The patch titled
block-document-blk-plug-v3
has been added to the -mm tree. Its filename is
block-document-blk-plug-v3.patch
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/SubmitChecklist when testing your code ***
See http://userweb.kernel.org/~akpm/stuff/added-to-mm.txt to find
out what to do about this
The current -mm tree may be found at http://userweb.kernel.org/~akpm/mmotm/
------------------------------------------------------
Subject: block-document-blk-plug-v3
From: Suresh Jayaraman <sjayaraman@xxxxxxx>
* clarify why we need not disable preemption while modifying the plug list.
Signed-off-by: Suresh Jayaraman <sjayaraman@xxxxxxx>
Cc: Jens Axboe <axboe@xxxxxxxxx>
Cc: Shaohua Li <shaohua.li@xxxxxxxxx>
Cc: Jonathan Corbet <corbet@xxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
block/blk-core.c | 2 +-
include/linux/blkdev.h | 16 ++++++++--------
2 files changed, 9 insertions(+), 9 deletions(-)
diff -puN block/blk-core.c~block-document-blk-plug-v3 block/blk-core.c
--- a/block/blk-core.c~block-document-blk-plug-v3
+++ a/block/blk-core.c
@@ -2641,7 +2641,7 @@ EXPORT_SYMBOL(kblockd_schedule_delayed_w
* for a memory allocation, memory reclaim could end up wanting to free a
* page belonging to that request that is currently residing in our private
* plug. By flushing the pending I/O when the process goes to sleep, we avoid
- * this kind of deadlock.
+ * this kind of deadlocks.
*/
void blk_start_plug(struct blk_plug *plug)
{
diff -puN include/linux/blkdev.h~block-document-blk-plug-v3 include/linux/blkdev.h
--- a/include/linux/blkdev.h~block-document-blk-plug-v3
+++ a/include/linux/blkdev.h
@@ -863,16 +863,16 @@ struct request_queue *blk_alloc_queue_no
extern void blk_put_queue(struct request_queue *);
/*
- * blk_plug permits building a queue of related requests by holding the I/O
+ * blk_plug allows to build up a queue of related requests by holding the I/O
* fragments for a short period. This allows merging of sequential requests
- * into single larger request. As the requests are moved from a per-task list to
- * the device's request_queue in a batch, this results in improved scalability
- * as the lock contention for request_queue lock is reduced.
+ * into single larger request. As the requests are moved from per-task list to
+ * the device's request_queue in a batch, this results in improved
+ * scalability as the lock contention for request_queue lock is reduced.
*
- * Note: code in between changing the blk_plug list/cb_list or an element of
- * these lists is preemptible, but such code cannot sleep (or must be very
- * careful), otherwise data is corrupted. For details, please check schedule()
- * where blk_schedule_flush_plug() is called.
+ * It is ok not to disable preemption when adding the request to the plug list
+ * or when attempting merge, because blk_schedule_flush_list() will only flush
+ * the plug list when the task sleeps by itself. For details, please check
+ * schedule() where blk_schedule_flush_plug() is called.
*/
struct blk_plug {
unsigned long magic; /* detect uninitialized use-cases */
_
Patches currently in -mm which might be from sjayaraman@xxxxxxx are
block-document-blk-plug.patch
block-document-blk-plug-v3.patch
block-document-blk-plug-v3-fix.patch
--
To unsubscribe from this list: send the line "unsubscribe mm-commits" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
