Add a helper to initialize a bvec based of a page pointer. This will help
removing various open code bvec initializations.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Link: https://lore.kernel.org/r/20230203150634.3199647-2-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
cgroup information only makes sense on a live gendisk that allows
file system I/O (which includes the raw block device). So move over
the cgroup related members.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Andreas Herrmann <aherrmann@suse.de>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://lore.kernel.org/r/20230203150400.3199230-20-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Pass a gendisk to blkg_lookup and use that to find the match as part
of phasing out usage of the request_queue in the blk-cgroup code.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Andreas Herrmann <aherrmann@suse.de>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://lore.kernel.org/r/20230203150400.3199230-19-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
No need to the request_queue here, pass a gendisk and extract the
node ids from that.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Andreas Herrmann <aherrmann@suse.de>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://lore.kernel.org/r/20230203150400.3199230-18-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Prepare for storing the blkcg information in the gendisk instead of
the request_queue.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Andreas Herrmann <aherrmann@suse.de>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://lore.kernel.org/r/20230203150400.3199230-17-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
This is what about half of the users already want, and it's only going to
grow more.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Andreas Herrmann <aherrmann@suse.de>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://lore.kernel.org/r/20230203150400.3199230-16-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
These op vectors are constant, so mark them const.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Andreas Herrmann <aherrmann@suse.de>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://lore.kernel.org/r/20230203150400.3199230-15-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Switch to passing a gendisk, and make rq_qos_add initialize all required
fields and drop the not required q argument from rq_qos_del.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Andreas Herrmann <aherrmann@suse.de>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://lore.kernel.org/r/20230203150400.3199230-14-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
These two functions are rather larger and not in a fast path, so move
them out of line.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://lore.kernel.org/r/20230203150400.3199230-13-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
wbt_queue_depth_changed just updates a field and calls another function.
Open code it in wbt_init, so that the local queue variable can be used
instead of the one stored in the rq_qos. This will allow delaying that
rq_qos->queue assignment in a subsequent patch.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Andreas Herrmann <aherrmann@suse.de>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://lore.kernel.org/r/20230203150400.3199230-12-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
A large part of blk-wbt.h is only used in blk-wbt.c, so move it there.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://lore.kernel.org/r/20230203150400.3199230-11-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Pass a gendisk to wbt_init to prepare for phasing out usage of the
request_queue in the blk-cgroup code.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Andreas Herrmann <aherrmann@suse.de>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://lore.kernel.org/r/20230203150400.3199230-10-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Pass a gendisk to wbt_enable_default and wbt_disable_default to
prepare for phasing out usage of the request_queue in the blk-cgroup
code.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Andreas Herrmann <aherrmann@suse.de>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://lore.kernel.org/r/20230203150400.3199230-9-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Switch from a request_queue pointer and reference to a gendisk once
for the throttle information in struct task_struct.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Andreas Herrmann <aherrmann@suse.de>
Link: https://lore.kernel.org/r/20230203150400.3199230-8-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Currently each blkcg_gq holds a request_queue reference, which is what
is used in the policies. But a lot of these interfaces will move over to
use a gendisk, so store a disk in struct blkcg_gq and hold a reference to
it.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Andreas Herrmann <aherrmann@suse.de>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://lore.kernel.org/r/20230203150400.3199230-7-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
There is no need to delay freeing a blkg to a workqueue when freeing it
after an initialization failure.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://lore.kernel.org/r/20230203150400.3199230-5-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Unwind only the previous initialization steps that happened in blkg_alloc
using goto based unwinding. This avoids the need for the !queue special
case in blkg_free and thus ensures that any blkg seens outside of
blkg_alloc is always fully constructed.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://lore.kernel.org/r/20230203150400.3199230-4-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
There is no need to initialize the cgroup code before the disk is marked
live. Moving the cgroup initialization earlier will help to have a
fully initialized struct device in the gendisk for the cgroup code to
use in the future. Similarly tear the cgroup information down in
del_gendisk to be symmetric and because none of the cgroup tracking is
needed once non-passthrough I/O stops.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Andreas Herrmann <aherrmann@suse.de>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://lore.kernel.org/r/20230203150400.3199230-3-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
blk_throtl_stat_add is called from blk_stat_add explicitly, unlike the
other stats that go through q->stats->callbacks. To prepare for cgroup
data moving to the gendisk, ensure blk_throtl_stat_add is only called
for the plain READ and WRITE commands that it actually handles internally,
as blk_stat_add can also be called for passthrough commands on queues that
do not have a gendisk associated with them.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Andreas Herrmann <aherrmann@suse.de>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://lore.kernel.org/r/20230203150400.3199230-2-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Pull MD updates from Song:
"Non-urgent fixes:
md: don't update recovery_cp when curr_resync is ACTIVE
md: Free writes_pending in md_stop
Performance optimization:
md: Change active_io to percpu"
* 'md-next' of https://git.kernel.org/pub/scm/linux/kernel/git/song/md:
md: use MD_RESYNC_* whenever possible
md: Free writes_pending in md_stop
md: Change active_io to percpu
md: Factor out is_md_suspended helper
md: don't update recovery_cp when curr_resync is ACTIVE
Just replace magic numbers by MD_RESYNC_* enumerations.
Signed-off-by: Hou Tao <houtao1@huawei.com>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Song Liu <song@kernel.org>
dm raid calls md_stop to stop the raid device. It needs to
free the writes_pending here.
Signed-off-by: Xiao Ni <xni@redhat.com>
Signed-off-by: Song Liu <song@kernel.org>
Now the type of active_io is atomic. It's used to count how many ios are
in the submitting process and it's added and decreased very time. But it
only needs to check if it's zero when suspending the raid. So we can
switch atomic to percpu to improve the performance.
After switching active_io to percpu type, we use the state of active_io
to judge if the raid device is suspended. And we don't need to wake up
->sb_wait in md_handle_request anymore. It's done in the callback function
which is registered when initing active_io. The argument mddev->suspended
is only used to count how many users are trying to set raid to suspend
state.
Signed-off-by: Xiao Ni <xni@redhat.com>
Signed-off-by: Song Liu <song@kernel.org>
This helper function will be used in next patch. It's easy for
understanding.
Signed-off-by: Xiao Ni <xni@redhat.com>
Signed-off-by: Song Liu <song@kernel.org>
Don't update recovery_cp when curr_resync is MD_RESYNC_ACTIVE, otherwise
md may skip the resync of the first 3 sectors if the resync procedure is
interrupted before the first calling of ->sync_request() as shown below:
md_do_sync thread control thread
// setup resync
mddev->recovery_cp = 0
j = 0
mddev->curr_resync = MD_RESYNC_ACTIVE
// e.g., set array as idle
set_bit(MD_RECOVERY_INTR, &&mddev_recovery)
// resync loop
// check INTR before calling sync_request
!test_bit(MD_RECOVERY_INTR, &mddev->recovery
// resync interrupted
// update recovery_cp from 0 to 3
// the resync of three 3 sectors will be skipped
mddev->recovery_cp = 3
Fixes: eac58d08d4 ("md: Use enum for overloaded magic numbers used by mddev->curr_resync")
Cc: stable@vger.kernel.org # 6.0+
Signed-off-by: Hou Tao <houtao1@huawei.com>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Song Liu <song@kernel.org>
Make the following minor changes which were reported by colleagues
while reviewing this code:
- Remove the parentheses from around the LOOP_DEFAULT_HW_Q_DEPTH
definition since these are superfluous.
- Accept other number formats than decimal, e.g. hexadecimal.
- Do not set hw_queue_depth to an out-of-range value, even if that value
won't be used.
- Use the LOOP_DEFAULT_HW_Q_DEPTH macro in the kernel module parameter
description to prevent that the description gets out of sync.
This patch has been tested as follows:
# modprobe -r loop
# modprobe loop hw_queue_depth=-1
modprobe: ERROR: could not insert 'loop': Invalid argument
# modprobe loop hw_queue_depth=0
modprobe: ERROR: could not insert 'loop': Invalid argument
# modprobe loop hw_queue_depth=1; cat /sys/module/loop/parameters/hw_queue_depth
1
# modprobe -r loop; modprobe loop; cat /sys/module/loop/parameters/hw_queue_depth hw_queue_depth=0x10
16
# modprobe -r loop; modprobe loop; cat /sys/module/loop/parameters/hw_queue_depth hw_queue_depth=128
128
# modprobe -r loop; modprobe loop hw_queue_depth=129; cat /sys/module/loop/parameters/hw_queue_depth
129
# modprobe -r loop; modprobe loop hw_queue_depth=$((1<<32))
modprobe: ERROR: could not insert 'loop': Numerical result out of range
See also commit ef44c50837 ("loop: allow user to set the queue
depth").
Cc: Chaitanya Kulkarni <kch@nvidia.com>
Cc: Himanshu Madhani <himanshu.madhani@oracle.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Link: https://lore.kernel.org/r/20230130211347.832110-1-bvanassche@acm.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
The nvme driver will freeze the IO queues in response to an admin
command with CSE bits set. These bits notify the host that the command
that's about to be executed needs to be done exclusively, hence the
freeze.
The Security Receive command is often reported by multiple vendors with
CSE bits set. The reason for this is that the result depends on the
previous Security Send. This has nothing to do with IO queues, though,
so the driver is taking an overly cautious response to seeing this
passthrough command, while unable to fufill the intended admin queue
action.
Rather than freeze IO during this harmless command, mask off the
effects. This freezing is observed to cause IO latency spikes when host
software periodically validates the security state of the drives.
Signed-off-by: Keith Busch <kbusch@kernel.org>
Reviewed-by: Jens Axboe <axboe@kernel.dk>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Instead of appending command effects flags per IO, set the known effects
flags the driver needs to react to just once during initial setup.
Signed-off-by: Keith Busch <kbusch@kernel.org>
Reviewed-by: Kanchan Joshi <joshi.k@samsung.com>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
This is due to the fact that the host is allowed to pass the controller
an sgl describing a buffer that is larger than the payload itself
Signed-off-by: Amit Engel <Amit.Engel@dell.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
This provides additional details about the rq/cmd that is timed out
example log if CONFIG_NVME_VERBOSE_ERRORS is configured:
"nvme nvme0: queue 2 timeout cid 0xd058 type 4 opc Write (0x1)"
example log if CONFIG_NVME_VERBOSE_ERRORS is not configured:
"nvme nvme0: queue 2 timeout cid 0xd058 type 4 opc I/O Cmd (0x1)"
Signed-off-by: Amit Engel <Amit.Engel@dell.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Christoph Hellwig <hch@lst.de>
nvme_opcode_str will handle io/admin/fabrics ops
This improves NVMe errors logging
Signed-off-by: Amit Engel <Amit.Engel@dell.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Christoph Hellwig <hch@lst.de>
After moving the nvme_passthru_end call to the callers of
nvme_execute_passthru_rq, this function has become quite pointless,
so remove it.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Jens Axboe <axboe@kernel.dk>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
The 'struct nvme_iod' space is appended at the end of the preallocated
'struct request', and padded to the cache line size. This leaves some
free memory (in most kernel configs) up for grabs.
Instead of appending the nvme data descriptor addresses after the
scatterlist, inline these for free within struct nvme_iod. There is now
enough space in the mempool for 128 possibe segments.
And without increasing the size of the preallocated requests, we can
hold up to 5 PRP descriptor elements, allowing the driver to increase
its max transfer size to 8MB.
Signed-off-by: Keith Busch <kbusch@kernel.org>
Signed-off-by: Christoph Hellwig <hch@lst.de>
The driver uses the dma entries for setting up its command's SGL/PRP
lists. The dma mapping might have fewer entries than the physical
segments, so check the dma mapped count to determine which nvme data
layout method is more optimal.
Signed-off-by: Keith Busch <kbusch@kernel.org>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
The max segments this driver can see is 127, well below the 256
threshold needed to add an nvme sgl segment descriptor. Remove all the
useless checks and dead code.
Signed-off-by: Keith Busch <kbusch@kernel.org>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
NVMe status codes are part of the wire protocol, and shouldn't be
fabricated in the stack. So with this patch the authentication code
is switched over to use error codes; as a side effect authentication
failures due to internal error won't be retried anymore.
But that shouldn't have happened anyway.
Signed-off-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Christoph Hellwig <hch@lst.de>
The NVMe 2.0 spec defines the ATR and ASCR bits in the AUTHREQ
connect response field to be mutually exclusive. So to clarify the
handling here switch the AUTHREQ handling to use the bit definitions
and check for both bits.
And while we're at it, add a message to the user that secure
concatenation is not supported (yet).
Suggested-by: Mark Lehrer <mark.lehrer@wdc.com>
Signed-off-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Christoph Hellwig <hch@lst.de>
This was originally added for the definition of nth_page(), but we no
longer use nth_page() in this header, so we can drop the heavyweight
mm.h now.
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Link: https://lore.kernel.org/r/20230131050132.2627124-1-willy@infradead.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Owner of one unprivileged ublk device could be one evil user, which
can grant this disk's privilege to other users deliberately, and
this way could be like making one trap and waiting for other users
to be caught.
So only owner to open unprivileged disk even though the owner
grants disk privilege to other user. This way is reasonable too
given anyone can create ublk disk, and no need other's grant.
Reported-by: Stefan Hajnoczi <stefanha@redhat.com>
Fixes: 4093cb5a06 ("ublk_drv: add mechanism for supporting unprivileged ublk device")
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Link: https://lore.kernel.org/r/20230131040446.214583-1-ming.lei@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Assuming that both Kconfig options, BLK_CGROUP and IOSCHED_BFQ are set, we
most likely want cgroup support for BFQ too (BFQ_GROUP_IOSCHED), so let's
make it default y.
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20230130121240.159456-1-ulf.hansson@linaro.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
bfqd->bfq_wr_max_time is set to 0 in bfq_init_queue and is never changed.
It is only used in bfq_wr_duration when bfq_wr_max_time > 0 which never
meets, so bfqd->bfq_wr_max_time is not used actually. Just remove it.
Signed-off-by: Kemeng Shi <shikemeng@huaweicloud.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20230116095153.3810101-9-shikemeng@huaweicloud.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We jump to tag only for returning current rq. Return directly to
remove this tag.
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Damien Le Moal <damien.lemoal@opensource.wdc.com>
Signed-off-by: Kemeng Shi <shikemeng@huaweicloud.com>
Link: https://lore.kernel.org/r/20230116095153.3810101-8-shikemeng@huaweicloud.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We have already avoided a circular list in bfq_setup_merge (see comments
in bfq_setup_merge() for details), so bfq_queue will not appear in it's
new_bfqq list. Just remove this check.
Signed-off-by: Kemeng Shi <shikemeng@huaweicloud.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20230116095153.3810101-7-shikemeng@huaweicloud.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Inject limit is updated or reset when time_is_before_eq_jiffies(
decrease_time_jif + several msecs) or think-time state changes.
decrease_time_jif is initialized to 0 and will be set to current jiffies
when inject limit is updated or reset. If the jiffies is slightly greater
than LONG_MAX, time_is_after_eq_jiffies(0) will keep for a long time, so as
time_is_after_eq_jiffies(decrease_time_jif + several msecs). If the
think-time state never chages, then the injection will not work as expected
for long time.
To be more specific:
Function bfq_update_inject_limit maybe triggered when jiffies pasts
decrease_time_jif + msecs_to_jiffies(10) in bfq_add_request by setting
bfqd->wait_dispatch to true.
Function bfq_reset_inject_limit are called in two conditions:
1. jiffies pasts bfqq->decrease_time_jif + msecs_to_jiffies(1000) in
function bfq_add_request.
2. jiffies pasts bfqq->decrease_time_jif + msecs_to_jiffies(100) or
bfq think-time state change from short to long.
Fix this by initializing bfqq->decrease_time_jif to current jiffies
to trigger service injection soon when service injection conditions
are met.
Signed-off-by: Kemeng Shi <shikemeng@huaweicloud.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20230116095153.3810101-4-shikemeng@huaweicloud.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Function bfq_choose_bfqq_for_injection may temporarily raise inject limit
to one request if current inject_limit is 0 before search of the source
queue for injection. However the search below will reset inject limit to
bfqd->in_service_queue which is zero for raised inject limit. Then the
temporarily raised inject limit never works as expected.
Assigment limit to bfqd->in_service_queue in search is needed as limit
maybe overwriten to min_t(unsigned int, 1, limit) for condition that
a large in-flight request is on non-rotational devices in found queue.
So we need to reset limit to bfqd->in_service_queue for normal case.
Actually, we have already make sure bfqd->rq_in_driver is < limit before
search, then
-Limit is >= 1 as bfqd->rq_in_driver is >= 0. Then min_t(unsigned int,
1, limit) is always 1. So we can simply check bfqd->rq_in_driver with
1 instead of result of min_t(unsigned int, 1, limit) for larget request in
non-rotational device case to avoid overwritting limit and the bug is gone.
-For normal case, we have already check bfqd->rq_in_driver is < limit,
so we can return found bfqq unconditionally to remove unncessary check.
Signed-off-by: Kemeng Shi <shikemeng@huaweicloud.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20230116095153.3810101-2-shikemeng@huaweicloud.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Commit 180dccb0db ("blk-mq: fix tag_get wait task can't be awakened")
mentioned that in case of shared tags, there could be just one real
active hctx(queue) because of lazy detection of tag idle. Then driver tag
allocation may wait forever on this real active hctx(queue) if wake_batch
is > hctx_max_depth where hctx_max_depth is available tags depth for the
actve hctx(queue). However, the condition wake_batch > hctx_max_depth is
not strong enough to avoid IO hung as the sbitmap_queue_wake_up will only
wake up one wait queue for each wake_batch even though there is only one
waiter in the woken wait queue. After this, there is only one tag to free
and wake_batch may not be reached anymore. Commit 180dccb0db ("blk-mq:
fix tag_get wait task can't be awakened") methioned that driver tag
allocation may wait forever. Actually, the inactive hctx(queue) will be
truely idle after at most 30 seconds and will call blk_mq_tag_wakeup_all
to wake one waiter per wait queue to break the hung. But IO hung for 30
seconds is also not acceptable. Set batch size to small enough that depth
of the shared hctx(queue) is enough to wake up all of the queues like
sbq_calc_wake_batch do to fix this potential IO hung.
Although hctx_max_depth will be clamped to at least 4 while wake_batch
recalculation does not do the clamp, the wake_batch will be always
recalculated to 1 when hctx_max_depth <= 4.
Fixes: 180dccb0db ("blk-mq: fix tag_get wait task can't be awakened")
Reviewed-by: Jan Kara <jack@suse.cz>
Signed-off-by: Kemeng Shi <shikemeng@huaweicloud.com>
Link: https://lore.kernel.org/r/20230116205059.3821738-6-shikemeng@huaweicloud.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>