ioc_pd_free() grabs irq-safe ioc->lock without ensuring that irq is disabled
when it can be called with irq disabled or enabled. This has a small chance
of causing A-A deadlocks and triggers lockdep splats. Use irqsave operations
instead.
Signed-off-by: Tejun Heo <tj@kernel.org>
Fixes: 7caa47151a ("blkcg: implement blk-iocost")
Cc: stable@vger.kernel.org # v5.4+
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We need to hold the whole device bd_mutex to protect against
other thread concurrently deleting out partition before we get
to it, and thus causing a use after free.
Fixes: cddae808ae ("block: pass a hd_struct to delete_partition")
Reported-by: syzbot+6448f3c229bc52b82f69@syzkaller.appspotmail.com
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Commit e8c7d14ac6 ("block: revert back to synchronous request_queue removal")
stops to release request queue from wq context because that commit
supposed all blk_put_queue() is called in context which is allowed
to sleep. However, this assumption isn't true because we release disk's
reference in partition's percpu_ref's ->release() which doesn't allow
to sleep, because the ->release() is run via call_rcu().
Fixes this issue by moving put disk reference into hd_struct_free_work()
Fixes: e8c7d14ac6 ("block: revert back to synchronous request_queue removal")
Reported-by: Ilya Dryomov <idryomov@gmail.com>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Tested-by: Ilya Dryomov <idryomov@gmail.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Cc: Luis Chamberlain <mcgrof@kernel.org>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
If a driver leaves the limit settings as the defaults, then we don't
initialize bdi->io_pages. This means that file systems may need to
work around bdi->io_pages == 0, which is somewhat messy.
Initialize the default value just like we do for ->ra_pages.
Cc: stable@vger.kernel.org
Fixes: 9491ae4aad ("mm: don't cap request size based on read-ahead setting")
Reported-by: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Pull NVMe fixes from Sagi:
"- instance leak and io boundary fixes from Keith
- fc locking fix from Christophe
- various tcp/rdma reset during traffic fixes from Me
- pci use-after-free fix from Tong
- tcp target null deref fix from Ziye"
* 'nvme-5.9-rc' of git://git.infradead.org/nvme:
nvme-pci: cancel nvme device request before disabling
nvme: only use power of two io boundaries
nvme: fix controller instance leak
nvmet-fc: Fix a missed _irqsave version of spin_lock in 'nvmet_fc_fod_op_done()'
nvme: Fix NULL dereference for pci nvme controllers
nvme-rdma: fix reset hang if controller died in the middle of a reset
nvme-rdma: fix timeout handler
nvme-rdma: serialize controller teardown sequences
nvme-tcp: fix reset hang if controller died in the middle of a reset
nvme-tcp: fix timeout handler
nvme-tcp: serialize controller teardown sequences
nvme: have nvme_wait_freeze_timeout return if it timed out
nvme-fabrics: don't check state NVME_CTRL_NEW for request acceptance
nvmet-tcp: Fix NULL dereference when a connect data comes in h2cdata pdu
The kernel requires a power of two for boundaries because that's the
only way it can efficiently split commands that cross them. A
controller, however, may report a non-power of two boundary.
The driver had been rounding the controller's value to one the kernel
can use, but splitting on the wrong boundary provides no benefit on the
device side, and incurs additional submission overhead from non-optimal
splits.
Don't provide any boundary hint if the controller's value can't be used
and log a warning when first scanning a disk's unreported IO boundary.
Since the chunk sector logic has grown, move it to a separate function.
Cc: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Keith Busch <kbusch@kernel.org>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
If the driver has to unbind from the controller for an early failure
before the subsystem has been set up, there won't be a subsystem holding
the controller's instance, so the controller needs to free its own
instance in this case.
Fixes: 733e4b69d5 ("nvme: Assign subsys instance from first ctrl")
Signed-off-by: Keith Busch <kbusch@kernel.org>
Reviewed-by: Chaitanya Kulkarni <chaitanya.kulkarni@wdc.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
The way 'spin_lock()' and 'spin_lock_irqsave()' are used is not consistent
in this function.
Use 'spin_lock_irqsave()' also here, as there is no guarantee that
interruptions are disabled at that point, according to surrounding code.
Fixes: a97ec51b37 ("nvmet_fc: Rework target side abort handling")
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
PCIe controllers do not have fabric opts, verify they exist before
showing ctrl_loss_tmo or reconnect_delay attributes.
Fixes: 764075fdcb ("nvme: expose reconnect_delay and ctrl_loss_tmo via sysfs")
Reported-by: Tobias Markus <tobias@markus-regensburg.de>
Reviewed-by: Keith Busch <kbusch@kernel.org>
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
If the controller becomes unresponsive in the middle of a reset, we
will hang because we are waiting for the freeze to complete, but that
cannot happen since we have commands that are inflight holding the
q_usage_counter, and we can't blindly fail requests that times out.
So give a timeout and if we cannot wait for queue freeze before
unfreezing, fail and have the error handling take care how to
proceed (either schedule a reconnect of remove the controller).
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
When a request times out in a LIVE state, we simply trigger error
recovery and let the error recovery handle the request cancellation,
however when a request times out in a non LIVE state, we make sure to
complete it immediately as it might block controller setup or teardown
and prevent forward progress.
However tearing down the entire set of I/O and admin queues causes
freeze/unfreeze imbalance (q->mq_freeze_depth) because and is really
an overkill to what we actually need, which is to just fence controller
teardown that may be running, stop the queue, and cancel the request if
it is not already completed.
Now that we have the controller teardown_lock, we can safely serialize
request cancellation. This addresses a hang caused by calling extra
queue freeze on controller namespaces, causing unfreeze to not complete
correctly.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: James Smart <james.smart@broadcom.com>
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
In the timeout handler we may need to complete a request because the
request that timed out may be an I/O that is a part of a serial sequence
of controller teardown or initialization. In order to complete the
request, we need to fence any other context that may compete with us
and complete the request that is timing out.
In this case, we could have a potential double completion in case
a hard-irq or a different competing context triggered error recovery
and is running inflight request cancellation concurrently with the
timeout handler.
Protect using a ctrl teardown_lock to serialize contexts that may
complete a cancelled request due to error recovery or a reset.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: James Smart <james.smart@broadcom.com>
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
If the controller becomes unresponsive in the middle of a reset, we will
hang because we are waiting for the freeze to complete, but that cannot
happen since we have commands that are inflight holding the
q_usage_counter, and we can't blindly fail requests that times out.
So give a timeout and if we cannot wait for queue freeze before
unfreezing, fail and have the error handling take care how to proceed
(either schedule a reconnect of remove the controller).
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
When a request times out in a LIVE state, we simply trigger error
recovery and let the error recovery handle the request cancellation,
however when a request times out in a non LIVE state, we make sure to
complete it immediately as it might block controller setup or teardown
and prevent forward progress.
However tearing down the entire set of I/O and admin queues causes
freeze/unfreeze imbalance (q->mq_freeze_depth) because and is really
an overkill to what we actually need, which is to just fence controller
teardown that may be running, stop the queue, and cancel the request if
it is not already completed.
Now that we have the controller teardown_lock, we can safely serialize
request cancellation. This addresses a hang caused by calling extra
queue freeze on controller namespaces, causing unfreeze to not complete
correctly.
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
In the timeout handler we may need to complete a request because the
request that timed out may be an I/O that is a part of a serial sequence
of controller teardown or initialization. In order to complete the
request, we need to fence any other context that may compete with us
and complete the request that is timing out.
In this case, we could have a potential double completion in case
a hard-irq or a different competing context triggered error recovery
and is running inflight request cancellation concurrently with the
timeout handler.
Protect using a ctrl teardown_lock to serialize contexts that may
complete a cancelled request due to error recovery or a reset.
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
Users can detect if the wait has completed or not and take appropriate
actions based on this information (e.g. weather to continue
initialization or rather fail and schedule another initialization
attempt).
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
NVME_CTRL_NEW should never see any I/O, because in order to start
initialization it has to transition to NVME_CTRL_CONNECTING and from
there it will never return to this state.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
When handling commands without in-capsule data, we assign the ttag
assuming we already have the queue commands array allocated (based
on the queue size information in the connect data payload). However
if the connect itself did not send the connect data in-capsule we
have yet to allocate the queue commands,and we will assign a bogus
ttag and suffer a NULL dereference when we receive the corresponding
h2cdata pdu.
Fix this by checking if we already allocated commands before
dereferencing it when handling h2cdata, if we didn't, its for sure a
connect and we should use the preallocated connect command.
Signed-off-by: Ziye Yang <ziye.yang@intel.com>
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
Commit 3b5408b98e ("md/raid5: support config stripe_size by sysfs
entry") make stripe_size as a configurable value. It just requires
stripe_size as multiple of 4KB.
In fact, we should make sure stripe_size as power of two. Otherwise,
stripe_shift which is the result of ilog2 can not represent the real
stripe_size. Then, stripe_hash() and stripe_hash_locks_hash() may
get unexpected value.
Fixes: 3b5408b98e ("md/raid5: support config stripe_size by sysfs entry")
Signed-off-by: Yufen Yu <yuyufen@huawei.com>
Signed-off-by: Song Liu <songliubraving@fb.com>
The device size calculation was done before processing the loop
configuration, which meant that the we set the size on the underlying
block device incorrectly in case lo_offset/lo_sizelimit were set in the
configuration. Delay computing the size until we've setup the device
parameters correctly.
Fixes: 3448914e8cc5("loop: Add LOOP_CONFIGURE ioctl")
Reported-by: Lennart Poettering <mzxreary@0pointer.de>
Tested-by: Yang Xu <xuyang2018.jy@cn.fujitsu.com>
Signed-off-by: Martijn Coenen <maco@android.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
If we configured io timeout of nbd0 to 100s. Later after we
finished using it, we configured nbd0 again and set the io
timeout to 0. We expect it would timeout after 30 seconds
and keep retry. But in fact we could not change the timeout
when we set it to 0. the timeout is still the original 100s.
So change the timeout to default 30s when we set it to zero.
It also behaves same as commit 2da22da573 ("nbd: fix zero
cmd timeout handling v2").
It becomes more important if we were reconfigure a nbd device
and the io timeout it set to zero. Because it could take 30s
to detect the new socket and thus io could be completed more
quickly compared to 100s.
Signed-off-by: Hou Pu <houpu@bytedance.com>
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
REQ_FUA should be checked using rq->cmd_flags instead of req_op().
Fixes: deb78b419d ("nullb: emulate cache")
Signed-off-by: Hou Pu <houpu@bytedance.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Based on nvme spec, when keep alive timeout is set to zero
the keep-alive timer should be disabled.
Signed-off-by: Amit Engel <amit.engel@dell.com>
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
If a command send through nvme-multipath failed on a dying queue, resend it
on another path.
Signed-off-by: Chao Leng <lengchao@huawei.com>
[hch: rebased on top of the completion refactoring]
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Check the SCT sub-field for a path related status instead of enumerating
invididual status code. As of NVMe 1.4 this adds "Internal Path Error"
and "Controller Pathing Error" to the list, but it also future proofs for
additional status codes added to the category.
Suggested-by: Chao Leng <lengchao@huawei.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Lift all the code to decide the dispostition of a completed command
from nvme_complete_rq and nvme_failover_req into a new helper, which
returns an emum of the potential actions. nvme_complete_rq then
just switches on those and calls the proper helper for the action.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
nvme_end_request is a bit misnamed, as it wraps around the
blk_mq_complete_* API. It's semantics also are non-trivial, so give it
a more descriptive name and add a comment explaining the semantics.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Zoned block devices reuse the chunk_sectors queue limit to define zone
boundaries. If a such a device happens to also report an optimal
boundary, do not use that to define the chunk_sectors as that may
intermittently interfere with io splitting and zone size queries.
Signed-off-by: Keith Busch <kbusch@kernel.org>
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
All operations are based on the controller, not the host page size.
Switch the dma pool to use the controller page size as well to avoid
massive overallocations on large page size systems.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Keith Busch <kbusch@kernel.org>
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
When locking the ctrl->lock spinlock IRQs need to be disabled to avoid a
dead lock. The new spin_lock() calls recently added produce the
following lockdep warning when running the blktest nvme/003:
================================
WARNING: inconsistent lock state
--------------------------------
inconsistent {SOFTIRQ-ON-W} -> {IN-SOFTIRQ-W} usage.
ksoftirqd/2/22 [HC0[0]:SC1[1]:HE0:SE0] takes:
ffff888276a8c4c0 (&ctrl->lock){+.?.}-{2:2}, at: nvme_keep_alive_end_io+0x50/0xc0
{SOFTIRQ-ON-W} state was registered at:
lock_acquire+0x164/0x500
_raw_spin_lock+0x28/0x40
nvme_get_effects_log+0x37/0x1c0
nvme_init_identify+0x9e4/0x14f0
nvme_reset_work+0xadd/0x2360
process_one_work+0x66b/0xb70
worker_thread+0x6e/0x6c0
kthread+0x1e7/0x210
ret_from_fork+0x22/0x30
irq event stamp: 1449221
hardirqs last enabled at (1449220): [<ffffffff81c58e69>] ktime_get+0xf9/0x140
hardirqs last disabled at (1449221): [<ffffffff83129665>] _raw_spin_lock_irqsave+0x25/0x60
softirqs last enabled at (1449210): [<ffffffff83400447>] __do_softirq+0x447/0x595
softirqs last disabled at (1449215): [<ffffffff81b489b5>] run_ksoftirqd+0x35/0x50
other info that might help us debug this:
Possible unsafe locking scenario:
CPU0
----
lock(&ctrl->lock);
<Interrupt>
lock(&ctrl->lock);
*** DEADLOCK ***
no locks held by ksoftirqd/2/22.
stack backtrace:
CPU: 2 PID: 22 Comm: ksoftirqd/2 Not tainted 5.8.0-rc4-eid-vmlocalyes-dbg-00157-g7236657c6b3a #1450
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.12.0-1 04/01/2014
Call Trace:
dump_stack+0xc8/0x11a
print_usage_bug.cold.63+0x235/0x23e
mark_lock+0xa9c/0xcf0
__lock_acquire+0xd9a/0x2b50
lock_acquire+0x164/0x500
_raw_spin_lock_irqsave+0x40/0x60
nvme_keep_alive_end_io+0x50/0xc0
blk_mq_end_request+0x158/0x210
nvme_complete_rq+0x146/0x500
nvme_loop_complete_rq+0x26/0x30 [nvme_loop]
blk_done_softirq+0x187/0x1e0
__do_softirq+0x118/0x595
run_ksoftirqd+0x35/0x50
smpboot_thread_fn+0x1d3/0x310
kthread+0x1e7/0x210
ret_from_fork+0x22/0x30
Fixes: be93e87e78 ("nvme: support for multiple Command Sets Supported and Effects log pages")
Signed-off-by: Logan Gunthorpe <logang@deltatee.com>
Reviewed-by: Keith Busch <kbusch@kernel.org>
Tested-by: Chaitanya Kulkarni <chaitanya.kulkarni@wdc.com>
Reviewed-by: Chaitanya Kulkarni <chaitanya.kulkarni@wdc.com>
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Instead of calling blk_put_request() which calls blk_mq_free_request(),
call blk_mq_free_request() directly for NVMeOF passthru. This is to
mainly avoid an extra function call in the completion path
nvmet_passthru_req_done().
Signed-off-by: Chaitanya Kulkarni <chaitanya.kulkarni@wdc.com>
Reviewed-by: Keith Busch <kbusch@kernel.org>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
In the current implementation before submitting the passthru cmd we
may come across error e.g. getting ns from passthru controller,
allocating a request from passthru controller, etc. For all the failure
cases it only uses single goto label fail_out.
In the target code, we follow the pattern to have a separate label for
each error out the case when setting up multiple things before the actual
action.
This patch follows the same pattern and renames generic fail_out label
to out_put_ns and updates the error out cases in the
nvmet_passthru_execute_cmd() where it is needed.
Signed-off-by: Chaitanya Kulkarni <chaitanya.kulkarni@wdc.com>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
If we find an optimized path, we quit the loop immediately. Thus we can use
just one variable for the next path, slighly simplifying the code.
Signed-off-by: Martin Wilck <mwilck@suse.com>
Reviewed-by: Keith Busch <kbusch@kernel.org>
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
If there's only one usable, non-optimized path, nvme_round_robin_path()
returns NULL, which is wrong. Fix it by falling back to "old", like in
the single optimized path case. Also, if the active path isn't changed,
there's no need to re-assign the pointer.
Fixes: 3f6e3246db ("nvme-multipath: fix logic for non-optimized paths")
Signed-off-by: Martin Wilck <mwilck@suse.com>
Signed-off-by: Martin George <marting@netapp.com>
Reviewed-by: Keith Busch <kbusch@kernel.org>
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
On an error exit path, a negative error code should be returned
instead of a positive return value.
Fixes: e399441de9 ("nvme-fabrics: Add host support for FC transport")
Cc: James Smart <jsmart2021@gmail.com>
Signed-off-by: Tianjia Zhang <tianjia.zhang@linux.alibaba.com>
Reviewed-by: Chaitanya Kulkarni <chaitanya.kulkarni@wdc.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Any command with a non-SGL flag set (like fuse flags) should be
rejected.
Fixes: c1fef73f79 ("nvmet: add passthru code to process commands")
Signed-off-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Normally, blkcg_iolatency_exit() will free related memory in iolatency
when cleanup queue. But if blk_throtl_init() return error and queue init
fail, blkcg_iolatency_exit() will not do that for us. Then it cause
memory leak.
Fixes: d706751215 ("block: introduce blk-iolatency io controller")
Signed-off-by: Yufen Yu <yuyufen@huawei.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
The various <linux/blk*.h> header files are part of the Block Layer.
Add them to the corresponding section in the MAINTAINERS file, so
scripts/get_maintainer.pl will pick them up.
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
A previous commit aligning splits to physical block sizes inadvertently
modified one return case such that that it now returns 0 length splits
when the number of sectors doesn't exceed the physical offset. This
later hits a BUG in bio_split(). Restore the previous working behavior.
Fixes: 9cc5169cd4 ("block: Improve physical block alignment of split bios")
Reported-by: Eric Deal <eric.deal@wdc.com>
Signed-off-by: Keith Busch <kbusch@kernel.org>
Cc: Bart Van Assche <bvanassche@acm.org>
Cc: stable@vger.kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
c616cbee97 ("blk-mq: punt failed direct issue to dispatch list") supposed
to add request which has been through ->queue_rq() to the hw queue dispatch
list, however it adds request running out of budget or driver tag to hw queue
too. This way basically bypasses request merge, and causes too many request
dispatched to LLD, and system% is unnecessary increased.
Fixes this issue by adding request not through ->queue_rq into sw/scheduler
queue, and this way is safe because no ->queue_rq is called on this request
yet.
High %system can be observed on Azure storvsc device, and even soft lock
is observed. This patch reduces %system during heavy sequential IO,
meantime decreases soft lockup risk.
Fixes: c616cbee97 ("blk-mq: punt failed direct issue to dispatch list")
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Bart Van Assche <bvanassche@acm.org>
Cc: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Clang warns:
drivers/block/rnbd/rnbd-srv.c:150:6: warning: variable 'err' is used
uninitialized whenever 'if' condition is true
[-Wsometimes-uninitialized]
if (IS_ERR(bio)) {
^~~~~~~~~~~
drivers/block/rnbd/rnbd-srv.c:177:9: note: uninitialized use occurs here
return err;
^~~
drivers/block/rnbd/rnbd-srv.c:150:2: note: remove the 'if' if its
condition is always false
if (IS_ERR(bio)) {
^~~~~~~~~~~~~~~~~~
drivers/block/rnbd/rnbd-srv.c:126:9: note: initialize the variable 'err'
to silence this warning
int err;
^
= 0
1 warning generated.
err is indeed uninitialized when this statement is taken. Ensure that it
is assigned the error value of bio before jumping to the error handling
label.
Fixes: 735d77d4fd ("rnbd: remove rnbd_dev_submit_io")
Reported-by: Brooke Basile <brookebasile@gmail.com>
Signed-off-by: Nathan Chancellor <natechancellor@gmail.com>
Acked-by: Jack Wang <jinpu.wang@cloud.ionos.com>
Link: https://github.com/ClangBuiltLinux/linux/issues/1134
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Changes from v1:
- update commit description with proper ref-accounting justification
commit db37a34c56 ("block, bfq: get a ref to a group when adding it to a service tree")
introduce leak forbfq_group and blkcg_gq objects because of get/put
imbalance.
In fact whole idea of original commit is wrong because bfq_group entity
can not dissapear under us because it is referenced by child bfq_queue's
entities from here:
-> bfq_init_entity()
->bfqg_and_blkg_get(bfqg);
->entity->parent = bfqg->my_entity
-> bfq_put_queue(bfqq)
FINAL_PUT
->bfqg_and_blkg_put(bfqq_group(bfqq))
->kmem_cache_free(bfq_pool, bfqq);
So parent entity can not disappear while child entity is in tree,
and child entities already has proper protection.
This patch revert commit db37a34c56 ("block, bfq: get a ref to a group when adding it to a service tree")
bfq_group leak trace caused by bad commit:
-> blkg_alloc
-> bfq_pq_alloc
-> bfqg_get (+1)
->bfq_activate_bfqq
->bfq_activate_requeue_entity
-> __bfq_activate_entity
->bfq_get_entity
->bfqg_and_blkg_get (+1) <==== : Note1
->bfq_del_bfqq_busy
->bfq_deactivate_entity+0x53/0xc0 [bfq]
->__bfq_deactivate_entity+0x1b8/0x210 [bfq]
-> bfq_forget_entity(is_in_service = true)
entity->on_st_or_in_serv = false <=== :Note2
if (is_in_service)
return; ==> do not touch reference
-> blkcg_css_offline
-> blkcg_destroy_blkgs
-> blkg_destroy
-> bfq_pd_offline
-> __bfq_deactivate_entity
if (!entity->on_st_or_in_serv) /* true, because (Note2)
return false;
-> bfq_pd_free
-> bfqg_put() (-1, byt bfqg->ref == 2) because of (Note2)
So bfq_group and blkcg_gq will leak forever, see test-case below.
##TESTCASE_BEGIN:
#!/bin/bash
max_iters=${1:-100}
#prep cgroup mounts
mount -t tmpfs cgroup_root /sys/fs/cgroup
mkdir /sys/fs/cgroup/blkio
mount -t cgroup -o blkio none /sys/fs/cgroup/blkio
# Prepare blkdev
grep blkio /proc/cgroups
truncate -s 1M img
losetup /dev/loop0 img
echo bfq > /sys/block/loop0/queue/scheduler
grep blkio /proc/cgroups
for ((i=0;i<max_iters;i++))
do
mkdir -p /sys/fs/cgroup/blkio/a
echo 0 > /sys/fs/cgroup/blkio/a/cgroup.procs
dd if=/dev/loop0 bs=4k count=1 of=/dev/null iflag=direct 2> /dev/null
echo 0 > /sys/fs/cgroup/blkio/cgroup.procs
rmdir /sys/fs/cgroup/blkio/a
grep blkio /proc/cgroups
done
##TESTCASE_END:
Fixes: db37a34c56 ("block, bfq: get a ref to a group when adding it to a service tree")
Tested-by: Oleksandr Natalenko <oleksandr@natalenko.name>
Signed-off-by: Dmitry Monakhov <dmtrmonakhov@yandex-team.ru>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
If we pass in an offset which is larger than PAGE_SIZE, then
page_is_mergeable() thinks it's not mergeable with the previous bio_vec,
leading to a large number of bio_vecs being used. Use a slightly more
obvious test that the two pages are compatible with each other.
Fixes: 52d52d1c98 ("block: only allow contiguous page structs in a bio_vec")
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
1f23816b8e ("virtio_blk: add discard and write zeroes support") starts
to support multi-range discard for virtio-blk. However, the virtio-blk
disk may report max discard segment as 1, at least that is exactly what
qemu is doing.
So far, block layer switches to normal request merge if max discard segment
limit is 1, and multiple bios can be merged to single segment. This way may
cause memory corruption in virtblk_setup_discard_write_zeroes().
Fix the issue by handling single max discard segment in straightforward
way.
Fixes: 1f23816b8e ("virtio_blk: add discard and write zeroes support")
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Cc: Changpeng Liu <changpeng.liu@intel.com>
Cc: Daniel Verkamp <dverkamp@chromium.org>
Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Stefan Hajnoczi <stefanha@redhat.com>
Cc: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
When queue_max_discard_segments(q) is 1, blk_discard_mergable() will
return false for discard request, then normal request merge is applied.
However, only queue_max_segments() is checked, so max discard segment
limit isn't respected.
Check max discard segment limit in the request merge code for fixing
the issue.
Discard request failure of virtio_blk is fixed.
Fixes: 6984046608 ("block: fix the DISCARD request merge")
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Cc: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Tejun Heo