summaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorDamien Le Moal <damien.lemoal@wdc.com>2019-07-11 01:18:31 +0900
committerJens Axboe <axboe@kernel.dk>2019-07-10 14:18:01 -0600
commitb49773e7bcf316f238f6709ad9e1999dcc3ed433 (patch)
tree8425dfcb1a9dc02d155d48a00c9c22903ac819ad
parent9305d5d721f2bd5e2eeb670035159b560ca211ca (diff)
downloadlinux-b49773e7bcf316f238f6709ad9e1999dcc3ed433.tar.bz2
block: Disable write plugging for zoned block devices
Simultaneously writing to a sequential zone of a zoned block device from multiple contexts requires mutual exclusion for BIO issuing to ensure that writes happen sequentially. However, even for a well behaved user correctly implementing such synchronization, BIO plugging may interfere and result in BIOs from the different contextx to be reordered if plugging is done outside of the mutual exclusion section, e.g. the plug was started by a function higher in the call chain than the function issuing BIOs. Context A Context B | blk_start_plug() | ... | seq_write_zone() | mutex_lock(zone) | bio-0->bi_iter.bi_sector = zone->wp | zone->wp += bio_sectors(bio-0) | submit_bio(bio-0) | bio-1->bi_iter.bi_sector = zone->wp | zone->wp += bio_sectors(bio-1) | submit_bio(bio-1) | mutex_unlock(zone) | return | -----------------------> | seq_write_zone() | mutex_lock(zone) | bio-2->bi_iter.bi_sector = zone->wp | zone->wp += bio_sectors(bio-2) | submit_bio(bio-2) | mutex_unlock(zone) | <------------------------- | | blk_finish_plug() In the above example, despite the mutex synchronization ensuring the correct BIO issuing order 0, 1, 2, context A BIOs 0 and 1 end up being issued after BIO 2 of context B, when the plug is released with blk_finish_plug(). While this problem can be addressed using the blk_flush_plug_list() function (in the above example, the call must be inserted before the zone mutex lock is released), a simple generic solution in the block layer avoid this additional code in all zoned block device user code. The simple generic solution implemented with this patch is to introduce the internal helper function blk_mq_plug() to access the current context plug on BIO submission. This helper returns the current plug only if the target device is not a zoned block device or if the BIO to be plugged is not a write operation. Otherwise, the caller context plug is ignored and NULL returned, resulting is all writes to zoned block device to never be plugged. Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com> Signed-off-by: Jens Axboe <axboe@kernel.dk>
-rw-r--r--block/blk-core.c2
-rw-r--r--block/blk-mq.c2
-rw-r--r--block/blk-mq.h32
3 files changed, 34 insertions, 2 deletions
diff --git a/block/blk-core.c b/block/blk-core.c
index 260e36a2c343..d0cc6e14d2f0 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -688,7 +688,7 @@ bool blk_attempt_plug_merge(struct request_queue *q, struct bio *bio,
struct request *rq;
struct list_head *plug_list;
- plug = current->plug;
+ plug = blk_mq_plug(q, bio);
if (!plug)
return false;
diff --git a/block/blk-mq.c b/block/blk-mq.c
index e5ef40c603ca..b038ec680e84 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -1973,7 +1973,7 @@ static blk_qc_t blk_mq_make_request(struct request_queue *q, struct bio *bio)
blk_mq_bio_to_request(rq, bio, nr_segs);
- plug = current->plug;
+ plug = blk_mq_plug(q, bio);
if (unlikely(is_flush_fua)) {
/* bypass scheduler for flush rq */
blk_insert_flush(rq);
diff --git a/block/blk-mq.h b/block/blk-mq.h
index f4bf5161333e..32c62c64e6c2 100644
--- a/block/blk-mq.h
+++ b/block/blk-mq.h
@@ -233,4 +233,36 @@ static inline void blk_mq_clear_mq_map(struct blk_mq_queue_map *qmap)
qmap->mq_map[cpu] = 0;
}
+/*
+ * blk_mq_plug() - Get caller context plug
+ * @q: request queue
+ * @bio : the bio being submitted by the caller context
+ *
+ * Plugging, by design, may delay the insertion of BIOs into the elevator in
+ * order to increase BIO merging opportunities. This however can cause BIO
+ * insertion order to change from the order in which submit_bio() is being
+ * executed in the case of multiple contexts concurrently issuing BIOs to a
+ * device, even if these context are synchronized to tightly control BIO issuing
+ * order. While this is not a problem with regular block devices, this ordering
+ * change can cause write BIO failures with zoned block devices as these
+ * require sequential write patterns to zones. Prevent this from happening by
+ * ignoring the plug state of a BIO issuing context if the target request queue
+ * is for a zoned block device and the BIO to plug is a write operation.
+ *
+ * Return current->plug if the bio can be plugged and NULL otherwise
+ */
+static inline struct blk_plug *blk_mq_plug(struct request_queue *q,
+ struct bio *bio)
+{
+ /*
+ * For regular block devices or read operations, use the context plug
+ * which may be NULL if blk_start_plug() was not executed.
+ */
+ if (!blk_queue_is_zoned(q) || !op_is_write(bio_op(bio)))
+ return current->plug;
+
+ /* Zoned block device write operation case: do not plug the BIO */
+ return NULL;
+}
+
#endif