Hello,
I would like to get your feedback on the following patch for improving
raid456 discard performance. This seem to improve discard performance
for raid456 quite significantly (just like raid10 discard optimization did).
Unfortunately, this patch is not in stable form right now. I do not have
very good understanding of raid456 code and would really love to get
your guys feedback.
I basically tried to incorporate raid0's optimzed discard code logic
into raid456 make_discard fn, but I am sure I am missing some things
as the raid456 layout is very different than that of raid0 or 10.
Would appreciate the feedback.
This patch improves discard performance with raid456 by sending
discard bio directly to the underlying disk just like how raid0/10
handle discard request. Currently, the discard request for raid456
gets sent to the disks on a per stripe basis which involves lots of
bio split/merge and makes it pretty slow performant vs. sending
the requests directly to the disks. This patch is intended to issue
discard request in the the similar way with patch
29efc390b (md/md0: optimize raid0 discard handling).
Signed-off-by: Vishal Verma <vverma@xxxxxxxxxxxxxxxx>
---
drivers/md/raid5.c | 184 ++++++++++++++++++++++++---------------------
1 file changed, 99 insertions(+), 85 deletions(-)
diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c
index 7c119208a214..2d57cf105471 100644
--- a/drivers/md/raid5.c
+++ b/drivers/md/raid5.c
@@ -5681,93 +5681,108 @@ static void release_stripe_plug(struct mddev
*mddev,
static void make_discard_request(struct mddev *mddev, struct bio *bi)
{
- struct r5conf *conf = mddev->private;
- sector_t logical_sector, last_sector;
- struct stripe_head *sh;
- int stripe_sectors;
-
- /* We need to handle this when io_uring supports discard/trim */
- if (WARN_ON_ONCE(bi->bi_opf & REQ_NOWAIT))
+ struct r5conf *conf = mddev->private;
+ sector_t bio_start, bio_end;
+ unsigned int start_disk_index, end_disk_index;
+ sector_t start_disk_offset, end_disk_offset;
+ sector_t first_stripe_index, last_stripe_index;
+ sector_t split_size;
+ struct bio *split;
+ unsigned int remainder;
+ int d;
+ int stripe_sectors;
+
+ /* We need to handle this when io_uring supports discard/trim */
+ if (WARN_ON_ONCE(bi->bi_opf & REQ_NOWAIT))
+ return;
+
+ if (mddev->reshape_position != MaxSector)
+ /* Skip discard while reshape is happening */
+ return;
+
+ stripe_sectors = conf->chunk_sectors *
+ (conf->raid_disks - conf->max_degraded);
+
+ if (bio_sectors(bi) < stripe_sectors * 2)
return;
- if (mddev->reshape_position != MaxSector)
- /* Skip discard while reshape is happening */
- return;
-
- logical_sector = bi->bi_iter.bi_sector &
~((sector_t)RAID5_STRIPE_SECTORS(conf)-1);
- last_sector = bio_end_sector(bi);
-
- bi->bi_next = NULL;
-
- stripe_sectors = conf->chunk_sectors *
- (conf->raid_disks - conf->max_degraded);
- logical_sector = DIV_ROUND_UP_SECTOR_T(logical_sector,
- stripe_sectors);
- sector_div(last_sector, stripe_sectors);
-
- logical_sector *= conf->chunk_sectors;
- last_sector *= conf->chunk_sectors;
-
- for (; logical_sector < last_sector;
- logical_sector += RAID5_STRIPE_SECTORS(conf)) {
- DEFINE_WAIT(w);
- int d;
- again:
- sh = raid5_get_active_stripe(conf, logical_sector, 0, 0, 0);
- prepare_to_wait(&conf->wait_for_overlap, &w,
- TASK_UNINTERRUPTIBLE);
- set_bit(R5_Overlap, &sh->dev[sh->pd_idx].flags);
- if (test_bit(STRIPE_SYNCING, &sh->state)) {
- raid5_release_stripe(sh);
- schedule();
- goto again;
- }
- clear_bit(R5_Overlap, &sh->dev[sh->pd_idx].flags);
- spin_lock_irq(&sh->stripe_lock);
- for (d = 0; d < conf->raid_disks; d++) {
- if (d == sh->pd_idx || d == sh->qd_idx)
- continue;
- if (sh->dev[d].towrite || sh->dev[d].toread) {
- set_bit(R5_Overlap, &sh->dev[d].flags);
- spin_unlock_irq(&sh->stripe_lock);
- raid5_release_stripe(sh);
- schedule();
- goto again;
- }
- }
- set_bit(STRIPE_DISCARD, &sh->state);
- finish_wait(&conf->wait_for_overlap, &w);
- sh->overwrite_disks = 0;
- for (d = 0; d < conf->raid_disks; d++) {
- if (d == sh->pd_idx || d == sh->qd_idx)
- continue;
- sh->dev[d].towrite = bi;
- set_bit(R5_OVERWRITE, &sh->dev[d].flags);
- bio_inc_remaining(bi);
- md_write_inc(mddev, bi);
- sh->overwrite_disks++;
- }
- spin_unlock_irq(&sh->stripe_lock);
- if (conf->mddev->bitmap) {
- for (d = 0;
- d < conf->raid_disks - conf->max_degraded;
- d++)
- md_bitmap_startwrite(mddev->bitmap,
- sh->sector,
- RAID5_STRIPE_SECTORS(conf),
- 0);
- sh->bm_seq = conf->seq_flush + 1;
- set_bit(STRIPE_BIT_DELAY, &sh->state);
- }
-
- set_bit(STRIPE_HANDLE, &sh->state);
- clear_bit(STRIPE_DELAYED, &sh->state);
- if (!test_and_set_bit(STRIPE_PREREAD_ACTIVE, &sh->state))
- atomic_inc(&conf->preread_active_stripes);
- release_stripe_plug(mddev, sh);
- }
-
- bio_endio(bi);
+ bio_start = bi->bi_iter.bi_sector &
~((sector_t)RAID5_STRIPE_SECTORS(conf)-1);
+ bio_end = bio_end_sector(bi);
+
+ /*
+ * Keep bio aligned with strip size.
+ */
+ div_u64_rem(bio_start, stripe_sectors, &remainder);
+ if (remainder) {
+ split_size = stripe_sectors - remainder;
+ split = bio_split(bi, split_size, GFP_NOIO, &conf->bio_split);
+ bio_chain(split, bi);
+ /* Resend the fist split part */
+ submit_bio_noacct(split);
+ }
+ div_u64_rem(bio_end-bio_start, stripe_sectors, &remainder);
+ if (remainder) {
+ split_size = bio_sectors(bi) - remainder;
+ split = bio_split(bi, split_size, GFP_NOIO, &conf->bio_split);
+ bio_chain(split, bi);
+ /* Resend the second split part */
+ submit_bio_noacct(bi);
+ bi = split;
+ }
+
+ bio_start = bi->bi_iter.bi_sector &
~((sector_t)RAID5_STRIPE_SECTORS(conf)-1);
+ bio_end = bio_end_sector(bi);
+
+ bi->bi_next = NULL;
+
+ first_stripe_index = bio_start;
+ sector_div(first_stripe_index, stripe_sectors);
+
+ last_stripe_index = bio_end;
+ sector_div(last_stripe_index, stripe_sectors);
+
+ start_disk_index = (int)(bio_start - first_stripe_index *
stripe_sectors) /
+ conf->chunk_sectors;
+ start_disk_offset = ((int)(bio_start - first_stripe_index *
stripe_sectors) %
+ conf->chunk_sectors) +
+ first_stripe_index * conf->chunk_sectors;
+ end_disk_index = (int)(bio_end - last_stripe_index * stripe_sectors) /
+ conf->chunk_sectors;
+ end_disk_offset = ((int)(bio_end - last_stripe_index * stripe_sectors) %
+ conf->chunk_sectors) +
+ last_stripe_index * conf->chunk_sectors;
+
+ for (d = 0; d < conf->raid_disks; d++) {
+ sector_t dev_start, dev_end;
+ struct md_rdev *rdev = READ_ONCE(conf->disks[d].rdev);
+
+ dev_start = bio_start;
+ dev_end = bio_end;
+
+ if (d < start_disk_index)
+ dev_start = (first_stripe_index + 1) *
+ conf->chunk_sectors;
+ else if (d > start_disk_index)
+ dev_start = first_stripe_index * conf->chunk_sectors;
+ else
+ dev_start = start_disk_offset;
+
+ if (d < end_disk_index)
+ dev_end = (last_stripe_index + 1) * conf->chunk_sectors;
+ else if (d > end_disk_index)
+ dev_end = last_stripe_index * conf->chunk_sectors;
+ else
+ dev_end = end_disk_offset;
+
+ if (dev_end <= dev_start)
+ continue;
+
+ md_submit_discard_bio(mddev, rdev, bi,
+ dev_start + rdev->data_offset,
+ dev_end - dev_start);
+ }
+
+ bio_endio(bi);
}
static bool raid5_make_request(struct mddev *mddev, struct bio * bi)
--
2.17.1
