry, umode_t mode)
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
{
ceph: add ceph_sb_to_mdsc helper support to parse the mdsc This will help simplify the code. [ jlayton: fix minor merge conflict in quota.c ] Signed-off-by: Xiubo Li <xiubli@redhat.com> Signed-off-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-09-03 13:01:39 +00:00
struct ceph_mds_client *mdsc = ceph_sb_to_mdsc(dir->i_sb);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
struct ceph_mds_request *req;
ceph: rename struct ceph_acls_info to ceph_acl_sec_ctx Also rename ceph_release_acls_info() to ceph_release_acl_sec_ctx(). And move their definitions to different files. This is preparation for security label support. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-05-26 07:35:39 +00:00
struct ceph_acl_sec_ctx as_ctx = {};
ceph: wait for the first reply of inflight async unlink In async unlink case the kclient won't wait for the first reply from MDS and just drop all the links and unhash the dentry and then succeeds immediately. For any new create/link/rename,etc requests followed by using the same file names we must wait for the first reply of the inflight unlink request, or the MDS possibly will fail these following requests with -EEXIST if the inflight async unlink request was delayed for some reasons. And the worst case is that for the none async openc request it will successfully open the file if the CDentry hasn't been unlinked yet, but later the previous delayed async unlink request will remove the CDenty. That means the just created file is possiblly deleted later by accident. We need to wait for the inflight async unlink requests to finish when creating new files/directories by using the same file names. Link: https://tracker.ceph.com/issues/55332 Signed-off-by: Xiubo Li <xiubli@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2022-05-10 01:47:01 +00:00
int err;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
int op;
ceph: wait for the first reply of inflight async unlink In async unlink case the kclient won't wait for the first reply from MDS and just drop all the links and unhash the dentry and then succeeds immediately. For any new create/link/rename,etc requests followed by using the same file names we must wait for the first reply of the inflight unlink request, or the MDS possibly will fail these following requests with -EEXIST if the inflight async unlink request was delayed for some reasons. And the worst case is that for the none async openc request it will successfully open the file if the CDentry hasn't been unlinked yet, but later the previous delayed async unlink request will remove the CDenty. That means the just created file is possiblly deleted later by accident. We need to wait for the inflight async unlink requests to finish when creating new files/directories by using the same file names. Link: https://tracker.ceph.com/issues/55332 Signed-off-by: Xiubo Li <xiubli@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2022-05-10 01:47:01 +00:00
err = ceph_wait_on_conflict_unlink(dentry);
if (err)
return err;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
if (ceph_snap(dir) == CEPH_SNAPDIR) {
/* mkdir .snap/foo is a MKSNAP */
op = CEPH_MDS_OP_MKSNAP;
assorted conversions to %p[dD] Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2014-10-22 00:11:25 +00:00
dout("mksnap dir %p snap '%pd' dn %p\n", dir,
dentry, dentry);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
} else if (ceph_snap(dir) == CEPH_NOSNAP) {
switch vfs_mkdir() and ->mkdir() to umode_t vfs_mkdir() gets int, but immediately drops everything that might not fit into umode_t and that's the only caller of ->mkdir()... Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2011-07-26 05:41:39 +00:00
dout("mkdir dir %p dn %p mode 0%ho\n", dir, dentry, mode);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
op = CEPH_MDS_OP_MKDIR;
} else {
ceph: wait for the first reply of inflight async unlink In async unlink case the kclient won't wait for the first reply from MDS and just drop all the links and unhash the dentry and then succeeds immediately. For any new create/link/rename,etc requests followed by using the same file names we must wait for the first reply of the inflight unlink request, or the MDS possibly will fail these following requests with -EEXIST if the inflight async unlink request was delayed for some reasons. And the worst case is that for the none async openc request it will successfully open the file if the CDentry hasn't been unlinked yet, but later the previous delayed async unlink request will remove the CDenty. That means the just created file is possiblly deleted later by accident. We need to wait for the inflight async unlink requests to finish when creating new files/directories by using the same file names. Link: https://tracker.ceph.com/issues/55332 Signed-off-by: Xiubo Li <xiubli@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2022-05-10 01:47:01 +00:00
err = -EROFS;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
goto out;
}
ceph: include the initial ACL in create/mkdir/mknod MDS requests Current code set new file/directory's initial ACL in a non-atomic manner. Client first sends request to MDS to create new file/directory, then set the initial ACL after the new file/directory is successfully created. The fix is include the initial ACL in create/mkdir/mknod MDS requests. So MDS can handle creating file/directory and setting the initial ACL in one request. Signed-off-by: Yan, Zheng <zyan@redhat.com> Reviewed-by: Sage Weil <sage@redhat.com>
2014-09-16 12:35:17 +00:00
ceph: don't check quota for snap inode snap inode's i_snap_realm is not pointing to ceph_snap_realm. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-01-12 08:26:17 +00:00
if (op == CEPH_MDS_OP_MKDIR &&
ceph_quota_is_max_files_exceeded(dir)) {
ceph: quota: support for ceph.quota.max_files This patch adds support for the max_files quota. It hooks into all the ceph functions that add new filesystem objects that need to be checked against the quota limits. When these limits are hit, -EDQUOT is returned. Note that we're not checking quotas on ceph_link(). ceph_link doesn't really create a new inode, and since the MDS doesn't update the directory statistics when a new (hard) link is created (only with symlinks), they are not accounted as a new file. Signed-off-by: Luis Henriques <lhenriques@suse.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-01-05 10:47:19 +00:00
err = -EDQUOT;
goto out;
}
ceph: include the initial ACL in create/mkdir/mknod MDS requests Current code set new file/directory's initial ACL in a non-atomic manner. Client first sends request to MDS to create new file/directory, then set the initial ACL after the new file/directory is successfully created. The fix is include the initial ACL in create/mkdir/mknod MDS requests. So MDS can handle creating file/directory and setting the initial ACL in one request. Signed-off-by: Yan, Zheng <zyan@redhat.com> Reviewed-by: Sage Weil <sage@redhat.com>
2014-09-16 12:35:17 +00:00
mode |= S_IFDIR;
ceph: rename struct ceph_acls_info to ceph_acl_sec_ctx Also rename ceph_release_acls_info() to ceph_release_acl_sec_ctx(). And move their definitions to different files. This is preparation for security label support. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-05-26 07:35:39 +00:00
err = ceph_pre_init_acls(dir, &mode, &as_ctx);
ceph: add selinux support When creating new file/directory, use security_dentry_init_security() to prepare selinux context for the new inode, then send openc/mkdir request to MDS, together with selinux xattr. security_dentry_init_security() only supports single security module and only selinux has dentry_init_security hook. So only selinux is supported for now. We can add support for other security modules once kernel has a generic version of dentry_init_security() Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-05-26 08:27:56 +00:00
if (err < 0)
goto out;
err = ceph_security_init_secctx(dentry, mode, &as_ctx);
ceph: include the initial ACL in create/mkdir/mknod MDS requests Current code set new file/directory's initial ACL in a non-atomic manner. Client first sends request to MDS to create new file/directory, then set the initial ACL after the new file/directory is successfully created. The fix is include the initial ACL in create/mkdir/mknod MDS requests. So MDS can handle creating file/directory and setting the initial ACL in one request. Signed-off-by: Yan, Zheng <zyan@redhat.com> Reviewed-by: Sage Weil <sage@redhat.com>
2014-09-16 12:35:17 +00:00
if (err < 0)
goto out;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
req = ceph_mdsc_create_request(mdsc, op, USE_AUTH_MDS);
if (IS_ERR(req)) {
err = PTR_ERR(req);
goto out;
}
req->r_dentry = dget(dentry);
req->r_num_caps = 2;
ceph: add a new flag to indicate whether parent is locked struct ceph_mds_request has an r_locked_dir pointer, which is set to indicate the parent inode and that its i_rwsem is locked. In some critical places, we need to be able to indicate the parent inode to the request handling code, even when its i_rwsem may not be locked. Most of the code that operates on r_locked_dir doesn't require that the i_rwsem be locked. We only really need it to handle manipulation of the dcache. The rest (filling of the inode, updating dentry leases, etc.) already has its own locking. Add a new r_req_flags bit that indicates whether the parent is locked when doing the request, and rename the pointer to "r_parent". For now, all the places that set r_parent also set this flag, but that will change in a later patch. Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Yan, Zheng <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-01-31 15:28:26 +00:00
req->r_parent = dir;
ceph: take reference to req->r_parent at point of assignment Currently, we set the r_parent pointer but then don't take a reference to it until we submit the request. If we end up freeing the req before that point, then we'll do a iput when we shouldn't. Instead, take the inode reference in the callers, so that it's always safe to call ceph_mdsc_put_request on the req, even before submission. Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: Luis Henriques <lhenriques@suse.de> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2021-06-18 17:05:06 +00:00
ihold(dir);
ceph: add a new flag to indicate whether parent is locked struct ceph_mds_request has an r_locked_dir pointer, which is set to indicate the parent inode and that its i_rwsem is locked. In some critical places, we need to be able to indicate the parent inode to the request handling code, even when its i_rwsem may not be locked. Most of the code that operates on r_locked_dir doesn't require that the i_rwsem be locked. We only really need it to handle manipulation of the dcache. The rest (filling of the inode, updating dentry leases, etc.) already has its own locking. Add a new r_req_flags bit that indicates whether the parent is locked when doing the request, and rename the pointer to "r_parent". For now, all the places that set r_parent also set this flag, but that will change in a later patch. Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Yan, Zheng <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-01-31 15:28:26 +00:00
set_bit(CEPH_MDS_R_PARENT_LOCKED, &req->r_req_flags);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
req->r_args.mkdir.mode = cpu_to_le32(mode);
ceph: voluntarily drop Ax cap for requests that create new inode MDS need to rdlock directory inode's authlock when handling these requests. Voluntarily dropping CEPH_CAP_AUTH_EXCL avoids a cap revoke message. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-11-23 09:47:15 +00:00
req->r_dentry_drop = CEPH_CAP_FILE_SHARED | CEPH_CAP_AUTH_EXCL;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
req->r_dentry_unless = CEPH_CAP_FILE_EXCL;
ceph: rename struct ceph_acls_info to ceph_acl_sec_ctx Also rename ceph_release_acls_info() to ceph_release_acl_sec_ctx(). And move their definitions to different files. This is preparation for security label support. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-05-26 07:35:39 +00:00
if (as_ctx.pagelist) {
req->r_pagelist = as_ctx.pagelist;
as_ctx.pagelist = NULL;
ceph: include the initial ACL in create/mkdir/mknod MDS requests Current code set new file/directory's initial ACL in a non-atomic manner. Client first sends request to MDS to create new file/directory, then set the initial ACL after the new file/directory is successfully created. The fix is include the initial ACL in create/mkdir/mknod MDS requests. So MDS can handle creating file/directory and setting the initial ACL in one request. Signed-off-by: Yan, Zheng <zyan@redhat.com> Reviewed-by: Sage Weil <sage@redhat.com>
2014-09-16 12:35:17 +00:00
}
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
err = ceph_mdsc_do_request(mdsc, dir, req);
ceph: fix mksnap crash mksnap reply only contain 'target', does not contain 'dentry'. So it's wrong to use req->r_reply_info.head->is_dentry to detect traceless reply. Signed-off-by: Yan, Zheng <zyan@redhat.com> Reviewed-by: Sage Weil <sage@redhat.com>
2014-12-10 08:17:31 +00:00
if (!err &&
!req->r_reply_info.head->is_target &&
!req->r_reply_info.head->is_dentry)
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
err = ceph_handle_notrace_create(dir, dentry);
ceph_mdsc_put_request(req);
out:
ceph: add missing init_acl() for mkdir() and atomic_open() Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
2014-02-11 04:55:05 +00:00
if (!err)
ceph: rename struct ceph_acls_info to ceph_acl_sec_ctx Also rename ceph_release_acls_info() to ceph_release_acl_sec_ctx(). And move their definitions to different files. This is preparation for security label support. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-05-26 07:35:39 +00:00
ceph_init_inode_acls(d_inode(dentry), &as_ctx);
ceph: add missing init_acl() for mkdir() and atomic_open() Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
2014-02-11 04:55:05 +00:00
else
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
d_drop(dentry);
ceph: rename struct ceph_acls_info to ceph_acl_sec_ctx Also rename ceph_release_acls_info() to ceph_release_acl_sec_ctx(). And move their definitions to different files. This is preparation for security label support. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-05-26 07:35:39 +00:00
ceph_release_acl_sec_ctx(&as_ctx);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
return err;
}
static int ceph_link(struct dentry *old_dentry, struct inode *dir,
struct dentry *dentry)
{
ceph: add ceph_sb_to_mdsc helper support to parse the mdsc This will help simplify the code. [ jlayton: fix minor merge conflict in quota.c ] Signed-off-by: Xiubo Li <xiubli@redhat.com> Signed-off-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-09-03 13:01:39 +00:00
struct ceph_mds_client *mdsc = ceph_sb_to_mdsc(dir->i_sb);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
struct ceph_mds_request *req;
int err;
ceph: wait for the first reply of inflight async unlink In async unlink case the kclient won't wait for the first reply from MDS and just drop all the links and unhash the dentry and then succeeds immediately. For any new create/link/rename,etc requests followed by using the same file names we must wait for the first reply of the inflight unlink request, or the MDS possibly will fail these following requests with -EEXIST if the inflight async unlink request was delayed for some reasons. And the worst case is that for the none async openc request it will successfully open the file if the CDentry hasn't been unlinked yet, but later the previous delayed async unlink request will remove the CDenty. That means the just created file is possiblly deleted later by accident. We need to wait for the inflight async unlink requests to finish when creating new files/directories by using the same file names. Link: https://tracker.ceph.com/issues/55332 Signed-off-by: Xiubo Li <xiubli@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2022-05-10 01:47:01 +00:00
err = ceph_wait_on_conflict_unlink(dentry);
if (err)
return err;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
if (ceph_snap(dir) != CEPH_NOSNAP)
return -EROFS;
dout("link in dir %p old_dentry %p dentry %p\n", dir,
old_dentry, dentry);
req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_LINK, USE_AUTH_MDS);
if (IS_ERR(req)) {
d_drop(dentry);
return PTR_ERR(req);
}
req->r_dentry = dget(dentry);
req->r_num_caps = 2;
ceph: do not set r_old_dentry_dir on link() This is racy--we do not know whather d_parent has changed out from underneath us because i_mutex is not held on the source inode's directory. Also, taking this reference is useless. Reported-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Sage Weil <sage@inktank.com> Reviewed-by: Yan, Zheng <zheng.z.yan@intel.com>
2013-02-05 21:41:23 +00:00
req->r_old_dentry = dget(old_dentry);
ceph: add a new flag to indicate whether parent is locked struct ceph_mds_request has an r_locked_dir pointer, which is set to indicate the parent inode and that its i_rwsem is locked. In some critical places, we need to be able to indicate the parent inode to the request handling code, even when its i_rwsem may not be locked. Most of the code that operates on r_locked_dir doesn't require that the i_rwsem be locked. We only really need it to handle manipulation of the dcache. The rest (filling of the inode, updating dentry leases, etc.) already has its own locking. Add a new r_req_flags bit that indicates whether the parent is locked when doing the request, and rename the pointer to "r_parent". For now, all the places that set r_parent also set this flag, but that will change in a later patch. Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Yan, Zheng <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-01-31 15:28:26 +00:00
req->r_parent = dir;
ceph: take reference to req->r_parent at point of assignment Currently, we set the r_parent pointer but then don't take a reference to it until we submit the request. If we end up freeing the req before that point, then we'll do a iput when we shouldn't. Instead, take the inode reference in the callers, so that it's always safe to call ceph_mdsc_put_request on the req, even before submission. Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: Luis Henriques <lhenriques@suse.de> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2021-06-18 17:05:06 +00:00
ihold(dir);
ceph: add a new flag to indicate whether parent is locked struct ceph_mds_request has an r_locked_dir pointer, which is set to indicate the parent inode and that its i_rwsem is locked. In some critical places, we need to be able to indicate the parent inode to the request handling code, even when its i_rwsem may not be locked. Most of the code that operates on r_locked_dir doesn't require that the i_rwsem be locked. We only really need it to handle manipulation of the dcache. The rest (filling of the inode, updating dentry leases, etc.) already has its own locking. Add a new r_req_flags bit that indicates whether the parent is locked when doing the request, and rename the pointer to "r_parent". For now, all the places that set r_parent also set this flag, but that will change in a later patch. Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Yan, Zheng <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-01-31 15:28:26 +00:00
set_bit(CEPH_MDS_R_PARENT_LOCKED, &req->r_req_flags);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
req->r_dentry_drop = CEPH_CAP_FILE_SHARED;
req->r_dentry_unless = CEPH_CAP_FILE_EXCL;
ceph: drop CAP_LINK_SHARED when sending "link" request to MDS To handle "link" request, the MDS need to xlock inode's linklock, which requires revoking any CAP_LINK_SHARED. Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com> Reviewed-by: Sage Weil <sage@inktank.com>
2013-07-21 12:25:25 +00:00
/* release LINK_SHARED on source inode (mds will lock it) */
ceph: voluntarily drop Lx cap for link/rename requests MDS need to xlock inode's linklock when handling link/rename requests. Voluntarily dropping CEPH_CAP_AUTH_EXCL avoids a cap revoke message. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-11-23 09:59:13 +00:00
req->r_old_inode_drop = CEPH_CAP_LINK_SHARED | CEPH_CAP_LINK_EXCL;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
err = ceph_mdsc_do_request(mdsc, dir, req);
ceph: use ihold when we already have an inode ref We should use ihold whenever we already have a stable inode ref, even when we aren't holding i_lock. This avoids adding new and unnecessary locking dependencies. Signed-off-by: Sage Weil <sage@newdream.net>
2011-05-27 16:24:26 +00:00
if (err) {
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
d_drop(dentry);
ceph: use ihold when we already have an inode ref We should use ihold whenever we already have a stable inode ref, even when we aren't holding i_lock. This avoids adding new and unnecessary locking dependencies. Signed-off-by: Sage Weil <sage@newdream.net>
2011-05-27 16:24:26 +00:00
} else if (!req->r_reply_info.head->is_dentry) {
VFS: normal filesystems (and lustre): d_inode() annotations that's the bulk of filesystem drivers dealing with inodes of their own Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2015-03-17 22:25:59 +00:00
ihold(d_inode(old_dentry));
d_instantiate(dentry, d_inode(old_dentry));
ceph: use ihold when we already have an inode ref We should use ihold whenever we already have a stable inode ref, even when we aren't holding i_lock. This avoids adding new and unnecessary locking dependencies. Signed-off-by: Sage Weil <sage@newdream.net>
2011-05-27 16:24:26 +00:00
}
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
ceph_mdsc_put_request(req);
return err;
}
ceph: perform asynchronous unlink if we have sufficient caps The MDS is getting a new lock-caching facility that will allow it to cache the necessary locks to allow asynchronous directory operations. Since the CEPH_CAP_FILE_* caps are currently unused on directories, we can repurpose those bits for this purpose. When performing an unlink, if we have Fx on the parent directory, and CEPH_CAP_DIR_UNLINK (aka Fr), and we know that the dentry being removed is the primary link, then then we can fire off an unlink request immediately and don't need to wait on reply before returning. In that situation, just fix up the dcache and link count and return immediately after issuing the call to the MDS. This does mean that we need to hold an extra reference to the inode being unlinked, and extra references to the caps to avoid races. Those references are put and error handling is done in the r_callback routine. If the operation ends up failing, then set a writeback error on the directory inode, and the inode itself that can be fetched later by an fsync on the dir. The behavior of dir caps is slightly different from caps on normal files. Because these are just considered an optimization, if the session is reconnected, we will not automatically reclaim them. They are instead considered lost until we do another synchronous op in the parent directory. Async dirops are enabled via the "nowsync" mount option, which is patterned after the xfs "wsync" mount option. For now, the default is "wsync", but eventually we may flip that. Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-04-02 19:35:56 +00:00
static void ceph_async_unlink_cb(struct ceph_mds_client *mdsc,
struct ceph_mds_request *req)
{
ceph: wait for the first reply of inflight async unlink In async unlink case the kclient won't wait for the first reply from MDS and just drop all the links and unhash the dentry and then succeeds immediately. For any new create/link/rename,etc requests followed by using the same file names we must wait for the first reply of the inflight unlink request, or the MDS possibly will fail these following requests with -EEXIST if the inflight async unlink request was delayed for some reasons. And the worst case is that for the none async openc request it will successfully open the file if the CDentry hasn't been unlinked yet, but later the previous delayed async unlink request will remove the CDenty. That means the just created file is possiblly deleted later by accident. We need to wait for the inflight async unlink requests to finish when creating new files/directories by using the same file names. Link: https://tracker.ceph.com/issues/55332 Signed-off-by: Xiubo Li <xiubli@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2022-05-10 01:47:01 +00:00
struct dentry *dentry = req->r_dentry;
struct ceph_fs_client *fsc = ceph_sb_to_client(dentry->d_sb);
struct ceph_dentry_info *di = ceph_dentry(dentry);
ceph: perform asynchronous unlink if we have sufficient caps The MDS is getting a new lock-caching facility that will allow it to cache the necessary locks to allow asynchronous directory operations. Since the CEPH_CAP_FILE_* caps are currently unused on directories, we can repurpose those bits for this purpose. When performing an unlink, if we have Fx on the parent directory, and CEPH_CAP_DIR_UNLINK (aka Fr), and we know that the dentry being removed is the primary link, then then we can fire off an unlink request immediately and don't need to wait on reply before returning. In that situation, just fix up the dcache and link count and return immediately after issuing the call to the MDS. This does mean that we need to hold an extra reference to the inode being unlinked, and extra references to the caps to avoid races. Those references are put and error handling is done in the r_callback routine. If the operation ends up failing, then set a writeback error on the directory inode, and the inode itself that can be fetched later by an fsync on the dir. The behavior of dir caps is slightly different from caps on normal files. Because these are just considered an optimization, if the session is reconnected, we will not automatically reclaim them. They are instead considered lost until we do another synchronous op in the parent directory. Async dirops are enabled via the "nowsync" mount option, which is patterned after the xfs "wsync" mount option. For now, the default is "wsync", but eventually we may flip that. Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-04-02 19:35:56 +00:00
int result = req->r_err ? req->r_err :
le32_to_cpu(req->r_reply_info.head->result);
ceph: wait for the first reply of inflight async unlink In async unlink case the kclient won't wait for the first reply from MDS and just drop all the links and unhash the dentry and then succeeds immediately. For any new create/link/rename,etc requests followed by using the same file names we must wait for the first reply of the inflight unlink request, or the MDS possibly will fail these following requests with -EEXIST if the inflight async unlink request was delayed for some reasons. And the worst case is that for the none async openc request it will successfully open the file if the CDentry hasn't been unlinked yet, but later the previous delayed async unlink request will remove the CDenty. That means the just created file is possiblly deleted later by accident. We need to wait for the inflight async unlink requests to finish when creating new files/directories by using the same file names. Link: https://tracker.ceph.com/issues/55332 Signed-off-by: Xiubo Li <xiubli@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2022-05-10 01:47:01 +00:00
if (!test_bit(CEPH_DENTRY_ASYNC_UNLINK_BIT, &di->flags))
pr_warn("%s dentry %p:%pd async unlink bit is not set\n",
__func__, dentry, dentry);
spin_lock(&fsc->async_unlink_conflict_lock);
hash_del_rcu(&di->hnode);
spin_unlock(&fsc->async_unlink_conflict_lock);
spin_lock(&dentry->d_lock);
di->flags &= ~CEPH_DENTRY_ASYNC_UNLINK;
wake_up_bit(&di->flags, CEPH_DENTRY_ASYNC_UNLINK_BIT);
spin_unlock(&dentry->d_lock);
synchronize_rcu();
ceph: perform asynchronous unlink if we have sufficient caps The MDS is getting a new lock-caching facility that will allow it to cache the necessary locks to allow asynchronous directory operations. Since the CEPH_CAP_FILE_* caps are currently unused on directories, we can repurpose those bits for this purpose. When performing an unlink, if we have Fx on the parent directory, and CEPH_CAP_DIR_UNLINK (aka Fr), and we know that the dentry being removed is the primary link, then then we can fire off an unlink request immediately and don't need to wait on reply before returning. In that situation, just fix up the dcache and link count and return immediately after issuing the call to the MDS. This does mean that we need to hold an extra reference to the inode being unlinked, and extra references to the caps to avoid races. Those references are put and error handling is done in the r_callback routine. If the operation ends up failing, then set a writeback error on the directory inode, and the inode itself that can be fetched later by an fsync on the dir. The behavior of dir caps is slightly different from caps on normal files. Because these are just considered an optimization, if the session is reconnected, we will not automatically reclaim them. They are instead considered lost until we do another synchronous op in the parent directory. Async dirops are enabled via the "nowsync" mount option, which is patterned after the xfs "wsync" mount option. For now, the default is "wsync", but eventually we may flip that. Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-04-02 19:35:56 +00:00
if (result == -EJUKEBOX)
goto out;
/* If op failed, mark everyone involved for errors */
if (result) {
ceph: fix potential bad pointer deref in async dirops cb's The new async dirops callback routines can pass ERR_PTR values to ceph_mdsc_free_path, which could cause an oops. Make ceph_mdsc_free_path ignore ERR_PTR values. Also, ensure that the pr_warn messages look sane even if ceph_mdsc_build_path fails. Reported-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: Ilya Dryomov <idryomov@gmail.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-04-08 12:41:38 +00:00
int pathlen = 0;
u64 base = 0;
ceph: wait for the first reply of inflight async unlink In async unlink case the kclient won't wait for the first reply from MDS and just drop all the links and unhash the dentry and then succeeds immediately. For any new create/link/rename,etc requests followed by using the same file names we must wait for the first reply of the inflight unlink request, or the MDS possibly will fail these following requests with -EEXIST if the inflight async unlink request was delayed for some reasons. And the worst case is that for the none async openc request it will successfully open the file if the CDentry hasn't been unlinked yet, but later the previous delayed async unlink request will remove the CDenty. That means the just created file is possiblly deleted later by accident. We need to wait for the inflight async unlink requests to finish when creating new files/directories by using the same file names. Link: https://tracker.ceph.com/issues/55332 Signed-off-by: Xiubo Li <xiubli@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2022-05-10 01:47:01 +00:00
char *path = ceph_mdsc_build_path(dentry, &pathlen,
ceph: perform asynchronous unlink if we have sufficient caps The MDS is getting a new lock-caching facility that will allow it to cache the necessary locks to allow asynchronous directory operations. Since the CEPH_CAP_FILE_* caps are currently unused on directories, we can repurpose those bits for this purpose. When performing an unlink, if we have Fx on the parent directory, and CEPH_CAP_DIR_UNLINK (aka Fr), and we know that the dentry being removed is the primary link, then then we can fire off an unlink request immediately and don't need to wait on reply before returning. In that situation, just fix up the dcache and link count and return immediately after issuing the call to the MDS. This does mean that we need to hold an extra reference to the inode being unlinked, and extra references to the caps to avoid races. Those references are put and error handling is done in the r_callback routine. If the operation ends up failing, then set a writeback error on the directory inode, and the inode itself that can be fetched later by an fsync on the dir. The behavior of dir caps is slightly different from caps on normal files. Because these are just considered an optimization, if the session is reconnected, we will not automatically reclaim them. They are instead considered lost until we do another synchronous op in the parent directory. Async dirops are enabled via the "nowsync" mount option, which is patterned after the xfs "wsync" mount option. For now, the default is "wsync", but eventually we may flip that. Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-04-02 19:35:56 +00:00
&base, 0);
/* mark error on parent + clear complete */
mapping_set_error(req->r_parent->i_mapping, result);
ceph_dir_clear_complete(req->r_parent);
/* drop the dentry -- we don't know its status */
ceph: wait for the first reply of inflight async unlink In async unlink case the kclient won't wait for the first reply from MDS and just drop all the links and unhash the dentry and then succeeds immediately. For any new create/link/rename,etc requests followed by using the same file names we must wait for the first reply of the inflight unlink request, or the MDS possibly will fail these following requests with -EEXIST if the inflight async unlink request was delayed for some reasons. And the worst case is that for the none async openc request it will successfully open the file if the CDentry hasn't been unlinked yet, but later the previous delayed async unlink request will remove the CDenty. That means the just created file is possiblly deleted later by accident. We need to wait for the inflight async unlink requests to finish when creating new files/directories by using the same file names. Link: https://tracker.ceph.com/issues/55332 Signed-off-by: Xiubo Li <xiubli@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2022-05-10 01:47:01 +00:00
if (!d_unhashed(dentry))
d_drop(dentry);
ceph: perform asynchronous unlink if we have sufficient caps The MDS is getting a new lock-caching facility that will allow it to cache the necessary locks to allow asynchronous directory operations. Since the CEPH_CAP_FILE_* caps are currently unused on directories, we can repurpose those bits for this purpose. When performing an unlink, if we have Fx on the parent directory, and CEPH_CAP_DIR_UNLINK (aka Fr), and we know that the dentry being removed is the primary link, then then we can fire off an unlink request immediately and don't need to wait on reply before returning. In that situation, just fix up the dcache and link count and return immediately after issuing the call to the MDS. This does mean that we need to hold an extra reference to the inode being unlinked, and extra references to the caps to avoid races. Those references are put and error handling is done in the r_callback routine. If the operation ends up failing, then set a writeback error on the directory inode, and the inode itself that can be fetched later by an fsync on the dir. The behavior of dir caps is slightly different from caps on normal files. Because these are just considered an optimization, if the session is reconnected, we will not automatically reclaim them. They are instead considered lost until we do another synchronous op in the parent directory. Async dirops are enabled via the "nowsync" mount option, which is patterned after the xfs "wsync" mount option. For now, the default is "wsync", but eventually we may flip that. Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-04-02 19:35:56 +00:00
/* mark inode itself for an error (since metadata is bogus) */
mapping_set_error(req->r_old_inode->i_mapping, result);
ceph: wait for the first reply of inflight async unlink In async unlink case the kclient won't wait for the first reply from MDS and just drop all the links and unhash the dentry and then succeeds immediately. For any new create/link/rename,etc requests followed by using the same file names we must wait for the first reply of the inflight unlink request, or the MDS possibly will fail these following requests with -EEXIST if the inflight async unlink request was delayed for some reasons. And the worst case is that for the none async openc request it will successfully open the file if the CDentry hasn't been unlinked yet, but later the previous delayed async unlink request will remove the CDenty. That means the just created file is possiblly deleted later by accident. We need to wait for the inflight async unlink requests to finish when creating new files/directories by using the same file names. Link: https://tracker.ceph.com/issues/55332 Signed-off-by: Xiubo Li <xiubli@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2022-05-10 01:47:01 +00:00
pr_warn("async unlink failure path=(%llx)%s result=%d!\n",
ceph: perform asynchronous unlink if we have sufficient caps The MDS is getting a new lock-caching facility that will allow it to cache the necessary locks to allow asynchronous directory operations. Since the CEPH_CAP_FILE_* caps are currently unused on directories, we can repurpose those bits for this purpose. When performing an unlink, if we have Fx on the parent directory, and CEPH_CAP_DIR_UNLINK (aka Fr), and we know that the dentry being removed is the primary link, then then we can fire off an unlink request immediately and don't need to wait on reply before returning. In that situation, just fix up the dcache and link count and return immediately after issuing the call to the MDS. This does mean that we need to hold an extra reference to the inode being unlinked, and extra references to the caps to avoid races. Those references are put and error handling is done in the r_callback routine. If the operation ends up failing, then set a writeback error on the directory inode, and the inode itself that can be fetched later by an fsync on the dir. The behavior of dir caps is slightly different from caps on normal files. Because these are just considered an optimization, if the session is reconnected, we will not automatically reclaim them. They are instead considered lost until we do another synchronous op in the parent directory. Async dirops are enabled via the "nowsync" mount option, which is patterned after the xfs "wsync" mount option. For now, the default is "wsync", but eventually we may flip that. Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-04-02 19:35:56 +00:00
base, IS_ERR(path) ? "<<bad>>" : path, result);
ceph_mdsc_free_path(path, pathlen);
}
out:
iput(req->r_old_inode);
ceph_mdsc_release_dir_caps(req);
}
static int get_caps_for_async_unlink(struct inode *dir, struct dentry *dentry)
{
struct ceph_inode_info *ci = ceph_inode(dir);
struct ceph_dentry_info *di;
int got = 0, want = CEPH_CAP_FILE_EXCL | CEPH_CAP_DIR_UNLINK;
spin_lock(&ci->i_ceph_lock);
if ((__ceph_caps_issued(ci, NULL) & want) == want) {
ceph_take_cap_refs(ci, want, false);
got = want;
}
spin_unlock(&ci->i_ceph_lock);
/* If we didn't get anything, return 0 */
if (!got)
return 0;
spin_lock(&dentry->d_lock);
di = ceph_dentry(dentry);
/*
* - We are holding Fx, which implies Fs caps.
* - Only support async unlink for primary linkage
*/
if (atomic_read(&ci->i_shared_gen) != di->lease_shared_gen ||
!(di->flags & CEPH_DENTRY_PRIMARY_LINK))
want = 0;
spin_unlock(&dentry->d_lock);
/* Do we still want what we've got? */
if (want == got)
return got;
ceph_put_cap_refs(ci, got);
return 0;
}
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
/*
* rmdir and unlink are differ only by the metadata op code
*/
static int ceph_unlink(struct inode *dir, struct dentry *dentry)
{
ceph: factor out libceph from Ceph file system This factors out protocol and low-level storage parts of ceph into a separate libceph module living in net/ceph and include/linux/ceph. This is mostly a matter of moving files around. However, a few key pieces of the interface change as well: - ceph_client becomes ceph_fs_client and ceph_client, where the latter captures the mon and osd clients, and the fs_client gets the mds client and file system specific pieces. - Mount option parsing and debugfs setup is correspondingly broken into two pieces. - The mon client gets a generic handler callback for otherwise unknown messages (mds map, in this case). - The basic supported/required feature bits can be expanded (and are by ceph_fs_client). No functional change, aside from some subtle error handling cases that got cleaned up in the refactoring process. Signed-off-by: Sage Weil <sage@newdream.net>
2010-04-06 22:14:15 +00:00
struct ceph_fs_client *fsc = ceph_sb_to_client(dir->i_sb);
struct ceph_mds_client *mdsc = fsc->mdsc;
VFS: normal filesystems (and lustre): d_inode() annotations that's the bulk of filesystem drivers dealing with inodes of their own Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2015-03-17 22:25:59 +00:00
struct inode *inode = d_inode(dentry);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
struct ceph_mds_request *req;
ceph: perform asynchronous unlink if we have sufficient caps The MDS is getting a new lock-caching facility that will allow it to cache the necessary locks to allow asynchronous directory operations. Since the CEPH_CAP_FILE_* caps are currently unused on directories, we can repurpose those bits for this purpose. When performing an unlink, if we have Fx on the parent directory, and CEPH_CAP_DIR_UNLINK (aka Fr), and we know that the dentry being removed is the primary link, then then we can fire off an unlink request immediately and don't need to wait on reply before returning. In that situation, just fix up the dcache and link count and return immediately after issuing the call to the MDS. This does mean that we need to hold an extra reference to the inode being unlinked, and extra references to the caps to avoid races. Those references are put and error handling is done in the r_callback routine. If the operation ends up failing, then set a writeback error on the directory inode, and the inode itself that can be fetched later by an fsync on the dir. The behavior of dir caps is slightly different from caps on normal files. Because these are just considered an optimization, if the session is reconnected, we will not automatically reclaim them. They are instead considered lost until we do another synchronous op in the parent directory. Async dirops are enabled via the "nowsync" mount option, which is patterned after the xfs "wsync" mount option. For now, the default is "wsync", but eventually we may flip that. Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-04-02 19:35:56 +00:00
bool try_async = ceph_test_mount_opt(fsc, ASYNC_DIROPS);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
int err = -EROFS;
int op;
if (ceph_snap(dir) == CEPH_SNAPDIR) {
/* rmdir .snap/foo is RMSNAP */
assorted conversions to %p[dD] Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2014-10-22 00:11:25 +00:00
dout("rmsnap dir %p '%pd' dn %p\n", dir, dentry, dentry);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
op = CEPH_MDS_OP_RMSNAP;
} else if (ceph_snap(dir) == CEPH_NOSNAP) {
dout("unlink/rmdir dir %p dn %p inode %p\n",
dir, dentry, inode);
VFS: (Scripted) Convert S_ISLNK/DIR/REG(dentry->d_inode) to d_is_*(dentry) Convert the following where appropriate: (1) S_ISLNK(dentry->d_inode) to d_is_symlink(dentry). (2) S_ISREG(dentry->d_inode) to d_is_reg(dentry). (3) S_ISDIR(dentry->d_inode) to d_is_dir(dentry). This is actually more complicated than it appears as some calls should be converted to d_can_lookup() instead. The difference is whether the directory in question is a real dir with a ->lookup op or whether it's a fake dir with a ->d_automount op. In some circumstances, we can subsume checks for dentry->d_inode not being NULL into this, provided we the code isn't in a filesystem that expects d_inode to be NULL if the dirent really *is* negative (ie. if we're going to use d_inode() rather than d_backing_inode() to get the inode pointer). Note that the dentry type field may be set to something other than DCACHE_MISS_TYPE when d_inode is NULL in the case of unionmount, where the VFS manages the fall-through from a negative dentry to a lower layer. In such a case, the dentry type of the negative union dentry is set to the same as the type of the lower dentry. However, if you know d_inode is not NULL at the call site, then you can use the d_is_xxx() functions even in a filesystem. There is one further complication: a 0,0 chardev dentry may be labelled DCACHE_WHITEOUT_TYPE rather than DCACHE_SPECIAL_TYPE. Strictly, this was intended for special directory entry types that don't have attached inodes. The following perl+coccinelle script was used: use strict; my @callers; open($fd, 'git grep -l \'S_IS[A-Z].*->d_inode\' |') || die "Can't grep for S_ISDIR and co. callers"; @callers = <$fd>; close($fd); unless (@callers) { print "No matches\n"; exit(0); } my @cocci = ( '@@', 'expression E;', '@@', '', '- S_ISLNK(E->d_inode->i_mode)', '+ d_is_symlink(E)', '', '@@', 'expression E;', '@@', '', '- S_ISDIR(E->d_inode->i_mode)', '+ d_is_dir(E)', '', '@@', 'expression E;', '@@', '', '- S_ISREG(E->d_inode->i_mode)', '+ d_is_reg(E)' ); my $coccifile = "tmp.sp.cocci"; open($fd, ">$coccifile") || die $coccifile; print($fd "$_\n") || die $coccifile foreach (@cocci); close($fd); foreach my $file (@callers) { chomp $file; print "Processing ", $file, "\n"; system("spatch", "--sp-file", $coccifile, $file, "--in-place", "--no-show-diff") == 0 || die "spatch failed"; } [AV: overlayfs parts skipped] Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2015-01-29 12:02:35 +00:00
op = d_is_dir(dentry) ?
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
CEPH_MDS_OP_RMDIR : CEPH_MDS_OP_UNLINK;
} else
goto out;
ceph: perform asynchronous unlink if we have sufficient caps The MDS is getting a new lock-caching facility that will allow it to cache the necessary locks to allow asynchronous directory operations. Since the CEPH_CAP_FILE_* caps are currently unused on directories, we can repurpose those bits for this purpose. When performing an unlink, if we have Fx on the parent directory, and CEPH_CAP_DIR_UNLINK (aka Fr), and we know that the dentry being removed is the primary link, then then we can fire off an unlink request immediately and don't need to wait on reply before returning. In that situation, just fix up the dcache and link count and return immediately after issuing the call to the MDS. This does mean that we need to hold an extra reference to the inode being unlinked, and extra references to the caps to avoid races. Those references are put and error handling is done in the r_callback routine. If the operation ends up failing, then set a writeback error on the directory inode, and the inode itself that can be fetched later by an fsync on the dir. The behavior of dir caps is slightly different from caps on normal files. Because these are just considered an optimization, if the session is reconnected, we will not automatically reclaim them. They are instead considered lost until we do another synchronous op in the parent directory. Async dirops are enabled via the "nowsync" mount option, which is patterned after the xfs "wsync" mount option. For now, the default is "wsync", but eventually we may flip that. Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-04-02 19:35:56 +00:00
retry:
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
req = ceph_mdsc_create_request(mdsc, op, USE_AUTH_MDS);
if (IS_ERR(req)) {
err = PTR_ERR(req);
goto out;
}
req->r_dentry = dget(dentry);
req->r_num_caps = 2;
ceph: add a new flag to indicate whether parent is locked struct ceph_mds_request has an r_locked_dir pointer, which is set to indicate the parent inode and that its i_rwsem is locked. In some critical places, we need to be able to indicate the parent inode to the request handling code, even when its i_rwsem may not be locked. Most of the code that operates on r_locked_dir doesn't require that the i_rwsem be locked. We only really need it to handle manipulation of the dcache. The rest (filling of the inode, updating dentry leases, etc.) already has its own locking. Add a new r_req_flags bit that indicates whether the parent is locked when doing the request, and rename the pointer to "r_parent". For now, all the places that set r_parent also set this flag, but that will change in a later patch. Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Yan, Zheng <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-01-31 15:28:26 +00:00
req->r_parent = dir;
ceph: take reference to req->r_parent at point of assignment Currently, we set the r_parent pointer but then don't take a reference to it until we submit the request. If we end up freeing the req before that point, then we'll do a iput when we shouldn't. Instead, take the inode reference in the callers, so that it's always safe to call ceph_mdsc_put_request on the req, even before submission. Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: Luis Henriques <lhenriques@suse.de> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2021-06-18 17:05:06 +00:00
ihold(dir);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
req->r_dentry_drop = CEPH_CAP_FILE_SHARED;
req->r_dentry_unless = CEPH_CAP_FILE_EXCL;
ceph: flush dirty caps of unlinked inode ASAP Client should release unlinked inode from its cache ASAP. But client can't release inode with dirty caps. Link: http://tracker.ceph.com/issues/22886 Signed-off-by: Zhi Zhang <zhang.david2011@gmail.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-01-24 13:24:33 +00:00
req->r_inode_drop = ceph_drop_caps_for_unlink(inode);
ceph: perform asynchronous unlink if we have sufficient caps The MDS is getting a new lock-caching facility that will allow it to cache the necessary locks to allow asynchronous directory operations. Since the CEPH_CAP_FILE_* caps are currently unused on directories, we can repurpose those bits for this purpose. When performing an unlink, if we have Fx on the parent directory, and CEPH_CAP_DIR_UNLINK (aka Fr), and we know that the dentry being removed is the primary link, then then we can fire off an unlink request immediately and don't need to wait on reply before returning. In that situation, just fix up the dcache and link count and return immediately after issuing the call to the MDS. This does mean that we need to hold an extra reference to the inode being unlinked, and extra references to the caps to avoid races. Those references are put and error handling is done in the r_callback routine. If the operation ends up failing, then set a writeback error on the directory inode, and the inode itself that can be fetched later by an fsync on the dir. The behavior of dir caps is slightly different from caps on normal files. Because these are just considered an optimization, if the session is reconnected, we will not automatically reclaim them. They are instead considered lost until we do another synchronous op in the parent directory. Async dirops are enabled via the "nowsync" mount option, which is patterned after the xfs "wsync" mount option. For now, the default is "wsync", but eventually we may flip that. Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-04-02 19:35:56 +00:00
if (try_async && op == CEPH_MDS_OP_UNLINK &&
(req->r_dir_caps = get_caps_for_async_unlink(dir, dentry))) {
ceph: wait for the first reply of inflight async unlink In async unlink case the kclient won't wait for the first reply from MDS and just drop all the links and unhash the dentry and then succeeds immediately. For any new create/link/rename,etc requests followed by using the same file names we must wait for the first reply of the inflight unlink request, or the MDS possibly will fail these following requests with -EEXIST if the inflight async unlink request was delayed for some reasons. And the worst case is that for the none async openc request it will successfully open the file if the CDentry hasn't been unlinked yet, but later the previous delayed async unlink request will remove the CDenty. That means the just created file is possiblly deleted later by accident. We need to wait for the inflight async unlink requests to finish when creating new files/directories by using the same file names. Link: https://tracker.ceph.com/issues/55332 Signed-off-by: Xiubo Li <xiubli@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2022-05-10 01:47:01 +00:00
struct ceph_dentry_info *di = ceph_dentry(dentry);
ceph: fix inode number handling on arches with 32-bit ino_t Tuan and Ulrich mentioned that they were hitting a problem on s390x, which has a 32-bit ino_t value, even though it's a 64-bit arch (for historical reasons). I think the current handling of inode numbers in the ceph driver is wrong. It tries to use 32-bit inode numbers on 32-bit arches, but that's actually not a problem. 32-bit arches can deal with 64-bit inode numbers just fine when userland code is compiled with LFS support (the common case these days). What we really want to do is just use 64-bit numbers everywhere, unless someone has mounted with the ino32 mount option. In that case, we want to ensure that we hash the inode number down to something that will fit in 32 bits before presenting the value to userland. Add new helper functions that do this, and only do the conversion before presenting these values to userland in getattr and readdir. The inode table hashvalue is changed to just cast the inode number to unsigned long, as low-order bits are the most likely to vary anyway. While it's not strictly required, we do want to put something in inode->i_ino. Instead of basing it on BITS_PER_LONG, however, base it on the size of the ino_t type. NOTE: This is a user-visible change on 32-bit arches: 1/ inode numbers will be seen to have changed between kernel versions. 32-bit arches will see large inode numbers now instead of the hashed ones they saw before. 2/ any really old software not built with LFS support may start failing stat() calls with -EOVERFLOW on inode numbers >2^32. Nothing much we can do about these, but hopefully the intersection of people running such code on ceph will be very small. The workaround for both problems is to mount with "-o ino32". [ idryomov: changelog tweak ] URL: https://tracker.ceph.com/issues/46828 Reported-by: Ulrich Weigand <Ulrich.Weigand@de.ibm.com> Reported-and-Tested-by: Tuan Hoang1 <Tuan.Hoang1@ibm.com> Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-08-18 12:03:48 +00:00
dout("async unlink on %llu/%.*s caps=%s", ceph_ino(dir),
ceph: perform asynchronous unlink if we have sufficient caps The MDS is getting a new lock-caching facility that will allow it to cache the necessary locks to allow asynchronous directory operations. Since the CEPH_CAP_FILE_* caps are currently unused on directories, we can repurpose those bits for this purpose. When performing an unlink, if we have Fx on the parent directory, and CEPH_CAP_DIR_UNLINK (aka Fr), and we know that the dentry being removed is the primary link, then then we can fire off an unlink request immediately and don't need to wait on reply before returning. In that situation, just fix up the dcache and link count and return immediately after issuing the call to the MDS. This does mean that we need to hold an extra reference to the inode being unlinked, and extra references to the caps to avoid races. Those references are put and error handling is done in the r_callback routine. If the operation ends up failing, then set a writeback error on the directory inode, and the inode itself that can be fetched later by an fsync on the dir. The behavior of dir caps is slightly different from caps on normal files. Because these are just considered an optimization, if the session is reconnected, we will not automatically reclaim them. They are instead considered lost until we do another synchronous op in the parent directory. Async dirops are enabled via the "nowsync" mount option, which is patterned after the xfs "wsync" mount option. For now, the default is "wsync", but eventually we may flip that. Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-04-02 19:35:56 +00:00
dentry->d_name.len, dentry->d_name.name,
ceph_cap_string(req->r_dir_caps));
set_bit(CEPH_MDS_R_ASYNC, &req->r_req_flags);
req->r_callback = ceph_async_unlink_cb;
req->r_old_inode = d_inode(dentry);
ihold(req->r_old_inode);
ceph: wait for the first reply of inflight async unlink In async unlink case the kclient won't wait for the first reply from MDS and just drop all the links and unhash the dentry and then succeeds immediately. For any new create/link/rename,etc requests followed by using the same file names we must wait for the first reply of the inflight unlink request, or the MDS possibly will fail these following requests with -EEXIST if the inflight async unlink request was delayed for some reasons. And the worst case is that for the none async openc request it will successfully open the file if the CDentry hasn't been unlinked yet, but later the previous delayed async unlink request will remove the CDenty. That means the just created file is possiblly deleted later by accident. We need to wait for the inflight async unlink requests to finish when creating new files/directories by using the same file names. Link: https://tracker.ceph.com/issues/55332 Signed-off-by: Xiubo Li <xiubli@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2022-05-10 01:47:01 +00:00
spin_lock(&dentry->d_lock);
di->flags |= CEPH_DENTRY_ASYNC_UNLINK;
spin_unlock(&dentry->d_lock);
spin_lock(&fsc->async_unlink_conflict_lock);
hash_add_rcu(fsc->async_unlink_conflict, &di->hnode,
dentry->d_name.hash);
spin_unlock(&fsc->async_unlink_conflict_lock);
ceph: perform asynchronous unlink if we have sufficient caps The MDS is getting a new lock-caching facility that will allow it to cache the necessary locks to allow asynchronous directory operations. Since the CEPH_CAP_FILE_* caps are currently unused on directories, we can repurpose those bits for this purpose. When performing an unlink, if we have Fx on the parent directory, and CEPH_CAP_DIR_UNLINK (aka Fr), and we know that the dentry being removed is the primary link, then then we can fire off an unlink request immediately and don't need to wait on reply before returning. In that situation, just fix up the dcache and link count and return immediately after issuing the call to the MDS. This does mean that we need to hold an extra reference to the inode being unlinked, and extra references to the caps to avoid races. Those references are put and error handling is done in the r_callback routine. If the operation ends up failing, then set a writeback error on the directory inode, and the inode itself that can be fetched later by an fsync on the dir. The behavior of dir caps is slightly different from caps on normal files. Because these are just considered an optimization, if the session is reconnected, we will not automatically reclaim them. They are instead considered lost until we do another synchronous op in the parent directory. Async dirops are enabled via the "nowsync" mount option, which is patterned after the xfs "wsync" mount option. For now, the default is "wsync", but eventually we may flip that. Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-04-02 19:35:56 +00:00
err = ceph_mdsc_submit_request(mdsc, dir, req);
if (!err) {
/*
* We have enough caps, so we assume that the unlink
* will succeed. Fix up the target inode and dcache.
*/
drop_nlink(inode);
d_delete(dentry);
ceph: wait for the first reply of inflight async unlink In async unlink case the kclient won't wait for the first reply from MDS and just drop all the links and unhash the dentry and then succeeds immediately. For any new create/link/rename,etc requests followed by using the same file names we must wait for the first reply of the inflight unlink request, or the MDS possibly will fail these following requests with -EEXIST if the inflight async unlink request was delayed for some reasons. And the worst case is that for the none async openc request it will successfully open the file if the CDentry hasn't been unlinked yet, but later the previous delayed async unlink request will remove the CDenty. That means the just created file is possiblly deleted later by accident. We need to wait for the inflight async unlink requests to finish when creating new files/directories by using the same file names. Link: https://tracker.ceph.com/issues/55332 Signed-off-by: Xiubo Li <xiubli@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2022-05-10 01:47:01 +00:00
} else {
spin_lock(&fsc->async_unlink_conflict_lock);
hash_del_rcu(&di->hnode);
spin_unlock(&fsc->async_unlink_conflict_lock);
spin_lock(&dentry->d_lock);
di->flags &= ~CEPH_DENTRY_ASYNC_UNLINK;
spin_unlock(&dentry->d_lock);
if (err == -EJUKEBOX) {
try_async = false;
ceph_mdsc_put_request(req);
goto retry;
}
ceph: perform asynchronous unlink if we have sufficient caps The MDS is getting a new lock-caching facility that will allow it to cache the necessary locks to allow asynchronous directory operations. Since the CEPH_CAP_FILE_* caps are currently unused on directories, we can repurpose those bits for this purpose. When performing an unlink, if we have Fx on the parent directory, and CEPH_CAP_DIR_UNLINK (aka Fr), and we know that the dentry being removed is the primary link, then then we can fire off an unlink request immediately and don't need to wait on reply before returning. In that situation, just fix up the dcache and link count and return immediately after issuing the call to the MDS. This does mean that we need to hold an extra reference to the inode being unlinked, and extra references to the caps to avoid races. Those references are put and error handling is done in the r_callback routine. If the operation ends up failing, then set a writeback error on the directory inode, and the inode itself that can be fetched later by an fsync on the dir. The behavior of dir caps is slightly different from caps on normal files. Because these are just considered an optimization, if the session is reconnected, we will not automatically reclaim them. They are instead considered lost until we do another synchronous op in the parent directory. Async dirops are enabled via the "nowsync" mount option, which is patterned after the xfs "wsync" mount option. For now, the default is "wsync", but eventually we may flip that. Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-04-02 19:35:56 +00:00
}
} else {
set_bit(CEPH_MDS_R_PARENT_LOCKED, &req->r_req_flags);
err = ceph_mdsc_do_request(mdsc, dir, req);
if (!err && !req->r_reply_info.head->is_dentry)
d_delete(dentry);
}
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
ceph_mdsc_put_request(req);
out:
return err;
}
fs: make helpers idmap mount aware Extend some inode methods with an additional user namespace argument. A filesystem that is aware of idmapped mounts will receive the user namespace the mount has been marked with. This can be used for additional permission checking and also to enable filesystems to translate between uids and gids if they need to. We have implemented all relevant helpers in earlier patches. As requested we simply extend the exisiting inode method instead of introducing new ones. This is a little more code churn but it's mostly mechanical and doesnt't leave us with additional inode methods. Link: https://lore.kernel.org/r/20210121131959.646623-25-christian.brauner@ubuntu.com Cc: Christoph Hellwig <hch@lst.de> Cc: David Howells <dhowells@redhat.com> Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: linux-fsdevel@vger.kernel.org Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
2021-01-21 13:19:43 +00:00
static int ceph_rename(struct user_namespace *mnt_userns, struct inode *old_dir,
struct dentry *old_dentry, struct inode *new_dir,
struct dentry *new_dentry, unsigned int flags)
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
{
ceph: add ceph_sb_to_mdsc helper support to parse the mdsc This will help simplify the code. [ jlayton: fix minor merge conflict in quota.c ] Signed-off-by: Xiubo Li <xiubli@redhat.com> Signed-off-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-09-03 13:01:39 +00:00
struct ceph_mds_client *mdsc = ceph_sb_to_mdsc(old_dir->i_sb);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
struct ceph_mds_request *req;
ceph: rename snapshot support Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-04-07 07:36:32 +00:00
int op = CEPH_MDS_OP_RENAME;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
int err;
fs: make remaining filesystems use .rename2 This is trivial to do: - add flags argument to foo_rename() - check if flags is zero - assign foo_rename() to .rename2 instead of .rename This doesn't mean it's impossible to support RENAME_NOREPLACE for these filesystems, but it is not trivial, like for local filesystems. RENAME_NOREPLACE must guarantee atomicity (i.e. it shouldn't be possible for a file to be created on one host while it is overwritten by rename on another host). Filesystems converted: 9p, afs, ceph, coda, ecryptfs, kernfs, lustre, ncpfs, nfs, ocfs2, orangefs. After this, we can get rid of the duplicate interfaces for rename. Signed-off-by: Miklos Szeredi <mszeredi@redhat.com> Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Acked-by: David Howells <dhowells@redhat.com> [AFS] Acked-by: Mike Marshall <hubcap@omnibond.com> Cc: Eric Van Hensbergen <ericvh@gmail.com> Cc: Ilya Dryomov <idryomov@gmail.com> Cc: Jan Harkes <jaharkes@cs.cmu.edu> Cc: Tyler Hicks <tyhicks@canonical.com> Cc: Oleg Drokin <oleg.drokin@intel.com> Cc: Trond Myklebust <trond.myklebust@primarydata.com> Cc: Mark Fasheh <mfasheh@suse.com>
2016-09-27 09:03:58 +00:00
if (flags)
return -EINVAL;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
if (ceph_snap(old_dir) != ceph_snap(new_dir))
return -EXDEV;
ceph: rename snapshot support Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-04-07 07:36:32 +00:00
if (ceph_snap(old_dir) != CEPH_NOSNAP) {
if (old_dir == new_dir && ceph_snap(old_dir) == CEPH_SNAPDIR)
op = CEPH_MDS_OP_RENAMESNAP;
else
return -EROFS;
}
Revert "ceph: allow rename operation under different quota realms" This reverts commit dffdcd71458e699e839f0bf47c3d42d64210b939. When doing a rename across quota realms, there's a corner case that isn't handled correctly. Here's a testcase: mkdir files limit truncate files/file -s 10G setfattr limit -n ceph.quota.max_bytes -v 1000000 mv files limit/ The above will succeed because ftruncate(2) won't immediately notify the MDSs with the new file size, and thus the quota realms stats won't be updated. Since the possible fixes for this issue would have a huge performance impact, the solution for now is to simply revert to returning -EXDEV when doing a cross quota realms rename. URL: https://tracker.ceph.com/issues/48203 Signed-off-by: Luis Henriques <lhenriques@suse.de> Reviewed-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: Ilya Dryomov <idryomov@gmail.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-11-12 15:23:21 +00:00
/* don't allow cross-quota renames */
if ((old_dir != new_dir) &&
(!ceph_quota_is_same_realm(old_dir, new_dir)))
return -EXDEV;
ceph: quota: don't allow cross-quota renames This patch changes ceph_rename so that -EXDEV is returned if an attempt is made to mv a file between two different dir trees with different quotas setup. Signed-off-by: Luis Henriques <lhenriques@suse.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-01-05 10:47:20 +00:00
ceph: wait for the first reply of inflight async unlink In async unlink case the kclient won't wait for the first reply from MDS and just drop all the links and unhash the dentry and then succeeds immediately. For any new create/link/rename,etc requests followed by using the same file names we must wait for the first reply of the inflight unlink request, or the MDS possibly will fail these following requests with -EEXIST if the inflight async unlink request was delayed for some reasons. And the worst case is that for the none async openc request it will successfully open the file if the CDentry hasn't been unlinked yet, but later the previous delayed async unlink request will remove the CDenty. That means the just created file is possiblly deleted later by accident. We need to wait for the inflight async unlink requests to finish when creating new files/directories by using the same file names. Link: https://tracker.ceph.com/issues/55332 Signed-off-by: Xiubo Li <xiubli@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2022-05-10 01:47:01 +00:00
err = ceph_wait_on_conflict_unlink(new_dentry);
if (err)
return err;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
dout("rename dir %p dentry %p to dir %p dentry %p\n",
old_dir, old_dentry, new_dir, new_dentry);
ceph: rename snapshot support Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-04-07 07:36:32 +00:00
req = ceph_mdsc_create_request(mdsc, op, USE_AUTH_MDS);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
if (IS_ERR(req))
return PTR_ERR(req);
ceph: avoid useless ceph_get_dentry_parent_inode() in ceph_rename() This is just old_dir; no reason to abuse the dcache pointers. Reported-by: Al Viro <viro.zeniv.linux.org.uk> Signed-off-by: Sage Weil <sage@inktank.com> Reviewed-by: Yan, Zheng <zheng.z.yan@intel.com>
2013-02-05 21:36:05 +00:00
ihold(old_dir);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
req->r_dentry = dget(new_dentry);
req->r_num_caps = 2;
req->r_old_dentry = dget(old_dentry);
ceph: avoid useless ceph_get_dentry_parent_inode() in ceph_rename() This is just old_dir; no reason to abuse the dcache pointers. Reported-by: Al Viro <viro.zeniv.linux.org.uk> Signed-off-by: Sage Weil <sage@inktank.com> Reviewed-by: Yan, Zheng <zheng.z.yan@intel.com>
2013-02-05 21:36:05 +00:00
req->r_old_dentry_dir = old_dir;
ceph: add a new flag to indicate whether parent is locked struct ceph_mds_request has an r_locked_dir pointer, which is set to indicate the parent inode and that its i_rwsem is locked. In some critical places, we need to be able to indicate the parent inode to the request handling code, even when its i_rwsem may not be locked. Most of the code that operates on r_locked_dir doesn't require that the i_rwsem be locked. We only really need it to handle manipulation of the dcache. The rest (filling of the inode, updating dentry leases, etc.) already has its own locking. Add a new r_req_flags bit that indicates whether the parent is locked when doing the request, and rename the pointer to "r_parent". For now, all the places that set r_parent also set this flag, but that will change in a later patch. Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Yan, Zheng <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-01-31 15:28:26 +00:00
req->r_parent = new_dir;
ceph: take reference to req->r_parent at point of assignment Currently, we set the r_parent pointer but then don't take a reference to it until we submit the request. If we end up freeing the req before that point, then we'll do a iput when we shouldn't. Instead, take the inode reference in the callers, so that it's always safe to call ceph_mdsc_put_request on the req, even before submission. Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: Luis Henriques <lhenriques@suse.de> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2021-06-18 17:05:06 +00:00
ihold(new_dir);
ceph: add a new flag to indicate whether parent is locked struct ceph_mds_request has an r_locked_dir pointer, which is set to indicate the parent inode and that its i_rwsem is locked. In some critical places, we need to be able to indicate the parent inode to the request handling code, even when its i_rwsem may not be locked. Most of the code that operates on r_locked_dir doesn't require that the i_rwsem be locked. We only really need it to handle manipulation of the dcache. The rest (filling of the inode, updating dentry leases, etc.) already has its own locking. Add a new r_req_flags bit that indicates whether the parent is locked when doing the request, and rename the pointer to "r_parent". For now, all the places that set r_parent also set this flag, but that will change in a later patch. Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Yan, Zheng <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-01-31 15:28:26 +00:00
set_bit(CEPH_MDS_R_PARENT_LOCKED, &req->r_req_flags);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
req->r_old_dentry_drop = CEPH_CAP_FILE_SHARED;
req->r_old_dentry_unless = CEPH_CAP_FILE_EXCL;
req->r_dentry_drop = CEPH_CAP_FILE_SHARED;
req->r_dentry_unless = CEPH_CAP_FILE_EXCL;
/* release LINK_RDCACHE on source inode (mds will lock it) */
ceph: voluntarily drop Lx cap for link/rename requests MDS need to xlock inode's linklock when handling link/rename requests. Voluntarily dropping CEPH_CAP_AUTH_EXCL avoids a cap revoke message. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-11-23 09:59:13 +00:00
req->r_old_inode_drop = CEPH_CAP_LINK_SHARED | CEPH_CAP_LINK_EXCL;
ceph: flush dirty caps of unlinked inode ASAP Client should release unlinked inode from its cache ASAP. But client can't release inode with dirty caps. Link: http://tracker.ceph.com/issues/22886 Signed-off-by: Zhi Zhang <zhang.david2011@gmail.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-01-24 13:24:33 +00:00
if (d_really_is_positive(new_dentry)) {
req->r_inode_drop =
ceph_drop_caps_for_unlink(d_inode(new_dentry));
}
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
err = ceph_mdsc_do_request(mdsc, old_dir, req);
if (!err && !req->r_reply_info.head->is_dentry) {
/*
* Normally d_move() is done by fill_trace (called by
* do_request, above). If there is no trace, we need
* to do it here.
*/
d_move(old_dentry, new_dentry);
}
ceph_mdsc_put_request(req);
return err;
}
ceph: periodically trim stale dentries Previous commit make VFS delete stale dentry when last reference is dropped. Lease also can become invalid when corresponding dentry has no reference. This patch make cephfs periodically scan lease list, delete corresponding dentry if lease is invalid. There are two types of lease, dentry lease and dir lease. dentry lease has life time and applies to singe dentry. Dentry lease is added to tail of a list when it's updated, leases at front of the list will expire first. Dir lease is CEPH_CAP_FILE_SHARED on directory inode, it applies to all dentries in the directory. Dentries have dir leases are added to another list. Dentries in the list are periodically checked in a round robin manner. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-01-31 08:55:51 +00:00
/*
* Move dentry to tail of mdsc->dentry_leases list when lease is updated.
* Leases at front of the list will expire first. (Assume all leases have
* similar duration)
*
* Called under dentry->d_lock.
*/
void __ceph_dentry_lease_touch(struct ceph_dentry_info *di)
{
struct dentry *dn = di->dentry;
struct ceph_mds_client *mdsc;
dout("dentry_lease_touch %p %p '%pd'\n", di, dn, dn);
di->flags |= CEPH_DENTRY_LEASE_LIST;
if (di->flags & CEPH_DENTRY_SHRINK_LIST) {
di->flags |= CEPH_DENTRY_REFERENCED;
return;
}
mdsc = ceph_sb_to_client(dn->d_sb)->mdsc;
spin_lock(&mdsc->dentry_list_lock);
list_move_tail(&di->lease_list, &mdsc->dentry_leases);
spin_unlock(&mdsc->dentry_list_lock);
}
static void __dentry_dir_lease_touch(struct ceph_mds_client* mdsc,
struct ceph_dentry_info *di)
{
di->flags &= ~(CEPH_DENTRY_LEASE_LIST | CEPH_DENTRY_REFERENCED);
di->lease_gen = 0;
di->time = jiffies;
list_move_tail(&di->lease_list, &mdsc->dentry_dir_leases);
}
/*
* When dir lease is used, add dentry to tail of mdsc->dentry_dir_leases
* list if it's not in the list, otherwise set 'referenced' flag.
*
* Called under dentry->d_lock.
*/
void __ceph_dentry_dir_lease_touch(struct ceph_dentry_info *di)
{
struct dentry *dn = di->dentry;
struct ceph_mds_client *mdsc;
ceph: print dentry offset in hex and fix xattr_version type In the debug logs about the di->offset or ctx->pos it is in hex format, but some others are using the dec format. It is a little hard to read. For the xattr version, it is u64 type, using a shorter type may truncate it. Signed-off-by: Xiubo Li <xiubli@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-12-19 02:15:18 +00:00
dout("dentry_dir_lease_touch %p %p '%pd' (offset 0x%llx)\n",
ceph: periodically trim stale dentries Previous commit make VFS delete stale dentry when last reference is dropped. Lease also can become invalid when corresponding dentry has no reference. This patch make cephfs periodically scan lease list, delete corresponding dentry if lease is invalid. There are two types of lease, dentry lease and dir lease. dentry lease has life time and applies to singe dentry. Dentry lease is added to tail of a list when it's updated, leases at front of the list will expire first. Dir lease is CEPH_CAP_FILE_SHARED on directory inode, it applies to all dentries in the directory. Dentries have dir leases are added to another list. Dentries in the list are periodically checked in a round robin manner. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-01-31 08:55:51 +00:00
di, dn, dn, di->offset);
if (!list_empty(&di->lease_list)) {
if (di->flags & CEPH_DENTRY_LEASE_LIST) {
/* don't remove dentry from dentry lease list
* if its lease is valid */
if (__dentry_lease_is_valid(di))
return;
} else {
di->flags |= CEPH_DENTRY_REFERENCED;
return;
}
}
if (di->flags & CEPH_DENTRY_SHRINK_LIST) {
di->flags |= CEPH_DENTRY_REFERENCED;
di->flags &= ~CEPH_DENTRY_LEASE_LIST;
return;
}
mdsc = ceph_sb_to_client(dn->d_sb)->mdsc;
spin_lock(&mdsc->dentry_list_lock);
__dentry_dir_lease_touch(mdsc, di),
spin_unlock(&mdsc->dentry_list_lock);
}
static void __dentry_lease_unlist(struct ceph_dentry_info *di)
{
struct ceph_mds_client *mdsc;
if (di->flags & CEPH_DENTRY_SHRINK_LIST)
return;
if (list_empty(&di->lease_list))
return;
mdsc = ceph_sb_to_client(di->dentry->d_sb)->mdsc;
spin_lock(&mdsc->dentry_list_lock);
list_del_init(&di->lease_list);
spin_unlock(&mdsc->dentry_list_lock);
}
enum {
KEEP = 0,
DELETE = 1,
TOUCH = 2,
STOP = 4,
};
struct ceph_lease_walk_control {
bool dir_lease;
ceph: add mount option to limit caps count If number of caps exceed the limit, ceph_trim_dentires() also trim dentries with valid leases. Trimming dentry releases references to associated inode, which may evict inode and release caps. By default, there is no limit for caps count. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-02-01 06:57:15 +00:00
bool expire_dir_lease;
ceph: periodically trim stale dentries Previous commit make VFS delete stale dentry when last reference is dropped. Lease also can become invalid when corresponding dentry has no reference. This patch make cephfs periodically scan lease list, delete corresponding dentry if lease is invalid. There are two types of lease, dentry lease and dir lease. dentry lease has life time and applies to singe dentry. Dentry lease is added to tail of a list when it's updated, leases at front of the list will expire first. Dir lease is CEPH_CAP_FILE_SHARED on directory inode, it applies to all dentries in the directory. Dentries have dir leases are added to another list. Dentries in the list are periodically checked in a round robin manner. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-01-31 08:55:51 +00:00
unsigned long nr_to_scan;
unsigned long dir_lease_ttl;
};
static unsigned long
__dentry_leases_walk(struct ceph_mds_client *mdsc,
struct ceph_lease_walk_control *lwc,
int (*check)(struct dentry*, void*))
{
struct ceph_dentry_info *di, *tmp;
struct dentry *dentry, *last = NULL;
struct list_head* list;
LIST_HEAD(dispose);
unsigned long freed = 0;
int ret = 0;
list = lwc->dir_lease ? &mdsc->dentry_dir_leases : &mdsc->dentry_leases;
spin_lock(&mdsc->dentry_list_lock);
list_for_each_entry_safe(di, tmp, list, lease_list) {
if (!lwc->nr_to_scan)
break;
--lwc->nr_to_scan;
dentry = di->dentry;
if (last == dentry)
break;
if (!spin_trylock(&dentry->d_lock))
continue;
ceph: don't open-code the check for dead lockref Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2019-06-28 02:25:23 +00:00
if (__lockref_is_dead(&dentry->d_lockref)) {
ceph: periodically trim stale dentries Previous commit make VFS delete stale dentry when last reference is dropped. Lease also can become invalid when corresponding dentry has no reference. This patch make cephfs periodically scan lease list, delete corresponding dentry if lease is invalid. There are two types of lease, dentry lease and dir lease. dentry lease has life time and applies to singe dentry. Dentry lease is added to tail of a list when it's updated, leases at front of the list will expire first. Dir lease is CEPH_CAP_FILE_SHARED on directory inode, it applies to all dentries in the directory. Dentries have dir leases are added to another list. Dentries in the list are periodically checked in a round robin manner. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-01-31 08:55:51 +00:00
list_del_init(&di->lease_list);
goto next;
}
ret = check(dentry, lwc);
if (ret & TOUCH) {
/* move it into tail of dir lease list */
__dentry_dir_lease_touch(mdsc, di);
if (!last)
last = dentry;
}
if (ret & DELETE) {
/* stale lease */
di->flags &= ~CEPH_DENTRY_REFERENCED;
if (dentry->d_lockref.count > 0) {
/* update_dentry_lease() will re-add
* it to lease list, or
* ceph_d_delete() will return 1 when
* last reference is dropped */
list_del_init(&di->lease_list);
} else {
di->flags |= CEPH_DENTRY_SHRINK_LIST;
list_move_tail(&di->lease_list, &dispose);
dget_dlock(dentry);
}
}
next:
spin_unlock(&dentry->d_lock);
if (ret & STOP)
break;
}
spin_unlock(&mdsc->dentry_list_lock);
while (!list_empty(&dispose)) {
di = list_first_entry(&dispose, struct ceph_dentry_info,
lease_list);
dentry = di->dentry;
spin_lock(&dentry->d_lock);
list_del_init(&di->lease_list);
di->flags &= ~CEPH_DENTRY_SHRINK_LIST;
if (di->flags & CEPH_DENTRY_REFERENCED) {
spin_lock(&mdsc->dentry_list_lock);
if (di->flags & CEPH_DENTRY_LEASE_LIST) {
list_add_tail(&di->lease_list,
&mdsc->dentry_leases);
} else {
__dentry_dir_lease_touch(mdsc, di);
}
spin_unlock(&mdsc->dentry_list_lock);
} else {
freed++;
}
spin_unlock(&dentry->d_lock);
/* ceph_d_delete() does the trick */
dput(dentry);
}
return freed;
}
static int __dentry_lease_check(struct dentry *dentry, void *arg)
{
struct ceph_dentry_info *di = ceph_dentry(dentry);
int ret;
if (__dentry_lease_is_valid(di))
return STOP;
ret = __dir_lease_try_check(dentry);
if (ret == -EBUSY)
return KEEP;
if (ret > 0)
return TOUCH;
return DELETE;
}
static int __dir_lease_check(struct dentry *dentry, void *arg)
{
struct ceph_lease_walk_control *lwc = arg;
struct ceph_dentry_info *di = ceph_dentry(dentry);
int ret = __dir_lease_try_check(dentry);
if (ret == -EBUSY)
return KEEP;
if (ret > 0) {
if (time_before(jiffies, di->time + lwc->dir_lease_ttl))
return STOP;
/* Move dentry to tail of dir lease list if we don't want
* to delete it. So dentries in the list are checked in a
* round robin manner */
ceph: add mount option to limit caps count If number of caps exceed the limit, ceph_trim_dentires() also trim dentries with valid leases. Trimming dentry releases references to associated inode, which may evict inode and release caps. By default, there is no limit for caps count. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-02-01 06:57:15 +00:00
if (!lwc->expire_dir_lease)
return TOUCH;
if (dentry->d_lockref.count > 0 ||
(di->flags & CEPH_DENTRY_REFERENCED))
return TOUCH;
/* invalidate dir lease */
di->lease_shared_gen = 0;
ceph: periodically trim stale dentries Previous commit make VFS delete stale dentry when last reference is dropped. Lease also can become invalid when corresponding dentry has no reference. This patch make cephfs periodically scan lease list, delete corresponding dentry if lease is invalid. There are two types of lease, dentry lease and dir lease. dentry lease has life time and applies to singe dentry. Dentry lease is added to tail of a list when it's updated, leases at front of the list will expire first. Dir lease is CEPH_CAP_FILE_SHARED on directory inode, it applies to all dentries in the directory. Dentries have dir leases are added to another list. Dentries in the list are periodically checked in a round robin manner. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-01-31 08:55:51 +00:00
}
return DELETE;
}
int ceph_trim_dentries(struct ceph_mds_client *mdsc)
{
struct ceph_lease_walk_control lwc;
ceph: add mount option to limit caps count If number of caps exceed the limit, ceph_trim_dentires() also trim dentries with valid leases. Trimming dentry releases references to associated inode, which may evict inode and release caps. By default, there is no limit for caps count. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-02-01 06:57:15 +00:00
unsigned long count;
ceph: periodically trim stale dentries Previous commit make VFS delete stale dentry when last reference is dropped. Lease also can become invalid when corresponding dentry has no reference. This patch make cephfs periodically scan lease list, delete corresponding dentry if lease is invalid. There are two types of lease, dentry lease and dir lease. dentry lease has life time and applies to singe dentry. Dentry lease is added to tail of a list when it's updated, leases at front of the list will expire first. Dir lease is CEPH_CAP_FILE_SHARED on directory inode, it applies to all dentries in the directory. Dentries have dir leases are added to another list. Dentries in the list are periodically checked in a round robin manner. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-01-31 08:55:51 +00:00
unsigned long freed;
ceph: add mount option to limit caps count If number of caps exceed the limit, ceph_trim_dentires() also trim dentries with valid leases. Trimming dentry releases references to associated inode, which may evict inode and release caps. By default, there is no limit for caps count. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-02-01 06:57:15 +00:00
spin_lock(&mdsc->caps_list_lock);
if (mdsc->caps_use_max > 0 &&
mdsc->caps_use_count > mdsc->caps_use_max)
count = mdsc->caps_use_count - mdsc->caps_use_max;
else
count = 0;
spin_unlock(&mdsc->caps_list_lock);
ceph: periodically trim stale dentries Previous commit make VFS delete stale dentry when last reference is dropped. Lease also can become invalid when corresponding dentry has no reference. This patch make cephfs periodically scan lease list, delete corresponding dentry if lease is invalid. There are two types of lease, dentry lease and dir lease. dentry lease has life time and applies to singe dentry. Dentry lease is added to tail of a list when it's updated, leases at front of the list will expire first. Dir lease is CEPH_CAP_FILE_SHARED on directory inode, it applies to all dentries in the directory. Dentries have dir leases are added to another list. Dentries in the list are periodically checked in a round robin manner. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-01-31 08:55:51 +00:00
lwc.dir_lease = false;
lwc.nr_to_scan = CEPH_CAPS_PER_RELEASE * 2;
freed = __dentry_leases_walk(mdsc, &lwc, __dentry_lease_check);
if (!lwc.nr_to_scan) /* more invalid leases */
return -EAGAIN;
if (lwc.nr_to_scan < CEPH_CAPS_PER_RELEASE)
lwc.nr_to_scan = CEPH_CAPS_PER_RELEASE;
lwc.dir_lease = true;
ceph: add mount option to limit caps count If number of caps exceed the limit, ceph_trim_dentires() also trim dentries with valid leases. Trimming dentry releases references to associated inode, which may evict inode and release caps. By default, there is no limit for caps count. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-02-01 06:57:15 +00:00
lwc.expire_dir_lease = freed < count;
lwc.dir_lease_ttl = mdsc->fsc->mount_options->caps_wanted_delay_max * HZ;
ceph: periodically trim stale dentries Previous commit make VFS delete stale dentry when last reference is dropped. Lease also can become invalid when corresponding dentry has no reference. This patch make cephfs periodically scan lease list, delete corresponding dentry if lease is invalid. There are two types of lease, dentry lease and dir lease. dentry lease has life time and applies to singe dentry. Dentry lease is added to tail of a list when it's updated, leases at front of the list will expire first. Dir lease is CEPH_CAP_FILE_SHARED on directory inode, it applies to all dentries in the directory. Dentries have dir leases are added to another list. Dentries in the list are periodically checked in a round robin manner. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-01-31 08:55:51 +00:00
freed +=__dentry_leases_walk(mdsc, &lwc, __dir_lease_check);
if (!lwc.nr_to_scan) /* more to check */
return -EAGAIN;
return freed > 0 ? 1 : 0;
}
ceph: invalidate affected dentry leases on aborted requests If we abort a request, we return to caller, but the request may still complete. And if we hold the dir FILE_EXCL bit, we may not release a lease when sending a request. A simple un-tar, control-c, un-tar again will reproduce the bug (manifested as a 'Cannot open: File exists'). Ensure we invalidate affected dentry leases (as well dir I_COMPLETE) so we don't have valid (but incorrect) leases. Do the same, consistently, at other sites where I_COMPLETE is similarly cleared. Signed-off-by: Sage Weil <sage@newdream.net>
2010-05-14 16:35:38 +00:00
/*
* Ensure a dentry lease will no longer revalidate.
*/
void ceph_invalidate_dentry_lease(struct dentry *dentry)
{
ceph: periodically trim stale dentries Previous commit make VFS delete stale dentry when last reference is dropped. Lease also can become invalid when corresponding dentry has no reference. This patch make cephfs periodically scan lease list, delete corresponding dentry if lease is invalid. There are two types of lease, dentry lease and dir lease. dentry lease has life time and applies to singe dentry. Dentry lease is added to tail of a list when it's updated, leases at front of the list will expire first. Dir lease is CEPH_CAP_FILE_SHARED on directory inode, it applies to all dentries in the directory. Dentries have dir leases are added to another list. Dentries in the list are periodically checked in a round robin manner. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-01-31 08:55:51 +00:00
struct ceph_dentry_info *di = ceph_dentry(dentry);
ceph: invalidate affected dentry leases on aborted requests If we abort a request, we return to caller, but the request may still complete. And if we hold the dir FILE_EXCL bit, we may not release a lease when sending a request. A simple un-tar, control-c, un-tar again will reproduce the bug (manifested as a 'Cannot open: File exists'). Ensure we invalidate affected dentry leases (as well dir I_COMPLETE) so we don't have valid (but incorrect) leases. Do the same, consistently, at other sites where I_COMPLETE is similarly cleared. Signed-off-by: Sage Weil <sage@newdream.net>
2010-05-14 16:35:38 +00:00
spin_lock(&dentry->d_lock);
ceph: periodically trim stale dentries Previous commit make VFS delete stale dentry when last reference is dropped. Lease also can become invalid when corresponding dentry has no reference. This patch make cephfs periodically scan lease list, delete corresponding dentry if lease is invalid. There are two types of lease, dentry lease and dir lease. dentry lease has life time and applies to singe dentry. Dentry lease is added to tail of a list when it's updated, leases at front of the list will expire first. Dir lease is CEPH_CAP_FILE_SHARED on directory inode, it applies to all dentries in the directory. Dentries have dir leases are added to another list. Dentries in the list are periodically checked in a round robin manner. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-01-31 08:55:51 +00:00
di->time = jiffies;
di->lease_shared_gen = 0;
ceph: track primary dentry link Newer versions of the MDS will flag a dentry as "primary". In later patches, we'll need to consult this info, so track it in di->flags. Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-02-18 19:12:32 +00:00
di->flags &= ~CEPH_DENTRY_PRIMARY_LINK;
ceph: periodically trim stale dentries Previous commit make VFS delete stale dentry when last reference is dropped. Lease also can become invalid when corresponding dentry has no reference. This patch make cephfs periodically scan lease list, delete corresponding dentry if lease is invalid. There are two types of lease, dentry lease and dir lease. dentry lease has life time and applies to singe dentry. Dentry lease is added to tail of a list when it's updated, leases at front of the list will expire first. Dir lease is CEPH_CAP_FILE_SHARED on directory inode, it applies to all dentries in the directory. Dentries have dir leases are added to another list. Dentries in the list are periodically checked in a round robin manner. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-01-31 08:55:51 +00:00
__dentry_lease_unlist(di);
ceph: invalidate affected dentry leases on aborted requests If we abort a request, we return to caller, but the request may still complete. And if we hold the dir FILE_EXCL bit, we may not release a lease when sending a request. A simple un-tar, control-c, un-tar again will reproduce the bug (manifested as a 'Cannot open: File exists'). Ensure we invalidate affected dentry leases (as well dir I_COMPLETE) so we don't have valid (but incorrect) leases. Do the same, consistently, at other sites where I_COMPLETE is similarly cleared. Signed-off-by: Sage Weil <sage@newdream.net>
2010-05-14 16:35:38 +00:00
spin_unlock(&dentry->d_lock);
}
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
/*
* Check if dentry lease is valid. If not, delete the lease. Try to
* renew if the least is more than half up.
*/
ceph: delete stale dentry when last reference is dropped introduce ceph_d_delete(), which checks if dentry has valid lease. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-01-28 12:43:55 +00:00
static bool __dentry_lease_is_valid(struct ceph_dentry_info *di)
{
struct ceph_mds_session *session;
if (!di->lease_gen)
return false;
session = di->lease_session;
if (session) {
u32 gen;
unsigned long ttl;
ceph: eliminate session->s_gen_ttl_lock Turn s_cap_gen field into an atomic_t, and just rely on the fact that we hold the s_mutex when changing the s_cap_ttl field. Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: Luis Henriques <lhenriques@suse.de> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2021-06-04 16:03:09 +00:00
gen = atomic_read(&session->s_cap_gen);
ceph: delete stale dentry when last reference is dropped introduce ceph_d_delete(), which checks if dentry has valid lease. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-01-28 12:43:55 +00:00
ttl = session->s_cap_ttl;
if (di->lease_gen == gen &&
time_before(jiffies, ttl) &&
time_before(jiffies, di->time))
return true;
}
di->lease_gen = 0;
return false;
}
ceph: ensure d_name/d_parent stability in ceph_mdsc_lease_send_msg() Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-05-23 02:45:24 +00:00
static int dentry_lease_is_valid(struct dentry *dentry, unsigned int flags)
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
{
struct ceph_dentry_info *di;
struct ceph_mds_session *session = NULL;
u32 seq = 0;
ceph: delete stale dentry when last reference is dropped introduce ceph_d_delete(), which checks if dentry has valid lease. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-01-28 12:43:55 +00:00
int valid = 0;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
spin_lock(&dentry->d_lock);
di = ceph_dentry(dentry);
ceph: delete stale dentry when last reference is dropped introduce ceph_d_delete(), which checks if dentry has valid lease. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-01-28 12:43:55 +00:00
if (di && __dentry_lease_is_valid(di)) {
valid = 1;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
ceph: delete stale dentry when last reference is dropped introduce ceph_d_delete(), which checks if dentry has valid lease. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-01-28 12:43:55 +00:00
if (di->lease_renew_after &&
time_after(jiffies, di->lease_renew_after)) {
/*
* We should renew. If we're in RCU walk mode
* though, we can't do that so just return
* -ECHILD.
*/
if (flags & LOOKUP_RCU) {
valid = -ECHILD;
} else {
session = ceph_get_mds_session(di->lease_session);
seq = di->lease_seq;
di->lease_renew_after = 0;
di->lease_renew_from = jiffies;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
}
}
}
spin_unlock(&dentry->d_lock);
if (session) {
ceph: ensure d_name/d_parent stability in ceph_mdsc_lease_send_msg() Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-05-23 02:45:24 +00:00
ceph_mdsc_lease_send_msg(session, dentry,
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
CEPH_MDS_LEASE_RENEW, seq);
ceph_put_mds_session(session);
}
dout("dentry_lease_is_valid - dentry %p = %d\n", dentry, valid);
return valid;
}
ceph: delete stale dentry when last reference is dropped introduce ceph_d_delete(), which checks if dentry has valid lease. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-01-28 12:43:55 +00:00
/*
* Called under dentry->d_lock.
*/
static int __dir_lease_try_check(const struct dentry *dentry)
{
struct ceph_dentry_info *di = ceph_dentry(dentry);
struct inode *dir;
struct ceph_inode_info *ci;
int valid = 0;
if (!di->lease_shared_gen)
return 0;
if (IS_ROOT(dentry))
return 0;
dir = d_inode(dentry->d_parent);
ci = ceph_inode(dir);
if (spin_trylock(&ci->i_ceph_lock)) {
if (atomic_read(&ci->i_shared_gen) == di->lease_shared_gen &&
__ceph_caps_issued_mask(ci, CEPH_CAP_FILE_SHARED, 0))
valid = 1;
spin_unlock(&ci->i_ceph_lock);
} else {
valid = -EBUSY;
}
if (!valid)
di->lease_shared_gen = 0;
return valid;
}
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
/*
* Check if directory-wide content lease/cap is valid.
*/
ceph: consider inode's last read/write when calculating wanted caps Add i_last_rd and i_last_wr to ceph_inode_info. These fields are used to track the last time the client acquired read/write caps for the inode. If there is no read/write on an inode for 'caps_wanted_delay_max' seconds, __ceph_caps_file_wanted() does not request caps for read/write even there are open files. Call __ceph_touch_fmode() for dir operations. __ceph_caps_file_wanted() calculates dir's wanted caps according to last dir read/modification. If there is recent dir read, dir inode wants CEPH_CAP_ANY_SHARED caps. If there is recent dir modification, also wants CEPH_CAP_FILE_EXCL. Readdir is a special case. Dir inode wants CEPH_CAP_FILE_EXCL after readdir, as with that, modifications do not need to release CEPH_CAP_FILE_SHARED or invalidate all dentry leases issued by readdir. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-03-05 12:21:00 +00:00
static int dir_lease_is_valid(struct inode *dir, struct dentry *dentry,
struct ceph_mds_client *mdsc)
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
{
struct ceph_inode_info *ci = ceph_inode(dir);
ceph: fix dir_lease_is_valid() It should call __ceph_dentry_dir_lease_touch() under dentry->d_lock. Besides, ceph_dentry(dentry) can be NULL when called by LOOKUP_RCU d_revalidate() Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-05-22 09:26:27 +00:00
int valid;
int shared_gen;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
ceph: use i_ceph_lock instead of i_lock We have been using i_lock to protect all kinds of data structures in the ceph_inode_info struct, including lists of inodes that we need to iterate over while avoiding races with inode destruction. That requires grabbing a reference to the inode with the list lock protected, but igrab() now takes i_lock to check the inode flags. Changing the list lock ordering would be a painful process. However, using a ceph-specific i_ceph_lock in the ceph inode instead of i_lock is a simple mechanical change and avoids the ordering constraints imposed by igrab(). Reported-by: Amon Ott <a.ott@m-privacy.de> Signed-off-by: Sage Weil <sage@newdream.net>
2011-11-30 17:47:09 +00:00
spin_lock(&ci->i_ceph_lock);
ceph: fix dir_lease_is_valid() It should call __ceph_dentry_dir_lease_touch() under dentry->d_lock. Besides, ceph_dentry(dentry) can be NULL when called by LOOKUP_RCU d_revalidate() Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-05-22 09:26:27 +00:00
valid = __ceph_caps_issued_mask(ci, CEPH_CAP_FILE_SHARED, 1);
ceph: consider inode's last read/write when calculating wanted caps Add i_last_rd and i_last_wr to ceph_inode_info. These fields are used to track the last time the client acquired read/write caps for the inode. If there is no read/write on an inode for 'caps_wanted_delay_max' seconds, __ceph_caps_file_wanted() does not request caps for read/write even there are open files. Call __ceph_touch_fmode() for dir operations. __ceph_caps_file_wanted() calculates dir's wanted caps according to last dir read/modification. If there is recent dir read, dir inode wants CEPH_CAP_ANY_SHARED caps. If there is recent dir modification, also wants CEPH_CAP_FILE_EXCL. Readdir is a special case. Dir inode wants CEPH_CAP_FILE_EXCL after readdir, as with that, modifications do not need to release CEPH_CAP_FILE_SHARED or invalidate all dentry leases issued by readdir. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-03-05 12:21:00 +00:00
if (valid) {
__ceph_touch_fmode(ci, mdsc, CEPH_FILE_MODE_RD);
shared_gen = atomic_read(&ci->i_shared_gen);
}
ceph: use i_ceph_lock instead of i_lock We have been using i_lock to protect all kinds of data structures in the ceph_inode_info struct, including lists of inodes that we need to iterate over while avoiding races with inode destruction. That requires grabbing a reference to the inode with the list lock protected, but igrab() now takes i_lock to check the inode flags. Changing the list lock ordering would be a painful process. However, using a ceph-specific i_ceph_lock in the ceph inode instead of i_lock is a simple mechanical change and avoids the ordering constraints imposed by igrab(). Reported-by: Amon Ott <a.ott@m-privacy.de> Signed-off-by: Sage Weil <sage@newdream.net>
2011-11-30 17:47:09 +00:00
spin_unlock(&ci->i_ceph_lock);
ceph: fix dir_lease_is_valid() It should call __ceph_dentry_dir_lease_touch() under dentry->d_lock. Besides, ceph_dentry(dentry) can be NULL when called by LOOKUP_RCU d_revalidate() Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-05-22 09:26:27 +00:00
if (valid) {
struct ceph_dentry_info *di;
spin_lock(&dentry->d_lock);
di = ceph_dentry(dentry);
if (dir == d_inode(dentry->d_parent) &&
di && di->lease_shared_gen == shared_gen)
__ceph_dentry_dir_lease_touch(di);
else
valid = 0;
spin_unlock(&dentry->d_lock);
}
dout("dir_lease_is_valid dir %p v%u dentry %p = %d\n",
dir, (unsigned)atomic_read(&ci->i_shared_gen), dentry, valid);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
return valid;
}
/*
* Check if cached dentry can be trusted.
*/
stop passing nameidata * to ->d_revalidate() Just the lookup flags. Die, bastard, die... Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2012-06-10 20:03:43 +00:00
static int ceph_d_revalidate(struct dentry *dentry, unsigned int flags)
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
{
ceph: protect d_parent access in ceph_d_revalidate Protect d_parent with d_lock. Carry a reference. Simplify the flow so that there is a single exit point and cleanup. Reviewed-by: Yehuda Sadeh <yehuda@hq.newdream.net> Signed-off-by: Sage Weil <sage@newdream.net>
2011-07-26 18:30:43 +00:00
int valid = 0;
ceph: kill ceph_get_dentry_parent_inode() use vfs helper dget_parent() instead Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-03-16 08:40:23 +00:00
struct dentry *parent;
ceph: fix RCU case handling in ceph_d_revalidate() For RCU case ->d_revalidate() is called with rcu_read_lock() and without pinning the dentry passed to it. Which means that it can't rely upon ->d_inode remaining stable; that's the reason for d_inode_rcu(), actually. Make sure we don't reload ->d_inode there. Cc: stable@vger.kernel.org Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-10-29 13:50:19 +00:00
struct inode *dir, *inode;
ceph: consider inode's last read/write when calculating wanted caps Add i_last_rd and i_last_wr to ceph_inode_info. These fields are used to track the last time the client acquired read/write caps for the inode. If there is no read/write on an inode for 'caps_wanted_delay_max' seconds, __ceph_caps_file_wanted() does not request caps for read/write even there are open files. Call __ceph_touch_fmode() for dir operations. __ceph_caps_file_wanted() calculates dir's wanted caps according to last dir read/modification. If there is recent dir read, dir inode wants CEPH_CAP_ANY_SHARED caps. If there is recent dir modification, also wants CEPH_CAP_FILE_EXCL. Readdir is a special case. Dir inode wants CEPH_CAP_FILE_EXCL after readdir, as with that, modifications do not need to release CEPH_CAP_FILE_SHARED or invalidate all dentry leases issued by readdir. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-03-05 12:21:00 +00:00
struct ceph_mds_client *mdsc;
fs: rcu-walk aware d_revalidate method Require filesystems be aware of .d_revalidate being called in rcu-walk mode (nd->flags & LOOKUP_RCU). For now do a simple push down, returning -ECHILD from all implementations. Signed-off-by: Nick Piggin <npiggin@kernel.dk>
2011-01-07 06:49:57 +00:00
ceph: handle LOOKUP_RCU in ceph_d_revalidate We can now handle the snapshot cases under RCU, as well as the non-snapshot case when we don't need to queue up a lease renewal allow LOOKUP_RCU walks to proceed under those conditions. Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Yan, Zheng <zyan@redhat.com>
2016-07-01 13:39:21 +00:00
if (flags & LOOKUP_RCU) {
ceph: cleanup ACCESS_ONCE -> READ_ONCE This removes the uses of ACCESS_ONCE in favor of READ_ONCE Signed-off-by: Seraphime Kirkovski <kirkseraph@gmail.com> Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-12-26 09:26:34 +00:00
parent = READ_ONCE(dentry->d_parent);
ceph: handle LOOKUP_RCU in ceph_d_revalidate We can now handle the snapshot cases under RCU, as well as the non-snapshot case when we don't need to queue up a lease renewal allow LOOKUP_RCU walks to proceed under those conditions. Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Yan, Zheng <zyan@redhat.com>
2016-07-01 13:39:21 +00:00
dir = d_inode_rcu(parent);
if (!dir)
return -ECHILD;
ceph: fix RCU case handling in ceph_d_revalidate() For RCU case ->d_revalidate() is called with rcu_read_lock() and without pinning the dentry passed to it. Which means that it can't rely upon ->d_inode remaining stable; that's the reason for d_inode_rcu(), actually. Make sure we don't reload ->d_inode there. Cc: stable@vger.kernel.org Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-10-29 13:50:19 +00:00
inode = d_inode_rcu(dentry);
ceph: handle LOOKUP_RCU in ceph_d_revalidate We can now handle the snapshot cases under RCU, as well as the non-snapshot case when we don't need to queue up a lease renewal allow LOOKUP_RCU walks to proceed under those conditions. Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Yan, Zheng <zyan@redhat.com>
2016-07-01 13:39:21 +00:00
} else {
parent = dget_parent(dentry);
dir = d_inode(parent);
ceph: fix RCU case handling in ceph_d_revalidate() For RCU case ->d_revalidate() is called with rcu_read_lock() and without pinning the dentry passed to it. Which means that it can't rely upon ->d_inode remaining stable; that's the reason for d_inode_rcu(), actually. Make sure we don't reload ->d_inode there. Cc: stable@vger.kernel.org Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-10-29 13:50:19 +00:00
inode = d_inode(dentry);
ceph: handle LOOKUP_RCU in ceph_d_revalidate We can now handle the snapshot cases under RCU, as well as the non-snapshot case when we don't need to queue up a lease renewal allow LOOKUP_RCU walks to proceed under those conditions. Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Yan, Zheng <zyan@redhat.com>
2016-07-01 13:39:21 +00:00
}
fs: rcu-walk aware d_revalidate method Require filesystems be aware of .d_revalidate being called in rcu-walk mode (nd->flags & LOOKUP_RCU). For now do a simple push down, returning -ECHILD from all implementations. Signed-off-by: Nick Piggin <npiggin@kernel.dk>
2011-01-07 06:49:57 +00:00
ceph: print dentry offset in hex and fix xattr_version type In the debug logs about the di->offset or ctx->pos it is in hex format, but some others are using the dec format. It is a little hard to read. For the xattr version, it is u64 type, using a shorter type may truncate it. Signed-off-by: Xiubo Li <xiubli@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-12-19 02:15:18 +00:00
dout("d_revalidate %p '%pd' inode %p offset 0x%llx\n", dentry,
ceph: fix RCU case handling in ceph_d_revalidate() For RCU case ->d_revalidate() is called with rcu_read_lock() and without pinning the dentry passed to it. Which means that it can't rely upon ->d_inode remaining stable; that's the reason for d_inode_rcu(), actually. Make sure we don't reload ->d_inode there. Cc: stable@vger.kernel.org Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-10-29 13:50:19 +00:00
dentry, inode, ceph_dentry(dentry)->offset);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
ceph: consider inode's last read/write when calculating wanted caps Add i_last_rd and i_last_wr to ceph_inode_info. These fields are used to track the last time the client acquired read/write caps for the inode. If there is no read/write on an inode for 'caps_wanted_delay_max' seconds, __ceph_caps_file_wanted() does not request caps for read/write even there are open files. Call __ceph_touch_fmode() for dir operations. __ceph_caps_file_wanted() calculates dir's wanted caps according to last dir read/modification. If there is recent dir read, dir inode wants CEPH_CAP_ANY_SHARED caps. If there is recent dir modification, also wants CEPH_CAP_FILE_EXCL. Readdir is a special case. Dir inode wants CEPH_CAP_FILE_EXCL after readdir, as with that, modifications do not need to release CEPH_CAP_FILE_SHARED or invalidate all dentry leases issued by readdir. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-03-05 12:21:00 +00:00
mdsc = ceph_sb_to_client(dir->i_sb)->mdsc;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
/* always trust cached snapped dentries, snapdir dentry */
if (ceph_snap(dir) != CEPH_NOSNAP) {
assorted conversions to %p[dD] Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2014-10-22 00:11:25 +00:00
dout("d_revalidate %p '%pd' inode %p is SNAPPED\n", dentry,
ceph: fix RCU case handling in ceph_d_revalidate() For RCU case ->d_revalidate() is called with rcu_read_lock() and without pinning the dentry passed to it. Which means that it can't rely upon ->d_inode remaining stable; that's the reason for d_inode_rcu(), actually. Make sure we don't reload ->d_inode there. Cc: stable@vger.kernel.org Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-10-29 13:50:19 +00:00
dentry, inode);
ceph: protect d_parent access in ceph_d_revalidate Protect d_parent with d_lock. Carry a reference. Simplify the flow so that there is a single exit point and cleanup. Reviewed-by: Yehuda Sadeh <yehuda@hq.newdream.net> Signed-off-by: Sage Weil <sage@newdream.net>
2011-07-26 18:30:43 +00:00
valid = 1;
ceph: fix RCU case handling in ceph_d_revalidate() For RCU case ->d_revalidate() is called with rcu_read_lock() and without pinning the dentry passed to it. Which means that it can't rely upon ->d_inode remaining stable; that's the reason for d_inode_rcu(), actually. Make sure we don't reload ->d_inode there. Cc: stable@vger.kernel.org Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-10-29 13:50:19 +00:00
} else if (inode && ceph_snap(inode) == CEPH_SNAPDIR) {
ceph: protect d_parent access in ceph_d_revalidate Protect d_parent with d_lock. Carry a reference. Simplify the flow so that there is a single exit point and cleanup. Reviewed-by: Yehuda Sadeh <yehuda@hq.newdream.net> Signed-off-by: Sage Weil <sage@newdream.net>
2011-07-26 18:30:43 +00:00
valid = 1;
ceph: allow dentry_lease_is_valid to work under RCU walk Under rcuwalk, we need to take extra care when dereferencing d_parent. We want to do that once and pass a pointer to dentry_lease_is_valid. Also, we must ensure that that function can handle the case where we're racing with d_release. Check whether "di" is NULL under the d_lock, and just return 0 if so. Finally, we still need to kick off a renewal job if the lease is getting close to expiration. If that's the case, then just drop out of rcuwalk mode since that could block. Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Yan, Zheng <zyan@redhat.com>
2016-07-01 13:39:20 +00:00
} else {
ceph: ensure d_name/d_parent stability in ceph_mdsc_lease_send_msg() Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-05-23 02:45:24 +00:00
valid = dentry_lease_is_valid(dentry, flags);
ceph: allow dentry_lease_is_valid to work under RCU walk Under rcuwalk, we need to take extra care when dereferencing d_parent. We want to do that once and pass a pointer to dentry_lease_is_valid. Also, we must ensure that that function can handle the case where we're racing with d_release. Check whether "di" is NULL under the d_lock, and just return 0 if so. Finally, we still need to kick off a renewal job if the lease is getting close to expiration. If that's the case, then just drop out of rcuwalk mode since that could block. Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Yan, Zheng <zyan@redhat.com>
2016-07-01 13:39:20 +00:00
if (valid == -ECHILD)
return valid;
ceph: consider inode's last read/write when calculating wanted caps Add i_last_rd and i_last_wr to ceph_inode_info. These fields are used to track the last time the client acquired read/write caps for the inode. If there is no read/write on an inode for 'caps_wanted_delay_max' seconds, __ceph_caps_file_wanted() does not request caps for read/write even there are open files. Call __ceph_touch_fmode() for dir operations. __ceph_caps_file_wanted() calculates dir's wanted caps according to last dir read/modification. If there is recent dir read, dir inode wants CEPH_CAP_ANY_SHARED caps. If there is recent dir modification, also wants CEPH_CAP_FILE_EXCL. Readdir is a special case. Dir inode wants CEPH_CAP_FILE_EXCL after readdir, as with that, modifications do not need to release CEPH_CAP_FILE_SHARED or invalidate all dentry leases issued by readdir. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-03-05 12:21:00 +00:00
if (valid || dir_lease_is_valid(dir, dentry, mdsc)) {
ceph: fix RCU case handling in ceph_d_revalidate() For RCU case ->d_revalidate() is called with rcu_read_lock() and without pinning the dentry passed to it. Which means that it can't rely upon ->d_inode remaining stable; that's the reason for d_inode_rcu(), actually. Make sure we don't reload ->d_inode there. Cc: stable@vger.kernel.org Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-10-29 13:50:19 +00:00
if (inode)
valid = ceph_is_any_caps(inode);
ceph: allow dentry_lease_is_valid to work under RCU walk Under rcuwalk, we need to take extra care when dereferencing d_parent. We want to do that once and pass a pointer to dentry_lease_is_valid. Also, we must ensure that that function can handle the case where we're racing with d_release. Check whether "di" is NULL under the d_lock, and just return 0 if so. Finally, we still need to kick off a renewal job if the lease is getting close to expiration. If that's the case, then just drop out of rcuwalk mode since that could block. Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Yan, Zheng <zyan@redhat.com>
2016-07-01 13:39:20 +00:00
else
valid = 1;
}
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
}
ceph: use lookup request to revalidate dentry If dentry has no lease, ceph_d_revalidate() previously return 0. This causes VFS to invalidate the dentry and create a new dentry for later lookup. Invalidating a dentry also detach any underneath mount points. So mount point inside cephfs can disapear mystically (even the mount point is not modified by other hosts). The fix is using lookup request to revalidate dentry without lease. This can partly solve the mount points disapear issue (as long as the mount point is not modified by other hosts) Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-03-17 06:41:59 +00:00
if (!valid) {
struct ceph_mds_request *req;
ceph: fix endianness of getattr mask in ceph_d_revalidate sparse says: fs/ceph/dir.c:1248:50: warning: incorrect type in assignment (different base types) fs/ceph/dir.c:1248:50: expected restricted __le32 [usertype] mask fs/ceph/dir.c:1248:50: got int [signed] [assigned] mask Fixes: 200fd27c8fa2 ("ceph: use lookup request to revalidate dentry") Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Sage Weil <sage@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-01-12 19:42:38 +00:00
int op, err;
u32 mask;
ceph: use lookup request to revalidate dentry If dentry has no lease, ceph_d_revalidate() previously return 0. This causes VFS to invalidate the dentry and create a new dentry for later lookup. Invalidating a dentry also detach any underneath mount points. So mount point inside cephfs can disapear mystically (even the mount point is not modified by other hosts). The fix is using lookup request to revalidate dentry without lease. This can partly solve the mount points disapear issue (as long as the mount point is not modified by other hosts) Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-03-17 06:41:59 +00:00
ceph: handle LOOKUP_RCU in ceph_d_revalidate We can now handle the snapshot cases under RCU, as well as the non-snapshot case when we don't need to queue up a lease renewal allow LOOKUP_RCU walks to proceed under those conditions. Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Yan, Zheng <zyan@redhat.com>
2016-07-01 13:39:21 +00:00
if (flags & LOOKUP_RCU)
return -ECHILD;
ceph: add dentry lease metric support For dentry leases, only count the hit/miss info triggered from the vfs calls. For the cases like request reply handling and ceph_trim_dentries, ignore them. For now, these are only viewable using debugfs. Future patches will allow the client to send the stats to the MDS. The output looks like: item total miss hit ------------------------------------------------- d_lease 11 7 141 URL: https://tracker.ceph.com/issues/43215 Signed-off-by: Xiubo Li <xiubli@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-03-20 03:44:59 +00:00
percpu_counter_inc(&mdsc->metric.d_lease_mis);
ceph: use lookup request to revalidate dentry If dentry has no lease, ceph_d_revalidate() previously return 0. This causes VFS to invalidate the dentry and create a new dentry for later lookup. Invalidating a dentry also detach any underneath mount points. So mount point inside cephfs can disapear mystically (even the mount point is not modified by other hosts). The fix is using lookup request to revalidate dentry without lease. This can partly solve the mount points disapear issue (as long as the mount point is not modified by other hosts) Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-03-17 06:41:59 +00:00
op = ceph_snap(dir) == CEPH_SNAPDIR ?
ceph: do a LOOKUP in d_revalidate instead of GETATTR In commit c3f4688a08f (ceph: don't set req->r_locked_dir in ceph_d_revalidate), we changed the code to do a GETATTR instead of a LOOKUP as the parent info isn't strictly necessary to revalidate the dentry. What we missed there though is that in order to update the lease on the dentry after revalidating it, we _do_ need parent info. Change ceph_d_revalidate back to doing a LOOKUP instead of a GETATTR so that we can get the parent info in order to update the lease from ceph_fill_trace. Note that we set req->r_parent here, but we cannot set the CEPH_MDS_R_PARENT_LOCKED flag as we can't guarantee that it is. Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Yan, Zheng <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-01-30 14:47:25 +00:00
CEPH_MDS_OP_LOOKUPSNAP : CEPH_MDS_OP_LOOKUP;
ceph: use lookup request to revalidate dentry If dentry has no lease, ceph_d_revalidate() previously return 0. This causes VFS to invalidate the dentry and create a new dentry for later lookup. Invalidating a dentry also detach any underneath mount points. So mount point inside cephfs can disapear mystically (even the mount point is not modified by other hosts). The fix is using lookup request to revalidate dentry without lease. This can partly solve the mount points disapear issue (as long as the mount point is not modified by other hosts) Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-03-17 06:41:59 +00:00
req = ceph_mdsc_create_request(mdsc, op, USE_ANY_MDS);
if (!IS_ERR(req)) {
req->r_dentry = dget(dentry);
ceph: do a LOOKUP in d_revalidate instead of GETATTR In commit c3f4688a08f (ceph: don't set req->r_locked_dir in ceph_d_revalidate), we changed the code to do a GETATTR instead of a LOOKUP as the parent info isn't strictly necessary to revalidate the dentry. What we missed there though is that in order to update the lease on the dentry after revalidating it, we _do_ need parent info. Change ceph_d_revalidate back to doing a LOOKUP instead of a GETATTR so that we can get the parent info in order to update the lease from ceph_fill_trace. Note that we set req->r_parent here, but we cannot set the CEPH_MDS_R_PARENT_LOCKED flag as we can't guarantee that it is. Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Yan, Zheng <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-01-30 14:47:25 +00:00
req->r_num_caps = 2;
req->r_parent = dir;
ceph: take reference to req->r_parent at point of assignment Currently, we set the r_parent pointer but then don't take a reference to it until we submit the request. If we end up freeing the req before that point, then we'll do a iput when we shouldn't. Instead, take the inode reference in the callers, so that it's always safe to call ceph_mdsc_put_request on the req, even before submission. Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: Luis Henriques <lhenriques@suse.de> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2021-06-18 17:05:06 +00:00
ihold(dir);
ceph: use lookup request to revalidate dentry If dentry has no lease, ceph_d_revalidate() previously return 0. This causes VFS to invalidate the dentry and create a new dentry for later lookup. Invalidating a dentry also detach any underneath mount points. So mount point inside cephfs can disapear mystically (even the mount point is not modified by other hosts). The fix is using lookup request to revalidate dentry without lease. This can partly solve the mount points disapear issue (as long as the mount point is not modified by other hosts) Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-03-17 06:41:59 +00:00
mask = CEPH_STAT_CAP_INODE | CEPH_CAP_AUTH_SHARED;
if (ceph_security_xattr_wanted(dir))
mask |= CEPH_CAP_XATTR_SHARED;
ceph: fix endianness of getattr mask in ceph_d_revalidate sparse says: fs/ceph/dir.c:1248:50: warning: incorrect type in assignment (different base types) fs/ceph/dir.c:1248:50: expected restricted __le32 [usertype] mask fs/ceph/dir.c:1248:50: got int [signed] [assigned] mask Fixes: 200fd27c8fa2 ("ceph: use lookup request to revalidate dentry") Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Sage Weil <sage@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-01-12 19:42:38 +00:00
req->r_args.getattr.mask = cpu_to_le32(mask);
ceph: use lookup request to revalidate dentry If dentry has no lease, ceph_d_revalidate() previously return 0. This causes VFS to invalidate the dentry and create a new dentry for later lookup. Invalidating a dentry also detach any underneath mount points. So mount point inside cephfs can disapear mystically (even the mount point is not modified by other hosts). The fix is using lookup request to revalidate dentry without lease. This can partly solve the mount points disapear issue (as long as the mount point is not modified by other hosts) Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-03-17 06:41:59 +00:00
err = ceph_mdsc_do_request(mdsc, NULL, req);
ceph: don't set req->r_locked_dir in ceph_d_revalidate This function sets req->r_locked_dir which is supposed to indicate to ceph_fill_trace that the parent's i_rwsem is locked for write. Unfortunately, there is no guarantee that the dir will be locked when d_revalidate is called, so we really don't want ceph_fill_trace to do any dcache manipulation from this context. Clear req->r_locked_dir since it's clearly not safe to do that. What we really want to know with d_revalidate is whether the dentry still points to the same inode. ceph_fill_trace installs a pointer to the inode in req->r_target_inode, so we can just compare that to d_inode(dentry) to see if it's the same one after the lookup. Also, since we aren't generally interested in the parent here, we can switch to using a GETATTR to hint that to the MDS, which also means that we only need to reserve one cap. Finally, just remove the d_unhashed check. That's really outside the purview of a filesystem's d_revalidate. If the thing became unhashed while we're checking it, then that's up to the VFS to handle anyway. Fixes: 200fd27c8fa2 ("ceph: use lookup request to revalidate dentry") Link: http://tracker.ceph.com/issues/18041 Reported-by: Donatas Abraitis <donatas.abraitis@gmail.com> Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2016-11-30 20:56:46 +00:00
switch (err) {
case 0:
if (d_really_is_positive(dentry) &&
d_inode(dentry) == req->r_target_inode)
valid = 1;
break;
case -ENOENT:
if (d_really_is_negative(dentry))
valid = 1;
treewide: Use fallthrough pseudo-keyword Replace the existing /* fall through */ comments and its variants with the new pseudo-keyword macro fallthrough[1]. Also, remove unnecessary fall-through markings when it is the case. [1] https://www.kernel.org/doc/html/v5.7/process/deprecated.html?highlight=fallthrough#implicit-switch-case-fall-through Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
2020-08-23 22:36:59 +00:00
fallthrough;
ceph: don't set req->r_locked_dir in ceph_d_revalidate This function sets req->r_locked_dir which is supposed to indicate to ceph_fill_trace that the parent's i_rwsem is locked for write. Unfortunately, there is no guarantee that the dir will be locked when d_revalidate is called, so we really don't want ceph_fill_trace to do any dcache manipulation from this context. Clear req->r_locked_dir since it's clearly not safe to do that. What we really want to know with d_revalidate is whether the dentry still points to the same inode. ceph_fill_trace installs a pointer to the inode in req->r_target_inode, so we can just compare that to d_inode(dentry) to see if it's the same one after the lookup. Also, since we aren't generally interested in the parent here, we can switch to using a GETATTR to hint that to the MDS, which also means that we only need to reserve one cap. Finally, just remove the d_unhashed check. That's really outside the purview of a filesystem's d_revalidate. If the thing became unhashed while we're checking it, then that's up to the VFS to handle anyway. Fixes: 200fd27c8fa2 ("ceph: use lookup request to revalidate dentry") Link: http://tracker.ceph.com/issues/18041 Reported-by: Donatas Abraitis <donatas.abraitis@gmail.com> Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2016-11-30 20:56:46 +00:00
default:
break;
ceph: use lookup request to revalidate dentry If dentry has no lease, ceph_d_revalidate() previously return 0. This causes VFS to invalidate the dentry and create a new dentry for later lookup. Invalidating a dentry also detach any underneath mount points. So mount point inside cephfs can disapear mystically (even the mount point is not modified by other hosts). The fix is using lookup request to revalidate dentry without lease. This can partly solve the mount points disapear issue (as long as the mount point is not modified by other hosts) Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-03-17 06:41:59 +00:00
}
ceph_mdsc_put_request(req);
dout("d_revalidate %p lookup result=%d\n",
dentry, err);
}
ceph: add dentry lease metric support For dentry leases, only count the hit/miss info triggered from the vfs calls. For the cases like request reply handling and ceph_trim_dentries, ignore them. For now, these are only viewable using debugfs. Future patches will allow the client to send the stats to the MDS. The output looks like: item total miss hit ------------------------------------------------- d_lease 11 7 141 URL: https://tracker.ceph.com/issues/43215 Signed-off-by: Xiubo Li <xiubli@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-03-20 03:44:59 +00:00
} else {
percpu_counter_inc(&mdsc->metric.d_lease_hit);
ceph: use lookup request to revalidate dentry If dentry has no lease, ceph_d_revalidate() previously return 0. This causes VFS to invalidate the dentry and create a new dentry for later lookup. Invalidating a dentry also detach any underneath mount points. So mount point inside cephfs can disapear mystically (even the mount point is not modified by other hosts). The fix is using lookup request to revalidate dentry without lease. This can partly solve the mount points disapear issue (as long as the mount point is not modified by other hosts) Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-03-17 06:41:59 +00:00
}
ceph: protect d_parent access in ceph_d_revalidate Protect d_parent with d_lock. Carry a reference. Simplify the flow so that there is a single exit point and cleanup. Reviewed-by: Yehuda Sadeh <yehuda@hq.newdream.net> Signed-off-by: Sage Weil <sage@newdream.net>
2011-07-26 18:30:43 +00:00
dout("d_revalidate %p %s\n", dentry, valid ? "valid" : "invalid");
ceph: periodically trim stale dentries Previous commit make VFS delete stale dentry when last reference is dropped. Lease also can become invalid when corresponding dentry has no reference. This patch make cephfs periodically scan lease list, delete corresponding dentry if lease is invalid. There are two types of lease, dentry lease and dir lease. dentry lease has life time and applies to singe dentry. Dentry lease is added to tail of a list when it's updated, leases at front of the list will expire first. Dir lease is CEPH_CAP_FILE_SHARED on directory inode, it applies to all dentries in the directory. Dentries have dir leases are added to another list. Dentries in the list are periodically checked in a round robin manner. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-01-31 08:55:51 +00:00
if (!valid)
ceph: check inode caps in ceph_d_revalidate Some inodes in readdir reply may have no caps. Getattr mds request for these inodes can return -ESTALE. The fix is consider dentry that links to inode with no caps as invalid. Invalid dentry causes a lookup request to send to the mds, the MDS will send caps back. Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
2013-11-30 04:47:41 +00:00
ceph_dir_clear_complete(dir);
ceph: kill ceph_get_dentry_parent_inode() use vfs helper dget_parent() instead Signed-off-by: Yan, Zheng <zyan@redhat.com>
2016-03-16 08:40:23 +00:00
ceph: handle LOOKUP_RCU in ceph_d_revalidate We can now handle the snapshot cases under RCU, as well as the non-snapshot case when we don't need to queue up a lease renewal allow LOOKUP_RCU walks to proceed under those conditions. Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Yan, Zheng <zyan@redhat.com>
2016-07-01 13:39:21 +00:00
if (!(flags & LOOKUP_RCU))
dput(parent);
ceph: protect d_parent access in ceph_d_revalidate Protect d_parent with d_lock. Carry a reference. Simplify the flow so that there is a single exit point and cleanup. Reviewed-by: Yehuda Sadeh <yehuda@hq.newdream.net> Signed-off-by: Sage Weil <sage@newdream.net>
2011-07-26 18:30:43 +00:00
return valid;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
}
ceph: delete stale dentry when last reference is dropped introduce ceph_d_delete(), which checks if dentry has valid lease. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-01-28 12:43:55 +00:00
/*
* Delete unused dentry that doesn't have valid lease
*
* Called under dentry->d_lock.
*/
static int ceph_d_delete(const struct dentry *dentry)
{
struct ceph_dentry_info *di;
/* won't release caps */
if (d_really_is_negative(dentry))
return 0;
if (ceph_snap(d_inode(dentry)) != CEPH_NOSNAP)
return 0;
/* vaild lease? */
di = ceph_dentry(dentry);
if (di) {
if (__dentry_lease_is_valid(di))
return 0;
if (__dir_lease_try_check(dentry))
return 0;
}
return 1;
}
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
/*
ceph: rename dentry_release -> d_release, fix comment Just for consistency's sake. Fix obsolete comment too. Signed-off-by: Sage Weil <sage@newdream.net>
2011-03-15 21:57:41 +00:00
* Release our ceph_dentry_info.
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
*/
ceph: rename dentry_release -> d_release, fix comment Just for consistency's sake. Fix obsolete comment too. Signed-off-by: Sage Weil <sage@newdream.net>
2011-03-15 21:57:41 +00:00
static void ceph_d_release(struct dentry *dentry)
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
{
struct ceph_dentry_info *di = ceph_dentry(dentry);
ceph: add dentry lease metric support For dentry leases, only count the hit/miss info triggered from the vfs calls. For the cases like request reply handling and ceph_trim_dentries, ignore them. For now, these are only viewable using debugfs. Future patches will allow the client to send the stats to the MDS. The output looks like: item total miss hit ------------------------------------------------- d_lease 11 7 141 URL: https://tracker.ceph.com/issues/43215 Signed-off-by: Xiubo Li <xiubli@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-03-20 03:44:59 +00:00
struct ceph_fs_client *fsc = ceph_sb_to_client(dentry->d_sb);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
ceph: rename dentry_release -> d_release, fix comment Just for consistency's sake. Fix obsolete comment too. Signed-off-by: Sage Weil <sage@newdream.net>
2011-03-15 21:57:41 +00:00
dout("d_release %p\n", dentry);
ceph: clear d_fsinfo pointer under d_lock To check for a valid dentry lease, we need to get at the ceph_dentry_info. Under rcuwalk though, we may end up with a dentry that is on its way to destruction. Since we need to take the d_lock in dentry_lease_is_valid already, we can just ensure that we clear the d_fsinfo pointer out under the same lock before destroying it. Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Yan, Zheng <zyan@redhat.com>
2016-07-01 13:39:20 +00:00
ceph: add dentry lease metric support For dentry leases, only count the hit/miss info triggered from the vfs calls. For the cases like request reply handling and ceph_trim_dentries, ignore them. For now, these are only viewable using debugfs. Future patches will allow the client to send the stats to the MDS. The output looks like: item total miss hit ------------------------------------------------- d_lease 11 7 141 URL: https://tracker.ceph.com/issues/43215 Signed-off-by: Xiubo Li <xiubli@redhat.com> Reviewed-by: Jeff Layton <jlayton@kernel.org> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-03-20 03:44:59 +00:00
atomic64_dec(&fsc->mdsc->metric.total_dentries);
ceph: clear d_fsinfo pointer under d_lock To check for a valid dentry lease, we need to get at the ceph_dentry_info. Under rcuwalk though, we may end up with a dentry that is on its way to destruction. Since we need to take the d_lock in dentry_lease_is_valid already, we can just ensure that we clear the d_fsinfo pointer out under the same lock before destroying it. Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Yan, Zheng <zyan@redhat.com>
2016-07-01 13:39:20 +00:00
spin_lock(&dentry->d_lock);
ceph: periodically trim stale dentries Previous commit make VFS delete stale dentry when last reference is dropped. Lease also can become invalid when corresponding dentry has no reference. This patch make cephfs periodically scan lease list, delete corresponding dentry if lease is invalid. There are two types of lease, dentry lease and dir lease. dentry lease has life time and applies to singe dentry. Dentry lease is added to tail of a list when it's updated, leases at front of the list will expire first. Dir lease is CEPH_CAP_FILE_SHARED on directory inode, it applies to all dentries in the directory. Dentries have dir leases are added to another list. Dentries in the list are periodically checked in a round robin manner. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-01-31 08:55:51 +00:00
__dentry_lease_unlist(di);
ceph: clear d_fsinfo pointer under d_lock To check for a valid dentry lease, we need to get at the ceph_dentry_info. Under rcuwalk though, we may end up with a dentry that is on its way to destruction. Since we need to take the d_lock in dentry_lease_is_valid already, we can just ensure that we clear the d_fsinfo pointer out under the same lock before destroying it. Signed-off-by: Jeff Layton <jlayton@redhat.com> Reviewed-by: Yan, Zheng <zyan@redhat.com>
2016-07-01 13:39:20 +00:00
dentry->d_fsdata = NULL;
spin_unlock(&dentry->d_lock);
ceph: allow ceph_put_mds_session to take NULL or ERR_PTR ...to simplify some error paths. Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: Luis Henriques <lhenriques@suse.de> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2021-06-09 18:09:52 +00:00
ceph_put_mds_session(di->lease_session);
ceph: remove unnecessary d_fsdata conditional checks We now set d_fsdata unconditionally on all dentries prior to setting up the d_ops, so all of these checks are unnecessary. Signed-off-by: Sage Weil <sage@newdream.net>
2011-11-11 17:48:53 +00:00
kmem_cache_free(ceph_dentry_cachep, di);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
}
ceph: clear parent D_COMPLETE flag when on dentry prune When the VFS prunes a dentry from the cache, clear the D_COMPLETE flag on the parent dentry. Do this for the live and snapshotted namespaces. Do not bother for the .snap dir contents, since we do not cache that. Signed-off-by: Sage Weil <sage@newdream.net>
2011-03-15 22:53:40 +00:00
/*
* When the VFS prunes a dentry from the cache, we need to clear the
* complete flag on the parent directory.
*
* Called under dentry->d_lock.
*/
static void ceph_d_prune(struct dentry *dentry)
{
ceph: avoid dereferencing invalid pointer during cached readdir Readdir cache keeps array of dentry pointers in page cache. If any dentry in readdir cache gets pruned, ceph_d_prune() disables readdir cache for later readdir syscall. The problem is that ceph_d_prune() ignores unhashed dentry. Ideally MDS should have already revoked CEPH_CAP_FILE_SHARED (which also disables readdir cache) when dentry gets unhashed. But if it is somehow MDS does not properly revoke CEPH_CAP_FILE_SHARED and the unhashed dentry gets pruned later, ceph_d_prune() will not disable readdir cache, later readdir may reference invalid dentry pointer. The fix is make ceph_d_prune() do extra check for unhashed dentry. Disable readdir cache if the unhashed dentry is still referenced by readdir cache. Another fix in this patch is handle d_splice_alias(). If a dentry gets spliced into new parent dentry, treat it as if it was pruned (call ceph_d_prune() for it). Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Acked-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-11-27 03:23:48 +00:00
struct ceph_inode_info *dir_ci;
struct ceph_dentry_info *di;
dout("ceph_d_prune %pd %p\n", dentry, dentry);
ceph: clear parent D_COMPLETE flag when on dentry prune When the VFS prunes a dentry from the cache, clear the D_COMPLETE flag on the parent dentry. Do this for the live and snapshotted namespaces. Do not bother for the .snap dir contents, since we do not cache that. Signed-off-by: Sage Weil <sage@newdream.net>
2011-03-15 22:53:40 +00:00
/* do we have a valid parent? */
ceph: clean up useless d_parent checks d_parent is never NULL, and IS_ROOT() is the proper way to check for a (non-self-referential) parent. Reported-by: Al Viro <viro@ZenIV.linux.org.uk> Signed-off-by: Sage Weil <sage@inktank.com>
2012-06-07 20:43:35 +00:00
if (IS_ROOT(dentry))
ceph: clear parent D_COMPLETE flag when on dentry prune When the VFS prunes a dentry from the cache, clear the D_COMPLETE flag on the parent dentry. Do this for the live and snapshotted namespaces. Do not bother for the .snap dir contents, since we do not cache that. Signed-off-by: Sage Weil <sage@newdream.net>
2011-03-15 22:53:40 +00:00
return;
ceph: avoid dereferencing invalid pointer during cached readdir Readdir cache keeps array of dentry pointers in page cache. If any dentry in readdir cache gets pruned, ceph_d_prune() disables readdir cache for later readdir syscall. The problem is that ceph_d_prune() ignores unhashed dentry. Ideally MDS should have already revoked CEPH_CAP_FILE_SHARED (which also disables readdir cache) when dentry gets unhashed. But if it is somehow MDS does not properly revoke CEPH_CAP_FILE_SHARED and the unhashed dentry gets pruned later, ceph_d_prune() will not disable readdir cache, later readdir may reference invalid dentry pointer. The fix is make ceph_d_prune() do extra check for unhashed dentry. Disable readdir cache if the unhashed dentry is still referenced by readdir cache. Another fix in this patch is handle d_splice_alias(). If a dentry gets spliced into new parent dentry, treat it as if it was pruned (call ceph_d_prune() for it). Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Acked-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-11-27 03:23:48 +00:00
/* we hold d_lock, so d_parent is stable */
dir_ci = ceph_inode(d_inode(dentry->d_parent));
if (dir_ci->i_vino.snap == CEPH_SNAPDIR)
ceph: clear parent D_COMPLETE flag when on dentry prune When the VFS prunes a dentry from the cache, clear the D_COMPLETE flag on the parent dentry. Do this for the live and snapshotted namespaces. Do not bother for the .snap dir contents, since we do not cache that. Signed-off-by: Sage Weil <sage@newdream.net>
2011-03-15 22:53:40 +00:00
return;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
ceph: avoid dereferencing invalid pointer during cached readdir Readdir cache keeps array of dentry pointers in page cache. If any dentry in readdir cache gets pruned, ceph_d_prune() disables readdir cache for later readdir syscall. The problem is that ceph_d_prune() ignores unhashed dentry. Ideally MDS should have already revoked CEPH_CAP_FILE_SHARED (which also disables readdir cache) when dentry gets unhashed. But if it is somehow MDS does not properly revoke CEPH_CAP_FILE_SHARED and the unhashed dentry gets pruned later, ceph_d_prune() will not disable readdir cache, later readdir may reference invalid dentry pointer. The fix is make ceph_d_prune() do extra check for unhashed dentry. Disable readdir cache if the unhashed dentry is still referenced by readdir cache. Another fix in this patch is handle d_splice_alias(). If a dentry gets spliced into new parent dentry, treat it as if it was pruned (call ceph_d_prune() for it). Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Acked-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-11-27 03:23:48 +00:00
/* who calls d_delete() should also disable dcache readdir */
if (d_really_is_negative(dentry))
ceph: unify dentry_operations instances Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2016-10-29 01:52:50 +00:00
return;
ceph: avoid dereferencing invalid pointer during cached readdir Readdir cache keeps array of dentry pointers in page cache. If any dentry in readdir cache gets pruned, ceph_d_prune() disables readdir cache for later readdir syscall. The problem is that ceph_d_prune() ignores unhashed dentry. Ideally MDS should have already revoked CEPH_CAP_FILE_SHARED (which also disables readdir cache) when dentry gets unhashed. But if it is somehow MDS does not properly revoke CEPH_CAP_FILE_SHARED and the unhashed dentry gets pruned later, ceph_d_prune() will not disable readdir cache, later readdir may reference invalid dentry pointer. The fix is make ceph_d_prune() do extra check for unhashed dentry. Disable readdir cache if the unhashed dentry is still referenced by readdir cache. Another fix in this patch is handle d_splice_alias(). If a dentry gets spliced into new parent dentry, treat it as if it was pruned (call ceph_d_prune() for it). Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Acked-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2017-11-27 03:23:48 +00:00
/* d_fsdata does not get cleared until d_release */
if (!d_unhashed(dentry)) {
__ceph_dir_clear_complete(dir_ci);
return;
}
/* Disable dcache readdir just in case that someone called d_drop()
* or d_invalidate(), but MDS didn't revoke CEPH_CAP_FILE_SHARED
* properly (dcache readdir is still enabled) */
di = ceph_dentry(dentry);
if (di->offset > 0 &&
di->lease_shared_gen == atomic_read(&dir_ci->i_shared_gen))
__ceph_dir_clear_ordered(dir_ci);
ceph: clear parent D_COMPLETE flag when on dentry prune When the VFS prunes a dentry from the cache, clear the D_COMPLETE flag on the parent dentry. Do this for the live and snapshotted namespaces. Do not bother for the .snap dir contents, since we do not cache that. Signed-off-by: Sage Weil <sage@newdream.net>
2011-03-15 22:53:40 +00:00
}
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
/*
* read() on a dir. This weird interface hack only works if mounted
* with '-o dirstat'.
*/
static ssize_t ceph_read_dir(struct file *file, char __user *buf, size_t size,
loff_t *ppos)
{
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
struct ceph_dir_file_info *dfi = file->private_data;
new helper: file_inode(file) Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2013-01-23 22:07:38 +00:00
struct inode *inode = file_inode(file);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
struct ceph_inode_info *ci = ceph_inode(inode);
int left;
ceph: use snprintf for dirstat content We allocate a buffer for rstats if the dirstat option is enabled. Use snprintf. Signed-off-by: Sage Weil <sage@newdream.net>
2011-05-12 21:28:05 +00:00
const int bufsize = 1024;
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
ceph: factor out libceph from Ceph file system This factors out protocol and low-level storage parts of ceph into a separate libceph module living in net/ceph and include/linux/ceph. This is mostly a matter of moving files around. However, a few key pieces of the interface change as well: - ceph_client becomes ceph_fs_client and ceph_client, where the latter captures the mon and osd clients, and the fs_client gets the mds client and file system specific pieces. - Mount option parsing and debugfs setup is correspondingly broken into two pieces. - The mon client gets a generic handler callback for otherwise unknown messages (mds map, in this case). - The basic supported/required feature bits can be expanded (and are by ceph_fs_client). No functional change, aside from some subtle error handling cases that got cleaned up in the refactoring process. Signed-off-by: Sage Weil <sage@newdream.net>
2010-04-06 22:14:15 +00:00
if (!ceph_test_mount_opt(ceph_sb_to_client(inode->i_sb), DIRSTAT))
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
return -EISDIR;
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
if (!dfi->dir_info) {
dfi->dir_info = kmalloc(bufsize, GFP_KERNEL);
if (!dfi->dir_info)
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
return -ENOMEM;
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
dfi->dir_info_len =
snprintf(dfi->dir_info, bufsize,
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
"entries: %20lld\n"
" files: %20lld\n"
" subdirs: %20lld\n"
"rentries: %20lld\n"
" rfiles: %20lld\n"
" rsubdirs: %20lld\n"
"rbytes: %20lld\n"
ceph: use timespec64 for inode timestamp Since the vfs structures are all using timespec64, we can now change the internal representation, using ceph_encode_timespec64 and ceph_decode_timespec64. In case of ceph_aux_inode however, we need to avoid doing a memcmp() on uninitialized padding data, so the members of the i_mtime field get copied individually into 64-bit integers. Signed-off-by: Arnd Bergmann <arnd@arndb.de> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-07-13 20:18:36 +00:00
"rctime: %10lld.%09ld\n",
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
ci->i_files + ci->i_subdirs,
ci->i_files,
ci->i_subdirs,
ci->i_rfiles + ci->i_rsubdirs,
ci->i_rfiles,
ci->i_rsubdirs,
ci->i_rbytes,
ceph: use timespec64 for inode timestamp Since the vfs structures are all using timespec64, we can now change the internal representation, using ceph_encode_timespec64 and ceph_decode_timespec64. In case of ceph_aux_inode however, we need to avoid doing a memcmp() on uninitialized padding data, so the members of the i_mtime field get copied individually into 64-bit integers. Signed-off-by: Arnd Bergmann <arnd@arndb.de> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-07-13 20:18:36 +00:00
ci->i_rctime.tv_sec,
ci->i_rctime.tv_nsec);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
}
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
if (*ppos >= dfi->dir_info_len)
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
return 0;
ceph: optimize memory usage In current code, regular file and directory use same struct ceph_file_info to store fs specific data so the struct has to include some fields which are only used for directory (e.g., readdir related info), when having plenty of regular files, it will lead to memory waste. This patch introduces dedicated ceph_dir_file_info cache for readdir related thins. So that regular file does not include those unused fields anymore. Signed-off-by: Chengguang Xu <cgxu519@gmx.com> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-03-13 02:42:44 +00:00
size = min_t(unsigned, size, dfi->dir_info_len-*ppos);
left = copy_to_user(buf, dfi->dir_info + *ppos, size);
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
if (left == size)
return -EFAULT;
*ppos += (size - left);
return size - left;
}
ceph: add dir_layout to inode Add a ceph_dir_layout to the inode, and calculate dentry hash values based on the parent directory's specified dir_hash function. This is needed because the old default Linux dcache hash function is extremely week and leads to a poor distribution of files among dir fragments. Signed-off-by: Sage Weil <sage@newdream.net>
2010-11-16 19:14:34 +00:00
/*
* Return name hash for a given dentry. This is dependent on
* the parent directory's hash function.
*/
ceph: avoid d_parent in ceph_dentry_hash; fix ceph_encode_fh() hashing bug Have caller pass in a safely-obtained reference to the parent directory for calculating a dentry's hash valud. While we're here, simpify the flow through ceph_encode_fh() so that there is a single exit point and cleanup. Also fix a bug with the dentry hash calculation: calculate the hash for the dentry we were given, not its parent. Reviewed-by: Yehuda Sadeh <yehuda@hq.newdream.net> Signed-off-by: Sage Weil <sage@newdream.net>
2011-07-26 18:30:55 +00:00
unsigned ceph_dentry_hash(struct inode *dir, struct dentry *dn)
ceph: add dir_layout to inode Add a ceph_dir_layout to the inode, and calculate dentry hash values based on the parent directory's specified dir_hash function. This is needed because the old default Linux dcache hash function is extremely week and leads to a poor distribution of files among dir fragments. Signed-off-by: Sage Weil <sage@newdream.net>
2010-11-16 19:14:34 +00:00
{
struct ceph_inode_info *dci = ceph_inode(dir);
ceph: ensure d_name stability in ceph_dentry_hash() Take the d_lock here to ensure that d_name doesn't change. Cc: stable@vger.kernel.org Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-04-17 16:58:28 +00:00
unsigned hash;
ceph: add dir_layout to inode Add a ceph_dir_layout to the inode, and calculate dentry hash values based on the parent directory's specified dir_hash function. This is needed because the old default Linux dcache hash function is extremely week and leads to a poor distribution of files among dir fragments. Signed-off-by: Sage Weil <sage@newdream.net>
2010-11-16 19:14:34 +00:00
switch (dci->i_dir_layout.dl_dir_hash) {
case 0: /* for backward compat */
case CEPH_STR_HASH_LINUX:
return dn->d_name.hash;
default:
ceph: ensure d_name stability in ceph_dentry_hash() Take the d_lock here to ensure that d_name doesn't change. Cc: stable@vger.kernel.org Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-04-17 16:58:28 +00:00
spin_lock(&dn->d_lock);
hash = ceph_str_hash(dci->i_dir_layout.dl_dir_hash,
ceph: add dir_layout to inode Add a ceph_dir_layout to the inode, and calculate dentry hash values based on the parent directory's specified dir_hash function. This is needed because the old default Linux dcache hash function is extremely week and leads to a poor distribution of files among dir fragments. Signed-off-by: Sage Weil <sage@newdream.net>
2010-11-16 19:14:34 +00:00
dn->d_name.name, dn->d_name.len);
ceph: ensure d_name stability in ceph_dentry_hash() Take the d_lock here to ensure that d_name doesn't change. Cc: stable@vger.kernel.org Signed-off-by: Jeff Layton <jlayton@kernel.org> Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-04-17 16:58:28 +00:00
spin_unlock(&dn->d_lock);
return hash;
ceph: add dir_layout to inode Add a ceph_dir_layout to the inode, and calculate dentry hash values based on the parent directory's specified dir_hash function. This is needed because the old default Linux dcache hash function is extremely week and leads to a poor distribution of files among dir fragments. Signed-off-by: Sage Weil <sage@newdream.net>
2010-11-16 19:14:34 +00:00
}
}
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
const struct file_operations ceph_dir_fops = {
.read = ceph_read_dir,
[readdir] convert ceph Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2013-05-17 20:52:26 +00:00
.iterate = ceph_readdir,
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
.llseek = ceph_dir_llseek,
.open = ceph_open,
.release = ceph_release,
.unlocked_ioctl = ceph_ioctl,
ceph: fix compat_ioctl for ceph_dir_operations The ceph_ioctl function is used both for files and directories, but only the files support doing that in 32-bit compat mode. On the s390 architecture, there is also a problem with invalid 31-bit pointers that need to be passed through compat_ptr(). Use the new compat_ptr_ioctl() to address both issues. Note: When backporting this patch to stable kernels, "compat_ioctl: add compat_ptr_ioctl()" is needed as well. Reviewed-by: "Yan, Zheng" <zyan@redhat.com> Cc: stable@vger.kernel.org Signed-off-by: Arnd Bergmann <arnd@arndb.de>
2018-09-11 18:47:23 +00:00
.compat_ioctl = compat_ptr_ioctl,
ceph: fix directory fsync fsync() on directory should flush dirty caps and wait for any uncommitted directory opertions to commit. But ceph_dir_fsync() only waits for uncommitted directory opertions. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-05-27 03:19:34 +00:00
.fsync = ceph_fsync,
ceph: support file lock on directory Link: http://tracker.ceph.com/issues/24028 Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2018-05-15 03:30:43 +00:00
.lock = ceph_lock,
.flock = ceph_flock,
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
};
ceph: provide seperate {inode,file}_operations for snapdir remove all unsupported operations from {inode,file}_operations. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-01-14 05:46:04 +00:00
const struct file_operations ceph_snapdir_fops = {
.iterate = ceph_readdir,
.llseek = ceph_dir_llseek,
.open = ceph_open,
.release = ceph_release,
};
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
const struct inode_operations ceph_dir_iops = {
.lookup = ceph_lookup,
.permission = ceph_permission,
.getattr = ceph_getattr,
.setattr = ceph_setattr,
.listxattr = ceph_listxattr,
fs: rename current get acl method The current way of setting and getting posix acls through the generic xattr interface is error prone and type unsafe. The vfs needs to interpret and fixup posix acls before storing or reporting it to userspace. Various hacks exist to make this work. The code is hard to understand and difficult to maintain in it's current form. Instead of making this work by hacking posix acls through xattr handlers we are building a dedicated posix acl api around the get and set inode operations. This removes a lot of hackiness and makes the codepaths easier to maintain. A lot of background can be found in [1]. The current inode operation for getting posix acls takes an inode argument but various filesystems (e.g., 9p, cifs, overlayfs) need access to the dentry. In contrast to the ->set_acl() inode operation we cannot simply extend ->get_acl() to take a dentry argument. The ->get_acl() inode operation is called from: acl_permission_check() -> check_acl() -> get_acl() which is part of generic_permission() which in turn is part of inode_permission(). Both generic_permission() and inode_permission() are called in the ->permission() handler of various filesystems (e.g., overlayfs). So simply passing a dentry argument to ->get_acl() would amount to also having to pass a dentry argument to ->permission(). We should avoid this unnecessary change. So instead of extending the existing inode operation rename it from ->get_acl() to ->get_inode_acl() and add a ->get_acl() method later that passes a dentry argument and which filesystems that need access to the dentry can implement instead of ->get_inode_acl(). Filesystems like cifs which allow setting and getting posix acls but not using them for permission checking during lookup can simply not implement ->get_inode_acl(). This is intended to be a non-functional change. Link: https://lore.kernel.org/all/20220801145520.1532837-1-brauner@kernel.org [1] Suggested-by/Inspired-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Christian Brauner (Microsoft) <brauner@kernel.org>
2022-09-22 15:17:00 +00:00
.get_inode_acl = ceph_get_acl,
ceph: fix posix ACL hooks The merge of commit 7221fe4c2ed7 ("ceph: add acl for cephfs") raced with upstream changes in the generic POSIX ACL code (eg commit 2aeccbe957d0 "fs: add generic xattr_acl handlers" and others). Some of the fallout was fixed in commit 4db658ea0ca ("ceph: Fix up after semantic merge conflict"), but it was incomplete: the set_acl inode_operation wasn't getting set, and the prototype needed to be adjusted a bit (it doesn't take a dentry anymore). Signed-off-by: Sage Weil <sage@inktank.com> Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-01-29 14:22:25 +00:00
.set_acl = ceph_set_acl,
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
.mknod = ceph_mknod,
.symlink = ceph_symlink,
.mkdir = ceph_mkdir,
.link = ceph_link,
.unlink = ceph_unlink,
.rmdir = ceph_unlink,
.rename = ceph_rename,
.create = ceph_create,
ceph: implement i_op->atomic_open() Add an ->atomic_open implementation which replaces the atomic lookup+open+create operation implemented via ->lookup and ->create operations. Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> CC: Sage Weil <sage@newdream.net> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2012-06-05 13:10:25 +00:00
.atomic_open = ceph_atomic_open,
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
};
ceph: provide seperate {inode,file}_operations for snapdir remove all unsupported operations from {inode,file}_operations. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-01-14 05:46:04 +00:00
const struct inode_operations ceph_snapdir_iops = {
.lookup = ceph_lookup,
.permission = ceph_permission,
.getattr = ceph_getattr,
.mkdir = ceph_mkdir,
.rmdir = ceph_unlink,
ceph: rename snapshot support Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-04-07 07:36:32 +00:00
.rename = ceph_rename,
ceph: provide seperate {inode,file}_operations for snapdir remove all unsupported operations from {inode,file}_operations. Signed-off-by: Yan, Zheng <zyan@redhat.com>
2015-01-14 05:46:04 +00:00
};
ceph: constify dentry_operations This makes checkpatch happy. Signed-off-by: Sage Weil <sage@newdream.net>
2010-08-03 17:25:30 +00:00
const struct dentry_operations ceph_dentry_ops = {
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
.d_revalidate = ceph_d_revalidate,
ceph: delete stale dentry when last reference is dropped introduce ceph_d_delete(), which checks if dentry has valid lease. Signed-off-by: "Yan, Zheng" <zyan@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-01-28 12:43:55 +00:00
.d_delete = ceph_d_delete,
ceph: rename dentry_release -> d_release, fix comment Just for consistency's sake. Fix obsolete comment too. Signed-off-by: Sage Weil <sage@newdream.net>
2011-03-15 21:57:41 +00:00
.d_release = ceph_d_release,
ceph: clear parent D_COMPLETE flag when on dentry prune When the VFS prunes a dentry from the cache, clear the D_COMPLETE flag on the parent dentry. Do this for the live and snapshotted namespaces. Do not bother for the .snap dir contents, since we do not cache that. Signed-off-by: Sage Weil <sage@newdream.net>
2011-03-15 22:53:40 +00:00
.d_prune = ceph_d_prune,
ceph: switch to use of ->d_init() Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2016-10-29 02:05:13 +00:00
.d_init = ceph_d_init,
ceph: directory operations Directory operations, including lookup, are defined here. We take advantage of lookup intents when possible. For the most part, we just need to build the proper requests for the metadata server(s) and pass things off to the mds_client. The results of most operations are normally incorporated into the client's cache when the reply is parsed by ceph_fill_trace(). However, if the MDS replies without a trace (e.g., when retrying an update after an MDS failure recovery), some operation-specific cleanup may be needed. We can validate cached dentries in two ways. A per-dentry lease may be issued by the MDS, or a per-directory cap may be issued that acts as a lease on the entire directory. In the latter case, a 'gen' value is used to determine which dentries belong to the currently leased directory contents. We normally prepopulate the dcache and icache with readdir results. This makes subsequent lookups and getattrs avoid any server interaction. It also lets us satisfy readdir operation by peeking at the dcache IFF we hold the per-directory cap/lease, previously performed a readdir, and haven't dropped any of the resulting dentries. Signed-off-by: Sage Weil <sage@newdream.net>
2009-10-06 18:31:08 +00:00
};
Reference in New Issue Copy Permalink