Revision 7ce9d5d1f3c8736511daa413c64985a05b2feee3 authored by Eric Sandeen on 04 March 2009, 23:38:18 UTC, committed by Theodore Ts'o on 04 March 2009, 23:38:18 UTC
I was seeing fsck errors on inode bitmaps after a 4 thread
dbench run on a 4 cpu machine:

Inode bitmap differences: -50736 -(50752--50753) etc...

I believe that this is because ext4_free_inode() uses atomic
bitops, and although ext4_new_inode() *used* to also use atomic 
bitops for synchronization, commit 
393418676a7602e1d7d3f6e560159c65c8cbd50e changed this to use
the sb_bgl_lock, so that we could also synchronize against
read_inode_bitmap and initialization of uninit inode tables.

However, that change left ext4_free_inode using atomic bitops,
which I think leaves no synchronization between setting & 
unsetting bits in the inode table.

The below patch fixes it for me, although I wonder if we're 
getting at all heavy-handed with this spinlock...

Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
1 parent fec6c6f
History
File Mode Size
i1480
wlp
Kconfig -rw-r--r-- 3.1 KB
Makefile -rw-r--r-- 479 bytes
address.c -rw-r--r-- 10.2 KB
allocator.c -rw-r--r-- 9.9 KB
beacon.c -rw-r--r-- 16.2 KB
driver.c -rw-r--r-- 3.7 KB
drp-avail.c -rw-r--r-- 8.6 KB
drp-ie.c -rw-r--r-- 9.5 KB
drp.c -rw-r--r-- 24.1 KB
est.c -rw-r--r-- 13.4 KB
hwa-rc.c -rw-r--r-- 27.7 KB
ie-rcv.c -rw-r--r-- 1.6 KB
ie.c -rw-r--r-- 9.6 KB
lc-dev.c -rw-r--r-- 12.7 KB
lc-rc.c -rw-r--r-- 10.8 KB
neh.c -rw-r--r-- 17.5 KB
pal.c -rw-r--r-- 2.3 KB
radio.c -rw-r--r-- 4.5 KB
reset.c -rw-r--r-- 10.9 KB
rsv.c -rw-r--r-- 26.9 KB
scan.c -rw-r--r-- 4.0 KB
umc-bus.c -rw-r--r-- 5.4 KB
umc-dev.c -rw-r--r-- 2.3 KB
umc-drv.c -rw-r--r-- 819 bytes
uwb-debug.c -rw-r--r-- 9.5 KB
uwb-internal.h -rw-r--r-- 11.1 KB
uwbd.c -rw-r--r-- 10.2 KB
whc-rc.c -rw-r--r-- 13.2 KB
whci.c -rw-r--r-- 6.4 KB

back to top