Revision be14eb619108fa8b7120eb2c42d66d5f623ae10e authored by Brian King on 10 September 2010, 07:03:21 UTC, committed by Jens Axboe on 10 September 2010, 07:03:21 UTC
While testing CPU DLPAR, the following problem was discovered.
We were DLPAR removing the first CPU, which in this case was
logical CPUs 0-3. CPUs 0-2 were already marked offline and
we were in the process of offlining CPU 3. After marking
the CPU inactive and offline in cpu_disable, but before the
cpu was completely idle (cpu_die), we ended up in __make_request
on CPU 3. There we looked at the topology map to see which CPU
to complete the I/O on and found no CPUs in the cpu_sibling_map.
This resulted in the block layer setting the completion cpu
to be NR_CPUS, which then caused an oops when we tried to
complete the I/O.

Fix this by sanity checking the value we return from blk_cpu_to_group
to be a valid cpu value.

Signed-off-by: Brian King <brking@linux.vnet.ibm.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
1 parent edce682
History
File Mode Size
Kconfig -rw-r--r-- 2.1 KB
Makefile -rw-r--r-- 374 bytes
budget.c -rw-r--r-- 23.8 KB
commit.c -rw-r--r-- 18.7 KB
compress.c -rw-r--r-- 6.6 KB
debug.c -rw-r--r-- 72.1 KB
debug.h -rw-r--r-- 17.9 KB
dir.c -rw-r--r-- 33.2 KB
file.c -rw-r--r-- 44.7 KB
find.c -rw-r--r-- 29.8 KB
gc.c -rw-r--r-- 26.7 KB
io.c -rw-r--r-- 26.6 KB
ioctl.c -rw-r--r-- 5.2 KB
journal.c -rw-r--r-- 41.7 KB
key.h -rw-r--r-- 14.1 KB
log.c -rw-r--r-- 20.3 KB
lprops.c -rw-r--r-- 36.2 KB
lpt.c -rw-r--r-- 58.6 KB
lpt_commit.c -rw-r--r-- 50.4 KB
master.c -rw-r--r-- 10.4 KB
misc.h -rw-r--r-- 9.0 KB
orphan.c -rw-r--r-- 24.8 KB
recovery.c -rw-r--r-- 38.5 KB
replay.c -rw-r--r-- 28.0 KB
sb.c -rw-r--r-- 19.0 KB
scan.c -rw-r--r-- 9.4 KB
shrinker.c -rw-r--r-- 9.3 KB
super.c -rw-r--r-- 58.1 KB
tnc.c -rw-r--r-- 87.0 KB
tnc_commit.c -rw-r--r-- 27.1 KB
tnc_misc.c -rw-r--r-- 12.6 KB
ubifs-media.h -rw-r--r-- 22.8 KB
ubifs.h -rw-r--r-- 59.0 KB
xattr.c -rw-r--r-- 15.5 KB

back to top