Revision c4f675cd40d955d539180506c09515c90169b15b authored by Sergei Trofimovich on 20 May 2011, 20:20:30 UTC, committed by Chris Mason on 23 May 2011, 17:24:14 UTC
Observed as a large delay when --mixed filesystem is filled up.
Test example:
1. create tiny --mixed FS:
   $ dd if=/dev/zero of=2G.img seek=$((2048 * 1024 * 1024 - 1)) count=1 bs=1
   $ mkfs.btrfs --mixed 2G.img
   $ mount -oloop 2G.img /mnt/ut/
2. Try to fill it up:
   $ dd if=/dev/urandom of=10M.file bs=10240 count=1024
   $ seq 1 256 | while read file_no; do echo $file_no; time cp 10M.file ${file_no}.copy; done

Up to '200.copy' it goes fast, but when disk fills-up each -ENOSPC
message takes 3 seconds to pop-up _every_ ENOSPC (and in usermode linux
it's even more: 30-60 seconds!). (Maybe, time depends on kernel's timer resolution).

No IO, no CPU load, just rescheduling. Some debugging revealed busy spinning
in shrink_delalloc.

Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org>
Reviewed-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
1 parent 0f3b708
History
File Mode Size
Kconfig -rw-r--r-- 8.9 KB
Makefile -rw-r--r-- 620 bytes
acorn.c -rw-r--r-- 12.3 KB
acorn.h -rw-r--r-- 488 bytes
amiga.c -rw-r--r-- 3.6 KB
amiga.h -rw-r--r-- 89 bytes
atari.c -rw-r--r-- 3.9 KB
atari.h -rw-r--r-- 1.0 KB
check.c -rw-r--r-- 16.6 KB
check.h -rw-r--r-- 1.0 KB
efi.c -rw-r--r-- 20.4 KB
efi.h -rw-r--r-- 4.3 KB
ibm.c -rw-r--r-- 6.8 KB
ibm.h -rw-r--r-- 47 bytes
karma.c -rw-r--r-- 1.1 KB
karma.h -rw-r--r-- 124 bytes
ldm.c -rw-r--r-- 44.0 KB
ldm.h -rw-r--r-- 5.4 KB
mac.c -rw-r--r-- 3.3 KB
mac.h -rw-r--r-- 1.2 KB
msdos.c -rw-r--r-- 15.2 KB
msdos.h -rw-r--r-- 124 bytes
osf.c -rw-r--r-- 1.9 KB
osf.h -rw-r--r-- 123 bytes
sgi.c -rw-r--r-- 2.2 KB
sgi.h -rw-r--r-- 128 bytes
sun.c -rw-r--r-- 3.7 KB
sun.h -rw-r--r-- 169 bytes
sysv68.c -rw-r--r-- 1.9 KB
sysv68.h -rw-r--r-- 62 bytes
ultrix.c -rw-r--r-- 1.2 KB
ultrix.h -rw-r--r-- 90 bytes

back to top