[ Upstream commit 99dc422335 ]
Currently, io_ticks is accounted based on sampling, specifically
update_io_ticks() will always account io_ticks by 1 jiffies from
bdev_start_io_acct()/blk_account_io_start(), and the result can be
inaccurate, for example(HZ is 250):
Test script:
fio -filename=/dev/sda -bs=4k -rw=write -direct=1 -name=test -thinktime=4ms
Test result: util is about 90%, while the disk is really idle.
This behaviour is introduced by commit 5b18b5a737 ("block: delete
part_round_stats and switch to less precise counting"), however, there
was a key point that is missed that this patch also improve performance
a lot:
Before the commit:
part_round_stats:
if (part->stamp != now)
stats |= 1;
part_in_flight()
-> there can be lots of task here in 1 jiffies.
part_round_stats_single()
__part_stat_add()
part->stamp = now;
After the commit:
update_io_ticks:
stamp = part->bd_stamp;
if (time_after(now, stamp))
if (try_cmpxchg())
__part_stat_add()
-> only one task can reach here in 1 jiffies.
Hence in order to account io_ticks precisely, we only need to know if
there are IO inflight at most once in one jiffies. Noted that for
rq-based device, iterating tags should not be used here because
'tags->lock' is grabbed in blk_mq_find_and_get_req(), hence
part_stat_lock_inc/dec() and part_in_flight() is used to trace inflight.
The additional overhead is quite little:
- per cpu add/dec for each IO for rq-based device;
- per cpu sum for each jiffies;
And it's verified by null-blk that there are no performance degration
under heavy IO pressure.
Fixes: 5b18b5a737 ("block: delete part_round_stats and switch to less precise counting")
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
Link: https://lore.kernel.org/r/20240509123717.3223892-2-yukuai1@huaweicloud.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 06965037ce ]
There is only one caller for __blk_account_io_done(), the function
is small enough to fit in its caller blk_account_io_done().
Remove the function and opencode in the its caller
blk_account_io_done().
Signed-off-by: Chaitanya Kulkarni <kch@nvidia.com>
Link: https://lore.kernel.org/r/20230327073427.4403-2-kch@nvidia.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Stable-dep-of: 99dc422335 ("block: support to account io_ticks precisely")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit e165fb4dd6 ]
There is only one caller for __blk_account_io_start(), the function
is small enough to fit in its caller blk_account_io_start().
Remove the function and opencode in the its caller
blk_account_io_start().
Signed-off-by: Chaitanya Kulkarni <kch@nvidia.com>
Link: https://lore.kernel.org/r/20230327073427.4403-2-kch@nvidia.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Stable-dep-of: 99dc422335 ("block: support to account io_ticks precisely")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit beaa51b360 ]
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:
<IRQ>
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 <riel@surriel.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Josef Bacik <josef@toxicpanda.com>
Cc: Jens Axboe <axboe@kernel.dk>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://lore.kernel.org/r/20240404123253.0f58010f@imladris.surriel.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit 55251fbdf0 upstream.
This reverts commit 748dc0b65e.
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 748dc0b65e ("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: 748dc0b65e ("block: fix partial zone append completion handling in req_bio_endio()")
Cc: stable@vger.kernel.org
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20240328004409.594888-2-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit e5c0ca1365 upstream.
Chuck reported [1] an IO hang problem on NFS exports that reside on SATA
devices and bisected to commit 615939a2ae ("blk-mq: defer to the normal
submission path for post-flush requests").
We analysed the IO hang problem, found there are two postflush requests
waiting for each other.
The first postflush request completed the REQ_FSEQ_DATA sequence, so go to
the REQ_FSEQ_POSTFLUSH sequence and added in the flush pending list, but
failed to blk_kick_flush() because of the second postflush request which
is inflight waiting in scheduler queue.
The second postflush waiting in scheduler queue can't be dispatched because
the first postflush hasn't released scheduler resource even though it has
completed by itself.
Fix it by releasing scheduler resource when the first postflush request
completed, so the second postflush can be dispatched and completed, then
make blk_kick_flush() succeed.
While at it, remove the check for e->ops.finish_request, as all
schedulers set that. Reaffirm this requirement by adding a WARN_ON_ONCE()
at scheduler registration time, just like we do for insert_requests and
dispatch_request.
[1] https://lore.kernel.org/all/7A57C7AE-A51A-4254-888B-FE15CA21F9E9@oracle.com/
Link: https://lore.kernel.org/linux-block/20230819031206.2744005-1-chengming.zhou@linux.dev/
Reported-by: kernel test robot <oliver.sang@intel.com>
Closes: https://lore.kernel.org/oe-lkp/202308172100.8ce4b853-oliver.sang@intel.com
Fixes: 615939a2ae ("blk-mq: defer to the normal submission path for post-flush requests")
Reported-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Chengming Zhou <zhouchengming@bytedance.com>
Tested-by: Chuck Lever <chuck.lever@oracle.com>
Link: https://lore.kernel.org/r/20230813152325.3017343-1-chengming.zhou@linux.dev
[axboe: folded in incremental fix and added tags]
Signed-off-by: Jens Axboe <axboe@kernel.dk>
[bvanassche: changed RQF_USE_SCHED into RQF_ELVPRIV; restored the
finish_request pointer check before calling finish_request and removed
the new warning from the elevator code. This patch fixes an I/O hang
when submitting a REQ_FUA request to a request queue for a zoned block
device for which FUA has been disabled (QUEUE_FLAG_FUA is not set).]
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit c8f6f88d25 ]
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: 3093a47972 ("block: inherit the zoned characteristics in blk_stack_limits")
Cc: stable@vger.kernel.org
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Link: https://lore.kernel.org/r/20240222131724.1803520-1-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 748dc0b65e ]
Partial completions of zone append request is not allowed but if a zone
append completion indicates a number of completed bytes different from
the original BIO size, only the BIO status is set to error. This leads
to bio_advance() not setting the BIO size to 0 and thus to not call
bio_endio() at the end of req_bio_endio().
Make sure a partially completed zone append is failed and completed
immediately by forcing the completed number of bytes (nbytes) to be
equal to the BIO size, thus ensuring that bio_endio() is called.
Fixes: 297db73184 ("block: fix req_bio_endio append error handling")
Cc: stable@kernel.vger.org
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Link: https://lore.kernel.org/r/20240110092942.442334-1-dlemoal@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit 33391eecd6 upstream.
We ran into an issue where a production workload would randomly grind to
a halt and not continue until the pending IO had timed out. This turned
out to be a complicated interaction between queue freezing and polled
IO:
1) You have an application that does polled IO. At any point in time,
there may be polled IO pending.
2) You have a monitoring application that issues a passthrough command,
which is marked with side effects such that it needs to freeze the
queue.
3) Passthrough command is started, which calls blk_freeze_queue_start()
on the device. At this point the queue is marked frozen, and any
attempt to enter the queue will fail (for non-blocking) or block.
4) Now the driver calls blk_mq_freeze_queue_wait(), which will return
when the queue is quiesced and pending IO has completed.
5) The pending IO is polled IO, but any attempt to poll IO through the
normal iocb_bio_iopoll() -> bio_poll() will fail when it gets to
bio_queue_enter() as the queue is frozen. Rather than poll and
complete IO, the polling threads will sit in a tight loop attempting
to poll, but failing to enter the queue to do so.
The end result is that progress for either application will be stalled
until all pending polled IO has timed out. This causes obvious huge
latency issues for the application doing polled IO, but also long delays
for passthrough command.
Fix this by treating queue enter for polled IO just like we do for
timeouts. This allows quick quiesce of the queue as we still poll and
complete this IO, while still disallowing queueing up new IO.
Reviewed-by: Keith Busch <kbusch@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 2a427b49d0 ]
When iocg_kick_delay() is called from a CPU different than the one which set
the delay, @now may be in the past of @iocg->delay_at leading to the
following warning:
UBSAN: shift-out-of-bounds in block/blk-iocost.c:1359:23
shift exponent 18446744073709 is too large for 64-bit type 'u64' (aka 'unsigned long long')
...
Call Trace:
<TASK>
dump_stack_lvl+0x79/0xc0
__ubsan_handle_shift_out_of_bounds+0x2ab/0x300
iocg_kick_delay+0x222/0x230
ioc_rqos_merge+0x1d7/0x2c0
__rq_qos_merge+0x2c/0x80
bio_attempt_back_merge+0x83/0x190
blk_attempt_plug_merge+0x101/0x150
blk_mq_submit_bio+0x2b1/0x720
submit_bio_noacct_nocheck+0x320/0x3e0
__swap_writepage+0x2ab/0x9d0
The underflow itself doesn't really affect the behavior in any meaningful
way; however, the past timestamp may exaggerate the delay amount calculated
later in the code, which shouldn't be a material problem given the nature of
the delay mechanism.
If @now is in the past, this CPU is racing another CPU which recently set up
the delay and there's nothing this CPU can contribute w.r.t. the delay.
Let's bail early from iocg_kick_delay() in such cases.
Reported-by: Breno Leitão <leitao@debian.org>
Signed-off-by: Tejun Heo <tj@kernel.org>
Fixes: 5160a5a53c ("blk-iocost: implement delay adjustment hysteresis")
Link: https://lore.kernel.org/r/ZVvc9L_CYk5LO1fT@slm.duckdns.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 5266caaf56 ]
In blk_mq_mark_tag_wait(), __add_wait_queue() may be re-ordered
with the following blk_mq_get_driver_tag() in case of getting driver
tag failure.
Then in __sbitmap_queue_wake_up(), waitqueue_active() may not observe
the added waiter in blk_mq_mark_tag_wait() and wake up nothing, meantime
blk_mq_mark_tag_wait() can't get driver tag successfully.
This issue can be reproduced by running the following test in loop, and
fio hang can be observed in < 30min when running it on my test VM
in laptop.
modprobe -r scsi_debug
modprobe scsi_debug delay=0 dev_size_mb=4096 max_queue=1 host_max_queue=1 submit_queues=4
dev=`ls -d /sys/bus/pseudo/drivers/scsi_debug/adapter*/host*/target*/*/block/* | head -1 | xargs basename`
fio --filename=/dev/"$dev" --direct=1 --rw=randrw --bs=4k --iodepth=1 \
--runtime=100 --numjobs=40 --time_based --name=test \
--ioengine=libaio
Fix the issue by adding one explicit barrier in blk_mq_mark_tag_wait(), which
is just fine in case of running out of tag.
Cc: Jan Kara <jack@suse.cz>
Cc: Kemeng Shi <shikemeng@huaweicloud.com>
Reported-by: Changhui Zhong <czhong@redhat.com>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Link: https://lore.kernel.org/r/20240112122626.4181044-1-ming.lei@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit 1b151e2435 upstream.
The special casing was originally added in pre-git history; reproducing
the commit log here:
> commit a318a92567d77
> Author: Andrew Morton <akpm@osdl.org>
> Date: Sun Sep 21 01:42:22 2003 -0700
>
> [PATCH] Speed up direct-io hugetlbpage handling
>
> This patch short-circuits all the direct-io page dirtying logic for
> higher-order pages. Without this, we pointlessly bounce BIOs up to
> keventd all the time.
In the last twenty years, compound pages have become used for more than
just hugetlb. Rewrite these functions to operate on folios instead
of pages and remove the special case for hugetlbfs; I don't think
it's needed any more (and if it is, we can put it back in as a call
to folio_test_hugetlb()).
This was found by inspection; as far as I can tell, this bug can lead
to pages used as the destination of a direct I/O read not being marked
as dirty. If those pages are then reclaimed by the MM without being
dirtied for some other reason, they won't be written out. Then when
they're faulted back in, they will not contain the data they should.
It'll take a pretty unusual setup to produce this problem with several
races all going the wrong way.
This problem predates the folio work; it could for example have been
triggered by mmaping a THP in tmpfs and using that as the target of an
O_DIRECT read.
Fixes: 800d8c63b2 ("shmem: add huge pages support")
Cc: <stable@vger.kernel.org>
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 7b4f36cd22 ]
q_usage_counter is the only thing preventing us from the limits changing
under us in __bio_split_to_limits, but blk_mq_submit_bio doesn't hold
it while calling into it.
Move the splitting inside the region where we know we've got a queue
reference. Ideally this could still remain a shared section of code, but
let's keep the fix simple and defer any refactoring here to later.
Reported-by: Christoph Hellwig <hch@lst.de>
Fixes: 900e080752 ("block: move queue enter logic into blk_mq_submit_bio()")
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit 6f64f866aa upstream.
Before calling add partition or resize partition, there is no check
on whether the length is aligned with the logical block size.
If the logical block size of the disk is larger than 512 bytes,
then the partition size maybe not the multiple of the logical block size,
and when the last sector is read, bio_truncate() will adjust the bio size,
resulting in an IO error if the size of the read command is smaller than
the logical block size.If integrity data is supported, this will also
result in a null pointer dereference when calling bio_integrity_free.
Cc: <stable@vger.kernel.org>
Signed-off-by: Min Li <min15.li@samsung.com>
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20230629142517.121241-1-min15.li@samsung.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 4c434392c4 ]
'first_minor' represents the starting minor number of disks, and
'minors' represents the number of partitions in the device. Neither
of them can be greater than MINORMASK + 1.
Commit e338924bd0 ("block: check minor range in device_add_disk()")
only added the check of 'first_minor + minors'. However, their sum might
be less than MINORMASK but their values are wrong. Complete the checks now.
Fixes: e338924bd0 ("block: check minor range in device_add_disk()")
Signed-off-by: Li Nan <linan122@huawei.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20231219075942.840255-1-linan666@huaweicloud.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 5fa3d1a00c ]
On the error path of device_add_disk(), device's memalloc_noio flag was
set but not cleared. As the comment of pm_runtime_set_memalloc_noio(),
"The function should be called between device_add() and device_del()".
Clear this flag before device_del() now.
Fixes: 25e823c8c3 ("block/genhd.c: apply pm_runtime_set_memalloc_noio on block devices")
Signed-off-by: Li Nan <linan122@huawei.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20231211075356.1839282-1-linan666@huaweicloud.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 0e4237ae8d ]
Request queue quiesce may interrupt flush sequence, and the original request
may have been marked as COMPLETE, but can't get finished because of
queue quiesce.
This way is fine from driver viewpoint, because flush sequence is block
layer concept, and it isn't related with driver.
However, driver(such as dm-rq) can call blk_mq_queue_inflight() to count &
drain inflight requests, then the wait & drain never gets done because
the completed & not-finished flush request is counted as inflight.
Fix this issue by not counting completed flush data request as inflight in
case of quiesce.
Cc: Mike Snitzer <snitzer@kernel.org>
Cc: David Jeffery <djeffery@redhat.com>
Cc: John Pittman <jpittman@redhat.com>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Link: https://lore.kernel.org/r/20231201085605.577730-1-ming.lei@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 1898efcdbe ]
Propagate the per-queue stable_write flags into each bdev inode in bdev_add.
This makes sure devices that require stable writes have it set for I/O
on the block device node as well.
Note that this doesn't cover the case of a flag changing on a live device
yet. We should handle that as well, but I plan to cover it as part of a
more general rework of how changing runtime paramters on block devices
works.
Fixes: 1cb039f3dc ("bdi: replace BDI_CAP_STABLE_WRITES with a queue and a sb flag")
Reported-by: Ilya Dryomov <idryomov@gmail.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20231025141020.192413-3-hch@lst.de
Tested-by: Ilya Dryomov <idryomov@gmail.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit b0077e269f ]
blk_integrity_unregister() can come if queue usage counter isn't held
for one bio with integrity prepared, so this request may be completed with
calling profile->complete_fn, then kernel panic.
Another constraint is that bio_integrity_prep() needs to be called
before bio merge.
Fix the issue by:
- call bio_integrity_prep() with one queue usage counter grabbed reliably
- call bio_integrity_prep() before bio merge
Fixes: 900e080752 ("block: move queue enter logic into blk_mq_submit_bio()")
Reported-by: Yi Zhang <yi.zhang@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Tested-by: Yi Zhang <yi.zhang@redhat.com>
Link: https://lore.kernel.org/r/20231113035231.2708053-1-ming.lei@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 1b0a151c10 ]
If one of the underlying disks of raid or dm is set to read-only, then
each io will generate new log, which will cause message storm. This
environment is indeed problematic, however we can't make sure our
naive custormer won't do this, hence use pr_warn_ratelimited() to
prevent message storm in this case.
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
Fixes: 57e95e4670 ("block: fix and cleanup bio_check_ro")
Signed-off-by: Ye Bin <yebin10@huawei.com>
Link: https://lore.kernel.org/r/20231107111247.2157820-1-yukuai1@huaweicloud.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit 2dd710d476 upstream.
Inexact, we may reject some not-overflowing values incorrectly, but
they'll be on the order of exabytes allowed anyways.
This fixes divide error crash on x86 if bps_limit is not configured or
is set too high in the rare case that jiffy_elapsed is greater than HZ.
Fixes: e8368b57c0 ("blk-throttle: use calculate_io/bytes_allowed() for throtl_trim_slice()")
Fixes: 8d6bbaada2 ("blk-throttle: prevent overflow while calculating wait time")
Signed-off-by: Khazhismel Kumykov <khazhy@google.com>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://lore.kernel.org/r/20231020223617.2739774-1-khazhy@google.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit eead005664 ]
Currently, 'carryover_ios/bytes' is not handled in throtl_trim_slice(),
for consequence, 'carryover_ios/bytes' will be used to throttle bio
multiple times, for example:
1) set iops limit to 100, and slice start is 0, slice end is 100ms;
2) current time is 0, and 10 ios are dispatched, those io won't be
throttled and io_disp is 10;
3) still at current time 0, update iops limit to 1000, carryover_ios is
updated to (0 - 10) = -10;
4) in this slice(0 - 100ms), io_allowed = 100 + (-10) = 90, which means
only 90 ios can be dispatched without waiting;
5) assume that io is throttled in slice(0 - 100ms), and
throtl_trim_slice() update silce to (100ms - 200ms). In this case,
'carryover_ios/bytes' is not cleared and still only 90 ios can be
dispatched between 100ms - 200ms.
Fix this problem by updating 'carryover_ios/bytes' in
throtl_trim_slice().
Fixes: a880ae93e5 ("blk-throttle: fix io hung due to configuration updates")
Reported-by: zhuxiaohui <zhuxiaohui.400@bytedance.com>
Link: https://lore.kernel.org/all/20230812072116.42321-1-zhuxiaohui.400@bytedance.com/
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://lore.kernel.org/r/20230816012708.1193747-5-yukuai1@huaweicloud.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit 1a721de848 upstream.
Commit a33df75c63 ("block: use an xarray for disk->part_tbl") remove
disk_expand_part_tbl() in add_partition(), which means all kinds of
devices will support extended dynamic `dev_t`.
However, some devices with GENHD_FL_NO_PART are not expected to add or
resize partition.
Fix this by adding check of GENHD_FL_NO_PART before add or resize
partition.
Fixes: a33df75c63 ("block: use an xarray for disk->part_tbl")
Signed-off-by: Li Lingfeng <lilingfeng3@huawei.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20230831075900.1725842-1-lilingfeng@huaweicloud.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit d47f9717e5 ]
The original formula was inaccurate:
dd->async_depth = max(1UL, 3 * q->nr_requests / 4);
For write requests, when we assign a tags from sched_tags,
data->shallow_depth will be passed to sbitmap_find_bit,
see the following code:
nr = sbitmap_find_bit_in_word(&sb->map[index],
min_t (unsigned int,
__map_depth(sb, index),
depth),
alloc_hint, wrap);
The smaller of data->shallow_depth and __map_depth(sb, index)
will be used as the maximum range when allocating bits.
For a mmc device (one hw queue, deadline I/O scheduler):
q->nr_requests = sched_tags = 128, so according to the previous
calculation method, dd->async_depth = data->shallow_depth = 96,
and the platform is 64bits with 8 cpus, sched_tags.bitmap_tags.sb.shift=5,
sb.maps[]=32/32/32/32, 32 is smaller than 96, whether it is a read or
a write I/O, tags can be allocated to the maximum range each time,
which has not throttling effect.
In addition, refer to the methods of bfg/kyber I/O scheduler,
limit ratiois are calculated base on sched_tags.bitmap_tags.sb.shift.
This patch can throttle write requests really.
Fixes: 07757588e5 ("block/mq-deadline: Reserve 25% of scheduler tags for synchronous requests")
Signed-off-by: Zhiguo Niu <zhiguo.niu@unisoc.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/1691061162-22898-1-git-send-email-zhiguo.niu@unisoc.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 43c9835b14 ]
Currently the write_cache attribute allows enabling the QUEUE_FLAG_WC
flag on devices that never claimed the capability.
Fix that by adding a QUEUE_FLAG_HW_WC flag that is set by
blk_queue_write_cache and guards re-enabling the cache through sysfs.
Note that any rescan that calls blk_queue_write_cache will still
re-enable the write cache as in the current code.
Fixes: 93e9d8e836 ("block: add ability to flag write back caching on a device")
Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20230707094239.107968-3-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit c984ff1423 upstream.
blk_crypto_profile_init() calls lockdep_register_key(), which warns and
does not register if the provided memory is a static object.
blk-crypto-fallback currently has a static blk_crypto_profile and calls
blk_crypto_profile_init() thereupon, resulting in the warning and
failure to register.
Fortunately it is simple enough to use a dynamically allocated profile
and make lockdep function correctly.
Fixes: 2fb48d88e7 ("blk-crypto: use dynamic lock class for blk_crypto_profile::lock")
Cc: stable@vger.kernel.org
Signed-off-by: Sweet Tea Dorminy <sweettea-kernel@dorminy.me>
Reviewed-by: Eric Biggers <ebiggers@google.com>
Link: https://lore.kernel.org/r/20230817141615.15387-1-sweettea-kernel@dorminy.me
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 7090426351 ]
We have seen rare IO stalls as follows:
* blk_mq_plug_issue_direct() is entered with an mq_list containing two
requests.
* For the first request, it sets last == false and enters the driver's
queue_rq callback.
* The driver queue_rq callback indirectly calls schedule() which calls
blk_flush_plug(). This may happen if the driver has the
BLK_MQ_F_BLOCKING flag set and is allowed to sleep in ->queue_rq.
* blk_flush_plug() handles the remaining request in the mq_list. mq_list
is now empty.
* The original call to queue_rq resumes (with last == false).
* The loop in blk_mq_plug_issue_direct() terminates because there are no
remaining requests in mq_list.
The IO is now stalled because the last request submitted to the driver
had last == false and there was no subsequent call to commit_rqs().
Fix this by returning early in blk_mq_flush_plug_list() if rq_count is 0
which it will be in the recursive case, rather than checking if the
mq_list is empty. At the same time, adjust one of the callers to skip
the mq_list empty check as it is not necessary.
Fixes: dc5fc361d8 ("block: attempt direct issue of plug list")
Signed-off-by: Ross Lagerwall <ross.lagerwall@citrix.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20230714101106.3635611-1-ross.lagerwall@citrix.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 2fb48d88e7 ]
When a device-mapper device is passing through the inline encryption
support of an underlying device, calls to blk_crypto_evict_key() take
the blk_crypto_profile::lock of the device-mapper device, then take the
blk_crypto_profile::lock of the underlying device (nested). This isn't
a real deadlock, but it causes a lockdep report because there is only
one lock class for all instances of this lock.
Lockdep subclasses don't really work here because the hierarchy of block
devices is dynamic and could have more than 2 levels.
Instead, register a dynamic lock class for each blk_crypto_profile, and
associate that with the lock.
This avoids false-positive lockdep reports like the following:
============================================
WARNING: possible recursive locking detected
6.4.0-rc5 #2 Not tainted
--------------------------------------------
fscryptctl/1421 is trying to acquire lock:
ffffff80829ca418 (&profile->lock){++++}-{3:3}, at: __blk_crypto_evict_key+0x44/0x1c0
but task is already holding lock:
ffffff8086b68ca8 (&profile->lock){++++}-{3:3}, at: __blk_crypto_evict_key+0xc8/0x1c0
other info that might help us debug this:
Possible unsafe locking scenario:
CPU0
----
lock(&profile->lock);
lock(&profile->lock);
*** DEADLOCK ***
May be due to missing lock nesting notation
Fixes: 1b26283970 ("block: Keyslot Manager for Inline Encryption")
Reported-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Eric Biggers <ebiggers@google.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20230610061139.212085-1-ebiggers@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit b90ecc0379 upstream.
Currently, associating a loop device with a different file descriptor
does not increment its diskseq. This allows the following race
condition:
1. Program X opens a loop device
2. Program X gets the diskseq of the loop device.
3. Program X associates a file with the loop device.
4. Program X passes the loop device major, minor, and diskseq to
something.
5. Program X exits.
6. Program Y detaches the file from the loop device.
7. Program Y attaches a different file to the loop device.
8. The opener finally gets around to opening the loop device and checks
that the diskseq is what it expects it to be. Even though the
diskseq is the expected value, the result is that the opener is
accessing the wrong file.
From discussions with Christoph Hellwig, it appears that
disk_force_media_change() was supposed to call inc_diskseq(), but in
fact it does not. Adding a Fixes: tag to indicate this. Christoph's
Reported-by is because he stated that disk_force_media_change()
calls inc_diskseq(), which is what led me to discover that it should but
does not.
Reported-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Demi Marie Obenour <demi@invisiblethingslab.com>
Fixes: e6138dc12d ("block: add a helper to raise a media changed event")
Cc: stable@vger.kernel.org # 5.15+
Reviewed-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20230607170837.1559-1-demi@invisiblethingslab.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit b6f3f28f60 upstream.
The Amiga partition parser module uses signed int for partition sector
address and count, which will overflow for disks larger than 1 TB.
Use u64 as type for sector address and size to allow using disks up to
2 TB without LBD support, and disks larger than 2 TB with LBD. The RBD
format allows to specify disk sizes up to 2^128 bytes (though native
OS limitations reduce this somewhat, to max 2^68 bytes), so check for
u64 overflow carefully to protect against overflowing sector_t.
Bail out if sector addresses overflow 32 bits on kernels without LBD
support.
This bug was reported originally in 2012, and the fix was created by
the RDB author, Joanne Dow <jdow@earthlink.net>. A patch had been
discussed and reviewed on linux-m68k at that time but never officially
submitted (now resubmitted as patch 1 in this series).
This patch adds additional error checking and warning messages.
Reported-by: Martin Steigerwald <Martin@lichtvoll.de>
Closes: https://bugzilla.kernel.org/show_bug.cgi?id=43511
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Message-ID: <201206192146.09327.Martin@lichtvoll.de>
Cc: <stable@vger.kernel.org> # 5.2
Signed-off-by: Michael Schmitz <schmitzmic@gmail.com>
Reviewed-by: Geert Uytterhoeven <geert@linux-m68k.org>
Reviewed-by: Christoph Hellwig <hch@infradead.org>
Link: https://lore.kernel.org/r/20230620201725.7020-4-schmitzmic@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit dd7de3704a ]
Commit 99d055b4fd ("block: remove per-disk debugfs files in
blk_unregister_queue") moves blk_trace_shutdown() from
blk_release_queue() to blk_unregister_queue(), this is safe if blktrace
is created through sysfs, however, there is a regression in corner
case.
blktrace can still be enabled after del_gendisk() through ioctl if
the disk is opened before del_gendisk(), and if blktrace is not shutdown
through ioctl before closing the disk, debugfs entries will be leaked.
Fix this problem by shutdown blktrace in disk_release(), this is safe
because blk_trace_remove() is reentrant.
Fixes: 99d055b4fd ("block: remove per-disk debugfs files in blk_unregister_queue")
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20230610022003.2557284-4-yukuai1@huaweicloud.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 4f1731df60 ]
In __blk_mq_tag_busy/idle(), updating 'active_queues' and calculating
'wake_batch' is not atomic:
t1: t2:
_blk_mq_tag_busy blk_mq_tag_busy
inc active_queues
// assume 1->2
inc active_queues
// 2 -> 3
blk_mq_update_wake_batch
// calculate based on 3
blk_mq_update_wake_batch
/* calculate based on 2, while active_queues is actually 3. */
Fix this problem by protecting them wih 'tags->lock', this is not a hot
path, so performance should not be concerned. And now that all writers
are inside the lock, switch 'actives_queues' from atomic to unsigned
int.
Fixes: 180dccb0db ("blk-mq: fix tag_get wait task can't be awakened")
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20230610023043.2559121-1-yukuai1@huaweicloud.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Yu Kuai
Chaitanya Kulkarni
Rik van Riel
Li Nan
Roman Smirnov
Damien Le Moal
Chengming Zhou
Tony Battersby
Bart Van Assche
Greg Joyce
Christian A. Ehrhardt
Jens Axboe
Tejun Heo
Ming Lei
Christoph Hellwig
Li Lingfeng
Matthew Wilcox (Oracle)
Min Li
Keith Busch
Christoph Hellwig
Sarthak Kukreti
Khazhismel Kumykov
Zhiguo Niu
Sweet Tea Dorminy
Ross Lagerwall
Eric Biggers
Michael Schmitz
Demi Marie Obenour