linprocfs: Properly reset error variable for mtab generation

[michaelo]

Summary

Both functions linprocfs_domtab() and linprocfs_doprocmountinfo() are
logically identical, but the former fails with ECANCELED because error
is not reset after the for loop.

[michaelo]

Please add others you think see fit.

In a jail first before patch, then after patch:

root@deblndw013x1j:/compat/linux/proc
# cat mounts
cat: mounts: Operation canceled

root@deblndw013x1j:/compat/linux/proc
# cat mounts
hp-large/bastille/jails/deblndw013x1j/root / zfs rw,noatime 0 0
devfs /dev devfs rw 0 0
tmpfs /dev/shm tmpfs rw 0 0
fdescfs /dev/fd fdescfs rw 0 0
proc /proc proc rw 0 0
/sys /sys sysfs rw 0 0
devfs /dev devfs rw 0 0
fdescfs /dev/fd fdescfs rw 0 0

Another jail on another host:

root@deblndw011x1j:/compat/linux/proc
# cat mounts
cat: mounts: Operation canceled

[michaelo]

Jailhost before patch:

root@deblndw013x:/compat/linux/proc/self
# cat mounts
cat: mounts: Operation canceled

Jailhost after patch:

root@deblndw013x:/compat/linux/proc/self
# cat mounts
zroot/ROOT/default / zfs rw,noatime 0 0
devfs /dev devfs rw 0 0
fdescfs /dev/fd fdescfs rw 0 0
hp-large/osipovmi /var/osipovmi zfs rw,noatime 0 0
zroot/var/mail /var/mail zfs rw 0 0
zroot/usr/src /usr/src zfs rw,noatime 0 0
zroot/compat /compat zfs rw,noatime 0 0
zroot/var/crash /var/crash zfs rw,noexec,nosuid,noatime 0 0
zroot/usr/local /usr/local zfs rw,noatime 0 0
zroot/usr/home /usr/home zfs rw,noatime 0 0
zroot/var/audit /var/audit zfs rw,noexec,nosuid,noatime 0 0
zroot /zroot zfs rw,noatime 0 0
zroot/var/tmp /var/tmp zfs rw,nosuid,noatime 0 0
zroot/var/log /var/log zfs rw,noexec,nosuid,noatime 0 0
hp-large/bhyve /usr/local/bhyve zfs rw,noatime 0 0
zroot/tmp /tmp zfs rw,nosuid,noatime 0 0
hp-large/bastille /usr/local/bastille zfs rw,noatime 0 0
zroot/usr/ports /usr/ports zfs rw,nosuid,noatime 0 0
zroot/usr/obj /usr/obj zfs rw,noatime 0 0
hp-large/bastille/logs /var/log/bastille zfs rw,noatime 0 0
hp-large/bastille/backups /usr/local/bastille/backups zfs rw,noatime 0 0
hp-large/bastille/cache /usr/local/bastille/cache zfs rw,noatime 0 0
hp-large/bastille/templates /usr/local/bastille/templates zfs rw,noatime 0 0
hp-large/bastille/jails /usr/local/bastille/jails zfs rw,noatime 0 0
hp-large/bastille/releases /usr/local/bastille/releases zfs rw,noatime 0 0
hp-large/bastille/cache/13.3-RELEASE /usr/local/bastille/cache/13.3-RELEASE zfs rw,noatime 0 0
hp-large/bastille/cache/13.4-RELEASE /usr/local/bastille/cache/13.4-RELEASE zfs rw,noatime 0 0
hp-large/bastille/releases/13.4-RELEASE /usr/local/bastille/releases/13.4-RELEASE zfs rw,noatime 0 0
hp-large/bastille/jails/deblndw013x2j /usr/local/bastille/jails/deblndw013x2j zfs rw,noatime 0 0
hp-large/bastille/releases/13.3-RELEASE /usr/local/bastille/releases/13.3-RELEASE zfs rw,noatime 0 0
hp-large/bastille/jails/deblndw013x2j/root /usr/local/bastille/jails/deblndw013x2j/root zfs rw,noatime 0 0
hp-large/bastille/jails/deblndw013x1j /usr/local/bastille/jails/deblndw013x1j zfs rw,noatime 0 0
hp-large/bastille/jails/deblndw013x1j/root /usr/local/bastille/jails/deblndw013x1j/root zfs rw,noatime 0 0
proc /proc proc rw 0 0
/sys /sys sysfs rw 0 0
devfs /dev devfs rw 0 0
fdescfs /dev/fd fdescfs rw 0 0
tmpfs /dev/shm tmpfs rw 0 0
-hosts /net autofs rw 0 0
automounts /net autofs rw 0 0
devfs /usr/local/bastille/jails/deblndw013x1j/root/compat/linux/dev devfs rw 0 0
tmpfs /usr/local/bastille/jails/deblndw013x1j/root/compat/linux/dev/shm tmpfs rw 0 0
fdescfs /usr/local/bastille/jails/deblndw013x1j/root/compat/linux/dev/fd fdescfs rw 0 0
proc /usr/local/bastille/jails/deblndw013x1j/root/compat/linux/proc proc rw 0 0
/sys /usr/local/bastille/jails/deblndw013x1j/root/compat/linux/sys sysfs rw 0 0
devfs /usr/local/bastille/jails/deblndw013x1j/root/dev devfs rw 0 0
fdescfs /usr/local/bastille/jails/deblndw013x1j/root/dev/fd fdescfs rw 0 0
devfs /usr/local/bastille/jails/deblndw013x2j/root/dev devfs rw 0 0
fdescfs /usr/local/bastille/jails/deblndw013x2j/root/dev/fd fdescfs rw 0 0

[michaelo]

Anyone willing to review?

[fluffy]

LGTM

[michaelo]

Is @fluffy's review enough to merge this? Does @fluffy have src commit bit?

[jrm]

You can see source committers here: https://cgit.freebsd.org/src/tree/access?h=internal/admin

[jrm]

Your change seems simple enough. If you're confident, go for it. If you want review from a src committer, you'll have to pester someone.

[michaelo]

In D47865#1096992, @jrm wrote:

Your change seems simple enough. If you're confident, go for it. If you want review from a src committer, you'll have to pester someone.

Quite, yes. Since now it always fails for me on both UFS and ZFS. Will then push.