Currently we dropped freeze protection of aio writes just after IO was
submitted. Thus aio write could be in flight while the filesystem was
frozen and that could result in unexpected situation like aio completion
wanting to convert extent type on frozen filesystem. Testcase from
Dmitry triggering this is like:
for ((i=0;i<60;i++));do fsfreeze -f /mnt ;sleep 1;fsfreeze -u /mnt;done &
fio --bs=4k --ioengine=libaio --iodepth=128 --size=1g --direct=1 \
--runtime=60 --filename=/mnt/file --name=rand-write --rw=randwrite
Fix the problem by dropping freeze protection only once IO is completed
in aio_complete().
Reported-by: Dmitry Monakhov <dmonakhov@openvz.org>
Signed-off-by: Jan Kara <jack@suse.cz>
[hch: forward ported on top of various VFS and aio changes]
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Pass the ABI iocb structure to aio_setup_rw and let it handle the
non-vectored I/O case as well. With that and a new helper for the AIO
return value handling we can now define new aio_read and aio_write
helpers that implement reads and writes in a self-contained way without
duplicating too much code.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Otherwise we might dereference an already freed file and/or inode
when aio_complete is called before we return from the read_iter or
write_iter method.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Using a local variable we can prevent gcc from reloading
aio_ring_file->f_inode->i_mapping twice, eliminating 2x2 dependent
loads.
Signed-off-by: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
This ensures that do_mmap() won't implicitly make AIO memory mappings
executable if the READ_IMPLIES_EXEC personality flag is set. Such
behavior is problematic because the security_mmap_file LSM hook doesn't
catch this case, potentially permitting an attacker to bypass a W^X
policy enforced by SELinux.
I have tested the patch on my machine.
To test the behavior, compile and run this:
#define _GNU_SOURCE
#include <unistd.h>
#include <sys/personality.h>
#include <linux/aio_abi.h>
#include <err.h>
#include <stdlib.h>
#include <stdio.h>
#include <sys/syscall.h>
int main(void) {
personality(READ_IMPLIES_EXEC);
aio_context_t ctx = 0;
if (syscall(__NR_io_setup, 1, &ctx))
err(1, "io_setup");
char cmd[1000];
sprintf(cmd, "cat /proc/%d/maps | grep -F '/[aio]'",
(int)getpid());
system(cmd);
return 0;
}
In the output, "rw-s" is good, "rwxs" is bad.
Signed-off-by: Jann Horn <jann@thejh.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
aio_setup_ring waits for mmap_sem in writable mode. If the waiting task
gets killed by the oom killer it would block oom_reaper from
asynchronous address space reclaim and reduce the chances of timely OOM
resolving. Wait for the lock in the killable mode and return with EINTR
if the task got killed while waiting. This will also expedite the
return to the userspace and do_exit.
Signed-off-by: Michal Hocko <mhocko@suse.com>
Acked-by: Jeff Moyer <jmoyer@redhat.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Benamin LaHaise <bcrl@kvack.org>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
vma->vm_ops->mremap() looks more natural and clean in move_vma(), and this
way ->mremap() can have more users. Say, vdso.
While at it, s/aio_ring_remap/aio_ring_mremap/.
Note: this is the minimal change before ->mremap() finds another user in
file_operations; this method should have more arguments, and it can be
used to kill arch_remap().
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Acked-by: Pavel Emelyanov <xemul@parallels.com>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Benjamin LaHaise <bcrl@kvack.org>
Cc: Hugh Dickins <hughd@google.com>
Cc: Jeff Moyer <jmoyer@redhat.com>
Cc: Laurent Dufour <ldufour@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull third hunk of vfs changes from Al Viro:
"This contains the ->direct_IO() changes from Omar + saner
generic_write_checks() + dealing with fcntl()/{read,write}() races
(mirroring O_APPEND/O_DIRECT into iocb->ki_flags and instead of
repeatedly looking at ->f_flags, which can be changed by fcntl(2),
check ->ki_flags - which cannot) + infrastructure bits for dhowells'
d_inode annotations + Christophs switch of /dev/loop to
vfs_iter_write()"
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs: (30 commits)
block: loop: switch to VFS ITER_BVEC
configfs: Fix inconsistent use of file_inode() vs file->f_path.dentry->d_inode
VFS: Make pathwalk use d_is_reg() rather than S_ISREG()
VFS: Fix up debugfs to use d_is_dir() in place of S_ISDIR()
VFS: Combine inode checks with d_is_negative() and d_is_positive() in pathwalk
NFS: Don't use d_inode as a variable name
VFS: Impose ordering on accesses of d_inode and d_flags
VFS: Add owner-filesystem positive/negative dentry checks
nfs: generic_write_checks() shouldn't be done on swapout...
ocfs2: use __generic_file_write_iter()
mirror O_APPEND and O_DIRECT into iocb->ki_flags
switch generic_write_checks() to iocb and iter
ocfs2: move generic_write_checks() before the alignment checks
ocfs2_file_write_iter: stop messing with ppos
udf_file_write_iter: reorder and simplify
fuse: ->direct_IO() doesn't need generic_write_checks()
ext4_file_write_iter: move generic_write_checks() up
xfs_file_aio_write_checks: switch to iocb/iov_iter
generic_write_checks(): drop isblk argument
blkdev_write_iter: expand generic_file_checks() call in there
...
Pull block layer core bits from Jens Axboe:
"This is the core pull request for 4.1. Not a lot of stuff in here for
this round, mostly little fixes or optimizations. This pull request
contains:
- An optimization that speeds up queue runs on blk-mq, especially for
the case where there's a large difference between nr_cpu_ids and
the actual mapped software queues on a hardware queue. From Chong
Yuan.
- Honor node local allocations for requests on legacy devices. From
David Rientjes.
- Cleanup of blk_mq_rq_to_pdu() from me.
- exit_aio() fixup from me, greatly speeding up exiting multiple IO
contexts off exit_group(). For my particular test case, fio exit
took ~6 seconds. A typical case of both exposing RCU grace periods
to user space, and serializing exit of them.
- Make blk_mq_queue_enter() honor the gfp mask passed in, so we only
wait if __GFP_WAIT is set. From Keith Busch.
- blk-mq exports and two added helpers from Mike Snitzer, which will
be used by the dm-mq code.
- Cleanups of blk-mq queue init from Wei Fang and Xiaoguang Wang"
* 'for-4.1/core' of git://git.kernel.dk/linux-block:
blk-mq: reduce unnecessary software queue looping
aio: fix serial draining in exit_aio()
blk-mq: cleanup blk_mq_rq_to_pdu()
blk-mq: put blk_queue_rq_timeout together in blk_mq_init_queue()
block: remove redundant check about 'set->nr_hw_queues' in blk_mq_alloc_tag_set()
block: allocate request memory local to request queue
blk-mq: don't wait in blk_mq_queue_enter() if __GFP_WAIT isn't set
blk-mq: export blk_mq_run_hw_queues
blk-mq: add blk_mq_init_allocated_queue and export blk_mq_register_disk
Pull second vfs update from Al Viro:
"Now that net-next went in... Here's the next big chunk - killing
->aio_read() and ->aio_write().
There'll be one more pile today (direct_IO changes and
generic_write_checks() cleanups/fixes), but I'd prefer to keep that
one separate"
* 'for-linus-2' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs: (37 commits)
->aio_read and ->aio_write removed
pcm: another weird API abuse
infinibad: weird APIs switched to ->write_iter()
kill do_sync_read/do_sync_write
fuse: use iov_iter_get_pages() for non-splice path
fuse: switch to ->read_iter/->write_iter
switch drivers/char/mem.c to ->read_iter/->write_iter
make new_sync_{read,write}() static
coredump: accept any write method
switch /dev/loop to vfs_iter_write()
serial2002: switch to __vfs_read/__vfs_write
ashmem: use __vfs_read()
export __vfs_read()
autofs: switch to __vfs_write()
new helper: __vfs_write()
switch hugetlbfs to ->read_iter()
coda: switch to ->read_iter/->write_iter
ncpfs: switch to ->read_iter/->write_iter
net/9p: remove (now-)unused helpers
p9_client_attach(): set fid->uid correctly
...
exit_aio() currently serializes killing io contexts. Each context
killing ends up having to do percpu_ref_kill(), which in turns has
to wait for an RCU grace period. This can take a long time, depending
on the number of contexts. And there's no point in doing them serially,
when we could be waiting for all of them in one fell swoop.
This patches makes my fio thread offload test case exit 0.2s instead
of almost 6s.
Reviewed-by: Jeff Moyer <jmoyer@redhat.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
Pull vfs update from Al Viro:
"Part one:
- struct filename-related cleanups
- saner iov_iter_init() replacements (and switching the syscalls to
use of those)
- ntfs switch to ->write_iter() (Anton)
- aio cleanups and splitting iocb into common and async parts
(Christoph)
- assorted fixes (me, bfields, Andrew Elble)
There's a lot more, including the completion of switchover to
->{read,write}_iter(), d_inode/d_backing_inode annotations, f_flags
race fixes, etc, but that goes after #for-davem merge. David has
pulled it, and once it's in I'll send the next vfs pull request"
* 'for-linus-1' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs: (35 commits)
sg_start_req(): use import_iovec()
sg_start_req(): make sure that there's not too many elements in iovec
blk_rq_map_user(): use import_single_range()
sg_io(): use import_iovec()
process_vm_access: switch to {compat_,}import_iovec()
switch keyctl_instantiate_key_common() to iov_iter
switch {compat_,}do_readv_writev() to {compat_,}import_iovec()
aio_setup_vectored_rw(): switch to {compat_,}import_iovec()
vmsplice_to_user(): switch to import_iovec()
kill aio_setup_single_vector()
aio: simplify arguments of aio_setup_..._rw()
aio: lift iov_iter_init() into aio_setup_..._rw()
lift iov_iter into {compat_,}do_readv_writev()
NFS: fix BUG() crash in notify_change() with patch to chown_common()
dcache: return -ESTALE not -EBUSY on distributed fs race
NTFS: Version 2.1.32 - Update file write from aio_write to write_iter.
VFS: Add iov_iter_fault_in_multipages_readable()
drop bogus check in file_open_root()
switch security_inode_getattr() to struct path *
constify tomoyo_realpath_from_path()
...
We check if ->ki_pos is positive. However, by that point we have
already done rw_verify_area(), which would have rejected such
unless the file had been one of /dev/mem, /dev/kmem and /proc/kcore.
All of which do not have vectored rw methods, so we would've bailed
out even earlier.
This check had been introduced before rw_verify_area() had been added there
- in fact, it was a subset of checks done on sync paths by rw_verify_area()
(back then the /dev/mem exception didn't exist at all). The rest of checks
(mandatory locking, etc.) hadn't been added until later. Unfortunately,
by the time the call of rw_verify_area() got added, the /dev/mem exception
had already appeared, so it wasn't obvious that the older explicit check
downstream had become dead code. It *is* a dead code, though, since the few
files for which the exception applies do not have ->aio_{read,write}() or
->{read,write}_iter() and for them we won't reach that check anyway.
What's more, even if we ever introduce vectored methods for /dev/mem
and friends, they'll have to cope with negative positions anyway, since
readv(2) and writev(2) are using the same checks as read(2) and write(2) -
i.e. rw_verify_area().
Let's bury it.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Way, way back kiocb used to be picked from arrays, so ioctx_alloc()
checked for multiplication overflow when calculating the size of
such array. By the time fs/aio.c went into the tree (in 2002) they
were already allocated one-by-one by kmem_cache_alloc(), so that
check had already become pointless. Let's bury it...
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
We don't need req in either of those. We don't need nr_segs in caller.
We don't really need len in caller either - iov_iter_count(&iter) will do.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
the only non-trivial detail is that we do it before rw_verify_area(),
so we'd better cap the length ourselves in aio_setup_single_rw()
case (for vectored case rw_copy_check_uvector() will do that for us).
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
If we fail past the aio_setup_ring(), we need to destroy the
mapping. We don't need to care about anybody having found ctx,
or added requests to it, since the last failure exit is exactly
the failure to make ctx visible to lookups.
Reproducer (based on one by Joe Mario <jmario@redhat.com>):
void count(char *p)
{
char s[80];
printf("%s: ", p);
fflush(stdout);
sprintf(s, "/bin/cat /proc/%d/maps|/bin/fgrep -c '/[aio] (deleted)'", getpid());
system(s);
}
int main()
{
io_context_t *ctx;
int created, limit, i, destroyed;
FILE *f;
count("before");
if ((f = fopen("/proc/sys/fs/aio-max-nr", "r")) == NULL)
perror("opening aio-max-nr");
else if (fscanf(f, "%d", &limit) != 1)
fprintf(stderr, "can't parse aio-max-nr\n");
else if ((ctx = calloc(limit, sizeof(io_context_t))) == NULL)
perror("allocating aio_context_t array");
else {
for (i = 0, created = 0; i < limit; i++) {
if (io_setup(1000, ctx + created) == 0)
created++;
}
for (i = 0, destroyed = 0; i < created; i++)
if (io_destroy(ctx[i]) == 0)
destroyed++;
printf("created %d, failed %d, destroyed %d\n",
created, limit - created, destroyed);
count("after");
}
}
Found-by: Joe Mario <jmario@redhat.com>
Cc: stable@vger.kernel.org
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
teach ->mremap() method to return an error and have it fail for
aio mappings in process of being killed
Note that in case of ->mremap() failure we need to undo move_page_tables()
we'd already done; we could call ->mremap() first, but then the failure of
move_page_tables() would require undoing whatever _successful_ ->mremap()
has done, which would be a lot more headache in general.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Most callers in the kernel want to perform synchronous file I/O, but
still have to bloat the stack with a full struct kiocb. Split out
the parts needed in filesystem code from those in the aio code, and
only allocate those needed to pass down argument on the stack. The
aio code embedds the generic iocb in the one it allocates and can
easily get back to it by using container_of.
Also add a ->ki_complete method to struct kiocb, this is used to call
into the aio code and thus removes the dependency on aio for filesystems
impementing asynchronous operations. It will also allow other callers
to substitute their own completion callback.
We also add a new ->ki_flags field to work around the nasty layering
violation recently introduced in commit 5e33f6 ("usb: gadget: ffs: add
eventfd notification about ffs events").
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
The AIO interface is fairly complex because it tries to allow
filesystems to always work async and then wakeup a synchronous
caller through aio_complete. It turns out that basically no one
was doing this to avoid the complexity and context switches,
and we've already fixed up the remaining users and can now
get rid of this case.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
There is no need to pass the total request length in the kiocb, as
we already get passed in through the iov_iter argument.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Have defined pr_fmt as below in fs/aio.c, so remove duplicate
function name in pr_debug message.
#define pr_fmt(fmt) "%s: " fmt, __func__
Signed-off-by: Kinglong Mee <kinglongmee@gmail.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Pull backing device changes from Jens Axboe:
"This contains a cleanup of how the backing device is handled, in
preparation for a rework of the life time rules. In this part, the
most important change is to split the unrelated nommu mmap flags from
it, but also removing a backing_dev_info pointer from the
address_space (and inode), and a cleanup of other various minor bits.
Christoph did all the work here, I just fixed an oops with pages that
have a swap backing. Arnd fixed a missing export, and Oleg killed the
lustre backing_dev_info from staging. Last patch was from Al,
unexporting parts that are now no longer needed outside"
* 'for-3.20/bdi' of git://git.kernel.dk/linux-block:
Make super_blocks and sb_lock static
mtd: export new mtd_mmap_capabilities
fs: make inode_to_bdi() handle NULL inode
staging/lustre/llite: get rid of backing_dev_info
fs: remove default_backing_dev_info
fs: don't reassign dirty inodes to default_backing_dev_info
nfs: don't call bdi_unregister
ceph: remove call to bdi_unregister
fs: remove mapping->backing_dev_info
fs: export inode_to_bdi and use it in favor of mapping->backing_dev_info
nilfs2: set up s_bdi like the generic mount_bdev code
block_dev: get bdev inode bdi directly from the block device
block_dev: only write bdev inode on close
fs: introduce f_op->mmap_capabilities for nommu mmap support
fs: kill BDI_CAP_SWAP_BACKED
fs: deduplicate noop_backing_dev_info
Under CONFIG_DEBUG_ATOMIC_SLEEP=y, aio_read_event_ring() will throw
warnings like the following due to being called from wait_event
context:
WARNING: CPU: 0 PID: 16006 at kernel/sched/core.c:7300 __might_sleep+0x7f/0x90()
do not call blocking ops when !TASK_RUNNING; state=1 set at [<ffffffff810d85a3>] prepare_to_wait_event+0x63/0x110
Modules linked in:
CPU: 0 PID: 16006 Comm: aio-dio-fcntl-r Not tainted 3.19.0-rc6-dgc+ #705
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011
ffffffff821c0372 ffff88003c117cd8 ffffffff81daf2bd 000000000000d8d8
ffff88003c117d28 ffff88003c117d18 ffffffff8109beda ffff88003c117cf8
ffffffff821c115e 0000000000000061 0000000000000000 00007ffffe4aa300
Call Trace:
[<ffffffff81daf2bd>] dump_stack+0x4c/0x65
[<ffffffff8109beda>] warn_slowpath_common+0x8a/0xc0
[<ffffffff8109bf56>] warn_slowpath_fmt+0x46/0x50
[<ffffffff810d85a3>] ? prepare_to_wait_event+0x63/0x110
[<ffffffff810d85a3>] ? prepare_to_wait_event+0x63/0x110
[<ffffffff810bdfcf>] __might_sleep+0x7f/0x90
[<ffffffff81db8344>] mutex_lock+0x24/0x45
[<ffffffff81216b7c>] aio_read_events+0x4c/0x290
[<ffffffff81216fac>] read_events+0x1ec/0x220
[<ffffffff810d8650>] ? prepare_to_wait_event+0x110/0x110
[<ffffffff810fdb10>] ? hrtimer_get_res+0x50/0x50
[<ffffffff8121899d>] SyS_io_getevents+0x4d/0xb0
[<ffffffff81dba5a9>] system_call_fastpath+0x12/0x17
---[ end trace bde69eaf655a4fea ]---
There is not actually a bug here, so annotate the code to tell the
debug logic that everything is just fine and not to fire a false
positive.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
Now that we never use the backing_dev_info pointer in struct address_space
we can simply remove it and save 4 to 8 bytes in every inode.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
Reviewed-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Signed-off-by: Jens Axboe <axboe@fb.com>
Since "BDI: Provide backing device capability information [try #3]" the
backing_dev_info structure also provides flags for the kind of mmap
operation available in a nommu environment, which is entirely unrelated
to it's original purpose.
Introduce a new nommu-only file operation to provide this information to
the nommu mmap code instead. Splitting this from the backing_dev_info
structure allows to remove lots of backing_dev_info instance that aren't
otherwise needed, and entirely gets rid of the concept of providing a
backing_dev_info for a character device. It also removes the need for
the mtd_inodefs filesystem.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Tejun Heo <tj@kernel.org>
Acked-by: Brian Norris <computersforpeace@gmail.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
In this case, it is basically a polling. Let's not involve timer at all
because that would hurt performance for application event loops.
In an arbitrary test I've done, io_getevents syscall elapsed time
reduces from 50000+ nanoseconds to a few hundereds.
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
There are actually two issues this patch addresses. Let me start with
the one I tried to solve in the beginning.
So, in the checkpoint-restore project (criu) we try to dump tasks'
state and restore one back exactly as it was. One of the tasks' state
bits is rings set up with io_setup() call. There's (almost) no problems
in dumping them, there's a problem restoring them -- if I dump a task
with aio ring originally mapped at address A, I want to restore one
back at exactly the same address A. Unfortunately, the io_setup() does
not allow for that -- it mmaps the ring at whatever place mm finds
appropriate (it calls do_mmap_pgoff() with zero address and without
the MAP_FIXED flag).
To make restore possible I'm going to mremap() the freshly created ring
into the address A (under which it was seen before dump). The problem is
that the ring's virtual address is passed back to the user-space as the
context ID and this ID is then used as search key by all the other io_foo()
calls. Reworking this ID to be just some integer doesn't seem to work, as
this value is already used by libaio as a pointer using which this library
accesses memory for aio meta-data.
So, to make restore work we need to make sure that
a) ring is mapped at desired virtual address
b) kioctx->user_id matches this value
Having said that, the patch makes mremap() on aio region update the
kioctx's user_id and mmap_base values.
Here appears the 2nd issue I mentioned in the beginning of this mail.
If (regardless of the C/R dances I do) someone creates an io context
with io_setup(), then mremap()-s the ring and then destroys the context,
the kill_ioctx() routine will call munmap() on wrong (old) address.
This will result in a) aio ring remaining in memory and b) some other
vma get unexpectedly unmapped.
What do you think?
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
Acked-by: Dmitry Monakhov <dmonakhov@openvz.org>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
Pull aio fix from Ben LaHaise:
"Dirty page accounting fix for aio"
* git://git.kvack.org/~bcrl/aio-fixes:
aio: fix uncorrent dirty pages accouting when truncating AIO ring buffer
https://bugzilla.kernel.org/show_bug.cgi?id=86831
Markus reported that when shutting down mysqld (with AIO support,
on a ext3 formatted Harddrive) leads to a negative number of dirty pages
(underrun to the counter). The negative number results in a drastic reduction
of the write performance because the page cache is not used, because the kernel
thinks it is still 2 ^ 32 dirty pages open.
Add a warn trace in __dec_zone_state will catch this easily:
static inline void __dec_zone_state(struct zone *zone, enum
zone_stat_item item)
{
atomic_long_dec(&zone->vm_stat[item]);
+ WARN_ON_ONCE(item == NR_FILE_DIRTY &&
atomic_long_read(&zone->vm_stat[item]) < 0);
atomic_long_dec(&vm_stat[item]);
}
[ 21.341632] ------------[ cut here ]------------
[ 21.346294] WARNING: CPU: 0 PID: 309 at include/linux/vmstat.h:242
cancel_dirty_page+0x164/0x224()
[ 21.355296] Modules linked in: wutbox_cp sata_mv
[ 21.359968] CPU: 0 PID: 309 Comm: kworker/0:1 Not tainted 3.14.21-WuT #80
[ 21.366793] Workqueue: events free_ioctx
[ 21.370760] [<c0016a64>] (unwind_backtrace) from [<c0012f88>]
(show_stack+0x20/0x24)
[ 21.378562] [<c0012f88>] (show_stack) from [<c03f8ccc>]
(dump_stack+0x24/0x28)
[ 21.385840] [<c03f8ccc>] (dump_stack) from [<c0023ae4>]
(warn_slowpath_common+0x84/0x9c)
[ 21.393976] [<c0023ae4>] (warn_slowpath_common) from [<c0023bb8>]
(warn_slowpath_null+0x2c/0x34)
[ 21.402800] [<c0023bb8>] (warn_slowpath_null) from [<c00c0688>]
(cancel_dirty_page+0x164/0x224)
[ 21.411524] [<c00c0688>] (cancel_dirty_page) from [<c00c080c>]
(truncate_inode_page+0x8c/0x158)
[ 21.420272] [<c00c080c>] (truncate_inode_page) from [<c00c0a94>]
(truncate_inode_pages_range+0x11c/0x53c)
[ 21.429890] [<c00c0a94>] (truncate_inode_pages_range) from
[<c00c0f6c>] (truncate_pagecache+0x88/0xac)
[ 21.439252] [<c00c0f6c>] (truncate_pagecache) from [<c00c0fec>]
(truncate_setsize+0x5c/0x74)
[ 21.447731] [<c00c0fec>] (truncate_setsize) from [<c013b3a8>]
(put_aio_ring_file.isra.14+0x34/0x90)
[ 21.456826] [<c013b3a8>] (put_aio_ring_file.isra.14) from
[<c013b424>] (aio_free_ring+0x20/0xcc)
[ 21.465660] [<c013b424>] (aio_free_ring) from [<c013b4f4>]
(free_ioctx+0x24/0x44)
[ 21.473190] [<c013b4f4>] (free_ioctx) from [<c003d8d8>]
(process_one_work+0x134/0x47c)
[ 21.481132] [<c003d8d8>] (process_one_work) from [<c003e988>]
(worker_thread+0x130/0x414)
[ 21.489350] [<c003e988>] (worker_thread) from [<c00448ac>]
(kthread+0xd4/0xec)
[ 21.496621] [<c00448ac>] (kthread) from [<c000ec18>]
(ret_from_fork+0x14/0x20)
[ 21.503884] ---[ end trace 79c4bf42c038c9a1 ]---
The cause is that we set the aio ring file pages as *DIRTY* via SetPageDirty
(bypasses the VFS dirty pages increment) when init, and aio fs uses
*default_backing_dev_info* as the backing dev, which does not disable
the dirty pages accounting capability.
So truncating aio ring file will contribute to accounting dirty pages (VFS
dirty pages decrement), then error occurs.
The original goal is keeping these pages in memory (can not be reclaimed
or swapped) in life-time via marking it dirty. But thinking more, we have
already pinned pages via elevating the page's refcount, which can already
achieve the goal, so the SetPageDirty seems unnecessary.
In order to fix the issue, using the __set_page_dirty_no_writeback instead
of the nop .set_page_dirty, and dropped the SetPageDirty (don't manually
set the dirty flags, don't disable set_page_dirty(), rely on default behaviour).
With the above change, the dirty pages accounting can work well. But as we
known, aio fs is an anonymous one, which should never cause any real write-back,
we can ignore the dirty pages (write back) accounting by disabling the dirty
pages (write back) accounting capability. So we introduce an aio private
backing dev info (disabled the ACCT_DIRTY/WRITEBACK/ACCT_WB capabilities) to
replace the default one.
Reported-by: Markus Königshaus <m.koenigshaus@wut.de>
Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com>
Cc: stable <stable@vger.kernel.org>
Acked-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
With the recent addition of percpu_ref_reinit(), percpu_ref now can be
used as a persistent switch which can be turned on and off repeatedly
where turning off maps to killing the ref and waiting for it to drain;
however, there currently isn't a way to initialize a percpu_ref in its
off (killed and drained) state, which can be inconvenient for certain
persistent switch use cases.
Similarly, percpu_ref_switch_to_atomic/percpu() allow dynamic
selection of operation mode; however, currently a newly initialized
percpu_ref is always in percpu mode making it impossible to avoid the
latency overhead of switching to atomic mode.
This patch adds @flags to percpu_ref_init() and implements the
following flags.
* PERCPU_REF_INIT_ATOMIC : start ref in atomic mode
* PERCPU_REF_INIT_DEAD : start ref killed and drained
These flags should be able to serve the above two use cases.
v2: target_core_tpg.c conversion was missing. Fixed.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Kent Overstreet <kmo@daterainc.com>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Johannes Weiner <hannes@cmpxchg.org>
This is to receive 0a30288da1 ("blk-mq, percpu_ref: implement a
kludge for SCSI blk-mq stall during probe") which implements
__percpu_ref_kill_expedited() to work around SCSI blk-mq stall. The
commit reverted and patches to implement proper fix will be added.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Kent Overstreet <kmo@daterainc.com>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Christoph Hellwig <hch@lst.de>
Percpu allocator now supports allocation mask. Add @gfp to
percpu_ref_init() so that !GFP_KERNEL allocation masks can be used
with percpu_refs too.
This patch doesn't make any functional difference.
v2: blk-mq conversion was missing. Updated.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Kent Overstreet <koverstreet@google.com>
Cc: Benjamin LaHaise <bcrl@kvack.org>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Nicholas A. Bellinger <nab@linux-iscsi.org>
Cc: Jens Axboe <axboe@kernel.dk>
It seems that exit_aio() also needs to wait for all iocbs to complete (like
io_destroy), but we missed the wait step in current implemention, so fix
it in the same way as we did in io_destroy.
Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
Cc: stable@vger.kernel.org
We ran into a case on ppc64 running mariadb where io_getevents would
return zeroed out I/O events. After adding instrumentation, it became
clear that there was some missing synchronization between reading the
tail pointer and the events themselves. This small patch fixes the
problem in testing.
Thanks to Zach for helping to look into this, and suggesting the fix.
Signed-off-by: Jeff Moyer <jmoyer@redhat.com>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
Cc: stable@vger.kernel.org
As reported by Dan Aloni, commit f8567a3845 ("aio: fix aio request
leak when events are reaped by userspace") introduces a regression when
user code attempts to perform io_submit() with more events than are
available in the ring buffer. Reverting that commit would reintroduce a
regression when user space event reaping is used.
Fixing this bug is a bit more involved than the previous attempts to fix
this regression. Since we do not have a single point at which we can
count events as being reaped by user space and io_getevents(), we have
to track event completion by looking at the number of events left in the
event ring. So long as there are as many events in the ring buffer as
there have been completion events generate, we cannot call
put_reqs_available(). The code to check for this is now placed in
refill_reqs_available().
A test program from Dan and modified by me for verifying this bug is available
at http://www.kvack.org/~bcrl/20140824-aio_bug.c .
Reported-by: Dan Aloni <dan@kernelim.com>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
Acked-by: Dan Aloni <dan@kernelim.com>
Cc: Kent Overstreet <kmo@daterainc.com>
Cc: Mateusz Guzik <mguzik@redhat.com>
Cc: Petr Matousek <pmatouse@redhat.com>
Cc: stable@vger.kernel.org # v3.16 and anything that f8567a3845 was backported to
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull aio updates from Ben LaHaise.
* git://git.kvack.org/~bcrl/aio-next:
aio: use iovec array rather than the single one
aio: fix some comments
aio: use the macro rather than the inline magic number
aio: remove the needless registration of ring file's private_data
aio: remove no longer needed preempt_disable()
aio: kill the misleading rcu read locks in ioctx_add_table() and kill_ioctx()
aio: change exit_aio() to load mm->ioctx_table once and avoid rcu_read_lock()
Pull percpu updates from Tejun Heo:
- Major reorganization of percpu header files which I think makes
things a lot more readable and logical than before.
- percpu-refcount is updated so that it requires explicit destruction
and can be reinitialized if necessary. This was pulled into the
block tree to replace the custom percpu refcnting implemented in
blk-mq.
- In the process, percpu and percpu-refcount got cleaned up a bit
* 'for-3.17' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/percpu: (21 commits)
percpu-refcount: implement percpu_ref_reinit() and percpu_ref_is_zero()
percpu-refcount: require percpu_ref to be exited explicitly
percpu-refcount: use unsigned long for pcpu_count pointer
percpu-refcount: add helpers for ->percpu_count accesses
percpu-refcount: one bit is enough for REF_STATUS
percpu-refcount, aio: use percpu_ref_cancel_init() in ioctx_alloc()
workqueue: stronger test in process_one_work()
workqueue: clear POOL_DISASSOCIATED in rebind_workers()
percpu: Use ALIGN macro instead of hand coding alignment calculation
percpu: invoke __verify_pcpu_ptr() from the generic part of accessors and operations
percpu: preffity percpu header files
percpu: use raw_cpu_*() to define __this_cpu_*()
percpu: reorder macros in percpu header files
percpu: move {raw|this}_cpu_*() definitions to include/linux/percpu-defs.h
percpu: move generic {raw|this}_cpu_*_N() definitions to include/asm-generic/percpu.h
percpu: only allow sized arch overrides for {raw|this}_cpu_*() ops
percpu: reorganize include/linux/percpu-defs.h
percpu: move accessors from include/linux/percpu.h to percpu-defs.h
percpu: include/asm-generic/percpu.h should contain only arch-overridable parts
percpu: introduce arch_raw_cpu_ptr()
...
Previously, we only offer a single iovec to handle all the read/write cases, so
the PREADV/PWRITEV request always need to alloc more iovec buffer when copying
user vectors.
If we use a tmp iovec array rather than the single one, some small PREADV/PWRITEV
workloads(vector size small than the tmp buffer) will not need to alloc more
iovec buffer when copying user vectors.
Reviewed-by: Jeff Moyer <jmoyer@redhat.com>
Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
The function comments of aio_run_iocb and aio_read_events are out of date, so
fix them here.
Reviewed-by: Jeff Moyer <jmoyer@redhat.com>
Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
Replace the inline magic number with the ready-made macro(AIO_RING_MAGIC),
just clean up.
Reviewed-by: Jeff Moyer <jmoyer@redhat.com>
Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
Remove the registration of ring file's private_data, we do not use
it.
Reviewed-by: Jeff Moyer <jmoyer@redhat.com>
Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
Based on feedback from Jens Axboe on 263782c1c9,
clean up get/put_reqs_available() to remove the no longer needed preempt_disable()
and preempt_enable() pair.
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
Cc: Jens Axboe <axboe@kernel.dk>
As of commit f8567a3845 it is now possible to
have put_reqs_available() called from irq context. While put_reqs_available()
is per cpu, it did not protect itself from interrupts on the same CPU. This
lead to aio_complete() corrupting the available io requests count when run
under a heavy O_DIRECT workloads as reported by Robert Elliott. Fix this by
disabling irq updates around the per cpu batch updates of reqs_available.
Many thanks to Robert and folks for testing and tracking this down.
Reported-by: Robert Elliot <Elliott@hp.com>
Tested-by: Robert Elliot <Elliott@hp.com>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
Cc: Jens Axboe <axboe@kernel.dk>, Christoph Hellwig <hch@infradead.org>
Cc: stable@vger.kenel.org
Currently, a percpu_ref undoes percpu_ref_init() automatically by
freeing the allocated percpu area when the percpu_ref is killed.
While seemingly convenient, this has the following niggles.
* It's impossible to re-init a released reference counter without
going through re-allocation.
* In the similar vein, it's impossible to initialize a percpu_ref
count with static percpu variables.
* We need and have an explicit destructor anyway for failure paths -
percpu_ref_cancel_init().
This patch removes the automatic percpu counter freeing in
percpu_ref_kill_rcu() and repurposes percpu_ref_cancel_init() into a
generic destructor now named percpu_ref_exit(). percpu_ref_destroy()
is considered but it gets confusing with percpu_ref_kill() while
"exit" clearly indicates that it's the counterpart of
percpu_ref_init().
All percpu_ref_cancel_init() users are updated to invoke
percpu_ref_exit() instead and explicit percpu_ref_exit() calls are
added to the destruction path of all percpu_ref users.
Signed-off-by: Tejun Heo <tj@kernel.org>
Acked-by: Benjamin LaHaise <bcrl@kvack.org>
Cc: Kent Overstreet <kmo@daterainc.com>
Cc: Christoph Lameter <cl@linux-foundation.org>
Cc: Benjamin LaHaise <bcrl@kvack.org>
Cc: Nicholas A. Bellinger <nab@linux-iscsi.org>
Cc: Li Zefan <lizefan@huawei.com>
ioctx_alloc() reaches inside percpu_ref and directly frees
->pcpu_count in its failure path, which is quite gross. percpu_ref
has been providing a proper interface to do this,
percpu_ref_cancel_init(), for quite some time now. Let's use that
instead.
This patch doesn't introduce any behavior changes.
Signed-off-by: Tejun Heo <tj@kernel.org>
Acked-by: Benjamin LaHaise <bcrl@kvack.org>
Cc: Kent Overstreet <kmo@daterainc.com>
ioctx_add_table() is the writer, it does not need rcu_read_lock() to
protect ->ioctx_table. It relies on mm->ioctx_lock and rcu locks just
add the confusion.
And it doesn't need rcu_dereference() by the same reason, it must see
any updates previously done under the same ->ioctx_lock. We could use
rcu_dereference_protected() but the patch uses rcu_dereference_raw(),
the function is simple enough.
The same for kill_ioctx(), although it does not update the pointer.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
On 04/30, Benjamin LaHaise wrote:
>
> > - ctx->mmap_size = 0;
> > -
> > - kill_ioctx(mm, ctx, NULL);
> > + if (ctx) {
> > + ctx->mmap_size = 0;
> > + kill_ioctx(mm, ctx, NULL);
> > + }
>
> Rather than indenting and moving the two lines changing mmap_size and the
> kill_ioctx() call, why not just do "if (!ctx) ... continue;"? That reduces
> the number of lines changed and avoid excessive indentation.
OK. To me the code looks better/simpler with "if (ctx)", but this is subjective
of course, I won't argue.
The patch still removes the empty line between mmap_size = 0 and kill_ioctx(),
we reset mmap_size only for kill_ioctx(). But feel free to remove this change.
-------------------------------------------------------------------------------
Subject: [PATCH v3 1/2] aio: change exit_aio() to load mm->ioctx_table once and avoid rcu_read_lock()
1. We can read ->ioctx_table only once and we do not read rcu_read_lock()
or even rcu_dereference().
This mm has no users, nobody else can play with ->ioctx_table. Otherwise
the code is buggy anyway, if we need rcu_read_lock() in a loop because
->ioctx_table can be updated then kfree(table) is obviously wrong.
2. Update the comment. "exit_mmap(mm) is coming" is the good reason to avoid
munmap(), but another reason is that we simply can't do vm_munmap() unless
current->mm == mm and this is not true in general, the caller is mmput().
3. We do not really need to nullify mm->ioctx_table before return, probably
the current code does this to catch the potential problems. But in this
case RCU_INIT_POINTER(NULL) looks better.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
A kernel memory disclosure was introduced in aio_read_events_ring() in v3.10
by commit a31ad380be. The changes made to
aio_read_events_ring() failed to correctly limit the index into
ctx->ring_pages[], allowing an attacked to cause the subsequent kmap() of
an arbitrary page with a copy_to_user() to copy the contents into userspace.
This vulnerability has been assigned CVE-2014-0206. Thanks to Mateusz and
Petr for disclosing this issue.
This patch applies to v3.12+. A separate backport is needed for 3.10/3.11.
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
Cc: Mateusz Guzik <mguzik@redhat.com>
Cc: Petr Matousek <pmatouse@redhat.com>
Cc: Kent Overstreet <kmo@daterainc.com>
Cc: Jeff Moyer <jmoyer@redhat.com>
Cc: stable@vger.kernel.org
The aio cleanups and optimizations by kmo that were merged into the 3.10
tree added a regression for userspace event reaping. Specifically, the
reference counts are not decremented if the event is reaped in userspace,
leading to the application being unable to submit further aio requests.
This patch applies to 3.12+. A separate backport is required for 3.10/3.11.
This issue was uncovered as part of CVE-2014-0206.
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
Cc: stable@vger.kernel.org
Cc: Kent Overstreet <kmo@daterainc.com>
Cc: Mateusz Guzik <mguzik@redhat.com>
Cc: Petr Matousek <pmatouse@redhat.com>
Pull aio fix and cleanups from Ben LaHaise:
"This consists of a couple of code cleanups plus a minor bug fix"
* git://git.kvack.org/~bcrl/aio-next:
aio: cleanup: flatten kill_ioctx()
aio: report error from io_destroy() when threads race in io_destroy()
fs/aio.c: Remove ctx parameter in kiocb_cancel
Beginning to introduce those. Just the callers for now, and it's
clumsier than it'll eventually become; once we finish converting
aio_read and aio_write instances, the things will get nicer.
For now, these guys are in parallel to ->aio_read() and ->aio_write();
they take iocb and iov_iter, with everything in iov_iter already
validated. File offset is passed in iocb->ki_pos, iov/nr_segs -
in iov_iter.
Main concerns in that series are stack footprint and ability to
split the damn thing cleanly.
[fix from Peter Ujfalusi <peter.ujfalusi@ti.com> folded]
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
iovec should be reclaimed whenever caller of rw_copy_check_uvector() returns,
but it doesn't hold when failure happens right after aio_setup_vectored_rw().
Fix that in a such way to avoid hairy goto.
Signed-off-by: Leon Yu <chianglungyu@gmail.com>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
Cc: stable@vger.kernel.org
As reported by Anatol Pomozov, io_destroy() fails to report an error when
it loses the race to destroy a given ioctx. Since there is a difference in
behaviour between the thread that wins the race (which blocks on outstanding
io requests) versus lthe thread that loses (which returns immediately), wire
up a return code from kill_ioctx() to the io_destroy() syscall.
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
Cc: Anatol Pomozov <anatol.pomozov@gmail.com>
ctx is no longer used in kiocb_cancel since
57282d8fd7 ("aio: Kill ki_users")
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Fabian Frederick <fabf@skynet.be>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
deletes aio context and all resources related to. It makes sense that
no IO operations connected to the context should be running after the context
is destroyed. As we removed io_context we have no chance to
get requests status or call io_getevents().
man page for io_destroy says that this function may block until
all context's requests are completed. Before kernel 3.11 io_destroy()
blocked indeed, but since aio refactoring in 3.11 it is not true anymore.
Here is a pseudo-code that shows a testcase for a race condition discovered
in 3.11:
initialize io_context
io_submit(read to buffer)
io_destroy()
// context is destroyed so we can free the resources
free(buffers);
// if the buffer is allocated by some other user he'll be surprised
// to learn that the buffer still filled by an outstanding operation
// from the destroyed io_context
The fix is straight-forward - add a completion struct and wait on it
in io_destroy, complete() should be called when number of in-fligh requests
reaches zero.
If two or more io_destroy() called for the same context simultaneously then
only the first one waits for IO completion, other calls behaviour is undefined.
Tested: ran http://pastebin.com/LrPsQ4RL testcase for several hours and
do not see the race condition anymore.
Signed-off-by: Anatol Pomozov <anatol.pomozov@gmail.com>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
As reported by Tang Chen, Gu Zheng and Yasuaki Isimatsu, the following issues
exist in the aio ring page migration support.
As a result, for example, we have the following problem:
thread 1 | thread 2
|
aio_migratepage() |
|-> take ctx->completion_lock |
|-> migrate_page_copy(new, old) |
| *NOW*, ctx->ring_pages[idx] == old |
|
| *NOW*, ctx->ring_pages[idx] == old
| aio_read_events_ring()
| |-> ring = kmap_atomic(ctx->ring_pages[0])
| |-> ring->head = head; *HERE, write to the old ring page*
| |-> kunmap_atomic(ring);
|
|-> ctx->ring_pages[idx] = new |
| *BUT NOW*, the content of |
| ring_pages[idx] is old. |
|-> release ctx->completion_lock |
As above, the new ring page will not be updated.
Fix this issue, as well as prevent races in aio_ring_setup() by holding
the ring_lock mutex during kioctx setup and page migration. This avoids
the overhead of taking another spinlock in aio_read_events_ring() as Tang's
and Gu's original fix did, pushing the overhead into the migration code.
Note that to handle the nesting of ring_lock inside of mmap_sem, the
migratepage operation uses mutex_trylock(). Page migration is not a 100%
critical operation in this case, so the ocassional failure can be
tolerated. This issue was reported by Sasha Levin.
Based on feedback from Linus, avoid the extra taking of ctx->completion_lock.
Instead, make page migration fully serialised by mapping->private_lock, and
have aio_free_ring() simply disconnect the kioctx from the mapping by calling
put_aio_ring_file() before touching ctx->ring_pages[]. This simplifies the
error handling logic in aio_migratepage(), and should improve robustness.
v4: always do mutex_unlock() in cases when kioctx setup fails.
Reported-by: Yasuaki Ishimatsu <isimatu.yasuaki@jp.fujitsu.com>
Reported-by: Sasha Levin <sasha.levin@oracle.com>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
Cc: Tang Chen <tangchen@cn.fujitsu.com>
Cc: Gu Zheng <guz.fnst@cn.fujitsu.com>
Cc: stable@vger.kernel.org
Pull AIO leak fixes from Ben LaHaise:
"I've put these two patches plus Linus's change through a round of
tests, and it passes millions of iterations of the aio numa
migratepage test, as well as a number of repetitions of a few simple
read and write tests.
The first patch fixes the memory leak Kent introduced, while the
second patch makes aio_migratepage() much more paranoid and robust"
* git://git.kvack.org/~bcrl/aio-next:
aio/migratepages: make aio migrate pages sane
aio: fix kioctx leak introduced by "aio: Fix a trinity splat"
Since commit 36bc08cc01 ("fs/aio: Add support to aio ring pages
migration") the aio ring setup code has used a special per-ring backing
inode for the page allocations, rather than just using random anonymous
pages.
However, rather than remembering the pages as it allocated them, it
would allocate the pages, insert them into the file mapping (dirty, so
that they couldn't be free'd), and then forget about them. And then to
look them up again, it would mmap the mapping, and then use
"get_user_pages()" to get back an array of the pages we just created.
Now, not only is that incredibly inefficient, it also leaked all the
pages if the mmap failed (which could happen due to excessive number of
mappings, for example).
So clean it all up, making it much more straightforward. Also remove
some left-overs of the previous (broken) mm_populate() usage that was
removed in commit d6c355c7da ("aio: fix race in ring buffer page
lookup introduced by page migration support") but left the pointless and
now misleading MAP_POPULATE flag around.
Tested-and-acked-by: Benjamin LaHaise <bcrl@kvack.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The arbitrary restriction on page counts offered by the core
migrate_page_move_mapping() code results in rather suspicious looking
fiddling with page reference counts in the aio_migratepage() operation.
To fix this, make migrate_page_move_mapping() take an extra_count parameter
that allows aio to tell the code about its own reference count on the page
being migrated.
While cleaning up aio_migratepage(), make it validate that the old page
being passed in is actually what aio_migratepage() expects to prevent
misbehaviour in the case of races.
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
e34ecee2ae reworked the percpu reference
counting to correct a bug trinity found. Unfortunately, the change lead
to kioctxes being leaked because there was no final reference count to
put. Add that reference count back in to fix things.
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
Cc: stable@vger.kernel.org
Pull aio fix from Benjamin LaHaise:
"AIO fix from Gu Zheng that fixes a GPF that Dave Jones uncovered with
trinity"
* git://git.kvack.org/~bcrl/aio-next:
aio: clean up aio ring in the fail path
Clean up the aio ring file in the fail path of aio_setup_ring
and ioctx_alloc. And maybe it can fix the GPF issue reported by
Dave Jones:
https://lkml.org/lkml/2013/11/25/898
Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
Pull aio fixes from Benjamin LaHaise.
* git://git.kvack.org/~bcrl/aio-next:
aio: nullify aio->ring_pages after freeing it
aio: prevent double free in ioctx_alloc
aio: Fix a trinity splat
After freeing ring_pages we leave it as is causing a dangling pointer. This
has already caused an issue so to help catching any issues in the future
NULL it out.
Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
ioctx_alloc() calls aio_setup_ring() to allocate a ring. If aio_setup_ring()
fails to do so it would call aio_free_ring() before returning, but
ioctx_alloc() would call aio_free_ring() again causing a double free of
the ring.
This is easily reproducible from userspace.
Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
alloc_anon_inode() returns an ERR_PTR(), it doesn't return NULL.
Fixes: 71ad7490c1 ('rework aio migrate pages to use aio fs')
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Don't abuse anon_inodes.c to host private files needed by aio;
we can bloody well declare a mini-fs of our own instead of
patching up what anon_inodes can create for us.
Tested-by: Benjamin LaHaise <bcrl@kvack.org>
Acked-by: Benjamin LaHaise <bcrl@kvack.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
aio kiocb refcounting was broken - it was relying on keeping track of
the number of available ring buffer entries, which it needs to do
anyways; then at shutdown time it'd wait for completions to be delivered
until the # of available ring buffer entries equalled what it was
initialized to.
Problem with that is that the ring buffer is mapped writable into
userspace, so userspace could futz with the head and tail pointers to
cause the kernel to see extra completions, and cause free_ioctx() to
return while there were still outstanding kiocbs. Which would be bad.
Fix is just to directly refcount the kiocbs - which is more
straightforward, and with the new percpu refcounting code doesn't cost
us any cacheline bouncing which was the whole point of the original
scheme.
Also clean up ioctx_alloc()'s error path and fix a bug where it wasn't
subtracting from aio_nr if ioctx_add_table() failed.
Signed-off-by: Kent Overstreet <kmo@daterainc.com>
Dmitry Vyukov managed to trigger a case where aio_migratepage can cause a
use-after-free during teardown of the aio ring buffer's mapping. This turns
out to be caused by access to the ioctx's ring_pages via the migratepage
operation which was not being protected by any locks during ioctx freeing.
Use the address_space's private_lock to protect use and updates of the mapping's
private_data, and make ioctx teardown unlink the ioctx from the address space.
Reported-by: Dmitry Vyukov <dvyukov@google.com>
Tested-by: Dmitry Vyukov <dvyukov@google.com>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
Prior to the introduction of page migration support in "fs/aio: Add support
to aio ring pages migration" / 36bc08cc01,
mapping of the ring buffer pages was done via get_user_pages() while
retaining mmap_sem held for write. This avoided possible races with userland
racing an munmap() or mremap(). The page migration patch, however, switched
to using mm_populate() to prime the page mapping. mm_populate() cannot be
called with mmap_sem held.
Instead of dropping the mmap_sem, revert to the old behaviour and simply
drop the use of mm_populate() since get_user_pages() will cause the pages to
get mapped anyways. Thanks to Al Viro for spotting this issue.
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
Sseveral sparse warnings were caused by missing rcu_dereference() annotations
for dereferencing mm->ioctx_table. Thankfully, none of those were actual bugs
as the deref was protected by a spin lock in all instances.
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
Reported-by: Fengguang Wu <fengguang.wu@intel.com>
The commit 36bc08cc01 ("fs/aio: Add support to aio ring pages migration")
added some debugging code that is not required and resulted in a build error
when 98474236f7 ("vfs: make the dentry cache use the lockref infrastructure")
was added to the tree. The code is not required, so just delete it.
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
Another shortcoming of the table lookup patch was revealed where the pointer
was not being tested before being dereferenced. Verify this to avoid the
NULL pointer dereference.
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
In the patch "aio: convert the ioctx list to table lookup v3", incorrect
handling in the ioctx_alloc() error path was introduced that lead to an
ioctx being added via ioctx_add_table() while freed when the ioctx_alloc()
call returned -EAGAIN due to hitting the aio_max_nr limit. Fix this by
only calling ioctx_add_table() as the last step in ioctx_alloc().
Also, several unnecessary rcu_dereference() calls were added that lead to
RCU warnings where the system was already protected by a spin lock for
accessing mm->ioctx_table.
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
In the event that an overflow/underflow occurs while calculating req_batch,
clamp the minimum at 1 request instead of doing a BUG_ON().
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
On Wed, Jun 12, 2013 at 11:14:40AM -0700, Kent Overstreet wrote:
> On Mon, Apr 15, 2013 at 02:40:55PM +0300, Octavian Purdila wrote:
> > When using a large number of threads performing AIO operations the
> > IOCTX list may get a significant number of entries which will cause
> > significant overhead. For example, when running this fio script:
> >
> > rw=randrw; size=256k ;directory=/mnt/fio; ioengine=libaio; iodepth=1
> > blocksize=1024; numjobs=512; thread; loops=100
> >
> > on an EXT2 filesystem mounted on top of a ramdisk we can observe up to
> > 30% CPU time spent by lookup_ioctx:
> >
> > 32.51% [guest.kernel] [g] lookup_ioctx
> > 9.19% [guest.kernel] [g] __lock_acquire.isra.28
> > 4.40% [guest.kernel] [g] lock_release
> > 4.19% [guest.kernel] [g] sched_clock_local
> > 3.86% [guest.kernel] [g] local_clock
> > 3.68% [guest.kernel] [g] native_sched_clock
> > 3.08% [guest.kernel] [g] sched_clock_cpu
> > 2.64% [guest.kernel] [g] lock_release_holdtime.part.11
> > 2.60% [guest.kernel] [g] memcpy
> > 2.33% [guest.kernel] [g] lock_acquired
> > 2.25% [guest.kernel] [g] lock_acquire
> > 1.84% [guest.kernel] [g] do_io_submit
> >
> > This patchs converts the ioctx list to a radix tree. For a performance
> > comparison the above FIO script was run on a 2 sockets 8 core
> > machine. This are the results (average and %rsd of 10 runs) for the
> > original list based implementation and for the radix tree based
> > implementation:
> >
> > cores 1 2 4 8 16 32
> > list 109376 ms 69119 ms 35682 ms 22671 ms 19724 ms 16408 ms
> > %rsd 0.69% 1.15% 1.17% 1.21% 1.71% 1.43%
> > radix 73651 ms 41748 ms 23028 ms 16766 ms 15232 ms 13787 ms
> > %rsd 1.19% 0.98% 0.69% 1.13% 0.72% 0.75%
> > % of radix
> > relative 66.12% 65.59% 66.63% 72.31% 77.26% 83.66%
> > to list
> >
> > To consider the impact of the patch on the typical case of having
> > only one ctx per process the following FIO script was run:
> >
> > rw=randrw; size=100m ;directory=/mnt/fio; ioengine=libaio; iodepth=1
> > blocksize=1024; numjobs=1; thread; loops=100
> >
> > on the same system and the results are the following:
> >
> > list 58892 ms
> > %rsd 0.91%
> > radix 59404 ms
> > %rsd 0.81%
> > % of radix
> > relative 100.87%
> > to list
>
> So, I was just doing some benchmarking/profiling to get ready to send
> out the aio patches I've got for 3.11 - and it looks like your patch is
> causing a ~1.5% throughput regression in my testing :/
... <snip>
I've got an alternate approach for fixing this wart in lookup_ioctx()...
Instead of using an rbtree, just use the reserved id in the ring buffer
header to index an array pointing the ioctx. It's not finished yet, and
it needs to be tidied up, but is most of the way there.
-ben
--
"Thought is the essence of where you are now."
--
kmo> And, a rework of Ben's code, but this was entirely his idea
kmo> -Kent
bcrl> And fix the code to use the right mm_struct in kill_ioctx(), actually
free memory.
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
With the changes to use percpu counters for aio event ring size calculation,
existing increases to aio_max_nr are now insufficient to allow for the
allocation of enough events. Double the value used for aio_max_nr to account
for the doubling introduced by the percpu slack.
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
sock_aio_dtor() is dead code - and stuff that does need to do cleanup
can simply do it before calling aio_complete().
Signed-off-by: Kent Overstreet <koverstreet@google.com>
Cc: Zach Brown <zab@redhat.com>
Cc: Felipe Balbi <balbi@ti.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Mark Fasheh <mfasheh@suse.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Asai Thambi S P <asamymuthupa@micron.com>
Cc: Selvan Mani <smani@micron.com>
Cc: Sam Bradshaw <sbradshaw@micron.com>
Cc: Jeff Moyer <jmoyer@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Benjamin LaHaise <bcrl@kvack.org>
Cc: Theodore Ts'o <tytso@mit.edu>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
The kiocb refcount is only needed for cancellation - to ensure a kiocb
isn't freed while a ki_cancel callback is running. But if we restrict
ki_cancel callbacks to not block (which they currently don't), we can
simply drop the refcount.
Signed-off-by: Kent Overstreet <koverstreet@google.com>
Cc: Zach Brown <zab@redhat.com>
Cc: Felipe Balbi <balbi@ti.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Mark Fasheh <mfasheh@suse.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Asai Thambi S P <asamymuthupa@micron.com>
Cc: Selvan Mani <smani@micron.com>
Cc: Sam Bradshaw <sbradshaw@micron.com>
Cc: Jeff Moyer <jmoyer@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Benjamin LaHaise <bcrl@kvack.org>
Cc: Theodore Ts'o <tytso@mit.edu>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
The old aio retry infrastucture needed to save the various arguments to
to aio operations. But with the retry infrastructure gone, we can trim
struct kiocb quite a bit.
Signed-off-by: Kent Overstreet <koverstreet@google.com>
Cc: Zach Brown <zab@redhat.com>
Cc: Felipe Balbi <balbi@ti.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Mark Fasheh <mfasheh@suse.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Asai Thambi S P <asamymuthupa@micron.com>
Cc: Selvan Mani <smani@micron.com>
Cc: Sam Bradshaw <sbradshaw@micron.com>
Cc: Jeff Moyer <jmoyer@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Benjamin LaHaise <bcrl@kvack.org>
Cc: Theodore Ts'o <tytso@mit.edu>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
This code doesn't serve any purpose anymore, since the aio retry
infrastructure has been removed.
This change should be safe because aio_read/write are also used for
synchronous IO, and called from do_sync_read()/do_sync_write() - and
there's no looping done in the sync case (the read and write syscalls).
Signed-off-by: Kent Overstreet <koverstreet@google.com>
Cc: Zach Brown <zab@redhat.com>
Cc: Felipe Balbi <balbi@ti.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Mark Fasheh <mfasheh@suse.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Asai Thambi S P <asamymuthupa@micron.com>
Cc: Selvan Mani <smani@micron.com>
Cc: Sam Bradshaw <sbradshaw@micron.com>
Cc: Jeff Moyer <jmoyer@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Benjamin LaHaise <bcrl@kvack.org>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
aio_complete() (arguably) needs to keep its own trusted copy of the tail
pointer, but io_getevents() doesn't have to use it - it's already using
the head pointer from the ring buffer.
So convert it to use the tail from the ring buffer so it touches fewer
cachelines and doesn't contend with the cacheline aio_complete() needs.
Signed-off-by: Kent Overstreet <koverstreet@google.com>
Cc: Zach Brown <zab@redhat.com>
Cc: Felipe Balbi <balbi@ti.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Mark Fasheh <mfasheh@suse.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Asai Thambi S P <asamymuthupa@micron.com>
Cc: Selvan Mani <smani@micron.com>
Cc: Sam Bradshaw <sbradshaw@micron.com>
Cc: Jeff Moyer <jmoyer@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Benjamin LaHaise <bcrl@kvack.org>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
Originally, io_event() was documented to return the io_event if
cancellation succeeded - the io_event wouldn't be delivered via the ring
buffer like it normally would.
But this isn't what the implementation was actually doing; the only
driver implementing cancellation, the usb gadget code, never returned an
io_event in its cancel function. And aio_complete() was recently changed
to no longer suppress event delivery if the kiocb had been cancelled.
This gets rid of the unused io_event argument to kiocb_cancel() and
kiocb->ki_cancel(), and changes io_cancel() to return -EINPROGRESS if
kiocb->ki_cancel() returned success.
Also tweak the refcounting in kiocb_cancel() to make more sense.
Signed-off-by: Kent Overstreet <koverstreet@google.com>
Cc: Zach Brown <zab@redhat.com>
Cc: Felipe Balbi <balbi@ti.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Mark Fasheh <mfasheh@suse.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Asai Thambi S P <asamymuthupa@micron.com>
Cc: Selvan Mani <smani@micron.com>
Cc: Sam Bradshaw <sbradshaw@micron.com>
Cc: Jeff Moyer <jmoyer@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Benjamin LaHaise <bcrl@kvack.org>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
This just converts the ioctx refcount to the new generic dynamic percpu
refcount code.
Signed-off-by: Kent Overstreet <koverstreet@google.com>
Cc: Zach Brown <zab@redhat.com>
Cc: Felipe Balbi <balbi@ti.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Mark Fasheh <mfasheh@suse.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Asai Thambi S P <asamymuthupa@micron.com>
Cc: Selvan Mani <smani@micron.com>
Cc: Sam Bradshaw <sbradshaw@micron.com>
Cc: Jeff Moyer <jmoyer@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Benjamin LaHaise <bcrl@kvack.org>
Reviewed-by: "Theodore Ts'o" <tytso@mit.edu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
See the previous patch ("aio: reqs_active -> reqs_available") for why we
want to do this - this basically implements a per cpu allocator for
reqs_available that doesn't actually allocate anything.
Note that we need to increase the size of the ringbuffer we allocate,
since a single thread won't necessarily be able to use all the
reqs_available slots - some (up to about half) might be on other per cpu
lists, unavailable for the current thread.
We size the ringbuffer based on the nr_events userspace passed to
io_setup(), so this is a slight behaviour change - but nr_events wasn't
being used as a hard limit before, it was being rounded up to the next
page before so this doesn't change the actual semantics.
Signed-off-by: Kent Overstreet <koverstreet@google.com>
Cc: Zach Brown <zab@redhat.com>
Cc: Felipe Balbi <balbi@ti.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Mark Fasheh <mfasheh@suse.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Asai Thambi S P <asamymuthupa@micron.com>
Cc: Selvan Mani <smani@micron.com>
Cc: Sam Bradshaw <sbradshaw@micron.com>
Cc: Jeff Moyer <jmoyer@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Benjamin LaHaise <bcrl@kvack.org>
Reviewed-by: "Theodore Ts'o" <tytso@mit.edu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
The number of outstanding kiocbs is one of the few shared things left that
has to be touched for every kiocb - it'd be nice to make it percpu.
We can make it per cpu by treating it like an allocation problem: we have
a maximum number of kiocbs that can be outstanding (i.e. slots) - then we
just allocate and free slots, and we know how to write per cpu allocators.
So as prep work for that, we convert reqs_active to reqs_available.
Signed-off-by: Kent Overstreet <koverstreet@google.com>
Cc: Zach Brown <zab@redhat.com>
Cc: Felipe Balbi <balbi@ti.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Mark Fasheh <mfasheh@suse.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Asai Thambi S P <asamymuthupa@micron.com>
Cc: Selvan Mani <smani@micron.com>
Cc: Sam Bradshaw <sbradshaw@micron.com>
Cc: Jeff Moyer <jmoyer@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Benjamin LaHaise <bcrl@kvack.org>
Reviewed-by: "Theodore Ts'o" <tytso@mit.edu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
When "fs/aio: Add support to aio ring pages migration" was applied, it
broke the build when CONFIG_MIGRATION was disabled. Wrap the migration
code with a test for CONFIG_MIGRATION to fix this and save a few bytes
when migration is disabled.
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
As the aio job will pin the ring pages, that will lead to mem migrated
failed. In order to fix this problem we use an anon inode to manage the aio ring
pages, and setup the migratepage callback in the anon inode's address space, so
that when mem migrating the aio ring pages will be moved to other mem node safely.
Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com>
Signed-off-by: Benjamin LaHaise <bcrl@kvack.org>
