This is a revised version of my series to remove all abuse of
bi_phys_segments from md/raid.
Most of the effort is for raid5 - raid1 and raid10 are
comparatively straight forward.
I've also include some optimization of md_write_start() which is now
used more heavily. This now uses percpu-refcount, which needed a
small modification.
As noted in some of these changelogs, a couple of bugs are fixed
in raid1/raid10. We should probably create fixes suitable for
-stable.
In this series there is a 'block' patch and a 'percpu-refcount' patch,
which should probably go in through other trees. I might send those
off as appropriate soonish.
Thanks,
NeilBrown
---
NeilBrown (15):
md/raid5: use md_write_start to count stripes, not bios
md/raid5: simplfy delaying of writes while metadata is updated.
md/raid5: call bio_endio() directly rather than queueing for later.
block: trace completion of all bios.
md/raid5: use bio_inc_remaining() instead of repurposing bi_phys_segments as a counter
md/raid5: remove over-loading of ->bi_phys_segments.
Revert "md/raid5: limit request size according to implementation limits"
md/raid1,raid10: move rXbio accounting closer to allocation.
md/raid10: stop using bi_phys_segments
md/raid1: stop using bi_phys_segment
md/raid5: don't test ->writes_pending in raid5_remove_disk
md: factor out set_in_sync()
md: close a race with setting mddev->in_sync
percpu-refcount: support synchronous switch to atomic mode.
MD: use per-cpu counter for writes_pending
block/bio.c | 3 +
drivers/md/dm.c | 1
drivers/md/md.c | 136 ++++++++++++++++++++-----------
drivers/md/md.h | 4 +
drivers/md/raid1.c | 95 ++++++----------------
drivers/md/raid10.c | 82 ++++++-------------
drivers/md/raid5-cache.c | 14 +--
drivers/md/raid5-log.h | 2
drivers/md/raid5.c | 170 +++++++++++----------------------------
drivers/md/raid5.h | 49 -----------
include/linux/percpu-refcount.h | 1
lib/percpu-refcount.c | 18 ++++
12 files changed, 216 insertions(+), 359 deletions(-)
--
Signature
--
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
