Il 03/07/2012 16:00, Vivek Goyal ha scritto:
>> >
>> > Ah, interesting, I missed it completely. I guess it's because queue/
>> > directories only exist for full disks, and the correct alignment varies
>> > for each partition. So this patch is unnecessary.
> Partition discard_alignments are available in
> /sys/block/<disk>/<partition>/discard_alignment.
Yes. If it were in queue/, it wouldn't be available.
> That raises an interesting question for patch3. If the discard is happening to
> a partition, shouldn't you be looking at partition discard_alignment
> instead of always looking at queue discard_alignment?
Good point! Like this?
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index ba43f40..3530764 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -1125,6 +1125,16 @@ static inline int queue_limit_discard_alignment(struct queue_limits *lim, sector
& (lim->discard_granularity - 1);
}
+static inline int bdev_discard_alignment(struct block_device *bdev)
+{
+ struct request_queue *q = bdev_get_queue(bdev);
+
+ if (bdev != bdev->bd_contains)
+ return bdev->bd_part->discard_alignment;
+
+ return q->limits.discard_alignment;
+}
+
static inline unsigned int queue_discard_zeroes_data(struct request_queue *q)
{
if (q->limits.max_discard_sectors && q->limits.discard_zeroes_data == 1)
diff --git a/block/blk-lib.c b/block/blk-lib.c
index b2bde5c..77d8869 100644
--- a/block/blk-lib.c
+++ b/block/blk-lib.c
@@ -58,7 +58,7 @@ int blkdev_issue_discard(struct block_device *bdev, sector_t sector,
/* Zero-sector (unknown) and one-sector granularities are the same. */
granularity = max(q->limits.discard_granularity >> 9, 1U);
mask = granularity - 1;
- alignment = (q->limits.discard_alignment >> 9) & mask;
+ alignment = bdev_discard_alignment(bdev) >> 9;
/*
* Ensure that max_discard_sectors is of the proper
Paolo
_______________________________________________
xfs mailing list
xfs@xxxxxxxxxxx
http://oss.sgi.com/mailman/listinfo/xfs
