Public bug reported:
With a Xenial kernel running in a VM (32 cores and 65536 MB RAM), the
following script hangs in the dd command:
==================================================================================
#!/bin/bash
dd if=/dev/zero of=in.bin bs=1 seek=6442506751 count=1
cp in.bin out.bin
IN=$(losetup --show -f in.bin)
OUT=$(losetup --show -f out.bin)
typeset -i i=1
while true; do
echo $i $(date)
dd if=$IN of=$OUT bs=128k
i=i+1
done
==================================================================================
The trace (sorry, I don't have the exact kernel version... I found this
mid-June 2019) looks like:
# cat /proc/5230/stack
[<0>] io_schedule+0x16/0x40
[<0>] wbt_wait+0x234/0x390
[<0>] blk_mq_make_request+0x103/0x590
[<0>] generic_make_request+0x122/0x2f0
[<0>] submit_bio+0x73/0x150
[<0>] submit_bh_wbc+0x17a/0x1b0
[<0>] __block_write_full_page+0x15b/0x400
[<0>] block_write_full_page+0x10c/0x130
[<0>] blkdev_writepage+0x18/0x20
[<0>] __writepage+0x1d/0x50
[<0>] write_cache_pages+0x228/0x4b0
[<0>] generic_writepages+0x5c/0x90
[<0>] blkdev_writepages+0x2f/0x40
[<0>] do_writepages+0x1f/0x70
[<0>] __filemap_fdatawrite_range+0xc6/0x100
[<0>] filemap_write_and_wait+0x31/0x90
[<0>] __blkdev_put+0x7a/0x210
[<0>] blkdev_put+0x4c/0xd0
[<0>] blkdev_close+0x34/0x70
[<0>] __fput+0xea/0x220
[<0>] ____fput+0xe/0x10
[<0>] task_work_run+0x9a/0xc0
[<0>] exit_to_usermode_loop+0xce/0xd0
[<0>] do_syscall_64+0x116/0x150
[<0>] entry_SYSCALL_64_after_hwframe+0x42/0xb7
[<0>] 0xffffffffffffffff
If we have used strace on the process, we find that dd is inside
"close(1)" as it closes the output file.
I don't know if this would reproduce on actual hardware, as I don't have
a machine that large to play with.
** Affects: linux (Ubuntu)
Importance: Undecided
Status: New
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/1839216
Title:
xenial VM hangs copying block device
Status in linux package in Ubuntu:
New
Bug description:
With a Xenial kernel running in a VM (32 cores and 65536 MB RAM), the
following script hangs in the dd command:
==================================================================================
#!/bin/bash
dd if=/dev/zero of=in.bin bs=1 seek=6442506751 count=1
cp in.bin out.bin
IN=$(losetup --show -f in.bin)
OUT=$(losetup --show -f out.bin)
typeset -i i=1
while true; do
echo $i $(date)
dd if=$IN of=$OUT bs=128k
i=i+1
done
==================================================================================
The trace (sorry, I don't have the exact kernel version... I found this
mid-June 2019) looks like:
# cat /proc/5230/stack
[<0>] io_schedule+0x16/0x40
[<0>] wbt_wait+0x234/0x390
[<0>] blk_mq_make_request+0x103/0x590
[<0>] generic_make_request+0x122/0x2f0
[<0>] submit_bio+0x73/0x150
[<0>] submit_bh_wbc+0x17a/0x1b0
[<0>] __block_write_full_page+0x15b/0x400
[<0>] block_write_full_page+0x10c/0x130
[<0>] blkdev_writepage+0x18/0x20
[<0>] __writepage+0x1d/0x50
[<0>] write_cache_pages+0x228/0x4b0
[<0>] generic_writepages+0x5c/0x90
[<0>] blkdev_writepages+0x2f/0x40
[<0>] do_writepages+0x1f/0x70
[<0>] __filemap_fdatawrite_range+0xc6/0x100
[<0>] filemap_write_and_wait+0x31/0x90
[<0>] __blkdev_put+0x7a/0x210
[<0>] blkdev_put+0x4c/0xd0
[<0>] blkdev_close+0x34/0x70
[<0>] __fput+0xea/0x220
[<0>] ____fput+0xe/0x10
[<0>] task_work_run+0x9a/0xc0
[<0>] exit_to_usermode_loop+0xce/0xd0
[<0>] do_syscall_64+0x116/0x150
[<0>] entry_SYSCALL_64_after_hwframe+0x42/0xb7
[<0>] 0xffffffffffffffff
If we have used strace on the process, we find that dd is inside
"close(1)" as it closes the output file.
I don't know if this would reproduce on actual hardware, as I don't
have a machine that large to play with.
To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1839216/+subscriptions
--
Mailing list: https://launchpad.net/~kernel-packages
Post to : kernel-packages@lists.launchpad.net
Unsubscribe : https://launchpad.net/~kernel-packages
More help : https://help.launchpad.net/ListHelp
