user/sven/linux.git/block, branch v6.6.31

blk-iocost: do not WARN if iocg was already offlined

2024-05-17T10:02:20Z

[ Upstream commit 01bc4fda9ea0a6b52f12326486f07a4910666cf6 ] In iocg_pay_debt(), warn is triggered if 'active_list' is empty, which is intended to confirm iocg is active when it has debt. However, warn can be triggered during a blkcg or disk removal, if iocg_waitq_timer_fn() is run at that time: WARNING: CPU: 0 PID: 2344971 at block/blk-iocost.c:1402 iocg_pay_debt+0x14c/0x190 Call trace: iocg_pay_debt+0x14c/0x190 iocg_kick_waitq+0x438/0x4c0 iocg_waitq_timer_fn+0xd8/0x130 __run_hrtimer+0x144/0x45c __hrtimer_run_queues+0x16c/0x244 hrtimer_interrupt+0x2cc/0x7b0 The warn in this situation is meaningless. Since this iocg is being removed, the state of the 'active_list' is irrelevant, and 'waitq_timer' is canceled after removing 'active_list' in ioc_pd_free(), which ensures iocg is freed after iocg_waitq_timer_fn() returns. Therefore, add the check if iocg was already offlined to avoid warn when removing a blkcg or disk. Signed-off-by: Li Nan Reviewed-by: Yu Kuai Acked-by: Tejun Heo Link: https://lore.kernel.org/r/20240419093257.3004211-1-linan666@huaweicloud.com Signed-off-by: Jens Axboe Signed-off-by: Sasha Levin

blk-iocost: avoid out of bounds shift

2024-05-17T10:02:15Z

[ Upstream commit beaa51b36012fad5a4d3c18b88a617aea7a9b96d ] UBSAN catches undefined behavior in blk-iocost, where sometimes iocg->delay is shifted right by a number that is too large, resulting in undefined behavior on some architectures. [ 186.556576] ------------[ cut here ]------------ UBSAN: shift-out-of-bounds in block/blk-iocost.c:1366:23 shift exponent 64 is too large for 64-bit type 'u64' (aka 'unsigned long long') CPU: 16 PID: 0 Comm: swapper/16 Tainted: G S E N 6.9.0-0_fbk700_debug_rc2_kbuilder_0_gc85af715cac0 #1 Hardware name: Quanta Twin Lakes MP/Twin Lakes Passive MP, BIOS F09_3A23 12/08/2020 Call Trace: dump_stack_lvl+0x8f/0xe0 __ubsan_handle_shift_out_of_bounds+0x22c/0x280 iocg_kick_delay+0x30b/0x310 ioc_timer_fn+0x2fb/0x1f80 __run_timer_base+0x1b6/0x250 ... Avoid that undefined behavior by simply taking the "delay = 0" branch if the shift is too large. I am not sure what the symptoms of an undefined value delay will be, but I suspect it could be more than a little annoying to debug. Signed-off-by: Rik van Riel Cc: Tejun Heo Cc: Josef Bacik Cc: Jens Axboe Acked-by: Tejun Heo Link: https://lore.kernel.org/r/20240404123253.0f58010f@imladris.surriel.com Signed-off-by: Jens Axboe Signed-off-by: Sasha Levin

block: fix overflow in blk_ioctl_discard()

2024-05-17T10:02:13Z

[ Upstream commit 22d24a544b0d49bbcbd61c8c0eaf77d3c9297155 ] There is no check for overflow of 'start + len' in blk_ioctl_discard(). Hung task occurs if submit an discard ioctl with the following param: start = 0x80000000000ff000, len = 0x8000000000fff000; Add the overflow validation now. Signed-off-by: Li Nan Reviewed-by: Christoph Hellwig Link: https://lore.kernel.org/r/20240329012319.2034550-1-linan666@huaweicloud.com Signed-off-by: Jens Axboe Signed-off-by: Sasha Levin

block: fix q->blkg_list corruption during disk rebind

2024-04-17T09:19:28Z

[ Upstream commit 8b8ace080319a866f5dfe9da8e665ae51d971c54 ] Multiple gendisk instances can allocated/added for single request queue in case of disk rebind. blkg may still stay in q->blkg_list when calling blkcg_init_disk() for rebind, then q->blkg_list becomes corrupted. Fix the list corruption issue by: - add blkg_init_queue() to initialize q->blkg_list & q->blkcg_mutex only - move calling blkg_init_queue() into blk_alloc_queue() The list corruption should be started since commit f1c006f1c685 ("blk-cgroup: synchronize pd_free_fn() from blkg_free_workfn() and blkcg_deactivate_policy()") which delays removing blkg from q->blkg_list into blkg_free_workfn(). Fixes: f1c006f1c685 ("blk-cgroup: synchronize pd_free_fn() from blkg_free_workfn() and blkcg_deactivate_policy()") Fixes: 1059699f87eb ("block: move blkcg initialization/destroy into disk allocation/release handler") Cc: Yu Kuai Cc: Tejun Heo Signed-off-by: Ming Lei Reviewed-by: Yu Kuai Link: https://lore.kernel.org/r/20240407125910.4053377-1-ming.lei@redhat.com Signed-off-by: Jens Axboe Signed-off-by: Sasha Levin

block: prevent division by zero in blk_rq_stat_sum()

2024-04-13T11:07:37Z

[ Upstream commit 93f52fbeaf4b676b21acfe42a5152620e6770d02 ] The expression dst->nr_samples + src->nr_samples may have zero value on overflow. It is necessary to add a check to avoid division by zero. Found by Linux Verification Center (linuxtesting.org) with Svace. Signed-off-by: Roman Smirnov Reviewed-by: Sergey Shtylyov Link: https://lore.kernel.org/r/20240305134509.23108-1-r.smirnov@omp.ru Signed-off-by: Jens Axboe Signed-off-by: Sasha Levin

block: Do not force full zone append completion in req_bio_endio()

2024-04-03T13:28:56Z

commit 55251fbdf0146c252ceff146a1bb145546f3e034 upstream. This reverts commit 748dc0b65ec2b4b7b3dbd7befcc4a54fdcac7988. Partial zone append completions cannot be supported as there is no guarantees that the fragmented data will be written sequentially in the same manner as with a full command. Commit 748dc0b65ec2 ("block: fix partial zone append completion handling in req_bio_endio()") changed req_bio_endio() to always advance a partially failed BIO by its full length, but this can lead to incorrect accounting. So revert this change and let low level device drivers handle this case by always failing completely zone append operations. With this revert, users will still see an IO error for a partially completed zone append BIO. Fixes: 748dc0b65ec2 ("block: fix partial zone append completion handling in req_bio_endio()") Cc: stable@vger.kernel.org Signed-off-by: Damien Le Moal Reviewed-by: Christoph Hellwig Link: https://lore.kernel.org/r/20240328004409.594888-2-dlemoal@kernel.org Signed-off-by: Jens Axboe Signed-off-by: Greg Kroah-Hartman

Revert "block/mq-deadline: use correct way to throttling write requests"

2024-04-03T13:28:33Z

[ Upstream commit 256aab46e31683d76d45ccbedc287b4d3f3e322b ] The code "max(1U, 3 * (1U << shift) / 4)" comes from the Kyber I/O scheduler. The Kyber I/O scheduler maintains one internal queue per hwq and hence derives its async_depth from the number of hwq tags. Using this approach for the mq-deadline scheduler is wrong since the mq-deadline scheduler maintains one internal queue for all hwqs combined. Hence this revert. Cc: stable@vger.kernel.org Cc: Damien Le Moal Cc: Harshit Mogalapalli Cc: Zhiguo Niu Fixes: d47f9717e5cf ("block/mq-deadline: use correct way to throttling write requests") Signed-off-by: Bart Van Assche Link: https://lore.kernel.org/r/20240313214218.1736147-1-bvanassche@acm.org Signed-off-by: Jens Axboe Signed-off-by: Sasha Levin

block: Fix page refcounts for unaligned buffers in __bio_release_pages()

2024-04-03T13:28:27Z

[ Upstream commit 38b43539d64b2fa020b3b9a752a986769f87f7a6 ] Fix an incorrect number of pages being released for buffers that do not start at the beginning of a page. Fixes: 1b151e2435fc ("block: Remove special-casing of compound pages") Cc: stable@vger.kernel.org Signed-off-by: Tony Battersby Tested-by: Greg Edwards Link: https://lore.kernel.org/r/86e592a9-98d4-4cff-a646-0c0084328356@cybernetics.com Signed-off-by: Jens Axboe Signed-off-by: Sasha Levin

block: Clear zone limits for a non-zoned stacked queue

2024-04-03T13:28:20Z

[ Upstream commit c8f6f88d25929ad2f290b428efcae3b526f3eab0 ] Device mapper may create a non-zoned mapped device out of a zoned device (e.g., the dm-zoned target). In such case, some queue limit such as the max_zone_append_sectors and zone_write_granularity endup being non zero values for a block device that is not zoned. Avoid this by clearing these limits in blk_stack_limits() when the stacked zoned limit is false. Fixes: 3093a479727b ("block: inherit the zoned characteristics in blk_stack_limits") Cc: stable@vger.kernel.org Signed-off-by: Damien Le Moal Link: https://lore.kernel.org/r/20240222131724.1803520-1-dlemoal@kernel.org Signed-off-by: Jens Axboe Signed-off-by: Sasha Levin

block: Provide bdev_open_* functions

2024-03-26T22:19:40Z

[ Upstream commit e719b4d156749f02eafed31a3c515f2aa9dcc72a ] Create struct bdev_handle that contains all parameters that need to be passed to blkdev_put() and provide bdev_open_* functions that return this structure instead of plain bdev pointer. This will eventually allow us to pass one more argument to blkdev_put() (renamed to bdev_release()) without too much hassle. Acked-by: Christoph Hellwig Reviewed-by: Christian Brauner Signed-off-by: Jan Kara Link: https://lore.kernel.org/r/20230927093442.25915-1-jack@suse.cz Signed-off-by: Christian Brauner Stable-dep-of: 0f28be64d132 ("erofs: fix lockdep false positives on initializing erofs_pseudo_mnt") Signed-off-by: Sasha Levin