When two md arrays on the same physical devices want to resync, they
must take turns, and currently the one that chooses to yield may end
up stealing all the CPU (on a uni-processor) and not give the other a
chance to get started.
This patches adds an explicit 'yield()' to avoid the problem.
----------- Diffstat output ------------
./drivers/md/md.c | 4 +++-
1 files changed, 3 insertions(+), 1 deletion(-)
diff ./drivers/md/md.c~current~ ./drivers/md/md.c
--- ./drivers/md/md.c~current~ 2002-12-09 15:47:05.000000000 +1100
+++ ./drivers/md/md.c 2002-12-09 15:47:05.000000000 +1100
@@ -2815,8 +2815,10 @@ static void md_do_sync(void *data)
printk(KERN_INFO "md: delaying resync of md%d until md%d "
"has finished resync (they share one or more physical units)\n",
mdidx(mddev), mdidx(mddev2));
- if (mddev < mddev2) /* arbitrarily yield */
+ if (mddev < mddev2) {/* arbitrarily yield */
mddev->curr_resync = 1;
+ yield();
+ }
if (wait_event_interruptible(resync_wait,
mddev2->curr_resync < 2)) {
flush_curr_signals();
-
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
