summaryrefslogtreecommitdiffstats
path: root/fs/bcachefs/io_write.c (follow)
AgeCommit message (Collapse)AuthorFilesLines
2025-06-02bcachefs: bch_err_throw()Kent Overstreet1-9/+10
Add a tracepoint for any time we return an error and unwind. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-06-01bcachefs: Replace rcu_read_lock() with guardsKent Overstreet1-5/+2
The new guard(), scoped_guard() allow for more natural code. Some of the uses with creative flow control have been left. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-05-21bcachefs: Call bch2_bkey_set_needs_rebalance() earlier in write pathKent Overstreet1-3/+9
There's no reason to be running this inside our transaction; it forces us to copy the key we're updating to a temporary, which we'd like to skip. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-05-21bcachefs: async objs now support bch_write_opsKent Overstreet1-0/+6
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-05-21bcachefs: Add comments for inode snapshot requirementsKent Overstreet1-0/+6
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-05-21bcachefs: bch_dev.io_ref -> enumerated_refKent Overstreet1-4/+11
Convert device IO refs to enumerated_refs, for easier debugging of refcount issues. Simple conversion: enumerate all users and convert to the new helpers. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-05-21bcachefs: bch_fs.writes -> enumerated_refsKent Overstreet1-2/+3
Drop the single-purpose write ref code in bcachefs.h, and convert to enumarated refs. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-05-21bcachefs: print_str_as_lines() -> print_str()Kent Overstreet1-1/+1
bch2_print_string_as_lines() is a low level helper that allows messages longer than 1k to be printed without truncation. But we should always be printing with the helpers that take a filesystem object, if we're in fsck they direct output to the userspace process controlling fsck instead of the dmesg log. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-05-21bcachefs: Simplify bch2_count_fsck_err()Kent Overstreet1-3/+1
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-05-21bcachefs: RO mounts now use less memoryKent Overstreet1-8/+0
Defer memory allocations only needed in RW mode until we actually go RW. This is part of improved support for RO images. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-05-01bcachefs: check for inode.bi_sectors underflowKent Overstreet1-0/+21
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-04-02bcachefs: Kill btree_iter.transKent Overstreet1-6/+6
This was planned to be done ages ago, now finally completed; there are places where we have quite a few btree_trans objects on the stack, so this reduces stack usage somewhat. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-04-02bcachefs: Split up bch_dev.io_refKent Overstreet1-2/+7
We now have separate per device io_refs for read and write access. This fixes a device removal bug where the discard workers were still running while we're removing alloc info for that device. It's also a bit of hardening; we no longer allow writes to devices that are read-only. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-03-31bcachefs: Fix null ptr deref in bch2_write_endio()Kent Overstreet1-6/+13
This was previously hard to hit since it requires racing with device removal, but splitting up io_ref uncovered it. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-03-26bcachefs: Don't unnecessarily decrypt data when movingKent Overstreet1-0/+3
There's various checks for "are we going to compress this" - but we're not going to compress if we know it's incompressible. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-03-25bcachefs: Fix duplicate checksum error messages in write pathKent Overstreet1-18/+20
Also, improve the message in prep_encoded_data() - it now prints good/bad checksums, and checksum type. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-03-25bcachefs: Fix nonce inconsistency in bch2_write_prep_encoded_data()Kent Overstreet1-1/+2
If we're moving an extent that was partially overwritten, bch2_write_rechecksum() will trim it to the currenty live range. If we then also want to compress it, it'll be decrypted - but the nonce has been advanced for the overwritten start of the extent that we dropped, and we were using the nonce we calculated before rechecksum(). Reported-by: Gabriel de Perthuis <g2p.code@gmail.com> Fixes: 127d90d2823e ("bcachefs: bch2_write_prep_encoded_data() now returns errcode") Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-03-24bcachefs: EIO cleanupKent Overstreet1-7/+7
Replace these with proper private error codes, so that when we get an error message we're not sifting through the entire codebase to see where it came from. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-03-24bcachefs: bch2_write_prep_encoded_data() now returns errcodeKent Overstreet1-88/+71
Prep work for killing off EIO and replacing them with proper private error codes. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-03-24bcachefs: Simplify bch2_write_op_error()Kent Overstreet1-78/+32
There's no reason for the caller to do the actual logging, it's all done the same. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-03-24bcachefs: trace_io_move_write_failKent Overstreet1-4/+10
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-03-16bcachefs: Debug params for data corruption injectionKent Overstreet1-0/+24
dm-flakey is busted, and this is simpler anyways - this lets us test the checksum error retry ptahs Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-03-14bcachefs: Finish bch2_account_io_completion() conversionsKent Overstreet1-5/+7
More prep work for automatically kicking devices out after too many IO errors. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-03-14bcachefs: bch2_write_op_error() now prints info about data updateKent Overstreet1-30/+62
A user has been seeing the "error verifying existing checksum while rewriting existing data (memory corruption?)" error. This generally indicates a hardware issue (and that may be the case here), but it might also indicate a bug, in which case we need more information to look for patterns. Reported-by: Roland Vet <vet.roland@protonmail.com> Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-03-14bcachefs: Rename BCH_WRITE flags fer consistency with other x-macros enumsKent Overstreet1-43/+43
The uppercase/lowercase style is nice for making the namespace explicit. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-03-14bcachefs: Separate running/runnable in wp statsKent Overstreet1-1/+9
We've got per-writepoint statistics to see how well the writepoint index update threads are pipelining; this separates running vs. runnable so we can see at a glance if they're blocking. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-03-14bcachefs: Don't inc io_(read|write) counters for movesKent Overstreet1-1/+2
This makes 'bcachefs fs top' more useful; we can now see at a glance whether the IO to the device is being done for user reads/writes, or copygc/rebalance. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-02-12bcachefs: Reuse transactionAlan Huang1-1/+11
bch2_nocow_write_convert_unwritten is already in transaction context: 00191 ========= TEST generic/648 00242 kernel BUG at fs/bcachefs/btree_iter.c:3332! 00242 Internal error: Oops - BUG: 00000000f2000800 [#1] PREEMPT SMP 00242 Modules linked in: 00242 CPU: 4 UID: 0 PID: 2593 Comm: fsstress Not tainted 6.13.0-rc3-ktest-g345af8f855b7 #14403 00242 Hardware name: linux,dummy-virt (DT) 00242 pstate: 60001005 (nZCv daif -PAN -UAO -TCO -DIT +SSBS BTYPE=--) 00242 pc : __bch2_trans_get+0x120/0x410 00242 lr : __bch2_trans_get+0xcc/0x410 00242 sp : ffffff80d89af600 00242 x29: ffffff80d89af600 x28: ffffff80ddb23000 x27: 00000000fffff705 00242 x26: ffffff80ddb23028 x25: ffffff80d8903fe0 x24: ffffff80ebb30168 00242 x23: ffffff80c8aeb500 x22: 000000000000005d x21: ffffff80d8904078 00242 x20: ffffff80d8900000 x19: ffffff80da9e8000 x18: 0000000000000000 00242 x17: 64747568735f6c61 x16: 6e72756f6a20726f x15: 0000000000000028 00242 x14: 0000000000000004 x13: 000000000000f787 x12: ffffffc081bbcdc8 00242 x11: 0000000000000000 x10: 0000000000000003 x9 : ffffffc08094efbc 00242 x8 : 000000001092c111 x7 : 000000000000000c x6 : ffffffc083c31fc4 00242 x5 : ffffffc083c31f28 x4 : ffffff80c8aeb500 x3 : ffffff80ebb30000 00242 x2 : 0000000000000001 x1 : 0000000000000a21 x0 : 000000000000028e 00242 Call trace: 00242 __bch2_trans_get+0x120/0x410 (P) 00242 bch2_inum_offset_err_msg+0x48/0xb0 00242 bch2_nocow_write_convert_unwritten+0x3d0/0x530 00242 bch2_nocow_write+0xeb0/0x1000 00242 __bch2_write+0x330/0x4e8 00242 bch2_write+0x1f0/0x530 00242 bch2_direct_write+0x530/0xc00 00242 bch2_write_iter+0x160/0xbe0 00242 vfs_write+0x1cc/0x360 00242 ksys_write+0x5c/0xf0 00242 __arm64_sys_write+0x20/0x30 00242 invoke_syscall.constprop.0+0x54/0xe8 00242 do_el0_svc+0x44/0xc0 00242 el0_svc+0x34/0xa0 00242 el0t_64_sync_handler+0x104/0x130 00242 el0t_64_sync+0x154/0x158 00242 Code: 6b01001f 54ffff01 79408460 3617fec0 (d4210000) 00242 ---[ end trace 0000000000000000 ]--- 00242 Kernel panic - not syncing: Oops - BUG: Fatal exception Signed-off-by: Alan Huang <mmpgouride@gmail.com> Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2025-01-25bcachefs: Improve decompression error messagesKent Overstreet1-2/+2
Ratelimit them, and use the new bch2_write_op_error() helper that prints path and file offset. Reported-by: https://github.com/koverstreet/bcachefs/issues/819 Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-12-21bcachefs: list_pop_entry()Kent Overstreet1-3/+1
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-12-21bcachefs: Convert write path errors to inum_to_path()Kent Overstreet1-36/+55
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-12-21bcachefs: add missing BTREE_ITER_intentKent Overstreet1-0/+1
this fixes excessive transaction restarts due to trans_commit having to upgrade Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-12-21bcachefs: Rename btree_iter_peek_upto() -> btree_iter_peek_max()Kent Overstreet1-2/+2
We'll be introducing btree_iter_peek_prev_min(), so rename for consistency. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-12-21bcachefs: io_opts_to_rebalance_opts()Kent Overstreet1-1/+1
New helper to simplify bch2_bkey_set_needs_rebalance() Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-11-07bcachefs: Fix null ptr deref in bucket_gen_get()Kent Overstreet1-5/+2
bucket_gen() checks if we're lookup up a valid bucket and returns NULL otherwise, but bucket_gen_get() was failing to check; other callers were correct. Also do a bit of cleanup on callers. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-10-18bcachefs: Don't use commit_do() unnecessarilyKent Overstreet1-2/+2
Using commit_do() to call alloc_sectors_start_trans() breaks when we're randomly injecting transaction restarts - the restart in the commit causes us to leak the lock that alloc_sectorS_start_trans() takes. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-27bcachefs: rename version -> bversionKent Overstreet1-2/+2
give bversions a more distinct name, to aid in grepping Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-09-21bcachefs: BCH_WRITE_ALLOC_NOWAIT no longer applies to open bucket allocationKent Overstreet1-3/+4
rebalance writes must be BCH_WRITE_ALLOC_NOWAIT because they don't allocate from the full filesystem - but we don't want spurious allocation failures due to open buckets. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-08-07bcachefs: Make allocator stuck timeout configurable, ratelimit messagesKent Overstreet1-4/+1
Limit these messages to once every 2 minutes to avoid spamming logs; with multiple devices the output can be quite significant. Also, up the default timeout to 30 seconds from 10 seconds. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-07-14bcachefs: Rename BCH_WRITE_DONE -> BCH_WRITE_SUBMITTEDKent Overstreet1-12/+12
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-07-14bcachefs: Ratelimit checksum error messagesKent Overstreet1-1/+4
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-07-14bcachefs: spelling fixKent Overstreet1-1/+1
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-07-14bcachefs: Use try_cmpxchg() family of functions instead of cmpxchg()Uros Bizjak1-4/+3
Use try_cmpxchg() family of functions instead of cmpxchg (*ptr, old, new) == old. x86 CMPXCHG instruction returns success in ZF flag, so this change saves a compare after cmpxchg (and related move instruction in front of cmpxchg). Also, try_cmpxchg() implicitly assigns old *ptr value to "old" when cmpxchg fails. There is no need to re-read the value in the loop. No functional change intended. Signed-off-by: Uros Bizjak <ubizjak@gmail.com> Cc: Kent Overstreet <kent.overstreet@linux.dev> Cc: Brian Foster <bfoster@redhat.com> Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-06-10bcachefs: Check for invalid bucket from bucket_gen(), gc_bucket()Kent Overstreet1-4/+15
Turn more asserts into proper recoverable error paths. Reported-by: syzbot+246b47da27f8e7e7d6fb@syzkaller.appspotmail.com Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-05-09bcachefs: bch2_dev_get_ioref() checks for device not presentKent Overstreet1-2/+2
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-05-09bcachefs: bch2_dev_get_ioref2(); io_write.cKent Overstreet1-10/+11
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-05-08bcachefs: Move nocow unlock to bch2_write_endio()Kent Overstreet1-19/+7
This fixes a lifetime issue; bch2_nocow_write_unlock() uses PTR_BUCKET_POS(), which needs the device - but we drop our ref to the device in bch2_write_endio(). Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-05-08bcachefs: bch2_dev_have_ref()Kent Overstreet1-2/+2
bch2_dev_bkey_exists() is going away; bch2_dev_have_ref() documents that we're looking up a device without checking if it's present because we have a reference to it already. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-05-08bcachefs: extent_ptr_durability() -> bch2_dev_rcu()Kent Overstreet1-1/+6
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2024-05-08bcachefs: move replica_set from bch_dev to bch_fsKent Overstreet1-4/+4
This is needed for the next patch - the write submit path has to be able to allocate a replica bio even when we weren't able to get a ref on the device. Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
generated by cgit v1.2.3 (git 2.39.1) at 2025-12-11 18:01:33 +0000