tmpfs leak when compress-and-save fails

[Robert Collins]

I have compress-and-save failing for some reason, but I notice that after the failure:

Filesystem Size Used Avail Use% Mounted on
/dev/mapper/lifelesshp-root 168G 155G 12G 94% /
none 4.0K 0 4.0K 0% /sys/fs/cgroup
udev 7.7G 12K 7.7G 1% /dev
tmpfs 1.6G 572K 1.6G 1% /run
none 5.0M 0 5.0M 0% /run/lock
none 7.8G 264K 7.8G 1% /run/shm
none 100M 24K 100M 1% /run/user
/dev/mapper/lifelesshp-root 168G 155G 12G 94% /home
/dev/sda1 228M 47M 170M 22% /boot
tmpfs 7.8G 2.5G 5.3G 32% /tmp/image.XAiRCgui

output from the failure to the end:

+ compress_and_save_image /home/robertc/.cache/image-create/image-file-cache/bc070c863ac465a7e335c9e9449448a29d0bf511.qcow2
+ test qcow2 '!=' qcow2
+ qemu-img convert -f raw /tmp/image.XAiRCgui/image.raw -O qcow2 /home/robertc/.cache/image-create/image-file-cache/bc070c863ac465a7e335c9e9449448a29d0bf511.qcow2-new
qemu-img: error while writing sector 42401792: No space left on device
+ cleanup
+ unmount_image
+ sync
+ unmount_dir /tmp/image.XAiRCgui/mnt
+ local pattern=/tmp/image.XAiRCgui/mnt mnts=
+ '[' -n /tmp/image.XAiRCgui/mnt ']'
++ awk '{print $2}'
++ grep '^/tmp/image.XAiRCgui/mnt'
++ sort -r
+ mnts=
+ '[' -n '' ']'
+ '[' -n 'detach_loopback /dev/loop0' ']'
+ detach_loopback /dev/loop0
+ local loopdev=/dev/loop0
++ echo /dev/loop0
++ sed 's/\/dev\///g'
+ loopdev_name=loop0
++ grep loop0
++ awk '{ print $1 }'
++ sudo dmsetup ls
+ mapper_name=
+ '[' '' ']'
++ seq 10 -1 1
+ for try in '$(seq 10 -1 1)'
+ sudo losetup -d /dev/loop0
loop: can't delete device /dev/loop0: No such device or address
+ echo /dev/loop0 may be busy, sleeping up to 10 more seconds...
/dev/loop0 may be busy, sleeping up to 10 more seconds...
+ sleep 1
+ for try in '$(seq 10 -1 1)'
+ sudo losetup -d /dev/loop0
loop: can't delete device /dev/loop0: No such device or address
+ echo /dev/loop0 may be busy, sleeping up to 9 more seconds...
/dev/loop0 may be busy, sleeping up to 9 more seconds...
+ sleep 1
+ for try in '$(seq 10 -1 1)'
+ sudo losetup -d /dev/loop0
loop: can't delete device /dev/loop0: No such device or address
+ echo /dev/loop0 may be busy, sleeping up to 8 more seconds...
/dev/loop0 may be busy, sleeping up to 8 more seconds...
+ sleep 1
+ for try in '$(seq 10 -1 1)'
+ sudo losetup -d /dev/loop0
loop: can't delete device /dev/loop0: No such device or address
+ echo /dev/loop0 may be busy, sleeping up to 7 more seconds...
/dev/loop0 may be busy, sleeping up to 7 more seconds...
+ sleep 1
+ for try in '$(seq 10 -1 1)'
+ sudo losetup -d /dev/loop0
loop: can't delete device /dev/loop0: No such device or address
+ echo /dev/loop0 may be busy, sleeping up to 6 more seconds...
/dev/loop0 may be busy, sleeping up to 6 more seconds...
+ sleep 1
+ for try in '$(seq 10 -1 1)'
+ sudo losetup -d /dev/loop0
loop: can't delete device /dev/loop0: No such device or address
+ echo /dev/loop0 may be busy, sleeping up to 5 more seconds...
/dev/loop0 may be busy, sleeping up to 5 more seconds...
+ sleep 1
+ for try in '$(seq 10 -1 1)'
+ sudo losetup -d /dev/loop0
loop: can't delete device /dev/loop0: No such device or address
+ echo /dev/loop0 may be busy, sleeping up to 4 more seconds...
/dev/loop0 may be busy, sleeping up to 4 more seconds...
+ sleep 1
+ for try in '$(seq 10 -1 1)'
+ sudo losetup -d /dev/loop0
loop: can't delete device /dev/loop0: No such device or address
+ echo /dev/loop0 may be busy, sleeping up to 3 more seconds...
/dev/loop0 may be busy, sleeping up to 3 more seconds...
+ sleep 1
+ for try in '$(seq 10 -1 1)'
+ sudo losetup -d /dev/loop0
loop: can't delete device /dev/loop0: No such device or address
+ echo /dev/loop0 may be busy, sleeping up to 2 more seconds...
/dev/loop0 may be busy, sleeping up to 2 more seconds...
+ sleep 1
+ for try in '$(seq 10 -1 1)'
+ sudo losetup -d /dev/loop0
loop: can't delete device /dev/loop0: No such device or address
+ echo /dev/loop0 may be busy, sleeping up to 1 more seconds...
/dev/loop0 may be busy, sleeping up to 1 more seconds...
+ sleep 1
+ echo Gave up trying to detach /dev/loop0
Gave up trying to detach /dev/loop0
+ return 1
cp: cannot stat ‘/home/robertc/.cache/image-create/image-file-cache/bc070c863ac465a7e335c9e9449448a29d0bf511.qcow2’: No such file or directory

[Ghe Rivero (ghe.rivero)]

Same origin problem as https://bugs.launchpad.net/diskimage-builder/+bug/1378033
I was able to reproduce it and confirm that is not longer a problem.
Solved by commit https://git.openstack.org/cgit/openstack/diskimage-builder/commit/?id=c7811157845fba6b31635815cfcf9689a5b32680