- Dec 20, 2021
-
-
Jens Axboe authored
Don't combine the task exiting and "already have io_context" case, we need to just abort if the task is marked as dead. Return -ESRCH, which is the documented value for ioprio_set() if the specified task could not be found. Reported-by:
Dan Carpenter <dan.carpenter@oracle.com> Reported-by:
<syzbot+8836466a79f4175961b0@syzkaller.appspotmail.com> Fixes: 5fc11eeb ("block: open code create_task_io_context in set_task_ioprio") Signed-off-by:
Jens Axboe <axboe@kernel.dk>
-
Keith Busch authored
The low level drivers don't expect to see new requests after a successful quiesce completes. Check the queue quiesce state within the rcu protected area prior to calling the driver's queue_rqs(). Fixes: 3c67d44d ("block: add mq_ops->queue_rqs hook") Signed-off-by:
Keith Busch <kbusch@kernel.org> Link: https://lore.kernel.org/r/20211220205919.180191-1-kbusch@kernel.org Signed-off-by:
-
Wander Lairson Costa authored
The running_trace_lock protects running_trace_list and is acquired within the tracepoint which implies disabled preemption. The spinlock_t typed lock can not be acquired with disabled preemption on PREEMPT_RT because it becomes a sleeping lock. The runtime of the tracepoint depends on the number of entries in running_trace_list and has no limit. The blk-tracer is considered debug code and higher latencies here are okay. Make running_trace_lock a raw_spinlock_t. Signed-off-by:
Wander Lairson Costa <wander@redhat.com> Link: https://lore.kernel.org/r/20211220192827.38297-1-wander@redhat.com Signed-off-by:
-
- Dec 16, 2021
-
-
Christoph Hellwig authored
Only bfq needs to code to track icq, so make it conditional. Signed-off-by:
Christoph Hellwig <hch@lst.de> Reviewed-by:
Jan Kara <jack@suse.cz> Link: https://lore.kernel.org/r/20211209063131.18537-12-hch@lst.de Signed-off-by:
-
Christoph Hellwig authored
Fold create_task_io_context into the only remaining caller. Signed-off-by:
-
Christoph Hellwig authored
The flow in set_task_ioprio can be simplified by simply open coding create_task_io_context, which removes a refcount roundtrip on the I/O context. Signed-off-by:
-
Christoph Hellwig authored
Fold get_task_io_context into its only caller, and simplify the code as no reference to the I/O context is required to just set the ioprio field. Signed-off-by:
-
Christoph Hellwig authored
Keep set_task_ioprio with the other low-level code that accesses the io_context structure. Signed-off-by:
-
Christoph Hellwig authored
Fold __ioc_clear_queue into ioc_clear_queue and switch to always use plain _irq locking instead of the more expensive _irqsave that is not needed here. Signed-off-by:
-
Christoph Hellwig authored
Move the code to delay freeing the icqs into a separate helper. Signed-off-by:
-
Christoph Hellwig authored
No caller passes in a NULL pointer, so remove the check. Signed-off-by:
-
Christoph Hellwig authored
Factor out a ioc_exit_icqs helper to tear down the icqs and the fold the rest of put_iocontext_active into exit_io_context. Signed-off-by:
-
Christoph Hellwig authored
Don't hold a reference to ->refcount for each active reference, but just one for all active references. Signed-off-by:
-
Christoph Hellwig authored
Nothing ever looks at ->nr_tasks, so remove it. Signed-off-by:
-
Jens Axboe authored
This enables the block layer to send us a full plug list of requests that need submitting. The block layer guarantees that they all belong to the same queue, but we do have to check the hardware queue mapping for each request. If errors are encountered, leave them in the passed in list. Then the block layer will handle them individually. This is good for about a 4% improvement in peak performance, taking us from 9.6M to 10M IOPS/core. Reviewed-by:
Hannes Reinecke <hare@suse.de> Reviewed-by:
-
Jens Axboe authored
Add a nvme_prep_rq() helper to setup a command, and nvme_queue_rq() is adapted to use this helper. Reviewed-by:
-
Jens Axboe authored
We'll need it for batched submit as well. Since we now have a copy helper, get rid of the nvme_submit_cmd() wrapper. Reviewed-by:
Chaitanya Kulkarni <kch@nvidia.com> Reviewed-by:
Max Gurtovoy <mgurtovoy@nvidia.com> Reviewed-by:
-
Jens Axboe authored
If we have a list of requests in our plug list, send it to the driver in one go, if possible. The driver must set mq_ops->queue_rqs() to support this, if not the usual one-by-one path is used. Reviewed-by:
-
Jens Axboe authored
Pointless to maintain a head/tail for the list, as we never need to access the tail. Entries are always LIFO for cache hotness reasons. Reviewed-by:
-
Jens Axboe authored
The batched completions only deal with non-partial requests anyway, and it doesn't deal with any requests that have errors. Add a completion handler that assumes it's a full request and that it's all being ended successfully. Reviewed-by:
-
- Dec 14, 2021
-
-
Jens Axboe authored
kyber turns on IO statistics when it is loaded on a queue, which means that even if kyber is then later unloaded, we're still stuck with stats enabled on the queue. Change the account enabled from a bool to an int, and pair the enable call with the equivalent disable call. This ensures that stats gets turned off again appropriately. Reviewed-by:
Omar Sandoval <osandov@fb.com> Signed-off-by:
-
- Dec 13, 2021
-
-
Matthew Wilcox (Oracle) authored
Add a Context section and rewrite the rest to be clearer. Signed-off-by:
Matthew Wilcox (Oracle) <willy@infradead.org> Reviewed-by:
Nikolay Borisov <nborisov@suse.com> Link: https://lore.kernel.org/r/20211213171113.3097631-1-willy@infradead.org Signed-off-by:
-
- Dec 12, 2021
-
-
Christoph Hellwig authored
mtdblock / mtdblock_ro set part_bits to 0 and thus nevever scanned partitions. Restore that behavior by setting the GENHD_FL_NO_PART flag. Fixes: 1ebe2e5f ("block: remove GENHD_FL_EXT_DEVT") Reported-by:
Geert Uytterhoeven <geert+renesas@glider.be> Signed-off-by:
-
- Dec 06, 2021
-
-
John Garry authored
Kashyap reports high CPU usage in blk_mq_queue_tag_busy_iter() and callees using megaraid SAS RAID card since moving to shared tags [0]. Previously, when shared tags was shared sbitmap, this function was less than optimum since we would iter through all tags for all hctx's, yet only ever match upto tagset depth number of rqs. Since the change to shared tags, things are even less efficient if we have parallel callers of blk_mq_queue_tag_busy_iter(). This is because in bt_iter() -> blk_mq_find_and_get_req() there would be more contention on accessing each request ref and tags->lock since they are now shared among all HW queues. Optimise by having separate calls to bt_for_each() for when we're using shared tags. In this case no longer pass a hctx, as it is no longer relevant, and teach bt_iter() about this. Ming suggested something along the lines of this change, apart from a different implementation. [0] https://lore.kernel.org/linux-block/e4e92abbe9d52bcba6b8cc6c91c442cc@mail.gmail.com/ Signed-off-by:
John Garry <john.garry@huawei.com> Reviewed-by:
Ming Lei <ming.lei@redhat.com> Reported-and-tested-by:
Kashyap Desai <kashyap.desai@broadcom.com> Fixes: e155b0c2 ("blk-mq: Use shared tags for shared sbitmap support") Link: https://lore.kernel.org/r/1638794990-137490-4-git-send-email-john.garry@huawei.com Signed-off-by:
-
John Garry authored
Typedefs busy_iter_fn and busy_tag_iter_fn are now identical, so delete busy_iter_fn to reduce duplication. It would be nicer to delete busy_tag_iter_fn, as the name busy_iter_fn is less specific. However busy_tag_iter_fn is used in many different parts of the tree, unlike busy_iter_fn which is just use in block/, so just take the straightforward path now, so that we could rename later treewide. Signed-off-by:
-
John Garry authored
The only user of blk_mq_hw_ctx blk_mq_hw_ctx argument is blk_mq_rq_inflight(). Function blk_mq_rq_inflight() uses the hctx to find the associated request queue to match against the request. However this same check is already done in caller bt_iter(), so drop this check. With that change there are no more users of busy_iter_fn blk_mq_hw_ctx argument, so drop the argument. Reviewed-by Hannes Reinecke <hare@suse.de> Signed-off-by:
-
Ming Lei authored
blk_mq_run_dispatch_ops() is defined as one macro, and plug->mq_list will be changed when running 'dispatch_ops', so add one local variable for holding request queue. Reported-and-tested-by:
Yi Zhang <yi.zhang@redhat.com> Fixes: 4cafe86c ("blk-mq: run dispatch lock once in case of issuing from list") Signed-off-by:
-
Ming Lei authored
The operation protected via blk_mq_run_dispatch_ops() in blk_mq_run_hw_queue won't sleep, so don't run might_sleep() for it. Reported-and-tested-by:
Marek Szyprowski <m.szyprowski@samsung.com> Signed-off-by:
-
- Dec 03, 2021
-
-
Ming Lei authored
It isn't necessary to call blk_mq_run_dispatch_ops() once for issuing single request directly, and enough to do it one time when issuing from whole list. Signed-off-by:
-
Ming Lei authored
We have switched to allocate srcu into request queue, so it is fine to pass request queue to blk_mq_run_dispatch_ops(). Signed-off-by:
-
Ming Lei authored
In case of BLK_MQ_F_BLOCKING, per-hctx srcu is used to protect dispatch critical area. However, this srcu instance stays at the end of hctx, and it often takes standalone cacheline, often cold. Inside srcu_read_lock() and srcu_read_unlock(), WRITE is always done on the indirect percpu variable which is allocated from heap instead of being embedded, srcu->srcu_idx is read only in srcu_read_lock(). It doesn't matter if srcu structure stays in hctx or request queue. So switch to per-request-queue srcu for protecting dispatch, and this way simplifies quiesce a lot, not mention quiesce is always done on the request queue wide. Signed-off-by:
-
Ming Lei authored
Remove hctx_lock and hctx_unlock, and add one helper of blk_mq_run_dispatch_ops() to run code block defined in dispatch_ops with rcu/srcu read held. Compared with hctx_lock()/hctx_unlock(): 1) remove 2 branch to 1, so we just need to check (hctx->flags & BLK_MQ_F_BLOCKING) once when running one dispatch_ops 2) srcu_idx needn't to be touched in case of non-blocking 3) might_sleep_if() can be moved to the blocking branch Also put the added blk_mq_run_dispatch_ops() in private header, so that the following patch can use it out of blk-mq.c. Signed-off-by:
-
Jens Axboe authored
refcount_t is not as expensive as it used to be, but it's still more expensive than the io_uring method of using atomic_t and just checking for potential over/underflow. This borrows that same implementation, which in turn is based on the mm implementation from Linus. Reviewed-by:
-
Jens Axboe authored
Don't call into generic_file_read_iter() if we know it's O_DIRECT, just set it up ourselves and call our own handler. This avoids an indirect call for O_DIRECT. Fall back to filemap_read() if we fail. Signed-off-by: -
Jens Axboe authored
No functional changes in this patch, just in preparation for efficiently calling this light function from the block O_DIRECT handling. Reviewed-by:
-
- Dec 02, 2021
-
-
Jens Axboe authored
When we attempt to merge off the cached request path, we return NULL if successful. This makes the caller believe that it's should allocate a new request, and hence we end up with the bio both merged and associated with a new request. This, predictably, leads to all sorts of crashes. Pass in a pointer to the bio pointer, and clear it for the merge case. Then the caller knows that the bio is already queued, and no new requests need to get allocated. Fixes: 5b13bc8a ("blk-mq: cleanup request allocation") Reviewed-by:
-
Jens Axboe authored
Expected case is returning a request, just check for success and return the request rather than having an error label. Signed-off-by: -
Ming Lei authored
Without checking q->poll_stat in queue_poll_stat_show(), kernel panic may be caused if q->poll_stat isn't allocated. Fixes: 48b5c1fb ("block: only allocate poll_stats if there's a user of them") Signed-off-by:
-
- Nov 29, 2021
-
-
Ye Bin authored
We do test with inject error fault base on v4.19, after test some time we found sync /dev/sda always failed. [root@localhost] sync /dev/sda sync: error syncing '/dev/sda': Input/output error scsi log as follows: [19069.812296] sd 0:0:0:0: [sda] tag#64 Send: scmd 0x00000000d03a0b6b [19069.812302] sd 0:0:0:0: [sda] tag#64 CDB: Synchronize Cache(10) 35 00 00 00 00 00 00 00 00 00 [19069.812533] sd 0:0:0:0: [sda] tag#64 Done: SUCCESS Result: hostbyte=DID_OK driverbyte=DRIVER_OK [19069.812536] sd 0:0:0:0: [sda] tag#64 CDB: Synchronize Cache(10) 35 00 00 00 00 00 00 00 00 00 [19069.812539] sd 0:0:0:0: [sda] tag#64 scsi host busy 1 failed 0 [19069.812542] sd 0:0:0:0: Notifying upper driver of completion (result 0) [19069.812546] sd 0:0:0:0: [sda] tag#64 sd_done: completed 0 of 0 bytes [19069.812549] sd 0:0:0:0: [sda] tag#64 0 sectors total, 0 bytes done. [19069.812564] print_req_error: I/O error, dev sda, sector 0 ftrace log as follows: rep-306069 [007] .... 19654.923315: block_bio_queue: 8,0 FWS 0 + 0 [rep] rep-306069 [007] .... 19654.923333: block_getrq: 8,0 FWS 0 + 0 [rep] kworker/7:1H-250 [007] .... 19654.923352: block_rq_issue: 8,0 FF 0 () 0 + 0 [kworker/7:1H] <idle>-0 [007] ..s. 19654.923562: block_rq_complete: 8,0 FF () 18446744073709551615 + 0 [0] <idle>-0 [007] d.s. 19654.923576: block_rq_complete: 8,0 WS () 0 + 0 [-5] As 8d699663 introduce 'fq->rq_status', this data only update when 'flush_rq' reference count isn't zero. If flush request once failed and record error code in 'fq->rq_status'. If there is no chance to update 'fq->rq_status',then do fsync will always failed. To address this issue reset 'fq->rq_status' after return error code to upper layer. Fixes: 8d699663 ("block: fix null pointer dereference in blk_mq_rq_timed_out()") Signed-off-by:
Ye Bin <yebin10@huawei.com> Reviewed-by:
-
Christoph Hellwig authored
Now that blk_execute_rq does not take a gendisk argument there is no need to pass it through the scsi_ioctl callchain either. Signed-off-by:
Martin K. Petersen <martin.petersen@oracle.com> Link: https://lore.kernel.org/r/20211126121802.2090656-6-hch@lst.de Signed-off-by:
-
