linux-raid.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Christoph Hellwig <hch@lst.de>
To: Shaohua Li <shli@fb.com>, neilb@suse.de
Cc: dan.j.williams@intel.com, Kernel-team@fb.com, linux-raid@vger.kernel.org
Subject: [PATCH 04/11] raid5-cache: simplify state machine when caches flushes are not needed
Date: Mon,  5 Oct 2015 09:31:09 +0200	[thread overview]
Message-ID: <1444030276-30850-5-git-send-email-hch@lst.de> (raw)
In-Reply-To: <1444030276-30850-1-git-send-email-hch@lst.de>

For devices without a volatile write cache we don't need to send a FLUSH
command to ensure writes are stable on disk, and thus can avoid the whole
step of batching up bios for processing by the MD thread.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 drivers/md/raid5-cache.c | 33 +++++++++++++++++++++++++++++----
 1 file changed, 29 insertions(+), 4 deletions(-)

diff --git a/drivers/md/raid5-cache.c b/drivers/md/raid5-cache.c
index b04e908..0d18ed7 100644
--- a/drivers/md/raid5-cache.c
+++ b/drivers/md/raid5-cache.c
@@ -83,6 +83,8 @@ struct r5l_log {
 
 	struct list_head no_space_stripes; /* pending stripes, log has no space */
 	spinlock_t no_space_stripes_lock;
+
+	bool need_cache_flush;
 };
 
 /*
@@ -205,6 +207,22 @@ static void r5l_io_run_stripes(struct r5l_io_unit *io)
 	}
 }
 
+static void r5l_log_run_stripes(struct r5l_log *log)
+{
+	struct r5l_io_unit *io, *next;
+
+	assert_spin_locked(&log->io_list_lock);
+
+	list_for_each_entry_safe(io, next, &log->running_ios, log_sibling) {
+		/* don't change list order */
+		if (io->state < IO_UNIT_IO_END)
+			break;
+
+		list_move_tail(&io->log_sibling, &log->finished_ios);
+		r5l_io_run_stripes(io);
+	}
+}
+
 static void r5l_log_endio(struct bio *bio)
 {
 	struct r5l_io_unit *io = bio->bi_private;
@@ -221,11 +239,15 @@ static void r5l_log_endio(struct bio *bio)
 
 	spin_lock_irqsave(&log->io_list_lock, flags);
 	__r5l_set_io_unit_state(io, IO_UNIT_IO_END);
-	r5l_move_io_unit_list(&log->running_ios, &log->io_end_ios,
-			IO_UNIT_IO_END);
+	if (log->need_cache_flush)
+		r5l_move_io_unit_list(&log->running_ios, &log->io_end_ios,
+				      IO_UNIT_IO_END);
+	else
+		r5l_log_run_stripes(log);
 	spin_unlock_irqrestore(&log->io_list_lock, flags);
 
-	md_wakeup_thread(log->rdev->mddev->thread);
+	if (log->need_cache_flush)
+		md_wakeup_thread(log->rdev->mddev->thread);
 }
 
 static void r5l_submit_current_io(struct r5l_log *log)
@@ -626,7 +648,8 @@ static void r5l_log_flush_endio(struct bio *bio)
 void r5l_flush_stripe_to_raid(struct r5l_log *log)
 {
 	bool do_flush;
-	if (!log)
+
+	if (!log || !log->need_cache_flush)
 		return;
 
 	spin_lock_irq(&log->io_list_lock);
@@ -1115,6 +1138,8 @@ int r5l_init_log(struct r5conf *conf, struct md_rdev *rdev)
 		return -ENOMEM;
 	log->rdev = rdev;
 
+	log->need_cache_flush = (rdev->bdev->bd_disk->queue->flush_flags != 0);
+
 	log->uuid_checksum = crc32_le(~0, (void *)rdev->mddev->uuid,
 				      sizeof(rdev->mddev->uuid));
 
-- 
1.9.1


  parent reply	other threads:[~2015-10-05  7:31 UTC|newest]

Thread overview: 14+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-10-05  7:31 raid5-cache I/O path improvements V3 Christoph Hellwig
2015-10-05  7:31 ` [PATCH 01/11] raid5-cache: free I/O units earlier Christoph Hellwig
2015-10-05  7:31 ` [PATCH 02/11] raid5-cache: rename flushed_ios to finished_ios Christoph Hellwig
2015-10-05  7:31 ` [PATCH 03/11] raid5-cache: factor out a helper to run all stripes for an I/O unit Christoph Hellwig
2015-10-05  7:31 ` Christoph Hellwig [this message]
2015-10-05  7:31 ` [PATCH 05/11] raid5-cache: clean up r5l_get_meta Christoph Hellwig
2015-10-05  7:31 ` [PATCH 06/11] raid5-cache: refactor bio allocation Christoph Hellwig
2015-10-05  7:31 ` [PATCH 07/11] raid5-cache: take rdev->data_offset into account early on Christoph Hellwig
2015-10-05  7:31 ` [PATCH 08/11] raid5-cache: inline r5l_alloc_io_unit into r5l_new_meta Christoph Hellwig
2015-10-05  7:31 ` [PATCH 09/11] raid5-cache: new helper: r5_reserve_log_entry Christoph Hellwig
2015-10-05  7:31 ` [PATCH 10/11] raid5-cache: small log->seq cleanup Christoph Hellwig
2015-10-05  7:31 ` [PATCH 11/11] raid5-cache: use bio chaining Christoph Hellwig
2015-10-06 18:14 ` raid5-cache I/O path improvements V3 Shaohua Li
2015-10-08  2:45   ` Neil Brown

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1444030276-30850-5-git-send-email-hch@lst.de \
    --to=hch@lst.de \
    --cc=Kernel-team@fb.com \
    --cc=dan.j.williams@intel.com \
    --cc=linux-raid@vger.kernel.org \
    --cc=neilb@suse.de \
    --cc=shli@fb.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).