[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-ID: <20160526180813.GA49039@shli-mbp.local>
Date: Thu, 26 May 2016 11:08:14 -0700
From: Shaohua Li <shli@...com>
To: <linux-block@...r.kernel.org>, <linux-kernel@...r.kernel.org>
CC: <sitsofe@...oo.com>, <snitzer@...hat.com>, <axboe@...com>,
<martin.petersen@...cle.com>, <Kernel-team@...com>
Subject: [PATCH] block: correctly fallback for zeroout
blkdev_issue_zeroout try discard/writesame first, if they fail, zeroout
fallback to regular write. The problem is discard/writesame doesn't
return error for -EOPNOTSUPP, then zeroout can't do fallback and leave
disk data not changed. zeroout should have guaranteed zero-fill
behavior.
BTW, I saw several callers of blkdev_issue_discard can handle
-EOPNOTSUPP, not sure why blkdev_issue_discard not returns -EOPNOTSUPP.
The same story for blkdev_issue_write_same.
https://bugzilla.kernel.org/show_bug.cgi?id=118581
Cc: Sitsofe Wheeler <sitsofe@...oo.com>
Cc: Mike Snitzer <snitzer@...hat.com>
Cc: Jens Axboe <axboe@...com>
Cc: Martin K. Petersen <martin.petersen@...cle.com>
Signed-off-by: Shaohua Li <shli@...com>
---
block/blk-lib.c | 35 ++++++++++++++++++++++++++---------
1 file changed, 26 insertions(+), 9 deletions(-)
diff --git a/block/blk-lib.c b/block/blk-lib.c
index 23d7f30..232f9ea 100644
--- a/block/blk-lib.c
+++ b/block/blk-lib.c
@@ -95,8 +95,9 @@ EXPORT_SYMBOL(__blkdev_issue_discard);
* Description:
* Issue a discard request for the sectors in question.
*/
-int blkdev_issue_discard(struct block_device *bdev, sector_t sector,
- sector_t nr_sects, gfp_t gfp_mask, unsigned long flags)
+static int do_blkdev_issue_discard(struct block_device *bdev, sector_t sector,
+ sector_t nr_sects, gfp_t gfp_mask, unsigned long flags,
+ bool ignore_nosupport)
{
int type = REQ_WRITE | REQ_DISCARD;
struct bio *bio = NULL;
@@ -111,13 +112,20 @@ int blkdev_issue_discard(struct block_device *bdev, sector_t sector,
&bio);
if (!ret && bio) {
ret = submit_bio_wait(type, bio);
- if (ret == -EOPNOTSUPP)
+ if (ignore_nosupport && ret == -EOPNOTSUPP)
ret = 0;
}
blk_finish_plug(&plug);
return ret;
}
+
+int blkdev_issue_discard(struct block_device *bdev, sector_t sector,
+ sector_t nr_sects, gfp_t gfp_mask, unsigned long flags)
+{
+ return do_blkdev_issue_discard(bdev, sector, nr_sects, gfp_mask,
+ flags, true);
+}
EXPORT_SYMBOL(blkdev_issue_discard);
/**
@@ -131,9 +139,9 @@ EXPORT_SYMBOL(blkdev_issue_discard);
* Description:
* Issue a write same request for the sectors in question.
*/
-int blkdev_issue_write_same(struct block_device *bdev, sector_t sector,
+static int do_blkdev_issue_write_same(struct block_device *bdev, sector_t sector,
sector_t nr_sects, gfp_t gfp_mask,
- struct page *page)
+ struct page *page, bool ignore_nosupport)
{
struct request_queue *q = bdev_get_queue(bdev);
unsigned int max_write_same_sectors;
@@ -167,7 +175,15 @@ int blkdev_issue_write_same(struct block_device *bdev, sector_t sector,
if (bio)
ret = submit_bio_wait(REQ_WRITE | REQ_WRITE_SAME, bio);
- return ret != -EOPNOTSUPP ? ret : 0;
+ return (ret != -EOPNOTSUPP || !ignore_nosupport) ? ret : 0;
+}
+
+int blkdev_issue_write_same(struct block_device *bdev, sector_t sector,
+ sector_t nr_sects, gfp_t gfp_mask,
+ struct page *page)
+{
+ return do_blkdev_issue_write_same(bdev, sector, nr_sects, gfp_mask,
+ page, true);
}
EXPORT_SYMBOL(blkdev_issue_write_same);
@@ -238,12 +254,13 @@ int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
struct request_queue *q = bdev_get_queue(bdev);
if (discard && blk_queue_discard(q) && q->limits.discard_zeroes_data &&
- blkdev_issue_discard(bdev, sector, nr_sects, gfp_mask, 0) == 0)
+ do_blkdev_issue_discard(bdev, sector, nr_sects, gfp_mask, 0,
+ false) == 0)
return 0;
if (bdev_write_same(bdev) &&
- blkdev_issue_write_same(bdev, sector, nr_sects, gfp_mask,
- ZERO_PAGE(0)) == 0)
+ do_blkdev_issue_write_same(bdev, sector, nr_sects, gfp_mask,
+ ZERO_PAGE(0), false) == 0)
return 0;
return __blkdev_issue_zeroout(bdev, sector, nr_sects, gfp_mask);
--
2.8.0.rc2
Powered by blists - more mailing lists