Messages in this thread | | | Date | Sat, 21 Dec 2013 20:56:28 -0700 | From | Jens Axboe <> | Subject | Re: [HANG 3.13-rc3] blk-mq/virtio: mkfs.ext4 hangs in blk_mq_wait_for_tags |
| |
On Fri, Dec 13 2013, Dave Chinner wrote: > On Fri, Dec 13, 2013 at 09:57:48AM +0800, Ming Lei wrote: > > On Tue, Dec 10, 2013 at 6:39 AM, Dave Chinner <david@fromorbit.com> wrote: > > > Hi Jens, > > > > > > Another day, another blkmq/virtio problem. Running mkfs.ext4 on a > > > sparse 100TB VM file image, it hangs hard while writing superblock > > > information: > > > > > > $ tests/fsmark-50-test-ext4.sh > > > mke2fs 1.43-WIP (20-Jun-2013) > > > Filesystem label= > > > OS type: Linux > > > Block size=4096 (log=2) > > > Fragment size=4096 (log=2) > > > Stride=0 blocks, Stripe width=0 blocks > > > 1677721600 inodes, 26843545600 blocks > > > 1342177280 blocks (5.00%) reserved for the super user > > > First data block=0 > > > 819200 block groups > > > 32768 blocks per group, 32768 fragments per group > > > 2048 inodes per group > > > Superblock backups stored on blocks: > > > 32768, 98304, 163840, 229376, 294912, 819200, 884736, 1605632, 2654208, > > > 4096000, 7962624, 11239424, 20480000, 23887872, 71663616, 78675968, > > > 102400000, 214990848, 512000000, 550731776, 644972544, 1934917632, > > > 2560000000, 3855122432, 5804752896, 12800000000, 17414258688 > > > > > > Allocating group tables: done > > > Writing inode tables: done > > > Creating journal (32768 blocks): done > > > Writing superblocks and filesystem accounting information: > > > > > > It writes a few superblocks, then hangs. Immediately after it stops > > > updating that last line, I see this: > > > > > > root@test4:~# echo w > /proc/sysrq-trigger > > > > It might be helpful to do below and post the result before the sysrq-trigger: > > > > cat /sys/class/block/vda/mq/0/tags > > I would, but for some reason I can't reproduce it now. I'm running a > slightly more recent kernel than a few days ago, and it isn't > hanging now....
Dave, I haven't found anything through testing. Is that the case at your end too?
-- Jens Axboe
| |