Minki/linux
1
0
Fork 1
mirror of https://github.com/torvalds/linux.git synced 2024-11-23 04:31:50 +00:00
Code Issues Packages Projects Releases Wiki Activity
b9fd2affe4
linux/fs/btrfs/bio.c

910 lines
25 KiB
C
Raw Normal View History

btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
// SPDX-License-Identifier: GPL-2.0
/*
* Copyright (C) 2007 Oracle. All rights reserved.
* Copyright (C) 2022 Christoph Hellwig.
*/
#include <linux/bio.h>
#include "bio.h"
#include "ctree.h"
#include "volumes.h"
#include "raid56.h"
#include "async-thread.h"
#include "dev-replace.h"
#include "zoned.h"
btrfs: pre-load data checksum for reads in btrfs_submit_bio Instead of calling btrfs_lookup_bio_sums in every caller of btrfs_submit_bio that reads data, do the call once in btrfs_submit_bio. Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:05 +00:00
#include "file-item.h"
btrfs: add support for inserting raid stripe extents Add support for inserting stripe extents into the raid stripe tree on completion of every write that needs an extra logical-to-physical translation when using RAID. Inserting the stripe extents happens after the data I/O has completed, this is done to a) support zone-append and b) rule out the possibility of a RAID-write-hole. Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-09-14 16:06:58 +00:00
#include "raid-stripe-tree.h"
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
static struct bio_set btrfs_bioset;
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
static struct bio_set btrfs_clone_bioset;
btrfs: handle checksum validation and repair at the storage layer Currently btrfs handles checksum validation and repair in the end I/O handler for the btrfs_bio. This leads to a lot of duplicate code plus issues with varying semantics or bugs, e.g. - the until recently broken repair for compressed extents - the fact that encoded reads validate the checksums but do not kick of read repair - the inconsistent checking of the BTRFS_FS_STATE_NO_CSUMS flag This commit revamps the checksum validation and repair code to instead work below the btrfs_submit_bio interfaces. In case of a checksum failure (or a plain old I/O error), the repair is now kicked off before the upper level ->end_io handler is invoked. Progress of an in-progress repair is tracked by a small structure that is allocated using a mempool for each original bio with failed sectors, which holds a reference to the original bio. This new structure is allocated using a mempool to guarantee forward progress even under memory pressure. The mempool will be replenished when the repair completes, just as the mempools backing the bios. There is one significant behavior change here: If repair fails or is impossible to start with, the whole bio will be failed to the upper layer. This is the behavior that all I/O submitters except for buffered I/O already emulated in their end_io handler. For buffered I/O this now means that a large readahead request can fail due to a single bad sector, but as readahead errors are ignored the following readpage if the sector is actually accessed will still be able to read. This also matches the I/O failure handling in other file systems. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:07 +00:00
static struct bio_set btrfs_repair_bioset;
static mempool_t btrfs_failed_bio_pool;
struct btrfs_failed_bio {
struct btrfs_bio *bbio;
int num_copies;
atomic_t repair_count;
};
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
btrfs: add a is_data_bbio helper Add a helper to check for that a btrfs_bio has a valid inode, and that it is a data inode to key off all the special handling for data path checksumming. Note that this uses is_data_inode instead of REQ_META as REQ_META is only set directly before submission in submit_one_bio and we'll also want to use this helper for error handling where REQ_META isn't set yet. Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Josef Bacik <josef@toxicpanda.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-05-31 07:54:00 +00:00
/* Is this a data path I/O that needs storage layer checksum and repair? */
static inline bool is_data_bbio(struct btrfs_bio *bbio)
{
return bbio->inode && is_data_inode(&bbio->inode->vfs_inode);
}
btrfs: add an ordered_extent pointer to struct btrfs_bio Add a pointer to the ordered_extent to the existing union in struct btrfs_bio, so all code dealing with data write bios can just use a pointer dereference to retrieve the ordered_extent instead of doing multiple rbtree lookups per I/O. The reference to this ordered_extent is dropped at end I/O time, which implies that an extra one must be acquired when the bio is split. This also requires moving the btrfs_extract_ordered_extent call into btrfs_split_bio so that the invariant of always having a valid ordered_extent reference for the btrfs_bio is kept. Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Josef Bacik <josef@toxicpanda.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-05-31 07:54:02 +00:00
static bool bbio_has_ordered_extent(struct btrfs_bio *bbio)
{
return is_data_bbio(bbio) && btrfs_op(&bbio->bio) == BTRFS_MAP_WRITE;
}
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
/*
* Initialize a btrfs_bio structure. This skips the embedded bio itself as it
* is already initialized by the block layer.
*/
btrfs: introduce btrfs_bio::fs_info member Currently we're doing a lot of work for btrfs_bio: - Checksum verification for data read bios - Bio splits if it crosses stripe boundary - Read repair for data read bios However for the incoming scrub patches, we don't want this extra functionality at all, just plain logical + mirror -> physical mapping ability. Thus here we do the following changes: - Introduce btrfs_bio::fs_info This is for the new scrub specific btrfs_bio, which would not populate btrfs_bio::inode. Thus we need such new member to grab a fs_info This new member will always be populated. - Replace @inode argument with @fs_info for btrfs_bio_init() and its caller Since @inode is no longer a mandatory member, replace it with @fs_info, and let involved users populate @inode. - Skip checksum verification and generation if @bbio->inode is NULL - Add extra ASSERT()s To make sure: * bbio->inode is properly set for involved read repair path * if @file_offset is set, bbio->inode is also populated - Grab @fs_info from @bbio directly We can no longer go @bbio->inode->root->fs_info, as bbio->inode can be NULL. This involves: * btrfs_simple_end_io() * should_async_write() * btrfs_wq_submit_bio() * btrfs_use_zone_append() Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-23 09:01:20 +00:00
void btrfs_bio_init(struct btrfs_bio *bbio, struct btrfs_fs_info *fs_info,
btrfs: pass the iomap bio to btrfs_submit_bio Now that btrfs_submit_bio splits the bio when crossing stripe boundaries, there is no need for the higher level code to do that manually. For direct I/O this is really helpful, as btrfs_submit_io can now simply take the bio allocated by iomap and send it on to btrfs_submit_bio instead of allocating clones. For that to work, the bio embedded into struct btrfs_dio_private needs to become a full btrfs_bio as expected by btrfs_submit_bio. With this change there is a single work item to offload the entire iomap bio so the heuristics to skip async processing for bios that were split isn't needed anymore either. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:21 +00:00
btrfs_bio_end_io_t end_io, void *private)
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
{
memset(bbio, 0, offsetof(struct btrfs_bio, bio));
btrfs: introduce btrfs_bio::fs_info member Currently we're doing a lot of work for btrfs_bio: - Checksum verification for data read bios - Bio splits if it crosses stripe boundary - Read repair for data read bios However for the incoming scrub patches, we don't want this extra functionality at all, just plain logical + mirror -> physical mapping ability. Thus here we do the following changes: - Introduce btrfs_bio::fs_info This is for the new scrub specific btrfs_bio, which would not populate btrfs_bio::inode. Thus we need such new member to grab a fs_info This new member will always be populated. - Replace @inode argument with @fs_info for btrfs_bio_init() and its caller Since @inode is no longer a mandatory member, replace it with @fs_info, and let involved users populate @inode. - Skip checksum verification and generation if @bbio->inode is NULL - Add extra ASSERT()s To make sure: * bbio->inode is properly set for involved read repair path * if @file_offset is set, bbio->inode is also populated - Grab @fs_info from @bbio directly We can no longer go @bbio->inode->root->fs_info, as bbio->inode can be NULL. This involves: * btrfs_simple_end_io() * should_async_write() * btrfs_wq_submit_bio() * btrfs_use_zone_append() Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-23 09:01:20 +00:00
bbio->fs_info = fs_info;
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
bbio->end_io = end_io;
bbio->private = private;
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
atomic_set(&bbio->pending_ios, 1);
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
}
/*
* Allocate a btrfs_bio structure. The btrfs_bio is the main I/O container for
* btrfs, and is used for all I/O submitted through btrfs_submit_bio.
*
* Just like the underlying bio_alloc_bioset it will not fail as it is backed by
* a mempool.
*/
btrfs: return a btrfs_bio from btrfs_bio_alloc Return the containing struct btrfs_bio instead of the less type safe struct bio from btrfs_bio_alloc. Reviewed-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-07 16:39:44 +00:00
struct btrfs_bio *btrfs_bio_alloc(unsigned int nr_vecs, blk_opf_t opf,
btrfs: introduce btrfs_bio::fs_info member Currently we're doing a lot of work for btrfs_bio: - Checksum verification for data read bios - Bio splits if it crosses stripe boundary - Read repair for data read bios However for the incoming scrub patches, we don't want this extra functionality at all, just plain logical + mirror -> physical mapping ability. Thus here we do the following changes: - Introduce btrfs_bio::fs_info This is for the new scrub specific btrfs_bio, which would not populate btrfs_bio::inode. Thus we need such new member to grab a fs_info This new member will always be populated. - Replace @inode argument with @fs_info for btrfs_bio_init() and its caller Since @inode is no longer a mandatory member, replace it with @fs_info, and let involved users populate @inode. - Skip checksum verification and generation if @bbio->inode is NULL - Add extra ASSERT()s To make sure: * bbio->inode is properly set for involved read repair path * if @file_offset is set, bbio->inode is also populated - Grab @fs_info from @bbio directly We can no longer go @bbio->inode->root->fs_info, as bbio->inode can be NULL. This involves: * btrfs_simple_end_io() * should_async_write() * btrfs_wq_submit_bio() * btrfs_use_zone_append() Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-23 09:01:20 +00:00
struct btrfs_fs_info *fs_info,
btrfs: return a btrfs_bio from btrfs_bio_alloc Return the containing struct btrfs_bio instead of the less type safe struct bio from btrfs_bio_alloc. Reviewed-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-07 16:39:44 +00:00
btrfs_bio_end_io_t end_io, void *private)
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
{
btrfs: return a btrfs_bio from btrfs_bio_alloc Return the containing struct btrfs_bio instead of the less type safe struct bio from btrfs_bio_alloc. Reviewed-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-07 16:39:44 +00:00
struct btrfs_bio *bbio;
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
struct bio *bio;
bio = bio_alloc_bioset(NULL, nr_vecs, opf, GFP_NOFS, &btrfs_bioset);
btrfs: return a btrfs_bio from btrfs_bio_alloc Return the containing struct btrfs_bio instead of the less type safe struct bio from btrfs_bio_alloc. Reviewed-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-07 16:39:44 +00:00
bbio = btrfs_bio(bio);
btrfs: introduce btrfs_bio::fs_info member Currently we're doing a lot of work for btrfs_bio: - Checksum verification for data read bios - Bio splits if it crosses stripe boundary - Read repair for data read bios However for the incoming scrub patches, we don't want this extra functionality at all, just plain logical + mirror -> physical mapping ability. Thus here we do the following changes: - Introduce btrfs_bio::fs_info This is for the new scrub specific btrfs_bio, which would not populate btrfs_bio::inode. Thus we need such new member to grab a fs_info This new member will always be populated. - Replace @inode argument with @fs_info for btrfs_bio_init() and its caller Since @inode is no longer a mandatory member, replace it with @fs_info, and let involved users populate @inode. - Skip checksum verification and generation if @bbio->inode is NULL - Add extra ASSERT()s To make sure: * bbio->inode is properly set for involved read repair path * if @file_offset is set, bbio->inode is also populated - Grab @fs_info from @bbio directly We can no longer go @bbio->inode->root->fs_info, as bbio->inode can be NULL. This involves: * btrfs_simple_end_io() * should_async_write() * btrfs_wq_submit_bio() * btrfs_use_zone_append() Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-23 09:01:20 +00:00
btrfs_bio_init(bbio, fs_info, end_io, private);
btrfs: return a btrfs_bio from btrfs_bio_alloc Return the containing struct btrfs_bio instead of the less type safe struct bio from btrfs_bio_alloc. Reviewed-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-07 16:39:44 +00:00
return bbio;
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
}
btrfs: make btrfs_split_bio work on struct btrfs_bio btrfs_split_bio expects a btrfs_bio as argument and always allocates one. Type both the orig_bio argument and the return value as struct btrfs_bio to improve type safety. Reviewed-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-07 16:39:45 +00:00
static struct btrfs_bio *btrfs_split_bio(struct btrfs_fs_info *fs_info,
struct btrfs_bio *orig_bbio,
u64 map_length, bool use_append)
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
{
btrfs: make btrfs_split_bio work on struct btrfs_bio btrfs_split_bio expects a btrfs_bio as argument and always allocates one. Type both the orig_bio argument and the return value as struct btrfs_bio to improve type safety. Reviewed-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-07 16:39:45 +00:00
struct btrfs_bio *bbio;
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
struct bio *bio;
btrfs: split zone append bios in btrfs_submit_bio The current btrfs zoned device support is a little cumbersome in the data I/O path as it requires the callers to not issue I/O larger than the supported ZONE_APPEND size of the underlying device. This leads to a lot of extra accounting. Instead change btrfs_submit_bio so that it can take write bios of arbitrary size and form from the upper layers, and just split them internally to the ZONE_APPEND queue limits. Then remove all the upper layer warts catering to limited write sized on zoned devices, including the extra refcount in the compressed_bio. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:30 +00:00
if (use_append) {
unsigned int nr_segs;
btrfs: make btrfs_split_bio work on struct btrfs_bio btrfs_split_bio expects a btrfs_bio as argument and always allocates one. Type both the orig_bio argument and the return value as struct btrfs_bio to improve type safety. Reviewed-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-07 16:39:45 +00:00
bio = bio_split_rw(&orig_bbio->bio, &fs_info->limits, &nr_segs,
btrfs: split zone append bios in btrfs_submit_bio The current btrfs zoned device support is a little cumbersome in the data I/O path as it requires the callers to not issue I/O larger than the supported ZONE_APPEND size of the underlying device. This leads to a lot of extra accounting. Instead change btrfs_submit_bio so that it can take write bios of arbitrary size and form from the upper layers, and just split them internally to the ZONE_APPEND queue limits. Then remove all the upper layer warts catering to limited write sized on zoned devices, including the extra refcount in the compressed_bio. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:30 +00:00
&btrfs_clone_bioset, map_length);
} else {
btrfs: make btrfs_split_bio work on struct btrfs_bio btrfs_split_bio expects a btrfs_bio as argument and always allocates one. Type both the orig_bio argument and the return value as struct btrfs_bio to improve type safety. Reviewed-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-07 16:39:45 +00:00
bio = bio_split(&orig_bbio->bio, map_length >> SECTOR_SHIFT,
GFP_NOFS, &btrfs_clone_bioset);
btrfs: split zone append bios in btrfs_submit_bio The current btrfs zoned device support is a little cumbersome in the data I/O path as it requires the callers to not issue I/O larger than the supported ZONE_APPEND size of the underlying device. This leads to a lot of extra accounting. Instead change btrfs_submit_bio so that it can take write bios of arbitrary size and form from the upper layers, and just split them internally to the ZONE_APPEND queue limits. Then remove all the upper layer warts catering to limited write sized on zoned devices, including the extra refcount in the compressed_bio. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:30 +00:00
}
btrfs: make btrfs_split_bio work on struct btrfs_bio btrfs_split_bio expects a btrfs_bio as argument and always allocates one. Type both the orig_bio argument and the return value as struct btrfs_bio to improve type safety. Reviewed-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-07 16:39:45 +00:00
bbio = btrfs_bio(bio);
btrfs: introduce btrfs_bio::fs_info member Currently we're doing a lot of work for btrfs_bio: - Checksum verification for data read bios - Bio splits if it crosses stripe boundary - Read repair for data read bios However for the incoming scrub patches, we don't want this extra functionality at all, just plain logical + mirror -> physical mapping ability. Thus here we do the following changes: - Introduce btrfs_bio::fs_info This is for the new scrub specific btrfs_bio, which would not populate btrfs_bio::inode. Thus we need such new member to grab a fs_info This new member will always be populated. - Replace @inode argument with @fs_info for btrfs_bio_init() and its caller Since @inode is no longer a mandatory member, replace it with @fs_info, and let involved users populate @inode. - Skip checksum verification and generation if @bbio->inode is NULL - Add extra ASSERT()s To make sure: * bbio->inode is properly set for involved read repair path * if @file_offset is set, bbio->inode is also populated - Grab @fs_info from @bbio directly We can no longer go @bbio->inode->root->fs_info, as bbio->inode can be NULL. This involves: * btrfs_simple_end_io() * should_async_write() * btrfs_wq_submit_bio() * btrfs_use_zone_append() Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-23 09:01:20 +00:00
btrfs_bio_init(bbio, fs_info, NULL, orig_bbio);
bbio->inode = orig_bbio->inode;
btrfs: make btrfs_split_bio work on struct btrfs_bio btrfs_split_bio expects a btrfs_bio as argument and always allocates one. Type both the orig_bio argument and the return value as struct btrfs_bio to improve type safety. Reviewed-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-07 16:39:45 +00:00
bbio->file_offset = orig_bbio->file_offset;
btrfs: fix file_offset for REQ_BTRFS_ONE_ORDERED bios that get split If a bio gets split, it needs to have a proper file_offset for checksum validation and repair to work properly. Based on feedback from Josef, commit 852eee62d31a ("btrfs: allow btrfs_submit_bio to split bios") skipped this adjustment for ONE_ORDERED bios. But if we actually ever need to split a ONE_ORDERED read bio, this will lead to a wrong file offset in the repair code. Right now the only user of the file_offset is logging of an error message so this is mostly harmless, but the wrong offset might be more problematic for additional users in the future. Fixes: 852eee62d31a ("btrfs: allow btrfs_submit_bio to split bios") Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Josef Bacik <josef@toxicpanda.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-05-31 07:53:54 +00:00
orig_bbio->file_offset += map_length;
btrfs: add an ordered_extent pointer to struct btrfs_bio Add a pointer to the ordered_extent to the existing union in struct btrfs_bio, so all code dealing with data write bios can just use a pointer dereference to retrieve the ordered_extent instead of doing multiple rbtree lookups per I/O. The reference to this ordered_extent is dropped at end I/O time, which implies that an extra one must be acquired when the bio is split. This also requires moving the btrfs_extract_ordered_extent call into btrfs_split_bio so that the invariant of always having a valid ordered_extent reference for the btrfs_bio is kept. Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Josef Bacik <josef@toxicpanda.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-05-31 07:54:02 +00:00
if (bbio_has_ordered_extent(bbio)) {
refcount_inc(&orig_bbio->ordered->refs);
bbio->ordered = orig_bbio->ordered;
}
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
atomic_inc(&orig_bbio->pending_ios);
btrfs: make btrfs_split_bio work on struct btrfs_bio btrfs_split_bio expects a btrfs_bio as argument and always allocates one. Type both the orig_bio argument and the return value as struct btrfs_bio to improve type safety. Reviewed-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-07 16:39:45 +00:00
return bbio;
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
}
btrfs: add an ordered_extent pointer to struct btrfs_bio Add a pointer to the ordered_extent to the existing union in struct btrfs_bio, so all code dealing with data write bios can just use a pointer dereference to retrieve the ordered_extent instead of doing multiple rbtree lookups per I/O. The reference to this ordered_extent is dropped at end I/O time, which implies that an extra one must be acquired when the bio is split. This also requires moving the btrfs_extract_ordered_extent call into btrfs_split_bio so that the invariant of always having a valid ordered_extent reference for the btrfs_bio is kept. Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Josef Bacik <josef@toxicpanda.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-05-31 07:54:02 +00:00
/* Free a bio that was never submitted to the underlying device. */
static void btrfs_cleanup_bio(struct btrfs_bio *bbio)
{
if (bbio_has_ordered_extent(bbio))
btrfs_put_ordered_extent(bbio->ordered);
bio_put(&bbio->bio);
}
static void __btrfs_bio_end_io(struct btrfs_bio *bbio)
{
if (bbio_has_ordered_extent(bbio)) {
struct btrfs_ordered_extent *ordered = bbio->ordered;
bbio->end_io(bbio);
btrfs_put_ordered_extent(ordered);
} else {
bbio->end_io(bbio);
}
}
void btrfs_bio_end_io(struct btrfs_bio *bbio, blk_status_t status)
{
bbio->bio.bi_status = status;
__btrfs_bio_end_io(bbio);
}
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
static void btrfs_orig_write_end_io(struct bio *bio);
static void btrfs_bbio_propagate_error(struct btrfs_bio *bbio,
struct btrfs_bio *orig_bbio)
{
/*
* For writes we tolerate nr_mirrors - 1 write failures, so we can't
* just blindly propagate a write failure here. Instead increment the
* error count in the original I/O context so that it is guaranteed to
* be larger than the error tolerance.
*/
if (bbio->bio.bi_end_io == &btrfs_orig_write_end_io) {
struct btrfs_io_stripe *orig_stripe = orig_bbio->bio.bi_private;
struct btrfs_io_context *orig_bioc = orig_stripe->bioc;
atomic_add(orig_bioc->max_errors, &orig_bioc->error);
} else {
orig_bbio->bio.bi_status = bbio->bio.bi_status;
}
}
static void btrfs_orig_bbio_end_io(struct btrfs_bio *bbio)
{
if (bbio->bio.bi_pool == &btrfs_clone_bioset) {
struct btrfs_bio *orig_bbio = bbio->private;
if (bbio->bio.bi_status)
btrfs_bbio_propagate_error(bbio, orig_bbio);
btrfs: add an ordered_extent pointer to struct btrfs_bio Add a pointer to the ordered_extent to the existing union in struct btrfs_bio, so all code dealing with data write bios can just use a pointer dereference to retrieve the ordered_extent instead of doing multiple rbtree lookups per I/O. The reference to this ordered_extent is dropped at end I/O time, which implies that an extra one must be acquired when the bio is split. This also requires moving the btrfs_extract_ordered_extent call into btrfs_split_bio so that the invariant of always having a valid ordered_extent reference for the btrfs_bio is kept. Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Josef Bacik <josef@toxicpanda.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-05-31 07:54:02 +00:00
btrfs_cleanup_bio(bbio);
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
bbio = orig_bbio;
}
if (atomic_dec_and_test(&bbio->pending_ios))
btrfs: add an ordered_extent pointer to struct btrfs_bio Add a pointer to the ordered_extent to the existing union in struct btrfs_bio, so all code dealing with data write bios can just use a pointer dereference to retrieve the ordered_extent instead of doing multiple rbtree lookups per I/O. The reference to this ordered_extent is dropped at end I/O time, which implies that an extra one must be acquired when the bio is split. This also requires moving the btrfs_extract_ordered_extent call into btrfs_split_bio so that the invariant of always having a valid ordered_extent reference for the btrfs_bio is kept. Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Josef Bacik <josef@toxicpanda.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-05-31 07:54:02 +00:00
__btrfs_bio_end_io(bbio);
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
}
btrfs: handle checksum validation and repair at the storage layer Currently btrfs handles checksum validation and repair in the end I/O handler for the btrfs_bio. This leads to a lot of duplicate code plus issues with varying semantics or bugs, e.g. - the until recently broken repair for compressed extents - the fact that encoded reads validate the checksums but do not kick of read repair - the inconsistent checking of the BTRFS_FS_STATE_NO_CSUMS flag This commit revamps the checksum validation and repair code to instead work below the btrfs_submit_bio interfaces. In case of a checksum failure (or a plain old I/O error), the repair is now kicked off before the upper level ->end_io handler is invoked. Progress of an in-progress repair is tracked by a small structure that is allocated using a mempool for each original bio with failed sectors, which holds a reference to the original bio. This new structure is allocated using a mempool to guarantee forward progress even under memory pressure. The mempool will be replenished when the repair completes, just as the mempools backing the bios. There is one significant behavior change here: If repair fails or is impossible to start with, the whole bio will be failed to the upper layer. This is the behavior that all I/O submitters except for buffered I/O already emulated in their end_io handler. For buffered I/O this now means that a large readahead request can fail due to a single bad sector, but as readahead errors are ignored the following readpage if the sector is actually accessed will still be able to read. This also matches the I/O failure handling in other file systems. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:07 +00:00
static int next_repair_mirror(struct btrfs_failed_bio *fbio, int cur_mirror)
{
if (cur_mirror == fbio->num_copies)
return cur_mirror + 1 - fbio->num_copies;
return cur_mirror + 1;
}
static int prev_repair_mirror(struct btrfs_failed_bio *fbio, int cur_mirror)
{
if (cur_mirror == 1)
return fbio->num_copies;
return cur_mirror - 1;
}
static void btrfs_repair_done(struct btrfs_failed_bio *fbio)
{
if (atomic_dec_and_test(&fbio->repair_count)) {
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
btrfs_orig_bbio_end_io(fbio->bbio);
btrfs: handle checksum validation and repair at the storage layer Currently btrfs handles checksum validation and repair in the end I/O handler for the btrfs_bio. This leads to a lot of duplicate code plus issues with varying semantics or bugs, e.g. - the until recently broken repair for compressed extents - the fact that encoded reads validate the checksums but do not kick of read repair - the inconsistent checking of the BTRFS_FS_STATE_NO_CSUMS flag This commit revamps the checksum validation and repair code to instead work below the btrfs_submit_bio interfaces. In case of a checksum failure (or a plain old I/O error), the repair is now kicked off before the upper level ->end_io handler is invoked. Progress of an in-progress repair is tracked by a small structure that is allocated using a mempool for each original bio with failed sectors, which holds a reference to the original bio. This new structure is allocated using a mempool to guarantee forward progress even under memory pressure. The mempool will be replenished when the repair completes, just as the mempools backing the bios. There is one significant behavior change here: If repair fails or is impossible to start with, the whole bio will be failed to the upper layer. This is the behavior that all I/O submitters except for buffered I/O already emulated in their end_io handler. For buffered I/O this now means that a large readahead request can fail due to a single bad sector, but as readahead errors are ignored the following readpage if the sector is actually accessed will still be able to read. This also matches the I/O failure handling in other file systems. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:07 +00:00
mempool_free(fbio, &btrfs_failed_bio_pool);
}
}
static void btrfs_end_repair_bio(struct btrfs_bio *repair_bbio,
struct btrfs_device *dev)
{
struct btrfs_failed_bio *fbio = repair_bbio->private;
struct btrfs_inode *inode = repair_bbio->inode;
struct btrfs_fs_info *fs_info = inode->root->fs_info;
struct bio_vec *bv = bio_first_bvec_all(&repair_bbio->bio);
int mirror = repair_bbio->mirror_num;
btrfs: migrate btrfs_repair_io_failure() to folio interfaces [BUG] Test case btrfs/124 failed if larger metadata folio is enabled, the dying message looks like this: BTRFS error (device dm-2): bad tree block start, mirror 2 want 31686656 have 0 BTRFS info (device dm-2): read error corrected: ino 0 off 31686656 (dev /dev/mapper/test-scratch2 sector 20928) BUG: kernel NULL pointer dereference, address: 0000000000000020 #PF: supervisor read access in kernel mode #PF: error_code(0x0000) - not-present page CPU: 6 PID: 350881 Comm: btrfs Tainted: G OE 6.7.0-rc3-custom+ #128 Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS unknown 2/2/2022 RIP: 0010:btrfs_read_extent_buffer+0x106/0x180 [btrfs] PKRU: 55555554 Call Trace: <TASK> read_tree_block+0x33/0xb0 [btrfs] read_block_for_search+0x23e/0x340 [btrfs] btrfs_search_slot+0x2f9/0xe60 [btrfs] btrfs_lookup_csum+0x75/0x160 [btrfs] btrfs_lookup_bio_sums+0x21a/0x560 [btrfs] btrfs_submit_chunk+0x152/0x680 [btrfs] btrfs_submit_bio+0x1c/0x50 [btrfs] submit_one_bio+0x40/0x80 [btrfs] submit_extent_page+0x158/0x390 [btrfs] btrfs_do_readpage+0x330/0x740 [btrfs] extent_readahead+0x38d/0x6c0 [btrfs] read_pages+0x94/0x2c0 page_cache_ra_unbounded+0x12d/0x190 relocate_file_extent_cluster+0x7c1/0x9d0 [btrfs] relocate_block_group+0x2d3/0x560 [btrfs] btrfs_relocate_block_group+0x2c7/0x4b0 [btrfs] btrfs_relocate_chunk+0x4c/0x1a0 [btrfs] btrfs_balance+0x925/0x13c0 [btrfs] btrfs_ioctl+0x19f1/0x25d0 [btrfs] __x64_sys_ioctl+0x90/0xd0 do_syscall_64+0x3f/0xf0 entry_SYSCALL_64_after_hwframe+0x6e/0x76 [CAUSE] The dying line is at btrfs_repair_io_failure() call inside btrfs_repair_eb_io_failure(). The function is still relying on the extent buffer using page sized folios. When the extent buffer is using larger folio, we go into the 2nd slot of folios[], and triggered the NULL pointer dereference. [FIX] Migrate btrfs_repair_io_failure() to folio interfaces. So that when we hit a larger folio, we just submit the whole folio in one go. This also affects data repair path through btrfs_end_repair_bio(), thankfully data is still fully page based, we can just add an ASSERT(), and use page_folio() to convert the page to folio. Signed-off-by: Qu Wenruo <wqu@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-12-12 05:24:10 +00:00
/*
* We can only trigger this for data bio, which doesn't support larger
* folios yet.
*/
ASSERT(folio_order(page_folio(bv->bv_page)) == 0);
btrfs: handle checksum validation and repair at the storage layer Currently btrfs handles checksum validation and repair in the end I/O handler for the btrfs_bio. This leads to a lot of duplicate code plus issues with varying semantics or bugs, e.g. - the until recently broken repair for compressed extents - the fact that encoded reads validate the checksums but do not kick of read repair - the inconsistent checking of the BTRFS_FS_STATE_NO_CSUMS flag This commit revamps the checksum validation and repair code to instead work below the btrfs_submit_bio interfaces. In case of a checksum failure (or a plain old I/O error), the repair is now kicked off before the upper level ->end_io handler is invoked. Progress of an in-progress repair is tracked by a small structure that is allocated using a mempool for each original bio with failed sectors, which holds a reference to the original bio. This new structure is allocated using a mempool to guarantee forward progress even under memory pressure. The mempool will be replenished when the repair completes, just as the mempools backing the bios. There is one significant behavior change here: If repair fails or is impossible to start with, the whole bio will be failed to the upper layer. This is the behavior that all I/O submitters except for buffered I/O already emulated in their end_io handler. For buffered I/O this now means that a large readahead request can fail due to a single bad sector, but as readahead errors are ignored the following readpage if the sector is actually accessed will still be able to read. This also matches the I/O failure handling in other file systems. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:07 +00:00
if (repair_bbio->bio.bi_status ||
!btrfs_data_csum_ok(repair_bbio, dev, 0, bv)) {
bio_reset(&repair_bbio->bio, NULL, REQ_OP_READ);
btrfs: rename btrfs_bio::iter field Rename iter to saved_iter and move it next to the repair internals and nothing outside of bio.c should be touching it. Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:13 +00:00
repair_bbio->bio.bi_iter = repair_bbio->saved_iter;
btrfs: handle checksum validation and repair at the storage layer Currently btrfs handles checksum validation and repair in the end I/O handler for the btrfs_bio. This leads to a lot of duplicate code plus issues with varying semantics or bugs, e.g. - the until recently broken repair for compressed extents - the fact that encoded reads validate the checksums but do not kick of read repair - the inconsistent checking of the BTRFS_FS_STATE_NO_CSUMS flag This commit revamps the checksum validation and repair code to instead work below the btrfs_submit_bio interfaces. In case of a checksum failure (or a plain old I/O error), the repair is now kicked off before the upper level ->end_io handler is invoked. Progress of an in-progress repair is tracked by a small structure that is allocated using a mempool for each original bio with failed sectors, which holds a reference to the original bio. This new structure is allocated using a mempool to guarantee forward progress even under memory pressure. The mempool will be replenished when the repair completes, just as the mempools backing the bios. There is one significant behavior change here: If repair fails or is impossible to start with, the whole bio will be failed to the upper layer. This is the behavior that all I/O submitters except for buffered I/O already emulated in their end_io handler. For buffered I/O this now means that a large readahead request can fail due to a single bad sector, but as readahead errors are ignored the following readpage if the sector is actually accessed will still be able to read. This also matches the I/O failure handling in other file systems. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:07 +00:00
mirror = next_repair_mirror(fbio, mirror);
if (mirror == fbio->bbio->mirror_num) {
btrfs_debug(fs_info, "no mirror left");
fbio->bbio->bio.bi_status = BLK_STS_IOERR;
goto done;
}
btrfs: pass a btrfs_bio to btrfs_submit_bio btrfs_submit_bio expects the bio passed to it to be embedded into a btrfs_bio structure. Pass the btrfs_bio directly to increase type safety and make the code self-documenting. Reviewed-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-07 16:39:39 +00:00
btrfs_submit_bio(repair_bbio, mirror);
btrfs: handle checksum validation and repair at the storage layer Currently btrfs handles checksum validation and repair in the end I/O handler for the btrfs_bio. This leads to a lot of duplicate code plus issues with varying semantics or bugs, e.g. - the until recently broken repair for compressed extents - the fact that encoded reads validate the checksums but do not kick of read repair - the inconsistent checking of the BTRFS_FS_STATE_NO_CSUMS flag This commit revamps the checksum validation and repair code to instead work below the btrfs_submit_bio interfaces. In case of a checksum failure (or a plain old I/O error), the repair is now kicked off before the upper level ->end_io handler is invoked. Progress of an in-progress repair is tracked by a small structure that is allocated using a mempool for each original bio with failed sectors, which holds a reference to the original bio. This new structure is allocated using a mempool to guarantee forward progress even under memory pressure. The mempool will be replenished when the repair completes, just as the mempools backing the bios. There is one significant behavior change here: If repair fails or is impossible to start with, the whole bio will be failed to the upper layer. This is the behavior that all I/O submitters except for buffered I/O already emulated in their end_io handler. For buffered I/O this now means that a large readahead request can fail due to a single bad sector, but as readahead errors are ignored the following readpage if the sector is actually accessed will still be able to read. This also matches the I/O failure handling in other file systems. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:07 +00:00
return;
}
do {
mirror = prev_repair_mirror(fbio, mirror);
btrfs_repair_io_failure(fs_info, btrfs_ino(inode),
repair_bbio->file_offset, fs_info->sectorsize,
btrfs: rename btrfs_bio::iter field Rename iter to saved_iter and move it next to the repair internals and nothing outside of bio.c should be touching it. Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:13 +00:00
repair_bbio->saved_iter.bi_sector << SECTOR_SHIFT,
btrfs: migrate btrfs_repair_io_failure() to folio interfaces [BUG] Test case btrfs/124 failed if larger metadata folio is enabled, the dying message looks like this: BTRFS error (device dm-2): bad tree block start, mirror 2 want 31686656 have 0 BTRFS info (device dm-2): read error corrected: ino 0 off 31686656 (dev /dev/mapper/test-scratch2 sector 20928) BUG: kernel NULL pointer dereference, address: 0000000000000020 #PF: supervisor read access in kernel mode #PF: error_code(0x0000) - not-present page CPU: 6 PID: 350881 Comm: btrfs Tainted: G OE 6.7.0-rc3-custom+ #128 Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS unknown 2/2/2022 RIP: 0010:btrfs_read_extent_buffer+0x106/0x180 [btrfs] PKRU: 55555554 Call Trace: <TASK> read_tree_block+0x33/0xb0 [btrfs] read_block_for_search+0x23e/0x340 [btrfs] btrfs_search_slot+0x2f9/0xe60 [btrfs] btrfs_lookup_csum+0x75/0x160 [btrfs] btrfs_lookup_bio_sums+0x21a/0x560 [btrfs] btrfs_submit_chunk+0x152/0x680 [btrfs] btrfs_submit_bio+0x1c/0x50 [btrfs] submit_one_bio+0x40/0x80 [btrfs] submit_extent_page+0x158/0x390 [btrfs] btrfs_do_readpage+0x330/0x740 [btrfs] extent_readahead+0x38d/0x6c0 [btrfs] read_pages+0x94/0x2c0 page_cache_ra_unbounded+0x12d/0x190 relocate_file_extent_cluster+0x7c1/0x9d0 [btrfs] relocate_block_group+0x2d3/0x560 [btrfs] btrfs_relocate_block_group+0x2c7/0x4b0 [btrfs] btrfs_relocate_chunk+0x4c/0x1a0 [btrfs] btrfs_balance+0x925/0x13c0 [btrfs] btrfs_ioctl+0x19f1/0x25d0 [btrfs] __x64_sys_ioctl+0x90/0xd0 do_syscall_64+0x3f/0xf0 entry_SYSCALL_64_after_hwframe+0x6e/0x76 [CAUSE] The dying line is at btrfs_repair_io_failure() call inside btrfs_repair_eb_io_failure(). The function is still relying on the extent buffer using page sized folios. When the extent buffer is using larger folio, we go into the 2nd slot of folios[], and triggered the NULL pointer dereference. [FIX] Migrate btrfs_repair_io_failure() to folio interfaces. So that when we hit a larger folio, we just submit the whole folio in one go. This also affects data repair path through btrfs_end_repair_bio(), thankfully data is still fully page based, we can just add an ASSERT(), and use page_folio() to convert the page to folio. Signed-off-by: Qu Wenruo <wqu@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-12-12 05:24:10 +00:00
page_folio(bv->bv_page), bv->bv_offset, mirror);
btrfs: handle checksum validation and repair at the storage layer Currently btrfs handles checksum validation and repair in the end I/O handler for the btrfs_bio. This leads to a lot of duplicate code plus issues with varying semantics or bugs, e.g. - the until recently broken repair for compressed extents - the fact that encoded reads validate the checksums but do not kick of read repair - the inconsistent checking of the BTRFS_FS_STATE_NO_CSUMS flag This commit revamps the checksum validation and repair code to instead work below the btrfs_submit_bio interfaces. In case of a checksum failure (or a plain old I/O error), the repair is now kicked off before the upper level ->end_io handler is invoked. Progress of an in-progress repair is tracked by a small structure that is allocated using a mempool for each original bio with failed sectors, which holds a reference to the original bio. This new structure is allocated using a mempool to guarantee forward progress even under memory pressure. The mempool will be replenished when the repair completes, just as the mempools backing the bios. There is one significant behavior change here: If repair fails or is impossible to start with, the whole bio will be failed to the upper layer. This is the behavior that all I/O submitters except for buffered I/O already emulated in their end_io handler. For buffered I/O this now means that a large readahead request can fail due to a single bad sector, but as readahead errors are ignored the following readpage if the sector is actually accessed will still be able to read. This also matches the I/O failure handling in other file systems. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:07 +00:00
} while (mirror != fbio->bbio->mirror_num);
done:
btrfs_repair_done(fbio);
bio_put(&repair_bbio->bio);
}
/*
* Try to kick off a repair read to the next available mirror for a bad sector.
*
* This primarily tries to recover good data to serve the actual read request,
* but also tries to write the good data back to the bad mirror(s) when a
* read succeeded to restore the redundancy.
*/
static struct btrfs_failed_bio *repair_one_sector(struct btrfs_bio *failed_bbio,
u32 bio_offset,
struct bio_vec *bv,
struct btrfs_failed_bio *fbio)
{
struct btrfs_inode *inode = failed_bbio->inode;
struct btrfs_fs_info *fs_info = inode->root->fs_info;
const u32 sectorsize = fs_info->sectorsize;
btrfs: rename btrfs_bio::iter field Rename iter to saved_iter and move it next to the repair internals and nothing outside of bio.c should be touching it. Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:13 +00:00
const u64 logical = (failed_bbio->saved_iter.bi_sector << SECTOR_SHIFT);
btrfs: handle checksum validation and repair at the storage layer Currently btrfs handles checksum validation and repair in the end I/O handler for the btrfs_bio. This leads to a lot of duplicate code plus issues with varying semantics or bugs, e.g. - the until recently broken repair for compressed extents - the fact that encoded reads validate the checksums but do not kick of read repair - the inconsistent checking of the BTRFS_FS_STATE_NO_CSUMS flag This commit revamps the checksum validation and repair code to instead work below the btrfs_submit_bio interfaces. In case of a checksum failure (or a plain old I/O error), the repair is now kicked off before the upper level ->end_io handler is invoked. Progress of an in-progress repair is tracked by a small structure that is allocated using a mempool for each original bio with failed sectors, which holds a reference to the original bio. This new structure is allocated using a mempool to guarantee forward progress even under memory pressure. The mempool will be replenished when the repair completes, just as the mempools backing the bios. There is one significant behavior change here: If repair fails or is impossible to start with, the whole bio will be failed to the upper layer. This is the behavior that all I/O submitters except for buffered I/O already emulated in their end_io handler. For buffered I/O this now means that a large readahead request can fail due to a single bad sector, but as readahead errors are ignored the following readpage if the sector is actually accessed will still be able to read. This also matches the I/O failure handling in other file systems. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:07 +00:00
struct btrfs_bio *repair_bbio;
struct bio *repair_bio;
int num_copies;
int mirror;
btrfs_debug(fs_info, "repair read error: read error at %llu",
failed_bbio->file_offset + bio_offset);
num_copies = btrfs_num_copies(fs_info, logical, sectorsize);
if (num_copies == 1) {
btrfs_debug(fs_info, "no copy to repair from");
failed_bbio->bio.bi_status = BLK_STS_IOERR;
return fbio;
}
if (!fbio) {
fbio = mempool_alloc(&btrfs_failed_bio_pool, GFP_NOFS);
fbio->bbio = failed_bbio;
fbio->num_copies = num_copies;
atomic_set(&fbio->repair_count, 1);
}
atomic_inc(&fbio->repair_count);
repair_bio = bio_alloc_bioset(NULL, 1, REQ_OP_READ, GFP_NOFS,
&btrfs_repair_bioset);
btrfs: rename btrfs_bio::iter field Rename iter to saved_iter and move it next to the repair internals and nothing outside of bio.c should be touching it. Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:13 +00:00
repair_bio->bi_iter.bi_sector = failed_bbio->saved_iter.bi_sector;
btrfs: use __bio_add_page for adding a single page in repair_one_sector The btrfs repair bio submission code uses bio_add_page() to add a page to a newly created bio. bio_add_page() can fail, but the return value is never checked. Use __bio_add_page() as adding a single page to a newly created bio is guaranteed to succeed. This brings us a step closer to marking bio_add_page() as __must_check. Reviewed-by: Damien Le Moal <damien.lemoal@opensource.wdc.com> Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-30 10:43:50 +00:00
__bio_add_page(repair_bio, bv->bv_page, bv->bv_len, bv->bv_offset);
btrfs: handle checksum validation and repair at the storage layer Currently btrfs handles checksum validation and repair in the end I/O handler for the btrfs_bio. This leads to a lot of duplicate code plus issues with varying semantics or bugs, e.g. - the until recently broken repair for compressed extents - the fact that encoded reads validate the checksums but do not kick of read repair - the inconsistent checking of the BTRFS_FS_STATE_NO_CSUMS flag This commit revamps the checksum validation and repair code to instead work below the btrfs_submit_bio interfaces. In case of a checksum failure (or a plain old I/O error), the repair is now kicked off before the upper level ->end_io handler is invoked. Progress of an in-progress repair is tracked by a small structure that is allocated using a mempool for each original bio with failed sectors, which holds a reference to the original bio. This new structure is allocated using a mempool to guarantee forward progress even under memory pressure. The mempool will be replenished when the repair completes, just as the mempools backing the bios. There is one significant behavior change here: If repair fails or is impossible to start with, the whole bio will be failed to the upper layer. This is the behavior that all I/O submitters except for buffered I/O already emulated in their end_io handler. For buffered I/O this now means that a large readahead request can fail due to a single bad sector, but as readahead errors are ignored the following readpage if the sector is actually accessed will still be able to read. This also matches the I/O failure handling in other file systems. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:07 +00:00
repair_bbio = btrfs_bio(repair_bio);
btrfs: introduce btrfs_bio::fs_info member Currently we're doing a lot of work for btrfs_bio: - Checksum verification for data read bios - Bio splits if it crosses stripe boundary - Read repair for data read bios However for the incoming scrub patches, we don't want this extra functionality at all, just plain logical + mirror -> physical mapping ability. Thus here we do the following changes: - Introduce btrfs_bio::fs_info This is for the new scrub specific btrfs_bio, which would not populate btrfs_bio::inode. Thus we need such new member to grab a fs_info This new member will always be populated. - Replace @inode argument with @fs_info for btrfs_bio_init() and its caller Since @inode is no longer a mandatory member, replace it with @fs_info, and let involved users populate @inode. - Skip checksum verification and generation if @bbio->inode is NULL - Add extra ASSERT()s To make sure: * bbio->inode is properly set for involved read repair path * if @file_offset is set, bbio->inode is also populated - Grab @fs_info from @bbio directly We can no longer go @bbio->inode->root->fs_info, as bbio->inode can be NULL. This involves: * btrfs_simple_end_io() * should_async_write() * btrfs_wq_submit_bio() * btrfs_use_zone_append() Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-23 09:01:20 +00:00
btrfs_bio_init(repair_bbio, fs_info, NULL, fbio);
repair_bbio->inode = failed_bbio->inode;
btrfs: handle checksum validation and repair at the storage layer Currently btrfs handles checksum validation and repair in the end I/O handler for the btrfs_bio. This leads to a lot of duplicate code plus issues with varying semantics or bugs, e.g. - the until recently broken repair for compressed extents - the fact that encoded reads validate the checksums but do not kick of read repair - the inconsistent checking of the BTRFS_FS_STATE_NO_CSUMS flag This commit revamps the checksum validation and repair code to instead work below the btrfs_submit_bio interfaces. In case of a checksum failure (or a plain old I/O error), the repair is now kicked off before the upper level ->end_io handler is invoked. Progress of an in-progress repair is tracked by a small structure that is allocated using a mempool for each original bio with failed sectors, which holds a reference to the original bio. This new structure is allocated using a mempool to guarantee forward progress even under memory pressure. The mempool will be replenished when the repair completes, just as the mempools backing the bios. There is one significant behavior change here: If repair fails or is impossible to start with, the whole bio will be failed to the upper layer. This is the behavior that all I/O submitters except for buffered I/O already emulated in their end_io handler. For buffered I/O this now means that a large readahead request can fail due to a single bad sector, but as readahead errors are ignored the following readpage if the sector is actually accessed will still be able to read. This also matches the I/O failure handling in other file systems. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:07 +00:00
repair_bbio->file_offset = failed_bbio->file_offset + bio_offset;
mirror = next_repair_mirror(fbio, failed_bbio->mirror_num);
btrfs_debug(fs_info, "submitting repair read to mirror %d", mirror);
btrfs: pass a btrfs_bio to btrfs_submit_bio btrfs_submit_bio expects the bio passed to it to be embedded into a btrfs_bio structure. Pass the btrfs_bio directly to increase type safety and make the code self-documenting. Reviewed-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-07 16:39:39 +00:00
btrfs_submit_bio(repair_bbio, mirror);
btrfs: handle checksum validation and repair at the storage layer Currently btrfs handles checksum validation and repair in the end I/O handler for the btrfs_bio. This leads to a lot of duplicate code plus issues with varying semantics or bugs, e.g. - the until recently broken repair for compressed extents - the fact that encoded reads validate the checksums but do not kick of read repair - the inconsistent checking of the BTRFS_FS_STATE_NO_CSUMS flag This commit revamps the checksum validation and repair code to instead work below the btrfs_submit_bio interfaces. In case of a checksum failure (or a plain old I/O error), the repair is now kicked off before the upper level ->end_io handler is invoked. Progress of an in-progress repair is tracked by a small structure that is allocated using a mempool for each original bio with failed sectors, which holds a reference to the original bio. This new structure is allocated using a mempool to guarantee forward progress even under memory pressure. The mempool will be replenished when the repair completes, just as the mempools backing the bios. There is one significant behavior change here: If repair fails or is impossible to start with, the whole bio will be failed to the upper layer. This is the behavior that all I/O submitters except for buffered I/O already emulated in their end_io handler. For buffered I/O this now means that a large readahead request can fail due to a single bad sector, but as readahead errors are ignored the following readpage if the sector is actually accessed will still be able to read. This also matches the I/O failure handling in other file systems. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:07 +00:00
return fbio;
}
static void btrfs_check_read_bio(struct btrfs_bio *bbio, struct btrfs_device *dev)
{
struct btrfs_inode *inode = bbio->inode;
struct btrfs_fs_info *fs_info = inode->root->fs_info;
u32 sectorsize = fs_info->sectorsize;
btrfs: rename btrfs_bio::iter field Rename iter to saved_iter and move it next to the repair internals and nothing outside of bio.c should be touching it. Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:13 +00:00
struct bvec_iter *iter = &bbio->saved_iter;
btrfs: handle checksum validation and repair at the storage layer Currently btrfs handles checksum validation and repair in the end I/O handler for the btrfs_bio. This leads to a lot of duplicate code plus issues with varying semantics or bugs, e.g. - the until recently broken repair for compressed extents - the fact that encoded reads validate the checksums but do not kick of read repair - the inconsistent checking of the BTRFS_FS_STATE_NO_CSUMS flag This commit revamps the checksum validation and repair code to instead work below the btrfs_submit_bio interfaces. In case of a checksum failure (or a plain old I/O error), the repair is now kicked off before the upper level ->end_io handler is invoked. Progress of an in-progress repair is tracked by a small structure that is allocated using a mempool for each original bio with failed sectors, which holds a reference to the original bio. This new structure is allocated using a mempool to guarantee forward progress even under memory pressure. The mempool will be replenished when the repair completes, just as the mempools backing the bios. There is one significant behavior change here: If repair fails or is impossible to start with, the whole bio will be failed to the upper layer. This is the behavior that all I/O submitters except for buffered I/O already emulated in their end_io handler. For buffered I/O this now means that a large readahead request can fail due to a single bad sector, but as readahead errors are ignored the following readpage if the sector is actually accessed will still be able to read. This also matches the I/O failure handling in other file systems. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:07 +00:00
blk_status_t status = bbio->bio.bi_status;
struct btrfs_failed_bio *fbio = NULL;
u32 offset = 0;
btrfs: introduce btrfs_bio::fs_info member Currently we're doing a lot of work for btrfs_bio: - Checksum verification for data read bios - Bio splits if it crosses stripe boundary - Read repair for data read bios However for the incoming scrub patches, we don't want this extra functionality at all, just plain logical + mirror -> physical mapping ability. Thus here we do the following changes: - Introduce btrfs_bio::fs_info This is for the new scrub specific btrfs_bio, which would not populate btrfs_bio::inode. Thus we need such new member to grab a fs_info This new member will always be populated. - Replace @inode argument with @fs_info for btrfs_bio_init() and its caller Since @inode is no longer a mandatory member, replace it with @fs_info, and let involved users populate @inode. - Skip checksum verification and generation if @bbio->inode is NULL - Add extra ASSERT()s To make sure: * bbio->inode is properly set for involved read repair path * if @file_offset is set, bbio->inode is also populated - Grab @fs_info from @bbio directly We can no longer go @bbio->inode->root->fs_info, as bbio->inode can be NULL. This involves: * btrfs_simple_end_io() * should_async_write() * btrfs_wq_submit_bio() * btrfs_use_zone_append() Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-23 09:01:20 +00:00
/* Read-repair requires the inode field to be set by the submitter. */
ASSERT(inode);
btrfs: handle checksum validation and repair at the storage layer Currently btrfs handles checksum validation and repair in the end I/O handler for the btrfs_bio. This leads to a lot of duplicate code plus issues with varying semantics or bugs, e.g. - the until recently broken repair for compressed extents - the fact that encoded reads validate the checksums but do not kick of read repair - the inconsistent checking of the BTRFS_FS_STATE_NO_CSUMS flag This commit revamps the checksum validation and repair code to instead work below the btrfs_submit_bio interfaces. In case of a checksum failure (or a plain old I/O error), the repair is now kicked off before the upper level ->end_io handler is invoked. Progress of an in-progress repair is tracked by a small structure that is allocated using a mempool for each original bio with failed sectors, which holds a reference to the original bio. This new structure is allocated using a mempool to guarantee forward progress even under memory pressure. The mempool will be replenished when the repair completes, just as the mempools backing the bios. There is one significant behavior change here: If repair fails or is impossible to start with, the whole bio will be failed to the upper layer. This is the behavior that all I/O submitters except for buffered I/O already emulated in their end_io handler. For buffered I/O this now means that a large readahead request can fail due to a single bad sector, but as readahead errors are ignored the following readpage if the sector is actually accessed will still be able to read. This also matches the I/O failure handling in other file systems. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:07 +00:00
/*
* Hand off repair bios to the repair code as there is no upper level
* submitter for them.
*/
if (bbio->bio.bi_pool == &btrfs_repair_bioset) {
btrfs_end_repair_bio(bbio, dev);
return;
}
/* Clear the I/O error. A failed repair will reset it. */
bbio->bio.bi_status = BLK_STS_OK;
while (iter->bi_size) {
struct bio_vec bv = bio_iter_iovec(&bbio->bio, *iter);
bv.bv_len = min(bv.bv_len, sectorsize);
if (status || !btrfs_data_csum_ok(bbio, dev, offset, &bv))
fbio = repair_one_sector(bbio, offset, &bv, fbio);
bio_advance_iter_single(&bbio->bio, iter, sectorsize);
offset += sectorsize;
}
btrfs: open code btrfs_bio_free_csum btrfs_bio_free_csum has only one caller left, and that caller is always for an data inode and doesn't need zeroing of the csum pointer as that pointer will never be touched again. Just open code the conditional kfree there. Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:08 +00:00
if (bbio->csum != bbio->csum_inline)
kfree(bbio->csum);
btrfs: handle checksum validation and repair at the storage layer Currently btrfs handles checksum validation and repair in the end I/O handler for the btrfs_bio. This leads to a lot of duplicate code plus issues with varying semantics or bugs, e.g. - the until recently broken repair for compressed extents - the fact that encoded reads validate the checksums but do not kick of read repair - the inconsistent checking of the BTRFS_FS_STATE_NO_CSUMS flag This commit revamps the checksum validation and repair code to instead work below the btrfs_submit_bio interfaces. In case of a checksum failure (or a plain old I/O error), the repair is now kicked off before the upper level ->end_io handler is invoked. Progress of an in-progress repair is tracked by a small structure that is allocated using a mempool for each original bio with failed sectors, which holds a reference to the original bio. This new structure is allocated using a mempool to guarantee forward progress even under memory pressure. The mempool will be replenished when the repair completes, just as the mempools backing the bios. There is one significant behavior change here: If repair fails or is impossible to start with, the whole bio will be failed to the upper layer. This is the behavior that all I/O submitters except for buffered I/O already emulated in their end_io handler. For buffered I/O this now means that a large readahead request can fail due to a single bad sector, but as readahead errors are ignored the following readpage if the sector is actually accessed will still be able to read. This also matches the I/O failure handling in other file systems. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:07 +00:00
if (fbio)
btrfs_repair_done(fbio);
else
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
btrfs_orig_bbio_end_io(bbio);
btrfs: handle checksum validation and repair at the storage layer Currently btrfs handles checksum validation and repair in the end I/O handler for the btrfs_bio. This leads to a lot of duplicate code plus issues with varying semantics or bugs, e.g. - the until recently broken repair for compressed extents - the fact that encoded reads validate the checksums but do not kick of read repair - the inconsistent checking of the BTRFS_FS_STATE_NO_CSUMS flag This commit revamps the checksum validation and repair code to instead work below the btrfs_submit_bio interfaces. In case of a checksum failure (or a plain old I/O error), the repair is now kicked off before the upper level ->end_io handler is invoked. Progress of an in-progress repair is tracked by a small structure that is allocated using a mempool for each original bio with failed sectors, which holds a reference to the original bio. This new structure is allocated using a mempool to guarantee forward progress even under memory pressure. The mempool will be replenished when the repair completes, just as the mempools backing the bios. There is one significant behavior change here: If repair fails or is impossible to start with, the whole bio will be failed to the upper layer. This is the behavior that all I/O submitters except for buffered I/O already emulated in their end_io handler. For buffered I/O this now means that a large readahead request can fail due to a single bad sector, but as readahead errors are ignored the following readpage if the sector is actually accessed will still be able to read. This also matches the I/O failure handling in other file systems. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:07 +00:00
}
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
static void btrfs_log_dev_io_error(struct bio *bio, struct btrfs_device *dev)
{
if (!dev || !dev->bdev)
return;
if (bio->bi_status != BLK_STS_IOERR && bio->bi_status != BLK_STS_TARGET)
return;
if (btrfs_op(bio) == BTRFS_MAP_WRITE)
btrfs_dev_stat_inc_and_print(dev, BTRFS_DEV_STAT_WRITE_ERRS);
btrfs: fix unnecessary increment of read error stat on write error Current btrfs_log_dev_io_error() increases the read error count even if the erroneous IO is a WRITE request. This is because it forget to use "else if", and all the error WRITE requests counts as READ error as there is (of course) no REQ_RAHEAD bit set. Fixes: c3a62baf21ad ("btrfs: use chained bios when cloning") CC: stable@vger.kernel.org # 6.1+ Reviewed-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-02-13 05:10:38 +00:00
else if (!(bio->bi_opf & REQ_RAHEAD))
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
btrfs_dev_stat_inc_and_print(dev, BTRFS_DEV_STAT_READ_ERRS);
if (bio->bi_opf & REQ_PREFLUSH)
btrfs_dev_stat_inc_and_print(dev, BTRFS_DEV_STAT_FLUSH_ERRS);
}
static struct workqueue_struct *btrfs_end_io_wq(struct btrfs_fs_info *fs_info,
struct bio *bio)
{
if (bio->bi_opf & REQ_META)
return fs_info->endio_meta_workers;
return fs_info->endio_workers;
}
static void btrfs_end_bio_work(struct work_struct *work)
{
struct btrfs_bio *bbio = container_of(work, struct btrfs_bio, end_io_work);
btrfs: handle checksum validation and repair at the storage layer Currently btrfs handles checksum validation and repair in the end I/O handler for the btrfs_bio. This leads to a lot of duplicate code plus issues with varying semantics or bugs, e.g. - the until recently broken repair for compressed extents - the fact that encoded reads validate the checksums but do not kick of read repair - the inconsistent checking of the BTRFS_FS_STATE_NO_CSUMS flag This commit revamps the checksum validation and repair code to instead work below the btrfs_submit_bio interfaces. In case of a checksum failure (or a plain old I/O error), the repair is now kicked off before the upper level ->end_io handler is invoked. Progress of an in-progress repair is tracked by a small structure that is allocated using a mempool for each original bio with failed sectors, which holds a reference to the original bio. This new structure is allocated using a mempool to guarantee forward progress even under memory pressure. The mempool will be replenished when the repair completes, just as the mempools backing the bios. There is one significant behavior change here: If repair fails or is impossible to start with, the whole bio will be failed to the upper layer. This is the behavior that all I/O submitters except for buffered I/O already emulated in their end_io handler. For buffered I/O this now means that a large readahead request can fail due to a single bad sector, but as readahead errors are ignored the following readpage if the sector is actually accessed will still be able to read. This also matches the I/O failure handling in other file systems. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:07 +00:00
/* Metadata reads are checked and repaired by the submitter. */
btrfs: add a is_data_bbio helper Add a helper to check for that a btrfs_bio has a valid inode, and that it is a data inode to key off all the special handling for data path checksumming. Note that this uses is_data_inode instead of REQ_META as REQ_META is only set directly before submission in submit_one_bio and we'll also want to use this helper for error handling where REQ_META isn't set yet. Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Josef Bacik <josef@toxicpanda.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-05-31 07:54:00 +00:00
if (is_data_bbio(bbio))
btrfs: remove struct btrfs_bio::device field The device field is only used by the simple end I/O handler, and for that it can simply be stored in the bi_private field of the bio, which is currently used for the fs_info that can be retrieved through bbio->inode as well. Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:11 +00:00
btrfs_check_read_bio(bbio, bbio->bio.bi_private);
btrfs: introduce btrfs_bio::fs_info member Currently we're doing a lot of work for btrfs_bio: - Checksum verification for data read bios - Bio splits if it crosses stripe boundary - Read repair for data read bios However for the incoming scrub patches, we don't want this extra functionality at all, just plain logical + mirror -> physical mapping ability. Thus here we do the following changes: - Introduce btrfs_bio::fs_info This is for the new scrub specific btrfs_bio, which would not populate btrfs_bio::inode. Thus we need such new member to grab a fs_info This new member will always be populated. - Replace @inode argument with @fs_info for btrfs_bio_init() and its caller Since @inode is no longer a mandatory member, replace it with @fs_info, and let involved users populate @inode. - Skip checksum verification and generation if @bbio->inode is NULL - Add extra ASSERT()s To make sure: * bbio->inode is properly set for involved read repair path * if @file_offset is set, bbio->inode is also populated - Grab @fs_info from @bbio directly We can no longer go @bbio->inode->root->fs_info, as bbio->inode can be NULL. This involves: * btrfs_simple_end_io() * should_async_write() * btrfs_wq_submit_bio() * btrfs_use_zone_append() Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-23 09:01:20 +00:00
else
btrfs: call btrfs_orig_bbio_end_io in btrfs_end_bio_work When I implemented the storage layer bio splitting, I was under the assumption that we'll never split metadata bios. But Qu reminded me that this can actually happen with very old file systems with unaligned metadata chunks and RAID0. I still haven't seen such a case in practice, but we better handled this case, especially as it is fairly easily to do not calling the ->end_іo method directly in btrfs_end_io_work, and using the proper btrfs_orig_bbio_end_io helper instead. In addition to the old file system with unaligned metadata chunks case documented in the commit log, the combination of the new scrub code with Johannes pending raid-stripe-tree also triggers this case. We spent some time debugging it and found that this patch solves the problem. Fixes: 103c19723c80 ("btrfs: split the bio submission path into a separate file") CC: stable@vger.kernel.org # 6.3+ Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Tested-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-05-15 09:18:21 +00:00
btrfs_orig_bbio_end_io(bbio);
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
}
static void btrfs_simple_end_io(struct bio *bio)
{
struct btrfs_bio *bbio = btrfs_bio(bio);
btrfs: remove struct btrfs_bio::device field The device field is only used by the simple end I/O handler, and for that it can simply be stored in the bi_private field of the bio, which is currently used for the fs_info that can be retrieved through bbio->inode as well. Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:11 +00:00
struct btrfs_device *dev = bio->bi_private;
btrfs: introduce btrfs_bio::fs_info member Currently we're doing a lot of work for btrfs_bio: - Checksum verification for data read bios - Bio splits if it crosses stripe boundary - Read repair for data read bios However for the incoming scrub patches, we don't want this extra functionality at all, just plain logical + mirror -> physical mapping ability. Thus here we do the following changes: - Introduce btrfs_bio::fs_info This is for the new scrub specific btrfs_bio, which would not populate btrfs_bio::inode. Thus we need such new member to grab a fs_info This new member will always be populated. - Replace @inode argument with @fs_info for btrfs_bio_init() and its caller Since @inode is no longer a mandatory member, replace it with @fs_info, and let involved users populate @inode. - Skip checksum verification and generation if @bbio->inode is NULL - Add extra ASSERT()s To make sure: * bbio->inode is properly set for involved read repair path * if @file_offset is set, bbio->inode is also populated - Grab @fs_info from @bbio directly We can no longer go @bbio->inode->root->fs_info, as bbio->inode can be NULL. This involves: * btrfs_simple_end_io() * should_async_write() * btrfs_wq_submit_bio() * btrfs_use_zone_append() Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-23 09:01:20 +00:00
struct btrfs_fs_info *fs_info = bbio->fs_info;
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
btrfs_bio_counter_dec(fs_info);
if (bio->bi_status)
btrfs: remove struct btrfs_bio::device field The device field is only used by the simple end I/O handler, and for that it can simply be stored in the bi_private field of the bio, which is currently used for the fs_info that can be retrieved through bbio->inode as well. Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:11 +00:00
btrfs_log_dev_io_error(bio, dev);
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
if (bio_op(bio) == REQ_OP_READ) {
INIT_WORK(&bbio->end_io_work, btrfs_end_bio_work);
queue_work(btrfs_end_io_wq(fs_info, bio), &bbio->end_io_work);
} else {
btrfs: don't call btrfs_record_physical_zoned for failed append When a zoned append command fails there is no written address reported, so don't try to record it. Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-05-24 15:03:05 +00:00
if (bio_op(bio) == REQ_OP_ZONE_APPEND && !bio->bi_status)
btrfs: handle recording of zoned writes in the storage layer Move the code that splits the ordered extents and records the physical location for them to the storage layer so that the higher level consumers don't have to care about physical block numbers at all. This will also allow to eventually remove accounting for the zone append write sizes in the upper layer with a little bit more block layer work. Reviewed-by: Naohiro Aota <naohiro.aota@wdc.com> Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:18 +00:00
btrfs_record_physical_zoned(bbio);
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
btrfs_orig_bbio_end_io(bbio);
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
}
}
static void btrfs_raid56_end_io(struct bio *bio)
{
struct btrfs_io_context *bioc = bio->bi_private;
struct btrfs_bio *bbio = btrfs_bio(bio);
btrfs_bio_counter_dec(bioc->fs_info);
bbio->mirror_num = bioc->mirror_num;
btrfs: add a is_data_bbio helper Add a helper to check for that a btrfs_bio has a valid inode, and that it is a data inode to key off all the special handling for data path checksumming. Note that this uses is_data_inode instead of REQ_META as REQ_META is only set directly before submission in submit_one_bio and we'll also want to use this helper for error handling where REQ_META isn't set yet. Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Josef Bacik <josef@toxicpanda.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-05-31 07:54:00 +00:00
if (bio_op(bio) == REQ_OP_READ && is_data_bbio(bbio))
btrfs: handle checksum validation and repair at the storage layer Currently btrfs handles checksum validation and repair in the end I/O handler for the btrfs_bio. This leads to a lot of duplicate code plus issues with varying semantics or bugs, e.g. - the until recently broken repair for compressed extents - the fact that encoded reads validate the checksums but do not kick of read repair - the inconsistent checking of the BTRFS_FS_STATE_NO_CSUMS flag This commit revamps the checksum validation and repair code to instead work below the btrfs_submit_bio interfaces. In case of a checksum failure (or a plain old I/O error), the repair is now kicked off before the upper level ->end_io handler is invoked. Progress of an in-progress repair is tracked by a small structure that is allocated using a mempool for each original bio with failed sectors, which holds a reference to the original bio. This new structure is allocated using a mempool to guarantee forward progress even under memory pressure. The mempool will be replenished when the repair completes, just as the mempools backing the bios. There is one significant behavior change here: If repair fails or is impossible to start with, the whole bio will be failed to the upper layer. This is the behavior that all I/O submitters except for buffered I/O already emulated in their end_io handler. For buffered I/O this now means that a large readahead request can fail due to a single bad sector, but as readahead errors are ignored the following readpage if the sector is actually accessed will still be able to read. This also matches the I/O failure handling in other file systems. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:07 +00:00
btrfs_check_read_bio(bbio, NULL);
else
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
btrfs_orig_bbio_end_io(bbio);
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
btrfs_put_bioc(bioc);
}
static void btrfs_orig_write_end_io(struct bio *bio)
{
struct btrfs_io_stripe *stripe = bio->bi_private;
struct btrfs_io_context *bioc = stripe->bioc;
struct btrfs_bio *bbio = btrfs_bio(bio);
btrfs_bio_counter_dec(bioc->fs_info);
if (bio->bi_status) {
atomic_inc(&bioc->error);
btrfs_log_dev_io_error(bio, stripe->dev);
}
/*
* Only send an error to the higher layers if it is beyond the tolerance
* threshold.
*/
if (atomic_read(&bioc->error) > bioc->max_errors)
bio->bi_status = BLK_STS_IOERR;
else
bio->bi_status = BLK_STS_OK;
btrfs: add support for inserting raid stripe extents Add support for inserting stripe extents into the raid stripe tree on completion of every write that needs an extra logical-to-physical translation when using RAID. Inserting the stripe extents happens after the data I/O has completed, this is done to a) support zone-append and b) rule out the possibility of a RAID-write-hole. Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-09-14 16:06:58 +00:00
if (bio_op(bio) == REQ_OP_ZONE_APPEND && !bio->bi_status)
stripe->physical = bio->bi_iter.bi_sector << SECTOR_SHIFT;
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
btrfs_orig_bbio_end_io(bbio);
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
btrfs_put_bioc(bioc);
}
static void btrfs_clone_write_end_io(struct bio *bio)
{
struct btrfs_io_stripe *stripe = bio->bi_private;
if (bio->bi_status) {
atomic_inc(&stripe->bioc->error);
btrfs_log_dev_io_error(bio, stripe->dev);
btrfs: add support for inserting raid stripe extents Add support for inserting stripe extents into the raid stripe tree on completion of every write that needs an extra logical-to-physical translation when using RAID. Inserting the stripe extents happens after the data I/O has completed, this is done to a) support zone-append and b) rule out the possibility of a RAID-write-hole. Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-09-14 16:06:58 +00:00
} else if (bio_op(bio) == REQ_OP_ZONE_APPEND) {
stripe->physical = bio->bi_iter.bi_sector << SECTOR_SHIFT;
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
}
/* Pass on control to the original bio this one was cloned from */
bio_endio(stripe->bioc->orig_bio);
bio_put(bio);
}
static void btrfs_submit_dev_bio(struct btrfs_device *dev, struct bio *bio)
{
if (!dev || !dev->bdev ||
test_bit(BTRFS_DEV_STATE_MISSING, &dev->dev_state) ||
(btrfs_op(bio) == BTRFS_MAP_WRITE &&
!test_bit(BTRFS_DEV_STATE_WRITEABLE, &dev->dev_state))) {
bio_io_error(bio);
return;
}
bio_set_dev(bio, dev->bdev);
/*
* For zone append writing, bi_sector must point the beginning of the
* zone
*/
if (bio_op(bio) == REQ_OP_ZONE_APPEND) {
u64 physical = bio->bi_iter.bi_sector << SECTOR_SHIFT;
btrfs: split zone append bios in btrfs_submit_bio The current btrfs zoned device support is a little cumbersome in the data I/O path as it requires the callers to not issue I/O larger than the supported ZONE_APPEND size of the underlying device. This leads to a lot of extra accounting. Instead change btrfs_submit_bio so that it can take write bios of arbitrary size and form from the upper layers, and just split them internally to the ZONE_APPEND queue limits. Then remove all the upper layer warts catering to limited write sized on zoned devices, including the extra refcount in the compressed_bio. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:30 +00:00
u64 zone_start = round_down(physical, dev->fs_info->zone_size);
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
btrfs: split zone append bios in btrfs_submit_bio The current btrfs zoned device support is a little cumbersome in the data I/O path as it requires the callers to not issue I/O larger than the supported ZONE_APPEND size of the underlying device. This leads to a lot of extra accounting. Instead change btrfs_submit_bio so that it can take write bios of arbitrary size and form from the upper layers, and just split them internally to the ZONE_APPEND queue limits. Then remove all the upper layer warts catering to limited write sized on zoned devices, including the extra refcount in the compressed_bio. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:30 +00:00
ASSERT(btrfs_dev_is_sequential(dev, physical));
bio->bi_iter.bi_sector = zone_start >> SECTOR_SHIFT;
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
}
btrfs_debug_in_rcu(dev->fs_info,
"%s: rw %d 0x%x, sector=%llu, dev=%lu (%s id %llu), size=%u",
__func__, bio_op(bio), bio->bi_opf, bio->bi_iter.bi_sector,
(unsigned long)dev->bdev->bd_dev, btrfs_dev_name(dev),
dev->devid, bio->bi_iter.bi_size);
btrfs, block: move REQ_CGROUP_PUNT to btrfs REQ_CGROUP_PUNT is a bit annoying as it is hard to follow and adds a branch to the bio submission hot path. To fix this, export blkcg_punt_bio_submit and let btrfs call it directly. Add a new REQ_FS_PRIVATE flag for btrfs to indicate to it's own low-level bio submission code that a punt to the cgroup submission helper is required. Reviewed-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-27 00:49:51 +00:00
if (bio->bi_opf & REQ_BTRFS_CGROUP_PUNT)
blkcg_punt_bio_submit(bio);
else
submit_bio(bio);
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
}
static void btrfs_submit_mirrored_bio(struct btrfs_io_context *bioc, int dev_nr)
{
struct bio *orig_bio = bioc->orig_bio, *bio;
ASSERT(bio_op(orig_bio) != REQ_OP_READ);
/* Reuse the bio embedded into the btrfs_bio for the last mirror */
if (dev_nr == bioc->num_stripes - 1) {
bio = orig_bio;
bio->bi_end_io = btrfs_orig_write_end_io;
} else {
bio = bio_alloc_clone(NULL, orig_bio, GFP_NOFS, &fs_bio_set);
bio_inc_remaining(orig_bio);
bio->bi_end_io = btrfs_clone_write_end_io;
}
bio->bi_private = &bioc->stripes[dev_nr];
bio->bi_iter.bi_sector = bioc->stripes[dev_nr].physical >> SECTOR_SHIFT;
bioc->stripes[dev_nr].bioc = bioc;
btrfs: add support for inserting raid stripe extents Add support for inserting stripe extents into the raid stripe tree on completion of every write that needs an extra logical-to-physical translation when using RAID. Inserting the stripe extents happens after the data I/O has completed, this is done to a) support zone-append and b) rule out the possibility of a RAID-write-hole. Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-09-14 16:06:58 +00:00
bioc->size = bio->bi_iter.bi_size;
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
btrfs_submit_dev_bio(bioc->stripes[dev_nr].dev, bio);
}
btrfs: handle checksum generation in the storage layer Instead of letting the callers of btrfs_submit_bio deal with checksumming the (meta)data in the bio and making decisions on when to offload the checksumming to the bio, leave that to btrfs_submit_bio. Do do so the existing btrfs_submit_bio function is split into an upper and a lower half, so that the lower half can be offloaded to a workqueue. Note that this changes the behavior for direct writes to raid56 volumes so that async checksum offloading is not skipped when more I/O is expected. This runs counter to the argument explaining why it was done, although I can't measure any affects of the change. Commits later in this series will make sure the entire direct writes is offloaded to the workqueue at once and thus make sure it is sent to the raid56 code from a single thread. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:17 +00:00
static void __btrfs_submit_bio(struct bio *bio, struct btrfs_io_context *bioc,
struct btrfs_io_stripe *smap, int mirror_num)
{
if (!bioc) {
/* Single mirror read/write fast path. */
btrfs_bio(bio)->mirror_num = mirror_num;
bio->bi_iter.bi_sector = smap->physical >> SECTOR_SHIFT;
btrfs: record orig_physical only for the original bio btrfs_submit_dev_bio is also called for clone bios that aren't embedded into a btrfs_bio structure, but previous commit "btrfs: optimize the logical to physical mapping for zoned writes" added code to assign btrfs_bio.orig_physical in it. This is harmless right now as only the single data profile can be used on zoned devices, but will blow up when the RAID stripe tree is added. Move it out into the single I/O specific branch in the caller. Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-06-09 05:27:04 +00:00
if (bio_op(bio) != REQ_OP_READ)
btrfs_bio(bio)->orig_physical = smap->physical;
btrfs: handle checksum generation in the storage layer Instead of letting the callers of btrfs_submit_bio deal with checksumming the (meta)data in the bio and making decisions on when to offload the checksumming to the bio, leave that to btrfs_submit_bio. Do do so the existing btrfs_submit_bio function is split into an upper and a lower half, so that the lower half can be offloaded to a workqueue. Note that this changes the behavior for direct writes to raid56 volumes so that async checksum offloading is not skipped when more I/O is expected. This runs counter to the argument explaining why it was done, although I can't measure any affects of the change. Commits later in this series will make sure the entire direct writes is offloaded to the workqueue at once and thus make sure it is sent to the raid56 code from a single thread. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:17 +00:00
bio->bi_private = smap->dev;
bio->bi_end_io = btrfs_simple_end_io;
btrfs_submit_dev_bio(smap->dev, bio);
} else if (bioc->map_type & BTRFS_BLOCK_GROUP_RAID56_MASK) {
/* Parity RAID write or read recovery. */
bio->bi_private = bioc;
bio->bi_end_io = btrfs_raid56_end_io;
if (bio_op(bio) == REQ_OP_READ)
raid56_parity_recover(bio, bioc, mirror_num);
else
raid56_parity_write(bio, bioc);
} else {
/* Write to multiple mirrors. */
int total_devs = bioc->num_stripes;
bioc->orig_bio = bio;
for (int dev_nr = 0; dev_nr < total_devs; dev_nr++)
btrfs_submit_mirrored_bio(bioc, dev_nr);
}
}
static blk_status_t btrfs_bio_csum(struct btrfs_bio *bbio)
{
if (bbio->bio.bi_opf & REQ_META)
btrfs: support cloned bios in btree_csum_one_bio To allow splitting bios in btrfs_submit_bio, btree_csum_one_bio needs to be able to handle cloned bios. As btree_csum_one_bio is always called before handing the bio to the block layer that is trivially done by using bio_for_each_segment instead of bio_for_each_segment_all. Also switch the function to take a btrfs_bio and use that to derive the fs_info. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:19 +00:00
return btree_csum_one_bio(bbio);
btrfs: handle checksum generation in the storage layer Instead of letting the callers of btrfs_submit_bio deal with checksumming the (meta)data in the bio and making decisions on when to offload the checksumming to the bio, leave that to btrfs_submit_bio. Do do so the existing btrfs_submit_bio function is split into an upper and a lower half, so that the lower half can be offloaded to a workqueue. Note that this changes the behavior for direct writes to raid56 volumes so that async checksum offloading is not skipped when more I/O is expected. This runs counter to the argument explaining why it was done, although I can't measure any affects of the change. Commits later in this series will make sure the entire direct writes is offloaded to the workqueue at once and thus make sure it is sent to the raid56 code from a single thread. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:17 +00:00
return btrfs_csum_one_bio(bbio);
}
/*
* Async submit bios are used to offload expensive checksumming onto the worker
* threads.
*/
struct async_submit_bio {
struct btrfs_bio *bbio;
struct btrfs_io_context *bioc;
struct btrfs_io_stripe smap;
int mirror_num;
struct btrfs_work work;
};
/*
* In order to insert checksums into the metadata in large chunks, we wait
* until bio submission time. All the pages in the bio are checksummed and
* sums are attached onto the ordered extent record.
*
* At IO completion time the csums attached on the ordered extent record are
* inserted into the btree.
*/
static void run_one_async_start(struct btrfs_work *work)
{
struct async_submit_bio *async =
container_of(work, struct async_submit_bio, work);
blk_status_t ret;
ret = btrfs_bio_csum(async->bbio);
if (ret)
async->bbio->bio.bi_status = ret;
}
/*
* In order to insert checksums into the metadata in large chunks, we wait
* until bio submission time. All the pages in the bio are checksummed and
* sums are attached onto the ordered extent record.
*
* At IO completion time the csums attached on the ordered extent record are
* inserted into the tree.
btrfs: merge ordered work callbacks in btrfs_work into one There are two callbacks defined in btrfs_work but only two actually make use of them, otherwise there are NULLs. We can get rid of the freeing callback making it a special case of the normal work. This reduces the size of btrfs_work by 8 bytes, final layout: struct btrfs_work { btrfs_func_t func; /* 0 8 */ btrfs_ordered_func_t ordered_func; /* 8 8 */ struct work_struct normal_work; /* 16 32 */ struct list_head ordered_list; /* 48 16 */ /* --- cacheline 1 boundary (64 bytes) --- */ struct btrfs_workqueue * wq; /* 64 8 */ long unsigned int flags; /* 72 8 */ /* size: 80, cachelines: 2, members: 6 */ /* last cacheline: 16 bytes */ }; This in turn reduces size of other structures (on a release config): - async_chunk 160 -> 152 - async_submit_bio 152 -> 144 - btrfs_async_delayed_work 104 -> 96 - btrfs_caching_control 176 -> 168 - btrfs_delalloc_work 144 -> 136 - btrfs_fs_info 3608 -> 3600 - btrfs_ordered_extent 440 -> 424 - btrfs_writepage_fixup 104 -> 96 Signed-off-by: David Sterba <dsterba@suse.com>
2023-09-19 16:49:23 +00:00
*
* If called with @do_free == true, then it will free the work struct.
btrfs: handle checksum generation in the storage layer Instead of letting the callers of btrfs_submit_bio deal with checksumming the (meta)data in the bio and making decisions on when to offload the checksumming to the bio, leave that to btrfs_submit_bio. Do do so the existing btrfs_submit_bio function is split into an upper and a lower half, so that the lower half can be offloaded to a workqueue. Note that this changes the behavior for direct writes to raid56 volumes so that async checksum offloading is not skipped when more I/O is expected. This runs counter to the argument explaining why it was done, although I can't measure any affects of the change. Commits later in this series will make sure the entire direct writes is offloaded to the workqueue at once and thus make sure it is sent to the raid56 code from a single thread. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:17 +00:00
*/
btrfs: merge ordered work callbacks in btrfs_work into one There are two callbacks defined in btrfs_work but only two actually make use of them, otherwise there are NULLs. We can get rid of the freeing callback making it a special case of the normal work. This reduces the size of btrfs_work by 8 bytes, final layout: struct btrfs_work { btrfs_func_t func; /* 0 8 */ btrfs_ordered_func_t ordered_func; /* 8 8 */ struct work_struct normal_work; /* 16 32 */ struct list_head ordered_list; /* 48 16 */ /* --- cacheline 1 boundary (64 bytes) --- */ struct btrfs_workqueue * wq; /* 64 8 */ long unsigned int flags; /* 72 8 */ /* size: 80, cachelines: 2, members: 6 */ /* last cacheline: 16 bytes */ }; This in turn reduces size of other structures (on a release config): - async_chunk 160 -> 152 - async_submit_bio 152 -> 144 - btrfs_async_delayed_work 104 -> 96 - btrfs_caching_control 176 -> 168 - btrfs_delalloc_work 144 -> 136 - btrfs_fs_info 3608 -> 3600 - btrfs_ordered_extent 440 -> 424 - btrfs_writepage_fixup 104 -> 96 Signed-off-by: David Sterba <dsterba@suse.com>
2023-09-19 16:49:23 +00:00
static void run_one_async_done(struct btrfs_work *work, bool do_free)
btrfs: handle checksum generation in the storage layer Instead of letting the callers of btrfs_submit_bio deal with checksumming the (meta)data in the bio and making decisions on when to offload the checksumming to the bio, leave that to btrfs_submit_bio. Do do so the existing btrfs_submit_bio function is split into an upper and a lower half, so that the lower half can be offloaded to a workqueue. Note that this changes the behavior for direct writes to raid56 volumes so that async checksum offloading is not skipped when more I/O is expected. This runs counter to the argument explaining why it was done, although I can't measure any affects of the change. Commits later in this series will make sure the entire direct writes is offloaded to the workqueue at once and thus make sure it is sent to the raid56 code from a single thread. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:17 +00:00
{
struct async_submit_bio *async =
container_of(work, struct async_submit_bio, work);
struct bio *bio = &async->bbio->bio;
btrfs: merge ordered work callbacks in btrfs_work into one There are two callbacks defined in btrfs_work but only two actually make use of them, otherwise there are NULLs. We can get rid of the freeing callback making it a special case of the normal work. This reduces the size of btrfs_work by 8 bytes, final layout: struct btrfs_work { btrfs_func_t func; /* 0 8 */ btrfs_ordered_func_t ordered_func; /* 8 8 */ struct work_struct normal_work; /* 16 32 */ struct list_head ordered_list; /* 48 16 */ /* --- cacheline 1 boundary (64 bytes) --- */ struct btrfs_workqueue * wq; /* 64 8 */ long unsigned int flags; /* 72 8 */ /* size: 80, cachelines: 2, members: 6 */ /* last cacheline: 16 bytes */ }; This in turn reduces size of other structures (on a release config): - async_chunk 160 -> 152 - async_submit_bio 152 -> 144 - btrfs_async_delayed_work 104 -> 96 - btrfs_caching_control 176 -> 168 - btrfs_delalloc_work 144 -> 136 - btrfs_fs_info 3608 -> 3600 - btrfs_ordered_extent 440 -> 424 - btrfs_writepage_fixup 104 -> 96 Signed-off-by: David Sterba <dsterba@suse.com>
2023-09-19 16:49:23 +00:00
if (do_free) {
kfree(container_of(work, struct async_submit_bio, work));
return;
}
btrfs: handle checksum generation in the storage layer Instead of letting the callers of btrfs_submit_bio deal with checksumming the (meta)data in the bio and making decisions on when to offload the checksumming to the bio, leave that to btrfs_submit_bio. Do do so the existing btrfs_submit_bio function is split into an upper and a lower half, so that the lower half can be offloaded to a workqueue. Note that this changes the behavior for direct writes to raid56 volumes so that async checksum offloading is not skipped when more I/O is expected. This runs counter to the argument explaining why it was done, although I can't measure any affects of the change. Commits later in this series will make sure the entire direct writes is offloaded to the workqueue at once and thus make sure it is sent to the raid56 code from a single thread. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:17 +00:00
/* If an error occurred we just want to clean up the bio and move on. */
if (bio->bi_status) {
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
btrfs_orig_bbio_end_io(async->bbio);
btrfs: handle checksum generation in the storage layer Instead of letting the callers of btrfs_submit_bio deal with checksumming the (meta)data in the bio and making decisions on when to offload the checksumming to the bio, leave that to btrfs_submit_bio. Do do so the existing btrfs_submit_bio function is split into an upper and a lower half, so that the lower half can be offloaded to a workqueue. Note that this changes the behavior for direct writes to raid56 volumes so that async checksum offloading is not skipped when more I/O is expected. This runs counter to the argument explaining why it was done, although I can't measure any affects of the change. Commits later in this series will make sure the entire direct writes is offloaded to the workqueue at once and thus make sure it is sent to the raid56 code from a single thread. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:17 +00:00
return;
}
/*
* All of the bios that pass through here are from async helpers.
btrfs, block: move REQ_CGROUP_PUNT to btrfs REQ_CGROUP_PUNT is a bit annoying as it is hard to follow and adds a branch to the bio submission hot path. To fix this, export blkcg_punt_bio_submit and let btrfs call it directly. Add a new REQ_FS_PRIVATE flag for btrfs to indicate to it's own low-level bio submission code that a punt to the cgroup submission helper is required. Reviewed-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-27 00:49:51 +00:00
* Use REQ_BTRFS_CGROUP_PUNT to issue them from the owning cgroup's
* context. This changes nothing when cgroups aren't in use.
btrfs: handle checksum generation in the storage layer Instead of letting the callers of btrfs_submit_bio deal with checksumming the (meta)data in the bio and making decisions on when to offload the checksumming to the bio, leave that to btrfs_submit_bio. Do do so the existing btrfs_submit_bio function is split into an upper and a lower half, so that the lower half can be offloaded to a workqueue. Note that this changes the behavior for direct writes to raid56 volumes so that async checksum offloading is not skipped when more I/O is expected. This runs counter to the argument explaining why it was done, although I can't measure any affects of the change. Commits later in this series will make sure the entire direct writes is offloaded to the workqueue at once and thus make sure it is sent to the raid56 code from a single thread. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:17 +00:00
*/
btrfs, block: move REQ_CGROUP_PUNT to btrfs REQ_CGROUP_PUNT is a bit annoying as it is hard to follow and adds a branch to the bio submission hot path. To fix this, export blkcg_punt_bio_submit and let btrfs call it directly. Add a new REQ_FS_PRIVATE flag for btrfs to indicate to it's own low-level bio submission code that a punt to the cgroup submission helper is required. Reviewed-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-27 00:49:51 +00:00
bio->bi_opf |= REQ_BTRFS_CGROUP_PUNT;
btrfs: handle checksum generation in the storage layer Instead of letting the callers of btrfs_submit_bio deal with checksumming the (meta)data in the bio and making decisions on when to offload the checksumming to the bio, leave that to btrfs_submit_bio. Do do so the existing btrfs_submit_bio function is split into an upper and a lower half, so that the lower half can be offloaded to a workqueue. Note that this changes the behavior for direct writes to raid56 volumes so that async checksum offloading is not skipped when more I/O is expected. This runs counter to the argument explaining why it was done, although I can't measure any affects of the change. Commits later in this series will make sure the entire direct writes is offloaded to the workqueue at once and thus make sure it is sent to the raid56 code from a single thread. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:17 +00:00
__btrfs_submit_bio(bio, async->bioc, &async->smap, async->mirror_num);
}
static bool should_async_write(struct btrfs_bio *bbio)
{
btrfs: introduce offload_csum_mode to tweak checksum offloading behavior We disable offloading checksum to workqueues and do it synchronously when the checksum algorithm is fast. However, as reported in the link below, RAID0 with multiple devices may suffer from the sync checksum, because "fast checksum" is still not fast enough to catch up with RAID0 writing. We don't have an effective way to determine whether to offload or not, for now add a sysfs knob so this can be debugged. This is intentionally under CONFIG_BTRFS_DEBUG so ti's not exposed to users as it may be removed in the future agin. Introduce fs_devices->offload_csum_mode, so that a btrfs developer can change the behavior by writing to /sys/fs/btrfs/<uuid>/offload_csum. The default is "auto" which is the same as the previous behavior. Or, you can set "on" or "off" (or "y" or "n" whatever kstrtobool() accepts) to always/never offload checksum. More benchmark need to be collected with this knob to implement a proper criteria to enable/disable checksum offloading. Link: https://lore.kernel.org/linux-btrfs/20230731152223.4EFB.409509F4@e16-tech.com/ Link: https://lore.kernel.org/linux-btrfs/p3vo3g7pqn664mhmdhlotu5dzcna6vjtcoc2hb2lsgo2fwct7k@xzaxclba5tae/ Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2024-02-05 13:01:16 +00:00
bool auto_csum_mode = true;
#ifdef CONFIG_BTRFS_DEBUG
struct btrfs_fs_devices *fs_devices = bbio->fs_info->fs_devices;
enum btrfs_offload_csum_mode csum_mode = READ_ONCE(fs_devices->offload_csum_mode);
if (csum_mode == BTRFS_OFFLOAD_CSUM_FORCE_OFF)
return false;
auto_csum_mode = (csum_mode == BTRFS_OFFLOAD_CSUM_AUTO);
#endif
btrfs: submit IO synchronously for fast checksum implementations Most modern hardware supports very fast accelerated crc32c calculation. If that is supported the CPU overhead of the checksum calculation is very limited, and offloading the calculation to special worker threads has a lot of overhead for no gain. E.g. on an Intel Optane device is actually very much slows down even 1M buffered writes with fio: Unpatched: write: IOPS=3316, BW=3316MiB/s (3477MB/s)(200GiB/61757msec); 0 zone resets With synchronous CRCs: write: IOPS=4882, BW=4882MiB/s (5119MB/s)(200GiB/41948msec); 0 zone resets With a lot of variation during the unpatched run going down as low as 1100MB/s, while the synchronous CRC version has about the same peak write speed but much lower dips, and fewer kworkers churning around. Both tests had fio saturated at 100% CPU. (thanks to Jens Axboe via Chris Mason for the benchmarking) Reviewed-by: Chris Mason <clm@fb.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-05-03 07:06:13 +00:00
/* Submit synchronously if the checksum implementation is fast. */
btrfs: introduce offload_csum_mode to tweak checksum offloading behavior We disable offloading checksum to workqueues and do it synchronously when the checksum algorithm is fast. However, as reported in the link below, RAID0 with multiple devices may suffer from the sync checksum, because "fast checksum" is still not fast enough to catch up with RAID0 writing. We don't have an effective way to determine whether to offload or not, for now add a sysfs knob so this can be debugged. This is intentionally under CONFIG_BTRFS_DEBUG so ti's not exposed to users as it may be removed in the future agin. Introduce fs_devices->offload_csum_mode, so that a btrfs developer can change the behavior by writing to /sys/fs/btrfs/<uuid>/offload_csum. The default is "auto" which is the same as the previous behavior. Or, you can set "on" or "off" (or "y" or "n" whatever kstrtobool() accepts) to always/never offload checksum. More benchmark need to be collected with this knob to implement a proper criteria to enable/disable checksum offloading. Link: https://lore.kernel.org/linux-btrfs/20230731152223.4EFB.409509F4@e16-tech.com/ Link: https://lore.kernel.org/linux-btrfs/p3vo3g7pqn664mhmdhlotu5dzcna6vjtcoc2hb2lsgo2fwct7k@xzaxclba5tae/ Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2024-02-05 13:01:16 +00:00
if (auto_csum_mode && test_bit(BTRFS_FS_CSUM_IMPL_FAST, &bbio->fs_info->flags))
btrfs: submit IO synchronously for fast checksum implementations Most modern hardware supports very fast accelerated crc32c calculation. If that is supported the CPU overhead of the checksum calculation is very limited, and offloading the calculation to special worker threads has a lot of overhead for no gain. E.g. on an Intel Optane device is actually very much slows down even 1M buffered writes with fio: Unpatched: write: IOPS=3316, BW=3316MiB/s (3477MB/s)(200GiB/61757msec); 0 zone resets With synchronous CRCs: write: IOPS=4882, BW=4882MiB/s (5119MB/s)(200GiB/41948msec); 0 zone resets With a lot of variation during the unpatched run going down as low as 1100MB/s, while the synchronous CRC version has about the same peak write speed but much lower dips, and fewer kworkers churning around. Both tests had fio saturated at 100% CPU. (thanks to Jens Axboe via Chris Mason for the benchmarking) Reviewed-by: Chris Mason <clm@fb.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-05-03 07:06:13 +00:00
return false;
btrfs: handle checksum generation in the storage layer Instead of letting the callers of btrfs_submit_bio deal with checksumming the (meta)data in the bio and making decisions on when to offload the checksumming to the bio, leave that to btrfs_submit_bio. Do do so the existing btrfs_submit_bio function is split into an upper and a lower half, so that the lower half can be offloaded to a workqueue. Note that this changes the behavior for direct writes to raid56 volumes so that async checksum offloading is not skipped when more I/O is expected. This runs counter to the argument explaining why it was done, although I can't measure any affects of the change. Commits later in this series will make sure the entire direct writes is offloaded to the workqueue at once and thus make sure it is sent to the raid56 code from a single thread. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:17 +00:00
/*
btrfs: determine synchronous writers from bio or writeback control The writeback_control structure already passes down the information about a writeback being synchronous from the core VM code, and thus information is propagated into the bio REQ_SYNC flag through the wbc_to_write_flags helper. Use that information to decide if checksums calculation is offloaded to a workqueue instead of btrfs_inode::sync_writers field that not only bloats the inode but also has too wide scope, being inode wide instead of limited to the actual writeback request. The sync writes were set in: - btrfs_do_write_iter - regular IO, sync status is set - start_ordered_ops - ordered write start, writeback with WB_SYNC_ALL mode - btrfs_write_marked_extents - write marked extents, writeback with WB_SYNC_ALL mode Reviewed-by: Chris Mason <clm@fb.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> [ update changelog ] Signed-off-by: David Sterba <dsterba@suse.com>
2023-05-03 07:06:14 +00:00
* Try to defer the submission to a workqueue to parallelize the
* checksum calculation unless the I/O is issued synchronously.
btrfs: handle checksum generation in the storage layer Instead of letting the callers of btrfs_submit_bio deal with checksumming the (meta)data in the bio and making decisions on when to offload the checksumming to the bio, leave that to btrfs_submit_bio. Do do so the existing btrfs_submit_bio function is split into an upper and a lower half, so that the lower half can be offloaded to a workqueue. Note that this changes the behavior for direct writes to raid56 volumes so that async checksum offloading is not skipped when more I/O is expected. This runs counter to the argument explaining why it was done, although I can't measure any affects of the change. Commits later in this series will make sure the entire direct writes is offloaded to the workqueue at once and thus make sure it is sent to the raid56 code from a single thread. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:17 +00:00
*/
btrfs: determine synchronous writers from bio or writeback control The writeback_control structure already passes down the information about a writeback being synchronous from the core VM code, and thus information is propagated into the bio REQ_SYNC flag through the wbc_to_write_flags helper. Use that information to decide if checksums calculation is offloaded to a workqueue instead of btrfs_inode::sync_writers field that not only bloats the inode but also has too wide scope, being inode wide instead of limited to the actual writeback request. The sync writes were set in: - btrfs_do_write_iter - regular IO, sync status is set - start_ordered_ops - ordered write start, writeback with WB_SYNC_ALL mode - btrfs_write_marked_extents - write marked extents, writeback with WB_SYNC_ALL mode Reviewed-by: Chris Mason <clm@fb.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> [ update changelog ] Signed-off-by: David Sterba <dsterba@suse.com>
2023-05-03 07:06:14 +00:00
if (op_is_sync(bbio->bio.bi_opf))
btrfs: handle checksum generation in the storage layer Instead of letting the callers of btrfs_submit_bio deal with checksumming the (meta)data in the bio and making decisions on when to offload the checksumming to the bio, leave that to btrfs_submit_bio. Do do so the existing btrfs_submit_bio function is split into an upper and a lower half, so that the lower half can be offloaded to a workqueue. Note that this changes the behavior for direct writes to raid56 volumes so that async checksum offloading is not skipped when more I/O is expected. This runs counter to the argument explaining why it was done, although I can't measure any affects of the change. Commits later in this series will make sure the entire direct writes is offloaded to the workqueue at once and thus make sure it is sent to the raid56 code from a single thread. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:17 +00:00
return false;
btrfs: submit IO synchronously for fast checksum implementations Most modern hardware supports very fast accelerated crc32c calculation. If that is supported the CPU overhead of the checksum calculation is very limited, and offloading the calculation to special worker threads has a lot of overhead for no gain. E.g. on an Intel Optane device is actually very much slows down even 1M buffered writes with fio: Unpatched: write: IOPS=3316, BW=3316MiB/s (3477MB/s)(200GiB/61757msec); 0 zone resets With synchronous CRCs: write: IOPS=4882, BW=4882MiB/s (5119MB/s)(200GiB/41948msec); 0 zone resets With a lot of variation during the unpatched run going down as low as 1100MB/s, while the synchronous CRC version has about the same peak write speed but much lower dips, and fewer kworkers churning around. Both tests had fio saturated at 100% CPU. (thanks to Jens Axboe via Chris Mason for the benchmarking) Reviewed-by: Chris Mason <clm@fb.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-05-03 07:06:13 +00:00
/* Zoned devices require I/O to be submitted in order. */
if ((bbio->bio.bi_opf & REQ_META) && btrfs_is_zoned(bbio->fs_info))
return false;
btrfs: handle checksum generation in the storage layer Instead of letting the callers of btrfs_submit_bio deal with checksumming the (meta)data in the bio and making decisions on when to offload the checksumming to the bio, leave that to btrfs_submit_bio. Do do so the existing btrfs_submit_bio function is split into an upper and a lower half, so that the lower half can be offloaded to a workqueue. Note that this changes the behavior for direct writes to raid56 volumes so that async checksum offloading is not skipped when more I/O is expected. This runs counter to the argument explaining why it was done, although I can't measure any affects of the change. Commits later in this series will make sure the entire direct writes is offloaded to the workqueue at once and thus make sure it is sent to the raid56 code from a single thread. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:17 +00:00
return true;
}
/*
* Submit bio to an async queue.
*
btrfs: fix typos found by codespell Signed-off-by: David Sterba <dsterba@suse.com>
2023-12-05 18:26:39 +00:00
* Return true if the work has been successfully submitted, else false.
btrfs: handle checksum generation in the storage layer Instead of letting the callers of btrfs_submit_bio deal with checksumming the (meta)data in the bio and making decisions on when to offload the checksumming to the bio, leave that to btrfs_submit_bio. Do do so the existing btrfs_submit_bio function is split into an upper and a lower half, so that the lower half can be offloaded to a workqueue. Note that this changes the behavior for direct writes to raid56 volumes so that async checksum offloading is not skipped when more I/O is expected. This runs counter to the argument explaining why it was done, although I can't measure any affects of the change. Commits later in this series will make sure the entire direct writes is offloaded to the workqueue at once and thus make sure it is sent to the raid56 code from a single thread. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:17 +00:00
*/
static bool btrfs_wq_submit_bio(struct btrfs_bio *bbio,
struct btrfs_io_context *bioc,
struct btrfs_io_stripe *smap, int mirror_num)
{
btrfs: introduce btrfs_bio::fs_info member Currently we're doing a lot of work for btrfs_bio: - Checksum verification for data read bios - Bio splits if it crosses stripe boundary - Read repair for data read bios However for the incoming scrub patches, we don't want this extra functionality at all, just plain logical + mirror -> physical mapping ability. Thus here we do the following changes: - Introduce btrfs_bio::fs_info This is for the new scrub specific btrfs_bio, which would not populate btrfs_bio::inode. Thus we need such new member to grab a fs_info This new member will always be populated. - Replace @inode argument with @fs_info for btrfs_bio_init() and its caller Since @inode is no longer a mandatory member, replace it with @fs_info, and let involved users populate @inode. - Skip checksum verification and generation if @bbio->inode is NULL - Add extra ASSERT()s To make sure: * bbio->inode is properly set for involved read repair path * if @file_offset is set, bbio->inode is also populated - Grab @fs_info from @bbio directly We can no longer go @bbio->inode->root->fs_info, as bbio->inode can be NULL. This involves: * btrfs_simple_end_io() * should_async_write() * btrfs_wq_submit_bio() * btrfs_use_zone_append() Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-23 09:01:20 +00:00
struct btrfs_fs_info *fs_info = bbio->fs_info;
btrfs: handle checksum generation in the storage layer Instead of letting the callers of btrfs_submit_bio deal with checksumming the (meta)data in the bio and making decisions on when to offload the checksumming to the bio, leave that to btrfs_submit_bio. Do do so the existing btrfs_submit_bio function is split into an upper and a lower half, so that the lower half can be offloaded to a workqueue. Note that this changes the behavior for direct writes to raid56 volumes so that async checksum offloading is not skipped when more I/O is expected. This runs counter to the argument explaining why it was done, although I can't measure any affects of the change. Commits later in this series will make sure the entire direct writes is offloaded to the workqueue at once and thus make sure it is sent to the raid56 code from a single thread. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:17 +00:00
struct async_submit_bio *async;
async = kmalloc(sizeof(*async), GFP_NOFS);
if (!async)
return false;
async->bbio = bbio;
async->bioc = bioc;
async->smap = *smap;
async->mirror_num = mirror_num;
btrfs: merge ordered work callbacks in btrfs_work into one There are two callbacks defined in btrfs_work but only two actually make use of them, otherwise there are NULLs. We can get rid of the freeing callback making it a special case of the normal work. This reduces the size of btrfs_work by 8 bytes, final layout: struct btrfs_work { btrfs_func_t func; /* 0 8 */ btrfs_ordered_func_t ordered_func; /* 8 8 */ struct work_struct normal_work; /* 16 32 */ struct list_head ordered_list; /* 48 16 */ /* --- cacheline 1 boundary (64 bytes) --- */ struct btrfs_workqueue * wq; /* 64 8 */ long unsigned int flags; /* 72 8 */ /* size: 80, cachelines: 2, members: 6 */ /* last cacheline: 16 bytes */ }; This in turn reduces size of other structures (on a release config): - async_chunk 160 -> 152 - async_submit_bio 152 -> 144 - btrfs_async_delayed_work 104 -> 96 - btrfs_caching_control 176 -> 168 - btrfs_delalloc_work 144 -> 136 - btrfs_fs_info 3608 -> 3600 - btrfs_ordered_extent 440 -> 424 - btrfs_writepage_fixup 104 -> 96 Signed-off-by: David Sterba <dsterba@suse.com>
2023-09-19 16:49:23 +00:00
btrfs_init_work(&async->work, run_one_async_start, run_one_async_done);
btrfs: remove hipri_workers workqueue Now that btrfs_wq_submit_bio is never called for synchronous I/O, the hipri_workers workqueue is not used anymore and can be removed. Reviewed-by: Chris Mason <clm@fb.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-05-03 07:06:15 +00:00
btrfs_queue_work(fs_info->workers, &async->work);
btrfs: handle checksum generation in the storage layer Instead of letting the callers of btrfs_submit_bio deal with checksumming the (meta)data in the bio and making decisions on when to offload the checksumming to the bio, leave that to btrfs_submit_bio. Do do so the existing btrfs_submit_bio function is split into an upper and a lower half, so that the lower half can be offloaded to a workqueue. Note that this changes the behavior for direct writes to raid56 volumes so that async checksum offloading is not skipped when more I/O is expected. This runs counter to the argument explaining why it was done, although I can't measure any affects of the change. Commits later in this series will make sure the entire direct writes is offloaded to the workqueue at once and thus make sure it is sent to the raid56 code from a single thread. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:17 +00:00
return true;
}
btrfs: pass a btrfs_bio to btrfs_submit_bio btrfs_submit_bio expects the bio passed to it to be embedded into a btrfs_bio structure. Pass the btrfs_bio directly to increase type safety and make the code self-documenting. Reviewed-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-07 16:39:39 +00:00
static bool btrfs_submit_chunk(struct btrfs_bio *bbio, int mirror_num)
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
{
btrfs: split zone append bios in btrfs_submit_bio The current btrfs zoned device support is a little cumbersome in the data I/O path as it requires the callers to not issue I/O larger than the supported ZONE_APPEND size of the underlying device. This leads to a lot of extra accounting. Instead change btrfs_submit_bio so that it can take write bios of arbitrary size and form from the upper layers, and just split them internally to the ZONE_APPEND queue limits. Then remove all the upper layer warts catering to limited write sized on zoned devices, including the extra refcount in the compressed_bio. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:30 +00:00
struct btrfs_inode *inode = bbio->inode;
btrfs: introduce btrfs_bio::fs_info member Currently we're doing a lot of work for btrfs_bio: - Checksum verification for data read bios - Bio splits if it crosses stripe boundary - Read repair for data read bios However for the incoming scrub patches, we don't want this extra functionality at all, just plain logical + mirror -> physical mapping ability. Thus here we do the following changes: - Introduce btrfs_bio::fs_info This is for the new scrub specific btrfs_bio, which would not populate btrfs_bio::inode. Thus we need such new member to grab a fs_info This new member will always be populated. - Replace @inode argument with @fs_info for btrfs_bio_init() and its caller Since @inode is no longer a mandatory member, replace it with @fs_info, and let involved users populate @inode. - Skip checksum verification and generation if @bbio->inode is NULL - Add extra ASSERT()s To make sure: * bbio->inode is properly set for involved read repair path * if @file_offset is set, bbio->inode is also populated - Grab @fs_info from @bbio directly We can no longer go @bbio->inode->root->fs_info, as bbio->inode can be NULL. This involves: * btrfs_simple_end_io() * should_async_write() * btrfs_wq_submit_bio() * btrfs_use_zone_append() Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-23 09:01:20 +00:00
struct btrfs_fs_info *fs_info = bbio->fs_info;
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
struct btrfs_bio *orig_bbio = bbio;
btrfs: pass a btrfs_bio to btrfs_submit_bio btrfs_submit_bio expects the bio passed to it to be embedded into a btrfs_bio structure. Pass the btrfs_bio directly to increase type safety and make the code self-documenting. Reviewed-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-07 16:39:39 +00:00
struct bio *bio = &bbio->bio;
btrfs: use SECTOR_SHIFT to convert LBA to physical offset Using SECTOR_SHIFT to convert LBA to physical address makes it more readable. Signed-off-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-04-15 11:32:38 +00:00
u64 logical = bio->bi_iter.bi_sector << SECTOR_SHIFT;
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
u64 length = bio->bi_iter.bi_size;
u64 map_length = length;
btrfs: pass a btrfs_bio to btrfs_use_append struct btrfs_bio has all the information needed for btrfs_use_append, so pass that instead of a btrfs_inode and file_offset. Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-12-12 07:37:21 +00:00
bool use_append = btrfs_use_zone_append(bbio);
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
struct btrfs_io_context *bioc = NULL;
struct btrfs_io_stripe smap;
btrfs: refactor error handling in btrfs_submit_bio Add a bbio local variable and to prepare for calling functions that return a blk_status_t, rename the existing int used for error handling so that ret can be reused for the blk_status_t, and a label that can be reused for failing the passed in bio. Reviewed-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:03 +00:00
blk_status_t ret;
int error;
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
btrfs: scrub: implement raid stripe tree support A filesystem that uses the raid stripe tree for logical to physical address translation can't use the regular scrub path, that reads all stripes and then checks if a sector is unused afterwards. When using the raid stripe tree, this will result in lookup errors, as the stripe tree doesn't know the requested logical addresses. In case we're scrubbing a filesystem which uses the RAID stripe tree for multi-device logical to physical address translation, perform an extra block mapping step to get the real on-disk stripe length from the stripe tree when scrubbing the sectors. This prevents a double completion of the btrfs_bio caused by splitting the underlying bio and ultimately a use-after-free. Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-09-14 16:07:01 +00:00
smap.is_scrub = !bbio->inode;
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
btrfs_bio_counter_inc_blocked(fs_info);
btrfs: rename __btrfs_map_block to btrfs_map_block Now that the old btrfs_map_block is gone, drop the leading underscores from __btrfs_map_block. Reviewed-by: Qu Wenruo <wqu@suse.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-05-31 04:17:37 +00:00
error = btrfs_map_block(fs_info, btrfs_op(bio), logical, &map_length,
btrfs: remove the need_raid_map parameter from btrfs_map_block() The parameter @need_raid_map is mostly a legacy from the old days where we don't yet have a solid definition on the @mirror_num, and only check-integrity was using that parameter, while all other call sites just pass 1 for that parameter. Now since we have removed check-integrity functionality, we can also remove the @need_raid_map parameter. This change will also remove the ability to read P/Q stripe directly when passing 0 as @need_raid_map. Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-09-17 10:06:21 +00:00
&bioc, &smap, &mirror_num);
btrfs: refactor error handling in btrfs_submit_bio Add a bbio local variable and to prepare for calling functions that return a blk_status_t, rename the existing int used for error handling so that ret can be reused for the blk_status_t, and a label that can be reused for failing the passed in bio. Reviewed-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:03 +00:00
if (error) {
ret = errno_to_blk_status(error);
goto fail;
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
}
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
map_length = min(map_length, length);
btrfs: split zone append bios in btrfs_submit_bio The current btrfs zoned device support is a little cumbersome in the data I/O path as it requires the callers to not issue I/O larger than the supported ZONE_APPEND size of the underlying device. This leads to a lot of extra accounting. Instead change btrfs_submit_bio so that it can take write bios of arbitrary size and form from the upper layers, and just split them internally to the ZONE_APPEND queue limits. Then remove all the upper layer warts catering to limited write sized on zoned devices, including the extra refcount in the compressed_bio. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:30 +00:00
if (use_append)
map_length = min(map_length, fs_info->max_zone_append_size);
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
if (map_length < length) {
btrfs: make btrfs_split_bio work on struct btrfs_bio btrfs_split_bio expects a btrfs_bio as argument and always allocates one. Type both the orig_bio argument and the return value as struct btrfs_bio to improve type safety. Reviewed-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-07 16:39:45 +00:00
bbio = btrfs_split_bio(fs_info, bbio, map_length, use_append);
bio = &bbio->bio;
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
}
btrfs: pre-load data checksum for reads in btrfs_submit_bio Instead of calling btrfs_lookup_bio_sums in every caller of btrfs_submit_bio that reads data, do the call once in btrfs_submit_bio. Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:05 +00:00
/*
* Save the iter for the end_io handler and preload the checksums for
* data reads.
*/
btrfs: add a is_data_bbio helper Add a helper to check for that a btrfs_bio has a valid inode, and that it is a data inode to key off all the special handling for data path checksumming. Note that this uses is_data_inode instead of REQ_META as REQ_META is only set directly before submission in submit_one_bio and we'll also want to use this helper for error handling where REQ_META isn't set yet. Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Josef Bacik <josef@toxicpanda.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-05-31 07:54:00 +00:00
if (bio_op(bio) == REQ_OP_READ && is_data_bbio(bbio)) {
btrfs: rename btrfs_bio::iter field Rename iter to saved_iter and move it next to the repair internals and nothing outside of bio.c should be touching it. Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:13 +00:00
bbio->saved_iter = bio->bi_iter;
btrfs: pre-load data checksum for reads in btrfs_submit_bio Instead of calling btrfs_lookup_bio_sums in every caller of btrfs_submit_bio that reads data, do the call once in btrfs_submit_bio. Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:05 +00:00
ret = btrfs_lookup_bio_sums(bbio);
if (ret)
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
goto fail_put_bio;
btrfs: pre-load data checksum for reads in btrfs_submit_bio Instead of calling btrfs_lookup_bio_sums in every caller of btrfs_submit_bio that reads data, do the call once in btrfs_submit_bio. Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:05 +00:00
}
btrfs: save the bio iter for checksum validation in common code All callers of btrfs_submit_bio that want to validate checksums currently have to store a copy of the iter in the btrfs_bio. Move the assignment into common code. Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:04 +00:00
btrfs: handle checksum generation in the storage layer Instead of letting the callers of btrfs_submit_bio deal with checksumming the (meta)data in the bio and making decisions on when to offload the checksumming to the bio, leave that to btrfs_submit_bio. Do do so the existing btrfs_submit_bio function is split into an upper and a lower half, so that the lower half can be offloaded to a workqueue. Note that this changes the behavior for direct writes to raid56 volumes so that async checksum offloading is not skipped when more I/O is expected. This runs counter to the argument explaining why it was done, although I can't measure any affects of the change. Commits later in this series will make sure the entire direct writes is offloaded to the workqueue at once and thus make sure it is sent to the raid56 code from a single thread. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:17 +00:00
if (btrfs_op(bio) == BTRFS_MAP_WRITE) {
btrfs: split zone append bios in btrfs_submit_bio The current btrfs zoned device support is a little cumbersome in the data I/O path as it requires the callers to not issue I/O larger than the supported ZONE_APPEND size of the underlying device. This leads to a lot of extra accounting. Instead change btrfs_submit_bio so that it can take write bios of arbitrary size and form from the upper layers, and just split them internally to the ZONE_APPEND queue limits. Then remove all the upper layer warts catering to limited write sized on zoned devices, including the extra refcount in the compressed_bio. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:30 +00:00
if (use_append) {
bio->bi_opf &= ~REQ_OP_WRITE;
bio->bi_opf |= REQ_OP_ZONE_APPEND;
btrfs: handle recording of zoned writes in the storage layer Move the code that splits the ordered extents and records the physical location for them to the storage layer so that the higher level consumers don't have to care about physical block numbers at all. This will also allow to eventually remove accounting for the zone append write sizes in the upper layer with a little bit more block layer work. Reviewed-by: Naohiro Aota <naohiro.aota@wdc.com> Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:18 +00:00
}
btrfs: add support for inserting raid stripe extents Add support for inserting stripe extents into the raid stripe tree on completion of every write that needs an extra logical-to-physical translation when using RAID. Inserting the stripe extents happens after the data I/O has completed, this is done to a) support zone-append and b) rule out the possibility of a RAID-write-hole. Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-09-14 16:06:58 +00:00
if (is_data_bbio(bbio) && bioc &&
btrfs_need_stripe_tree_update(bioc->fs_info, bioc->map_type)) {
/*
* No locking for the list update, as we only add to
* the list in the I/O submission path, and list
* iteration only happens in the completion path, which
* can't happen until after the last submission.
*/
btrfs_get_bioc(bioc);
list_add_tail(&bioc->rst_ordered_entry, &bbio->ordered->bioc_list);
}
btrfs: handle checksum generation in the storage layer Instead of letting the callers of btrfs_submit_bio deal with checksumming the (meta)data in the bio and making decisions on when to offload the checksumming to the bio, leave that to btrfs_submit_bio. Do do so the existing btrfs_submit_bio function is split into an upper and a lower half, so that the lower half can be offloaded to a workqueue. Note that this changes the behavior for direct writes to raid56 volumes so that async checksum offloading is not skipped when more I/O is expected. This runs counter to the argument explaining why it was done, although I can't measure any affects of the change. Commits later in this series will make sure the entire direct writes is offloaded to the workqueue at once and thus make sure it is sent to the raid56 code from a single thread. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:17 +00:00
/*
* Csum items for reloc roots have already been cloned at this
* point, so they are handled as part of the no-checksum case.
*/
btrfs: introduce btrfs_bio::fs_info member Currently we're doing a lot of work for btrfs_bio: - Checksum verification for data read bios - Bio splits if it crosses stripe boundary - Read repair for data read bios However for the incoming scrub patches, we don't want this extra functionality at all, just plain logical + mirror -> physical mapping ability. Thus here we do the following changes: - Introduce btrfs_bio::fs_info This is for the new scrub specific btrfs_bio, which would not populate btrfs_bio::inode. Thus we need such new member to grab a fs_info This new member will always be populated. - Replace @inode argument with @fs_info for btrfs_bio_init() and its caller Since @inode is no longer a mandatory member, replace it with @fs_info, and let involved users populate @inode. - Skip checksum verification and generation if @bbio->inode is NULL - Add extra ASSERT()s To make sure: * bbio->inode is properly set for involved read repair path * if @file_offset is set, bbio->inode is also populated - Grab @fs_info from @bbio directly We can no longer go @bbio->inode->root->fs_info, as bbio->inode can be NULL. This involves: * btrfs_simple_end_io() * should_async_write() * btrfs_wq_submit_bio() * btrfs_use_zone_append() Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-23 09:01:20 +00:00
if (inode && !(inode->flags & BTRFS_INODE_NODATASUM) &&
btrfs: handle checksum generation in the storage layer Instead of letting the callers of btrfs_submit_bio deal with checksumming the (meta)data in the bio and making decisions on when to offload the checksumming to the bio, leave that to btrfs_submit_bio. Do do so the existing btrfs_submit_bio function is split into an upper and a lower half, so that the lower half can be offloaded to a workqueue. Note that this changes the behavior for direct writes to raid56 volumes so that async checksum offloading is not skipped when more I/O is expected. This runs counter to the argument explaining why it was done, although I can't measure any affects of the change. Commits later in this series will make sure the entire direct writes is offloaded to the workqueue at once and thus make sure it is sent to the raid56 code from a single thread. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:17 +00:00
!test_bit(BTRFS_FS_STATE_NO_CSUMS, &fs_info->fs_state) &&
btrfs: split zone append bios in btrfs_submit_bio The current btrfs zoned device support is a little cumbersome in the data I/O path as it requires the callers to not issue I/O larger than the supported ZONE_APPEND size of the underlying device. This leads to a lot of extra accounting. Instead change btrfs_submit_bio so that it can take write bios of arbitrary size and form from the upper layers, and just split them internally to the ZONE_APPEND queue limits. Then remove all the upper layer warts catering to limited write sized on zoned devices, including the extra refcount in the compressed_bio. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:30 +00:00
!btrfs_is_data_reloc_root(inode->root)) {
btrfs: handle checksum generation in the storage layer Instead of letting the callers of btrfs_submit_bio deal with checksumming the (meta)data in the bio and making decisions on when to offload the checksumming to the bio, leave that to btrfs_submit_bio. Do do so the existing btrfs_submit_bio function is split into an upper and a lower half, so that the lower half can be offloaded to a workqueue. Note that this changes the behavior for direct writes to raid56 volumes so that async checksum offloading is not skipped when more I/O is expected. This runs counter to the argument explaining why it was done, although I can't measure any affects of the change. Commits later in this series will make sure the entire direct writes is offloaded to the workqueue at once and thus make sure it is sent to the raid56 code from a single thread. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:17 +00:00
if (should_async_write(bbio) &&
btrfs_wq_submit_bio(bbio, bioc, &smap, mirror_num))
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
goto done;
btrfs: handle checksum generation in the storage layer Instead of letting the callers of btrfs_submit_bio deal with checksumming the (meta)data in the bio and making decisions on when to offload the checksumming to the bio, leave that to btrfs_submit_bio. Do do so the existing btrfs_submit_bio function is split into an upper and a lower half, so that the lower half can be offloaded to a workqueue. Note that this changes the behavior for direct writes to raid56 volumes so that async checksum offloading is not skipped when more I/O is expected. This runs counter to the argument explaining why it was done, although I can't measure any affects of the change. Commits later in this series will make sure the entire direct writes is offloaded to the workqueue at once and thus make sure it is sent to the raid56 code from a single thread. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:17 +00:00
ret = btrfs_bio_csum(bbio);
if (ret)
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
goto fail_put_bio;
btrfs: zoned: allocate dummy checksums for zoned NODATASUM writes Shin'ichiro reported that when he's running fstests' test-case btrfs/167 on emulated zoned devices, he's seeing the following NULL pointer dereference in 'btrfs_zone_finish_endio()': Oops: general protection fault, probably for non-canonical address 0xdffffc0000000011: 0000 [#1] PREEMPT SMP KASAN NOPTI KASAN: null-ptr-deref in range [0x0000000000000088-0x000000000000008f] CPU: 4 PID: 2332440 Comm: kworker/u80:15 Tainted: G W 6.10.0-rc2-kts+ #4 Hardware name: Supermicro Super Server/X11SPi-TF, BIOS 3.3 02/21/2020 Workqueue: btrfs-endio-write btrfs_work_helper [btrfs] RIP: 0010:btrfs_zone_finish_endio.part.0+0x34/0x160 [btrfs] RSP: 0018:ffff88867f107a90 EFLAGS: 00010206 RAX: dffffc0000000000 RBX: 0000000000000000 RCX: ffffffff893e5534 RDX: 0000000000000011 RSI: 0000000000000004 RDI: 0000000000000088 RBP: 0000000000000002 R08: 0000000000000001 R09: ffffed1081696028 R10: ffff88840b4b0143 R11: ffff88834dfff600 R12: ffff88840b4b0000 R13: 0000000000020000 R14: 0000000000000000 R15: ffff888530ad5210 FS: 0000000000000000(0000) GS:ffff888e3f800000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00007f87223fff38 CR3: 00000007a7c6a002 CR4: 00000000007706f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 PKRU: 55555554 Call Trace: <TASK> ? __die_body.cold+0x19/0x27 ? die_addr+0x46/0x70 ? exc_general_protection+0x14f/0x250 ? asm_exc_general_protection+0x26/0x30 ? do_raw_read_unlock+0x44/0x70 ? btrfs_zone_finish_endio.part.0+0x34/0x160 [btrfs] btrfs_finish_one_ordered+0x5d9/0x19a0 [btrfs] ? __pfx_lock_release+0x10/0x10 ? do_raw_write_lock+0x90/0x260 ? __pfx_do_raw_write_lock+0x10/0x10 ? __pfx_btrfs_finish_one_ordered+0x10/0x10 [btrfs] ? _raw_write_unlock+0x23/0x40 ? btrfs_finish_ordered_zoned+0x5a9/0x850 [btrfs] ? lock_acquire+0x435/0x500 btrfs_work_helper+0x1b1/0xa70 [btrfs] ? __schedule+0x10a8/0x60b0 ? __pfx___might_resched+0x10/0x10 process_one_work+0x862/0x1410 ? __pfx_lock_acquire+0x10/0x10 ? __pfx_process_one_work+0x10/0x10 ? assign_work+0x16c/0x240 worker_thread+0x5e6/0x1010 ? __pfx_worker_thread+0x10/0x10 kthread+0x2c3/0x3a0 ? trace_irq_enable.constprop.0+0xce/0x110 ? __pfx_kthread+0x10/0x10 ret_from_fork+0x31/0x70 ? __pfx_kthread+0x10/0x10 ret_from_fork_asm+0x1a/0x30 </TASK> Enabling CONFIG_BTRFS_ASSERT revealed the following assertion to trigger: assertion failed: !list_empty(&ordered->list), in fs/btrfs/zoned.c:1815 This indicates, that we're missing the checksums list on the ordered_extent. As btrfs/167 is doing a NOCOW write this is to be expected. Further analysis with drgn confirmed the assumption: >>> inode = prog.crashed_thread().stack_trace()[11]['ordered'].inode >>> btrfs_inode = drgn.container_of(inode, "struct btrfs_inode", \ "vfs_inode") >>> print(btrfs_inode.flags) (u32)1 As zoned emulation mode simulates conventional zones on regular devices, we cannot use zone-append for writing. But we're only attaching dummy checksums if we're doing a zone-append write. So for NOCOW zoned data writes on conventional zones, also attach a dummy checksum. Reported-by: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> Fixes: cbfce4c7fbde ("btrfs: optimize the logical to physical mapping for zoned writes") CC: Naohiro Aota <Naohiro.Aota@wdc.com> # 6.6+ Tested-by: Shin'ichiro Kawasaki <shinichiro.kawasaki@wdc.com> Reviewed-by: Naohiro Aota <naohiro.aota@wdc.com> Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2024-06-07 11:27:48 +00:00
} else if (use_append ||
(btrfs_is_zoned(fs_info) && inode &&
inode->flags & BTRFS_INODE_NODATASUM)) {
btrfs: optimize the logical to physical mapping for zoned writes The current code to store the final logical to physical mapping for a zone append write in the extent tree is rather inefficient. It first has to split the ordered extent so that there is one ordered extent per bio, so that it can look up the ordered extent on I/O completion in btrfs_record_physical_zoned and store the physical LBA returned by the block driver in the ordered extent. btrfs_rewrite_logical_zoned then has to do a lookup in the chunk tree to see what physical address the logical address for this bio / ordered extent is mapped to, and then rewrite it in the extent tree. To optimize this process, we can store the physical address assigned in the chunk tree to the original logical address and a pointer to btrfs_ordered_sum structure the in the btrfs_bio structure, and then use this information to rewrite the logical address in the btrfs_ordered_sum structure directly at I/O completion time in btrfs_record_physical_zoned. btrfs_rewrite_logical_zoned then simply updates the logical address in the extent tree and the ordered_extent itself. The code in btrfs_rewrite_logical_zoned now runs for all data I/O completions in zoned file systems, which is fine as there is no remapping to do for non-append writes to conventional zones or for relocation, and the overhead for quickly breaking out of the loop is very low. Because zoned file systems now need the ordered_sums structure to record the actual write location returned by zone append, allocate dummy structures without the csum array for them when the I/O doesn't use checksums, and free them when completing the ordered_extent. Note that the btrfs_bio doesn't grow as the new field are places into a union that is so far not used for data writes and has plenty of space left in it. Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-05-24 15:03:08 +00:00
ret = btrfs_alloc_dummy_sum(bbio);
if (ret)
goto fail_put_bio;
btrfs: handle checksum generation in the storage layer Instead of letting the callers of btrfs_submit_bio deal with checksumming the (meta)data in the bio and making decisions on when to offload the checksumming to the bio, leave that to btrfs_submit_bio. Do do so the existing btrfs_submit_bio function is split into an upper and a lower half, so that the lower half can be offloaded to a workqueue. Note that this changes the behavior for direct writes to raid56 volumes so that async checksum offloading is not skipped when more I/O is expected. This runs counter to the argument explaining why it was done, although I can't measure any affects of the change. Commits later in this series will make sure the entire direct writes is offloaded to the workqueue at once and thus make sure it is sent to the raid56 code from a single thread. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:17 +00:00
}
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
}
btrfs: handle checksum generation in the storage layer Instead of letting the callers of btrfs_submit_bio deal with checksumming the (meta)data in the bio and making decisions on when to offload the checksumming to the bio, leave that to btrfs_submit_bio. Do do so the existing btrfs_submit_bio function is split into an upper and a lower half, so that the lower half can be offloaded to a workqueue. Note that this changes the behavior for direct writes to raid56 volumes so that async checksum offloading is not skipped when more I/O is expected. This runs counter to the argument explaining why it was done, although I can't measure any affects of the change. Commits later in this series will make sure the entire direct writes is offloaded to the workqueue at once and thus make sure it is sent to the raid56 code from a single thread. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:17 +00:00
__btrfs_submit_bio(bio, bioc, &smap, mirror_num);
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
done:
return map_length == length;
btrfs: refactor error handling in btrfs_submit_bio Add a bbio local variable and to prepare for calling functions that return a blk_status_t, rename the existing int used for error handling so that ret can be reused for the blk_status_t, and a label that can be reused for failing the passed in bio. Reviewed-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:03 +00:00
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
fail_put_bio:
if (map_length < length)
btrfs: add an ordered_extent pointer to struct btrfs_bio Add a pointer to the ordered_extent to the existing union in struct btrfs_bio, so all code dealing with data write bios can just use a pointer dereference to retrieve the ordered_extent instead of doing multiple rbtree lookups per I/O. The reference to this ordered_extent is dropped at end I/O time, which implies that an extra one must be acquired when the bio is split. This also requires moving the btrfs_extract_ordered_extent call into btrfs_split_bio so that the invariant of always having a valid ordered_extent reference for the btrfs_bio is kept. Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Josef Bacik <josef@toxicpanda.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-05-31 07:54:02 +00:00
btrfs_cleanup_bio(bbio);
btrfs: refactor error handling in btrfs_submit_bio Add a bbio local variable and to prepare for calling functions that return a blk_status_t, rename the existing int used for error handling so that ret can be reused for the blk_status_t, and a label that can be reused for failing the passed in bio. Reviewed-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:03 +00:00
fail:
btrfs_bio_counter_dec(fs_info);
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
btrfs_bio_end_io(orig_bbio, ret);
/* Do not submit another chunk */
return true;
}
btrfs: pass a btrfs_bio to btrfs_submit_bio btrfs_submit_bio expects the bio passed to it to be embedded into a btrfs_bio structure. Pass the btrfs_bio directly to increase type safety and make the code self-documenting. Reviewed-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-07 16:39:39 +00:00
void btrfs_submit_bio(struct btrfs_bio *bbio, int mirror_num)
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
{
btrfs: introduce btrfs_bio::fs_info member Currently we're doing a lot of work for btrfs_bio: - Checksum verification for data read bios - Bio splits if it crosses stripe boundary - Read repair for data read bios However for the incoming scrub patches, we don't want this extra functionality at all, just plain logical + mirror -> physical mapping ability. Thus here we do the following changes: - Introduce btrfs_bio::fs_info This is for the new scrub specific btrfs_bio, which would not populate btrfs_bio::inode. Thus we need such new member to grab a fs_info This new member will always be populated. - Replace @inode argument with @fs_info for btrfs_bio_init() and its caller Since @inode is no longer a mandatory member, replace it with @fs_info, and let involved users populate @inode. - Skip checksum verification and generation if @bbio->inode is NULL - Add extra ASSERT()s To make sure: * bbio->inode is properly set for involved read repair path * if @file_offset is set, bbio->inode is also populated - Grab @fs_info from @bbio directly We can no longer go @bbio->inode->root->fs_info, as bbio->inode can be NULL. This involves: * btrfs_simple_end_io() * should_async_write() * btrfs_wq_submit_bio() * btrfs_use_zone_append() Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-23 09:01:20 +00:00
/* If bbio->inode is not populated, its file_offset must be 0. */
ASSERT(bbio->inode || bbio->file_offset == 0);
btrfs: pass a btrfs_bio to btrfs_submit_bio btrfs_submit_bio expects the bio passed to it to be embedded into a btrfs_bio structure. Pass the btrfs_bio directly to increase type safety and make the code self-documenting. Reviewed-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-07 16:39:39 +00:00
while (!btrfs_submit_chunk(bbio, mirror_num))
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
;
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
}
btrfs: move repair_io_failure to bio.c repair_io_failure ties directly into all the glory low-level details of mapping a bio with a logic address to the actual physical location. Move it right below btrfs_submit_bio to keep all the related logic together. Also move btrfs_repair_eb_io_failure to its caller in disk-io.c now that repair_io_failure is available in a header. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:06 +00:00
/*
* Submit a repair write.
*
* This bypasses btrfs_submit_bio deliberately, as that writes all copies in a
* RAID setup. Here we only want to write the one bad copy, so we do the
* mapping ourselves and submit the bio directly.
*
btrfs: fix spelling mistakes found using codespell There quite a few spelling mistakes as found using codespell. Fix them. Signed-off-by: Colin Ian King <colin.i.king@gmail.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-17 10:03:21 +00:00
* The I/O is issued synchronously to block the repair read completion from
btrfs: move repair_io_failure to bio.c repair_io_failure ties directly into all the glory low-level details of mapping a bio with a logic address to the actual physical location. Move it right below btrfs_submit_bio to keep all the related logic together. Also move btrfs_repair_eb_io_failure to its caller in disk-io.c now that repair_io_failure is available in a header. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:06 +00:00
* freeing the bio.
*/
int btrfs_repair_io_failure(struct btrfs_fs_info *fs_info, u64 ino, u64 start,
btrfs: migrate btrfs_repair_io_failure() to folio interfaces [BUG] Test case btrfs/124 failed if larger metadata folio is enabled, the dying message looks like this: BTRFS error (device dm-2): bad tree block start, mirror 2 want 31686656 have 0 BTRFS info (device dm-2): read error corrected: ino 0 off 31686656 (dev /dev/mapper/test-scratch2 sector 20928) BUG: kernel NULL pointer dereference, address: 0000000000000020 #PF: supervisor read access in kernel mode #PF: error_code(0x0000) - not-present page CPU: 6 PID: 350881 Comm: btrfs Tainted: G OE 6.7.0-rc3-custom+ #128 Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS unknown 2/2/2022 RIP: 0010:btrfs_read_extent_buffer+0x106/0x180 [btrfs] PKRU: 55555554 Call Trace: <TASK> read_tree_block+0x33/0xb0 [btrfs] read_block_for_search+0x23e/0x340 [btrfs] btrfs_search_slot+0x2f9/0xe60 [btrfs] btrfs_lookup_csum+0x75/0x160 [btrfs] btrfs_lookup_bio_sums+0x21a/0x560 [btrfs] btrfs_submit_chunk+0x152/0x680 [btrfs] btrfs_submit_bio+0x1c/0x50 [btrfs] submit_one_bio+0x40/0x80 [btrfs] submit_extent_page+0x158/0x390 [btrfs] btrfs_do_readpage+0x330/0x740 [btrfs] extent_readahead+0x38d/0x6c0 [btrfs] read_pages+0x94/0x2c0 page_cache_ra_unbounded+0x12d/0x190 relocate_file_extent_cluster+0x7c1/0x9d0 [btrfs] relocate_block_group+0x2d3/0x560 [btrfs] btrfs_relocate_block_group+0x2c7/0x4b0 [btrfs] btrfs_relocate_chunk+0x4c/0x1a0 [btrfs] btrfs_balance+0x925/0x13c0 [btrfs] btrfs_ioctl+0x19f1/0x25d0 [btrfs] __x64_sys_ioctl+0x90/0xd0 do_syscall_64+0x3f/0xf0 entry_SYSCALL_64_after_hwframe+0x6e/0x76 [CAUSE] The dying line is at btrfs_repair_io_failure() call inside btrfs_repair_eb_io_failure(). The function is still relying on the extent buffer using page sized folios. When the extent buffer is using larger folio, we go into the 2nd slot of folios[], and triggered the NULL pointer dereference. [FIX] Migrate btrfs_repair_io_failure() to folio interfaces. So that when we hit a larger folio, we just submit the whole folio in one go. This also affects data repair path through btrfs_end_repair_bio(), thankfully data is still fully page based, we can just add an ASSERT(), and use page_folio() to convert the page to folio. Signed-off-by: Qu Wenruo <wqu@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-12-12 05:24:10 +00:00
u64 length, u64 logical, struct folio *folio,
unsigned int folio_offset, int mirror_num)
btrfs: move repair_io_failure to bio.c repair_io_failure ties directly into all the glory low-level details of mapping a bio with a logic address to the actual physical location. Move it right below btrfs_submit_bio to keep all the related logic together. Also move btrfs_repair_eb_io_failure to its caller in disk-io.c now that repair_io_failure is available in a header. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:06 +00:00
{
btrfs: introduce a new helper to submit write bio for repair Both scrub and read-repair are utilizing a special repair writes that: - Only writes back to a single device Even for read-repair on RAID56, we only update the corrupted data stripe itself, not triggering the full RMW path. - Requires a valid @mirror_num For RAID56 case, only @mirror_num == 1 is valid. For non-RAID56 cases, we need @mirror_num to locate our stripe. - No data csum generation needed These two call sites still have some differences though: - Read-repair goes plain bio It doesn't need a full btrfs_bio, and goes submit_bio_wait(). - New scrub repair would go btrfs_bio To simplify both read and write path. So here this patch would: - Introduce a common helper, btrfs_map_repair_block() Due to the single device nature, we can use an on-stack btrfs_io_stripe to pass device and its physical bytenr. - Introduce a new interface, btrfs_submit_repair_bio(), for later scrub code This is for the incoming scrub code. Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-20 02:12:49 +00:00
struct btrfs_io_stripe smap = { 0 };
btrfs: move repair_io_failure to bio.c repair_io_failure ties directly into all the glory low-level details of mapping a bio with a logic address to the actual physical location. Move it right below btrfs_submit_bio to keep all the related logic together. Also move btrfs_repair_eb_io_failure to its caller in disk-io.c now that repair_io_failure is available in a header. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:06 +00:00
struct bio_vec bvec;
struct bio bio;
int ret = 0;
ASSERT(!(fs_info->sb->s_flags & SB_RDONLY));
BUG_ON(!mirror_num);
if (btrfs_repair_one_zone(fs_info, logical))
return 0;
/*
* Avoid races with device replace and make sure our bioc has devices
* associated to its stripes that don't go away while we are doing the
* read repair operation.
*/
btrfs_bio_counter_inc_blocked(fs_info);
btrfs: introduce a new helper to submit write bio for repair Both scrub and read-repair are utilizing a special repair writes that: - Only writes back to a single device Even for read-repair on RAID56, we only update the corrupted data stripe itself, not triggering the full RMW path. - Requires a valid @mirror_num For RAID56 case, only @mirror_num == 1 is valid. For non-RAID56 cases, we need @mirror_num to locate our stripe. - No data csum generation needed These two call sites still have some differences though: - Read-repair goes plain bio It doesn't need a full btrfs_bio, and goes submit_bio_wait(). - New scrub repair would go btrfs_bio To simplify both read and write path. So here this patch would: - Introduce a common helper, btrfs_map_repair_block() Due to the single device nature, we can use an on-stack btrfs_io_stripe to pass device and its physical bytenr. - Introduce a new interface, btrfs_submit_repair_bio(), for later scrub code This is for the incoming scrub code. Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-20 02:12:49 +00:00
ret = btrfs_map_repair_block(fs_info, &smap, logical, length, mirror_num);
if (ret < 0)
goto out_counter_dec;
btrfs: move repair_io_failure to bio.c repair_io_failure ties directly into all the glory low-level details of mapping a bio with a logic address to the actual physical location. Move it right below btrfs_submit_bio to keep all the related logic together. Also move btrfs_repair_eb_io_failure to its caller in disk-io.c now that repair_io_failure is available in a header. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:06 +00:00
btrfs: introduce a new helper to submit write bio for repair Both scrub and read-repair are utilizing a special repair writes that: - Only writes back to a single device Even for read-repair on RAID56, we only update the corrupted data stripe itself, not triggering the full RMW path. - Requires a valid @mirror_num For RAID56 case, only @mirror_num == 1 is valid. For non-RAID56 cases, we need @mirror_num to locate our stripe. - No data csum generation needed These two call sites still have some differences though: - Read-repair goes plain bio It doesn't need a full btrfs_bio, and goes submit_bio_wait(). - New scrub repair would go btrfs_bio To simplify both read and write path. So here this patch would: - Introduce a common helper, btrfs_map_repair_block() Due to the single device nature, we can use an on-stack btrfs_io_stripe to pass device and its physical bytenr. - Introduce a new interface, btrfs_submit_repair_bio(), for later scrub code This is for the incoming scrub code. Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-20 02:12:49 +00:00
if (!smap.dev->bdev ||
!test_bit(BTRFS_DEV_STATE_WRITEABLE, &smap.dev->dev_state)) {
btrfs: move repair_io_failure to bio.c repair_io_failure ties directly into all the glory low-level details of mapping a bio with a logic address to the actual physical location. Move it right below btrfs_submit_bio to keep all the related logic together. Also move btrfs_repair_eb_io_failure to its caller in disk-io.c now that repair_io_failure is available in a header. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:06 +00:00
ret = -EIO;
goto out_counter_dec;
}
btrfs: introduce a new helper to submit write bio for repair Both scrub and read-repair are utilizing a special repair writes that: - Only writes back to a single device Even for read-repair on RAID56, we only update the corrupted data stripe itself, not triggering the full RMW path. - Requires a valid @mirror_num For RAID56 case, only @mirror_num == 1 is valid. For non-RAID56 cases, we need @mirror_num to locate our stripe. - No data csum generation needed These two call sites still have some differences though: - Read-repair goes plain bio It doesn't need a full btrfs_bio, and goes submit_bio_wait(). - New scrub repair would go btrfs_bio To simplify both read and write path. So here this patch would: - Introduce a common helper, btrfs_map_repair_block() Due to the single device nature, we can use an on-stack btrfs_io_stripe to pass device and its physical bytenr. - Introduce a new interface, btrfs_submit_repair_bio(), for later scrub code This is for the incoming scrub code. Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-20 02:12:49 +00:00
bio_init(&bio, smap.dev->bdev, &bvec, 1, REQ_OP_WRITE | REQ_SYNC);
bio.bi_iter.bi_sector = smap.physical >> SECTOR_SHIFT;
btrfs: migrate btrfs_repair_io_failure() to folio interfaces [BUG] Test case btrfs/124 failed if larger metadata folio is enabled, the dying message looks like this: BTRFS error (device dm-2): bad tree block start, mirror 2 want 31686656 have 0 BTRFS info (device dm-2): read error corrected: ino 0 off 31686656 (dev /dev/mapper/test-scratch2 sector 20928) BUG: kernel NULL pointer dereference, address: 0000000000000020 #PF: supervisor read access in kernel mode #PF: error_code(0x0000) - not-present page CPU: 6 PID: 350881 Comm: btrfs Tainted: G OE 6.7.0-rc3-custom+ #128 Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS unknown 2/2/2022 RIP: 0010:btrfs_read_extent_buffer+0x106/0x180 [btrfs] PKRU: 55555554 Call Trace: <TASK> read_tree_block+0x33/0xb0 [btrfs] read_block_for_search+0x23e/0x340 [btrfs] btrfs_search_slot+0x2f9/0xe60 [btrfs] btrfs_lookup_csum+0x75/0x160 [btrfs] btrfs_lookup_bio_sums+0x21a/0x560 [btrfs] btrfs_submit_chunk+0x152/0x680 [btrfs] btrfs_submit_bio+0x1c/0x50 [btrfs] submit_one_bio+0x40/0x80 [btrfs] submit_extent_page+0x158/0x390 [btrfs] btrfs_do_readpage+0x330/0x740 [btrfs] extent_readahead+0x38d/0x6c0 [btrfs] read_pages+0x94/0x2c0 page_cache_ra_unbounded+0x12d/0x190 relocate_file_extent_cluster+0x7c1/0x9d0 [btrfs] relocate_block_group+0x2d3/0x560 [btrfs] btrfs_relocate_block_group+0x2c7/0x4b0 [btrfs] btrfs_relocate_chunk+0x4c/0x1a0 [btrfs] btrfs_balance+0x925/0x13c0 [btrfs] btrfs_ioctl+0x19f1/0x25d0 [btrfs] __x64_sys_ioctl+0x90/0xd0 do_syscall_64+0x3f/0xf0 entry_SYSCALL_64_after_hwframe+0x6e/0x76 [CAUSE] The dying line is at btrfs_repair_io_failure() call inside btrfs_repair_eb_io_failure(). The function is still relying on the extent buffer using page sized folios. When the extent buffer is using larger folio, we go into the 2nd slot of folios[], and triggered the NULL pointer dereference. [FIX] Migrate btrfs_repair_io_failure() to folio interfaces. So that when we hit a larger folio, we just submit the whole folio in one go. This also affects data repair path through btrfs_end_repair_bio(), thankfully data is still fully page based, we can just add an ASSERT(), and use page_folio() to convert the page to folio. Signed-off-by: Qu Wenruo <wqu@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-12-12 05:24:10 +00:00
ret = bio_add_folio(&bio, folio, length, folio_offset);
ASSERT(ret);
btrfs: move repair_io_failure to bio.c repair_io_failure ties directly into all the glory low-level details of mapping a bio with a logic address to the actual physical location. Move it right below btrfs_submit_bio to keep all the related logic together. Also move btrfs_repair_eb_io_failure to its caller in disk-io.c now that repair_io_failure is available in a header. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:06 +00:00
ret = submit_bio_wait(&bio);
if (ret) {
/* try to remap that extent elsewhere? */
btrfs: introduce a new helper to submit write bio for repair Both scrub and read-repair are utilizing a special repair writes that: - Only writes back to a single device Even for read-repair on RAID56, we only update the corrupted data stripe itself, not triggering the full RMW path. - Requires a valid @mirror_num For RAID56 case, only @mirror_num == 1 is valid. For non-RAID56 cases, we need @mirror_num to locate our stripe. - No data csum generation needed These two call sites still have some differences though: - Read-repair goes plain bio It doesn't need a full btrfs_bio, and goes submit_bio_wait(). - New scrub repair would go btrfs_bio To simplify both read and write path. So here this patch would: - Introduce a common helper, btrfs_map_repair_block() Due to the single device nature, we can use an on-stack btrfs_io_stripe to pass device and its physical bytenr. - Introduce a new interface, btrfs_submit_repair_bio(), for later scrub code This is for the incoming scrub code. Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-20 02:12:49 +00:00
btrfs_dev_stat_inc_and_print(smap.dev, BTRFS_DEV_STAT_WRITE_ERRS);
btrfs: move repair_io_failure to bio.c repair_io_failure ties directly into all the glory low-level details of mapping a bio with a logic address to the actual physical location. Move it right below btrfs_submit_bio to keep all the related logic together. Also move btrfs_repair_eb_io_failure to its caller in disk-io.c now that repair_io_failure is available in a header. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:06 +00:00
goto out_bio_uninit;
}
btrfs_info_rl_in_rcu(fs_info,
"read error corrected: ino %llu off %llu (dev %s sector %llu)",
btrfs: introduce a new helper to submit write bio for repair Both scrub and read-repair are utilizing a special repair writes that: - Only writes back to a single device Even for read-repair on RAID56, we only update the corrupted data stripe itself, not triggering the full RMW path. - Requires a valid @mirror_num For RAID56 case, only @mirror_num == 1 is valid. For non-RAID56 cases, we need @mirror_num to locate our stripe. - No data csum generation needed These two call sites still have some differences though: - Read-repair goes plain bio It doesn't need a full btrfs_bio, and goes submit_bio_wait(). - New scrub repair would go btrfs_bio To simplify both read and write path. So here this patch would: - Introduce a common helper, btrfs_map_repair_block() Due to the single device nature, we can use an on-stack btrfs_io_stripe to pass device and its physical bytenr. - Introduce a new interface, btrfs_submit_repair_bio(), for later scrub code This is for the incoming scrub code. Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-20 02:12:49 +00:00
ino, start, btrfs_dev_name(smap.dev),
smap.physical >> SECTOR_SHIFT);
btrfs: move repair_io_failure to bio.c repair_io_failure ties directly into all the glory low-level details of mapping a bio with a logic address to the actual physical location. Move it right below btrfs_submit_bio to keep all the related logic together. Also move btrfs_repair_eb_io_failure to its caller in disk-io.c now that repair_io_failure is available in a header. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:06 +00:00
ret = 0;
out_bio_uninit:
bio_uninit(&bio);
out_counter_dec:
btrfs_bio_counter_dec(fs_info);
return ret;
}
btrfs: introduce a new helper to submit write bio for repair Both scrub and read-repair are utilizing a special repair writes that: - Only writes back to a single device Even for read-repair on RAID56, we only update the corrupted data stripe itself, not triggering the full RMW path. - Requires a valid @mirror_num For RAID56 case, only @mirror_num == 1 is valid. For non-RAID56 cases, we need @mirror_num to locate our stripe. - No data csum generation needed These two call sites still have some differences though: - Read-repair goes plain bio It doesn't need a full btrfs_bio, and goes submit_bio_wait(). - New scrub repair would go btrfs_bio To simplify both read and write path. So here this patch would: - Introduce a common helper, btrfs_map_repair_block() Due to the single device nature, we can use an on-stack btrfs_io_stripe to pass device and its physical bytenr. - Introduce a new interface, btrfs_submit_repair_bio(), for later scrub code This is for the incoming scrub code. Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-03-20 02:12:49 +00:00
/*
* Submit a btrfs_bio based repair write.
*
* If @dev_replace is true, the write would be submitted to dev-replace target.
*/
void btrfs_submit_repair_write(struct btrfs_bio *bbio, int mirror_num, bool dev_replace)
{
struct btrfs_fs_info *fs_info = bbio->fs_info;
u64 logical = bbio->bio.bi_iter.bi_sector << SECTOR_SHIFT;
u64 length = bbio->bio.bi_iter.bi_size;
struct btrfs_io_stripe smap = { 0 };
int ret;
ASSERT(fs_info);
ASSERT(mirror_num > 0);
ASSERT(btrfs_op(&bbio->bio) == BTRFS_MAP_WRITE);
ASSERT(!bbio->inode);
btrfs_bio_counter_inc_blocked(fs_info);
ret = btrfs_map_repair_block(fs_info, &smap, logical, length, mirror_num);
if (ret < 0)
goto fail;
if (dev_replace) {
ASSERT(smap.dev == fs_info->dev_replace.srcdev);
smap.dev = fs_info->dev_replace.tgtdev;
}
__btrfs_submit_bio(&bbio->bio, NULL, &smap, mirror_num);
return;
fail:
btrfs_bio_counter_dec(fs_info);
btrfs_bio_end_io(bbio, errno_to_blk_status(ret));
}
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
int __init btrfs_bioset_init(void)
{
if (bioset_init(&btrfs_bioset, BIO_POOL_SIZE,
offsetof(struct btrfs_bio, bio),
BIOSET_NEED_BVECS))
return -ENOMEM;
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
if (bioset_init(&btrfs_clone_bioset, BIO_POOL_SIZE,
offsetof(struct btrfs_bio, bio), 0))
goto out_free_bioset;
btrfs: handle checksum validation and repair at the storage layer Currently btrfs handles checksum validation and repair in the end I/O handler for the btrfs_bio. This leads to a lot of duplicate code plus issues with varying semantics or bugs, e.g. - the until recently broken repair for compressed extents - the fact that encoded reads validate the checksums but do not kick of read repair - the inconsistent checking of the BTRFS_FS_STATE_NO_CSUMS flag This commit revamps the checksum validation and repair code to instead work below the btrfs_submit_bio interfaces. In case of a checksum failure (or a plain old I/O error), the repair is now kicked off before the upper level ->end_io handler is invoked. Progress of an in-progress repair is tracked by a small structure that is allocated using a mempool for each original bio with failed sectors, which holds a reference to the original bio. This new structure is allocated using a mempool to guarantee forward progress even under memory pressure. The mempool will be replenished when the repair completes, just as the mempools backing the bios. There is one significant behavior change here: If repair fails or is impossible to start with, the whole bio will be failed to the upper layer. This is the behavior that all I/O submitters except for buffered I/O already emulated in their end_io handler. For buffered I/O this now means that a large readahead request can fail due to a single bad sector, but as readahead errors are ignored the following readpage if the sector is actually accessed will still be able to read. This also matches the I/O failure handling in other file systems. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:07 +00:00
if (bioset_init(&btrfs_repair_bioset, BIO_POOL_SIZE,
offsetof(struct btrfs_bio, bio),
BIOSET_NEED_BVECS))
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
goto out_free_clone_bioset;
btrfs: handle checksum validation and repair at the storage layer Currently btrfs handles checksum validation and repair in the end I/O handler for the btrfs_bio. This leads to a lot of duplicate code plus issues with varying semantics or bugs, e.g. - the until recently broken repair for compressed extents - the fact that encoded reads validate the checksums but do not kick of read repair - the inconsistent checking of the BTRFS_FS_STATE_NO_CSUMS flag This commit revamps the checksum validation and repair code to instead work below the btrfs_submit_bio interfaces. In case of a checksum failure (or a plain old I/O error), the repair is now kicked off before the upper level ->end_io handler is invoked. Progress of an in-progress repair is tracked by a small structure that is allocated using a mempool for each original bio with failed sectors, which holds a reference to the original bio. This new structure is allocated using a mempool to guarantee forward progress even under memory pressure. The mempool will be replenished when the repair completes, just as the mempools backing the bios. There is one significant behavior change here: If repair fails or is impossible to start with, the whole bio will be failed to the upper layer. This is the behavior that all I/O submitters except for buffered I/O already emulated in their end_io handler. For buffered I/O this now means that a large readahead request can fail due to a single bad sector, but as readahead errors are ignored the following readpage if the sector is actually accessed will still be able to read. This also matches the I/O failure handling in other file systems. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:07 +00:00
if (mempool_init_kmalloc_pool(&btrfs_failed_bio_pool, BIO_POOL_SIZE,
sizeof(struct btrfs_failed_bio)))
goto out_free_repair_bioset;
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
return 0;
btrfs: handle checksum validation and repair at the storage layer Currently btrfs handles checksum validation and repair in the end I/O handler for the btrfs_bio. This leads to a lot of duplicate code plus issues with varying semantics or bugs, e.g. - the until recently broken repair for compressed extents - the fact that encoded reads validate the checksums but do not kick of read repair - the inconsistent checking of the BTRFS_FS_STATE_NO_CSUMS flag This commit revamps the checksum validation and repair code to instead work below the btrfs_submit_bio interfaces. In case of a checksum failure (or a plain old I/O error), the repair is now kicked off before the upper level ->end_io handler is invoked. Progress of an in-progress repair is tracked by a small structure that is allocated using a mempool for each original bio with failed sectors, which holds a reference to the original bio. This new structure is allocated using a mempool to guarantee forward progress even under memory pressure. The mempool will be replenished when the repair completes, just as the mempools backing the bios. There is one significant behavior change here: If repair fails or is impossible to start with, the whole bio will be failed to the upper layer. This is the behavior that all I/O submitters except for buffered I/O already emulated in their end_io handler. For buffered I/O this now means that a large readahead request can fail due to a single bad sector, but as readahead errors are ignored the following readpage if the sector is actually accessed will still be able to read. This also matches the I/O failure handling in other file systems. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:07 +00:00
out_free_repair_bioset:
bioset_exit(&btrfs_repair_bioset);
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
out_free_clone_bioset:
bioset_exit(&btrfs_clone_bioset);
btrfs: handle checksum validation and repair at the storage layer Currently btrfs handles checksum validation and repair in the end I/O handler for the btrfs_bio. This leads to a lot of duplicate code plus issues with varying semantics or bugs, e.g. - the until recently broken repair for compressed extents - the fact that encoded reads validate the checksums but do not kick of read repair - the inconsistent checking of the BTRFS_FS_STATE_NO_CSUMS flag This commit revamps the checksum validation and repair code to instead work below the btrfs_submit_bio interfaces. In case of a checksum failure (or a plain old I/O error), the repair is now kicked off before the upper level ->end_io handler is invoked. Progress of an in-progress repair is tracked by a small structure that is allocated using a mempool for each original bio with failed sectors, which holds a reference to the original bio. This new structure is allocated using a mempool to guarantee forward progress even under memory pressure. The mempool will be replenished when the repair completes, just as the mempools backing the bios. There is one significant behavior change here: If repair fails or is impossible to start with, the whole bio will be failed to the upper layer. This is the behavior that all I/O submitters except for buffered I/O already emulated in their end_io handler. For buffered I/O this now means that a large readahead request can fail due to a single bad sector, but as readahead errors are ignored the following readpage if the sector is actually accessed will still be able to read. This also matches the I/O failure handling in other file systems. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:07 +00:00
out_free_bioset:
bioset_exit(&btrfs_bioset);
return -ENOMEM;
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
}
void __cold btrfs_bioset_exit(void)
{
btrfs: handle checksum validation and repair at the storage layer Currently btrfs handles checksum validation and repair in the end I/O handler for the btrfs_bio. This leads to a lot of duplicate code plus issues with varying semantics or bugs, e.g. - the until recently broken repair for compressed extents - the fact that encoded reads validate the checksums but do not kick of read repair - the inconsistent checking of the BTRFS_FS_STATE_NO_CSUMS flag This commit revamps the checksum validation and repair code to instead work below the btrfs_submit_bio interfaces. In case of a checksum failure (or a plain old I/O error), the repair is now kicked off before the upper level ->end_io handler is invoked. Progress of an in-progress repair is tracked by a small structure that is allocated using a mempool for each original bio with failed sectors, which holds a reference to the original bio. This new structure is allocated using a mempool to guarantee forward progress even under memory pressure. The mempool will be replenished when the repair completes, just as the mempools backing the bios. There is one significant behavior change here: If repair fails or is impossible to start with, the whole bio will be failed to the upper layer. This is the behavior that all I/O submitters except for buffered I/O already emulated in their end_io handler. For buffered I/O this now means that a large readahead request can fail due to a single bad sector, but as readahead errors are ignored the following readpage if the sector is actually accessed will still be able to read. This also matches the I/O failure handling in other file systems. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:07 +00:00
mempool_exit(&btrfs_failed_bio_pool);
bioset_exit(&btrfs_repair_bioset);
btrfs: allow btrfs_submit_bio to split bios Currently the I/O submitters have to split bios according to the chunk stripe boundaries. This leads to extra lookups in the extent trees and a lot of boilerplate code. To drop this requirement, split the bio when __btrfs_map_block returns a mapping that is smaller than the requested size and keep a count of pending bios in the original btrfs_bio so that the upper level completion is only invoked when all clones have completed. Based on a patch from Qu Wenruo. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Qu Wenruo <wqu@suse.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-21 06:50:20 +00:00
bioset_exit(&btrfs_clone_bioset);
btrfs: split the bio submission path into a separate file The code used by btrfs_submit_bio only interacts with the rest of volumes.c through __btrfs_map_block (which itself is a more generic version of two exported helpers) and does not really have anything to do with volumes.c. Create a new bio.c file and a bio.h header going along with it for the btrfs_bio-based storage layer, which will grow even more going forward. Also update the file with my copyright notice given that a large part of the moved code was written or rewritten by me. Reviewed-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2022-11-15 09:44:05 +00:00
bioset_exit(&btrfs_bioset);
}
Reference in New Issue Copy Permalink