[Bug 15906] serious performance regression in "umount" on ext4 over LVM

[bugzilla-daemon@bugzilla.kernel.org]

https://bugzilla.kernel.org/show_bug.cgi?id=15906





--- Comment #5 from Kees Cook <kees@outflux.net>  2010-05-04 21:11:03 ---
Hm, this was really a reduced test-case, as my initial situations where I
noticed this was umounting LVM snapshots, then later umounting aufs unions (in
an effort to get away from 3x slow-down of snapshot IO).

Doing this is just as slow as loopback for me:

VG=systemvg
lvcreate -n testlv -L1G $VG
mkfs.ext4 -F /dev/$VG/testlv
mkdir -p /mnt/test
mount -o loop /dev/$VG/testlv /mnt/test
seq 65536 | (cd /mnt/test; xargs touch)
time umount /mnt/test


Should 68db1961bbf4e16c220ccec4a780e966bc1fece3 be expected to fix it too?

-- 
Configure bugmail: https://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching the assignee of the bug.