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
File | Mode | Size |
---|---|---|
Kconfig | -rw-r--r-- | 521 bytes |
Makefile | -rw-r--r-- | 218 bytes |
crypto.c | -rw-r--r-- | 66.6 KB |
debug.c | -rw-r--r-- | 3.8 KB |
dentry.c | -rw-r--r-- | 2.9 KB |
ecryptfs_kernel.h | -rw-r--r-- | 27.7 KB |
file.c | -rw-r--r-- | 10.4 KB |
inode.c | -rw-r--r-- | 34.5 KB |
keystore.c | -rw-r--r-- | 76.6 KB |
kthread.c | -rw-r--r-- | 5.6 KB |
main.c | -rw-r--r-- | 25.5 KB |
messaging.c | -rw-r--r-- | 17.3 KB |
miscdev.c | -rw-r--r-- | 15.7 KB |
mmap.c | -rw-r--r-- | 16.0 KB |
read_write.c | -rw-r--r-- | 11.4 KB |
super.c | -rw-r--r-- | 6.1 KB |
Computing file changes ...