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
Makefile -rw-r--r-- 683 bytes
datagram.c -rw-r--r-- 18.1 KB
dev.c -rw-r--r-- 147.3 KB
dev_addr_lists.c -rw-r--r-- 17.3 KB
drop_monitor.c -rw-r--r-- 8.6 KB
dst.c -rw-r--r-- 8.1 KB
ethtool.c -rw-r--r-- 38.3 KB
fib_rules.c -rw-r--r-- 16.5 KB
filter.c -rw-r--r-- 14.7 KB
flow.c -rw-r--r-- 10.2 KB
gen_estimator.c -rw-r--r-- 8.6 KB
gen_stats.c -rw-r--r-- 6.7 KB
iovec.c -rw-r--r-- 5.6 KB
kmap_skb.h -rw-r--r-- 360 bytes
link_watch.c -rw-r--r-- 5.4 KB
neighbour.c -rw-r--r-- 65.5 KB
net-sysfs.c -rw-r--r-- 22.7 KB
net-sysfs.h -rw-r--r-- 190 bytes
net-traces.c -rw-r--r-- 771 bytes
net_namespace.c -rw-r--r-- 13.3 KB
netevent.c -rw-r--r-- 2.0 KB
netpoll.c -rw-r--r-- 21.7 KB
pktgen.c -rw-r--r-- 93.9 KB
request_sock.c -rw-r--r-- 3.5 KB
rtnetlink.c -rw-r--r-- 43.7 KB
scm.c -rw-r--r-- 7.6 KB
skbuff.c -rw-r--r-- 75.2 KB
sock.c -rw-r--r-- 63.4 KB
stream.c -rw-r--r-- 5.1 KB
sysctl_net_core.c -rw-r--r-- 5.2 KB
timestamping.c -rw-r--r-- 2.9 KB
user_dma.c -rw-r--r-- 3.3 KB
utils.c -rw-r--r-- 6.2 KB

back to top