kernel/linux.git/drivers/block/rbd.c, branch v7.1-rc5

rbd: eliminate a race in lock_dwork draining on unmap

2026-05-20T20:09:08+00:00

Given how rbd_lock_add_request() and rbd_img_exclusive_lock() are written, lock_dwork may be (re)queued more than it's actually needed: for example in case a new I/O request comes in while we are in the middle of rbd_acquire_lock() on behalf of another I/O request. This is expected and with rbd_release_lock() preemptively canceling lock_dwork is benign under normal operation. A more problematic example is maybe_kick_acquire(): if (have_requests || delayed_work_pending(&rbd_dev->lock_dwork)) { dout("%s rbd_dev %p kicking lock_dwork\n", __func__, rbd_dev); mod_delayed_work(rbd_dev->task_wq, &rbd_dev->lock_dwork, 0); } It's not unrealistic for lock_dwork to get canceled right after delayed_work_pending() returns true and for mod_delayed_work() to requeue it right there anyway. This is a classic TOCTOU race. When it comes to unmapping the image, there is an implicit assumption of no self-initiated exclusive lock activity past the point of return from rbd_dev_image_unlock() which unlocks the lock if it happens to be held. This unlock is assumed to be final and lock_dwork (as well as all other exclusive lock tasks, really) isn't expected to get queued again. However, lock_dwork is canceled only in cancel_tasks_sync() (i.e. later in the unmap sequence) and on top of that the cancellation can get in effect nullified by maybe_kick_acquire(). This may result in rbd_acquire_lock() executing after rbd_dev_device_release() and rbd_dev_image_release() run and free and/or reset a bunch of things. One of the possible failure modes then is a violated rbd_assert(rbd_image_format_valid(rbd_dev->image_format)); in rbd_dev_header_info() which is called via rbd_dev_refresh() from rbd_post_acquire_action(). Redo exclusive lock task draining to provide saner semantics and try to meet the assumptions around rbd_dev_image_unlock(). Cc: stable@vger.kernel.org Signed-off-by: Ilya Dryomov Reviewed-by: Viacheslav Dubeyko

rbd: fix null-ptr-deref when device_add_disk() fails

2026-04-21T23:40:23+00:00

do_rbd_add() publishes the device with device_add() before calling device_add_disk(). If device_add_disk() fails after device_add() succeeds, the error path calls rbd_free_disk() directly and then later falls through to rbd_dev_device_release(), which calls rbd_free_disk() again. This double teardown can leave blk-mq cleanup operating on invalid state and trigger a null-ptr-deref in __blk_mq_free_map_and_rqs(), reached from blk_mq_free_tag_set(). Fix this by following the normal remove ordering: call device_del() before rbd_dev_device_release() when device_add_disk() fails after device_add(). That keeps the teardown sequence consistent and avoids re-entering disk cleanup through the wrong path. The bug was first flagged by an experimental analysis tool we are developing for kernel memory-management bugs while analyzing v6.13-rc1. The tool is still under development and is not yet publicly available. We reproduced the bug on v7.0 with a real Ceph backend and a QEMU x86_64 guest booted with KASAN and CONFIG_FAILSLAB enabled. The reproducer confines failslab injections to the __add_disk() range and injects fail-nth while mapping an RBD image through /sys/bus/rbd/add_single_major. On the unpatched kernel, fail-nth=4 reliably triggered the fault: Oops: general protection fault, probably for non-canonical address 0xdffffc0000000000: 0000 [#1] SMP KASAN NOPTI KASAN: null-ptr-deref in range [0x0000000000000000-0x0000000000000007] CPU: 0 UID: 0 PID: 273 Comm: bash Not tainted 7.0.0-01247-gd60bc1401583 #6 PREEMPT(lazy) Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.15.0-1 04/01/2014 RIP: 0010:__blk_mq_free_map_and_rqs+0x8c/0x240 Code: 00 00 48 8b 6b 60 41 89 f4 49 c1 e4 03 4c 01 e5 45 85 ed 0f 85 0a 01 00 00 48 b8 00 00 00 00 00 fc ff df 48 89 e9 48 c1 e9 03 <80> 3c 01 00 0f 85 31 01 00 00 4c 8b 6d 00 4d 85 ed 0f 84 e2 00 00 RSP: 0018:ff1100000ab0fac8 EFLAGS: 00000246 RAX: dffffc0000000000 RBX: ff1100000c4806a0 RCX: 0000000000000000 RDX: 0000000000000002 RSI: 0000000000000000 RDI: ff1100000c4806f4 RBP: 0000000000000000 R08: 0000000000000001 R09: ffe21c000189001b R10: ff1100000c4800df R11: ff1100006cf37be0 R12: 0000000000000000 R13: 0000000000000000 R14: ff1100000c480700 R15: ff1100000c480004 FS: 00007f0fbe8fe740(0000) GS:ff110000e5851000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00007fe53473b2e0 CR3: 0000000012eef000 CR4: 00000000007516f0 PKRU: 55555554 Call Trace: blk_mq_free_tag_set+0x77/0x460 do_rbd_add+0x1446/0x2b80 ? __pfx_do_rbd_add+0x10/0x10 ? lock_acquire+0x18c/0x300 ? find_held_lock+0x2b/0x80 ? sysfs_file_kobj+0xb6/0x1b0 ? __pfx_sysfs_kf_write+0x10/0x10 kernfs_fop_write_iter+0x2f4/0x4a0 vfs_write+0x98e/0x1000 ? expand_files+0x51f/0x850 ? __pfx_vfs_write+0x10/0x10 ksys_write+0xf2/0x1d0 ? __pfx_ksys_write+0x10/0x10 do_syscall_64+0x115/0x690 entry_SYSCALL_64_after_hwframe+0x77/0x7f RIP: 0033:0x7f0fbea15907 Code: 10 00 f7 d8 64 89 02 48 c7 c0 ff ff ff ff eb b7 0f 1f 00 f3 0f 1e fa 64 8b 04 25 18 00 00 00 85 c0 75 10 b8 01 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 51 c3 48 83 ec 28 48 89 54 24 18 48 89 74 24 RSP: 002b:00007ffe22346ea8 EFLAGS: 00000246 ORIG_RAX: 0000000000000001 RAX: ffffffffffffffda RBX: 0000000000000058 RCX: 00007f0fbea15907 RDX: 0000000000000058 RSI: 0000563ace6c0ef0 RDI: 0000000000000001 RBP: 0000563ace6c0ef0 R08: 0000563ace6c0ef0 R09: 6b6435726d694141 R10: 5250337279762f78 R11: 0000000000000246 R12: 0000000000000058 R13: 00007f0fbeb1c780 R14: ff1100000c480700 R15: ff1100000c480004 With this fix applied, rerunning the reproducer over fail-nth=1..256 yields no KASAN reports. [ idryomov: rename err_out_device_del -> err_out_device ] Cc: stable@vger.kernel.org Fixes: 27c97abc30e2 ("rbd: add add_disk() error handling") Signed-off-by: Zilin Guan Signed-off-by: Dawei Feng Reviewed-by: Ilya Dryomov Signed-off-by: Ilya Dryomov

Convert 'alloc_obj' family to use the new default GFP_KERNEL argument

2026-02-22T01:09:51+00:00

This was done entirely with mindless brute force, using git grep -l '\

treewide: Replace kmalloc with kmalloc_obj for non-scalar types

2026-02-21T09:02:28+00:00

This is the result of running the Coccinelle script from scripts/coccinelle/api/kmalloc_objs.cocci. The script is designed to avoid scalar types (which need careful case-by-case checking), and instead replace kmalloc-family calls that allocate struct or union object instances: Single allocations: kmalloc(sizeof(TYPE), ...) are replaced with: kmalloc_obj(TYPE, ...) Array allocations: kmalloc_array(COUNT, sizeof(TYPE), ...) are replaced with: kmalloc_objs(TYPE, COUNT, ...) Flex array allocations: kmalloc(struct_size(PTR, FAM, COUNT), ...) are replaced with: kmalloc_flex(*PTR, FAM, COUNT, ...) (where TYPE may also be *VAR) The resulting allocations no longer return "void *", instead returning "TYPE *". Signed-off-by: Kees Cook

rbd: check for EOD after exclusive lock is ensured to be held

2026-02-03T20:00:22+00:00

Similar to commit 870611e4877e ("rbd: get snapshot context after exclusive lock is ensured to be held"), move the "beyond EOD" check into the image request state machine so that it's performed after exclusive lock is ensured to be held. This avoids various race conditions which can arise when the image is shrunk under I/O (in practice, mostly readahead). In one such scenario rbd_assert(objno < rbd_dev->object_map_size); can be triggered if a close-to-EOD read gets queued right before the shrink is initiated and the EOD check is performed against an outdated mapping_size. After the resize is done on the server side and exclusive lock is (re)acquired bringing along the new (now shrunk) object map, the read starts going through the state machine and rbd_obj_may_exist() gets invoked on an object that is out of bounds of rbd_dev->object_map array. Cc: stable@vger.kernel.org Signed-off-by: Ilya Dryomov Reviewed-by: Dongsheng Yang

drivers/block: WQ_PERCPU added to alloc_workqueue users

2025-09-09T15:11:31+00:00

Currently if a user enqueue a work item using schedule_delayed_work() the used wq is "system_wq" (per-cpu wq) while queue_delayed_work() use WORK_CPU_UNBOUND (used when a cpu is not specified). The same applies to schedule_work() that is using system_wq and queue_work(), that makes use again of WORK_CPU_UNBOUND. This lack of consistentcy cannot be addressed without refactoring the API. alloc_workqueue() treats all queues as per-CPU by default, while unbound workqueues must opt-in via WQ_UNBOUND. This default is suboptimal: most workloads benefit from unbound queues, allowing the scheduler to place worker threads where they’re needed and reducing noise when CPUs are isolated. This default is suboptimal: most workloads benefit from unbound queues, allowing the scheduler to place worker threads where they’re needed and reducing noise when CPUs are isolated. This patch adds a new WQ_PERCPU flag to explicitly request the use of the per-CPU behavior. Both flags coexist for one release cycle to allow callers to transition their calls. Once migration is complete, WQ_UNBOUND can be removed and unbound will become the implicit default. With the introduction of the WQ_PERCPU flag (equivalent to !WQ_UNBOUND), any alloc_workqueue() caller that doesn’t explicitly specify WQ_UNBOUND must now use WQ_PERCPU. All existing users have been updated accordingly. Suggested-by: Tejun Heo Signed-off-by: Marco Crivellari Signed-off-by: Jens Axboe

block: force noio scope in blk_mq_freeze_queue

2025-01-31T14:20:08+00:00

When block drivers or the core block code perform allocations with a frozen queue, this could try to recurse into the block device to reclaim memory and deadlock. Thus all allocations done by a process that froze a queue need to be done without __GFP_IO and __GFP_FS. Instead of tying to track all of them down, force a noio scope as part of freezing the queue. Note that nvme is a bit of a mess here due to the non-owner freezes, and they will be addressed separately. Signed-off-by: Christoph Hellwig Link: https://lore.kernel.org/r/20250131120352.1315351-2-hch@lst.de Signed-off-by: Jens Axboe

block: remove BLK_MQ_F_SHOULD_MERGE

2024-12-23T15:17:23+00:00

BLK_MQ_F_SHOULD_MERGE is set for all tag_sets except those that purely process passthrough commands (bsg-lib, ufs tmf, various nvme admin queues) and thus don't even check the flag. Remove it to simplify the driver interface. Signed-off-by: Christoph Hellwig Link: https://lore.kernel.org/r/20241219060214.1928848-1-hch@lst.de Signed-off-by: Jens Axboe

rbd: unfreeze queue after marking disk as dead

2024-11-07T23:27:22+00:00

Unfreeze queue after returning from blk_mark_disk_dead(), this way at least allows us to verify queue freeze correctly with lockdep. Suggested-by: Christoph Hellwig Cc: Ilya Dryomov Signed-off-by: Ming Lei Link: https://lore.kernel.org/r/20241031133723.303835-3-ming.lei@redhat.com Signed-off-by: Jens Axboe

Merge tag 'ceph-for-6.11-rc1' of https://github.com/ceph/ceph-client

2024-07-26T17:34:42+00:00

Pull ceph updates from Ilya Dryomov: "A small patchset to address bogus I/O errors and ultimately an assertion failure in the face of watch errors with -o exclusive mappings in RBD marked for stable and some assorted CephFS fixes" * tag 'ceph-for-6.11-rc1' of https://github.com/ceph/ceph-client: rbd: don't assume rbd_is_lock_owner() for exclusive mappings rbd: don't assume RBD_LOCK_STATE_LOCKED for exclusive mappings rbd: rename RBD_LOCK_STATE_RELEASING and releasing_wait ceph: fix incorrect kmalloc size of pagevec mempool ceph: periodically flush the cap releases ceph: convert comma to semicolon in __ceph_dentry_dir_lease_touch() ceph: use cap_wait_list only if debugfs is enabled