On Fri, May 19, 2017 at 08:41:13AM -0600, Jens Axboe wrote:

On 05/16/2017 07:27 PM, Ming Lei wrote:

quoted

When one NVMe PCI device is being resetted and found reset failue,
nvme_remove_dead_ctrl() is called to handle the failure: blk-mq hw queues
are put into stopped first, then schedule .remove_work to release the driver.

Unfortunately if the driver is being released via sysfs store
just before the .remove_work is run, del_gendisk() from
nvme_remove() may hang forever because hw queues are stopped and
the submitted writeback IOs from fsync_bdev() can't be completed at all.

This patch fixes the following issue[1][2] by moving nvme_kill_queues()
into nvme_remove_dead_ctrl() to avoid the issue because nvme_remove()
flushs .reset_work, and this way is reasonable and safe because
nvme_dev_disable() has started to suspend queues and canceled requests
already.

[1] test script
	fio -filename=$NVME_DISK -iodepth=1 -thread -rw=randwrite -ioengine=psync \
	    -bssplit=5k/10:9k/10:13k/10:17k/10:21k/10:25k/10:29k/10:33k/10:37k/10:41k/10 \
	    -bs_unaligned -runtime=1200 -size=-group_reporting -name=mytest -numjobs=60

	sleep 35
	echo 1 > $SYSFS_NVME_PCI_PATH/rescan
	echo 1 > $SYSFS_NVME_PCI_PATH/reset
	echo 1 > $SYSFS_NVME_PCI_PATH/remove
	echo 1 > /sys/bus/pci/rescan

The patch looks good to me. But since you have a nice reproducer, how about
turning that into a blktests [1] test case?

This test has triggered several problems recently, and looks a good idea to
integrate into blktests.

I am a little busy recently, and may take a while to start that. If
anyone would like to integrate the case, please go ahead, and I am happy
to provide any details you need.


Thanks,
Ming