Commit be14eb61 authored Sep 10, 2010 by Brian King Committed by Jens Axboe Sep 10, 2010

block: Range check cpu in blk_cpu_to_group



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>

parent edce6820

block/blk.h

+6 −2

Original line number	Diff line number	Diff line
		@@ -142,14 +142,18 @@ static inline int queue_congestion_off_threshold(struct request_queue *q)

		static inline int blk_cpu_to_group(int cpu)
		{
		int group = NR_CPUS;
		#ifdef CONFIG_SCHED_MC
		const struct cpumask *mask = cpu_coregroup_mask(cpu);
		return cpumask_first(mask);
		group = cpumask_first(mask);
		#elif defined(CONFIG_SCHED_SMT)
		return cpumask_first(topology_thread_cpumask(cpu));
		group = cpumask_first(topology_thread_cpumask(cpu));
		#else
		return cpu;
		#endif
		if (likely(group < NR_CPUS))
		return group;
		return cpu;
		}

		/*

Admin message