md/raid456: downlevel multicore operations to raid_run_ops
The percpu conversion allowed a straightforward handoff of stripe
processing to the async subsytem that initially showed some modest gains
(+4%). However, this model is too simplistic and leads to stripes
bouncing between raid5d and the async thread pool for every invocation
of handle_stripe(). As reported by Holger this can fall into a
pathological situation severely impacting throughput (6x performance
loss).
By downleveling the parallelism to raid_run_ops the pathological
stripe_head bouncing is eliminated. This version still exhibits an
average 11% throughput loss for:
mdadm --create /dev/md0 /dev/sd[b-q] -n 16 -l 6
echo 1024 > /sys/block/md0/md/stripe_cache_size
dd if=/dev/zero of=/dev/md0 bs=1024k count=2048
...but the results are at least stable and can be used as a base for
further multicore experimentation.
Reported-by: Holger Kiehl <Holger.Kiehl@dwd.de>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
diff --git a/drivers/md/raid5.h b/drivers/md/raid5.h
index 2390e0e..dcefdc9 100644
--- a/drivers/md/raid5.h
+++ b/drivers/md/raid5.h
@@ -214,12 +214,20 @@
int disks; /* disks in stripe */
enum check_states check_state;
enum reconstruct_states reconstruct_state;
- /* stripe_operations
+ /**
+ * struct stripe_operations
* @target - STRIPE_OP_COMPUTE_BLK target
+ * @target2 - 2nd compute target in the raid6 case
+ * @zero_sum_result - P and Q verification flags
+ * @request - async service request flags for raid_run_ops
*/
struct stripe_operations {
int target, target2;
enum sum_check_flags zero_sum_result;
+ #ifdef CONFIG_MULTICORE_RAID456
+ unsigned long request;
+ wait_queue_head_t wait_for_ops;
+ #endif
} ops;
struct r5dev {
struct bio req;
@@ -294,6 +302,8 @@
#define STRIPE_FULL_WRITE 13 /* all blocks are set to be overwritten */
#define STRIPE_BIOFILL_RUN 14
#define STRIPE_COMPUTE_RUN 15
+#define STRIPE_OPS_REQ_PENDING 16
+
/*
* Operation request flags
*/