linux-raid.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: NeilBrown <neilb@suse.de>
To: linux-raid@vger.kernel.org
Cc: Jonathan Brassow <jbrassow@redhat.com>
Subject: [md PATCH 10/13] md: separate meta and data devs
Date: Wed, 12 Jan 2011 09:12:47 +1100	[thread overview]
Message-ID: <20110111221247.12732.87882.stgit@notabene.brown> (raw)
In-Reply-To: <20110111221050.12732.31647.stgit@notabene.brown>

From: Jonathan Brassow <jbrassow@redhat.com>

Allow the metadata to be on a separate device from the
data.

This doesn't mean the data and metadata will by on separate
physical devices - it simply gives device-mapper and userspace
tools more flexibility.

Signed-off-by: NeilBrown <neilb@suse.de>
---

 drivers/md/bitmap.c |    6 +++++-
 drivers/md/md.c     |   10 ++++++----
 drivers/md/md.h     |    6 ++++++
 3 files changed, 17 insertions(+), 5 deletions(-)

diff --git a/drivers/md/bitmap.c b/drivers/md/bitmap.c
index 1977765..6cf5871 100644
--- a/drivers/md/bitmap.c
+++ b/drivers/md/bitmap.c
@@ -264,14 +264,18 @@ static mdk_rdev_t *next_active_rdev(mdk_rdev_t *rdev, mddev_t *mddev)
 static int write_sb_page(struct bitmap *bitmap, struct page *page, int wait)
 {
 	mdk_rdev_t *rdev = NULL;
+	struct block_device *bdev;
 	mddev_t *mddev = bitmap->mddev;
 
 	while ((rdev = next_active_rdev(rdev, mddev)) != NULL) {
 		int size = PAGE_SIZE;
 		loff_t offset = mddev->bitmap_info.offset;
+
+		bdev = (rdev->meta_bdev) ? rdev->meta_bdev : rdev->bdev;
+
 		if (page->index == bitmap->file_pages-1)
 			size = roundup(bitmap->last_page_size,
-				       bdev_logical_block_size(rdev->bdev));
+				       bdev_logical_block_size(bdev));
 		/* Just make sure we aren't corrupting data or
 		 * metadata
 		 */
diff --git a/drivers/md/md.c b/drivers/md/md.c
index 0bc10cc..b98a85f 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -765,7 +765,7 @@ void md_super_write(mddev_t *mddev, mdk_rdev_t *rdev,
 	 */
 	struct bio *bio = bio_alloc_mddev(GFP_NOIO, 1, mddev);
 
-	bio->bi_bdev = rdev->bdev;
+	bio->bi_bdev = rdev->meta_bdev ? rdev->meta_bdev : rdev->bdev;
 	bio->bi_sector = sector;
 	bio_add_page(bio, page, size, 0);
 	bio->bi_private = rdev;
@@ -803,7 +803,8 @@ int sync_page_io(mdk_rdev_t *rdev, sector_t sector, int size,
 
 	rw |= REQ_SYNC | REQ_UNPLUG;
 
-	bio->bi_bdev = rdev->bdev;
+	bio->bi_bdev = (metadata_op && rdev->meta_bdev) ?
+		rdev->meta_bdev : rdev->bdev;
 	if (metadata_op)
 		bio->bi_sector = sector + rdev->sb_start;
 	else
@@ -4435,7 +4436,9 @@ int md_run(mddev_t *mddev)
 		 * We don't want the data to overlap the metadata,
 		 * Internal Bitmap issues have been handled elsewhere.
 		 */
-		if (rdev->data_offset < rdev->sb_start) {
+		if (rdev->meta_bdev) {
+			/* Nothing to check */;
+		} else if (rdev->data_offset < rdev->sb_start) {
 			if (mddev->dev_sectors &&
 			    rdev->data_offset + mddev->dev_sectors
 			    > rdev->sb_start) {
@@ -5532,7 +5535,6 @@ static int update_size(mddev_t *mddev, sector_t num_sectors)
 	 * sb_start or, if that is <data_offset, it must fit before the size
 	 * of each device.  If num_sectors is zero, we find the largest size
 	 * that fits.
-
 	 */
 	if (mddev->sync_thread)
 		return -EBUSY;
diff --git a/drivers/md/md.h b/drivers/md/md.h
index 7e4f358..eec517c 100644
--- a/drivers/md/md.h
+++ b/drivers/md/md.h
@@ -60,6 +60,12 @@ struct mdk_rdev_s
 	mddev_t *mddev;			/* RAID array if running */
 	int last_events;		/* IO event timestamp */
 
+	/*
+	 * If meta_bdev is non-NULL, it means that a separate device is
+	 * being used to store the metadata (superblock/bitmap) which
+	 * would otherwise be contained on the same device as the data (bdev).
+	 */
+	struct block_device *meta_bdev;
 	struct block_device *bdev;	/* block device handle */
 
 	struct page	*sb_page;



  parent reply	other threads:[~2011-01-11 22:12 UTC|newest]

Thread overview: 14+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-01-11 22:12 [md PATCH 00/13] md patches for the current merge window NeilBrown
2011-01-11 22:12 ` [md PATCH 05/13] md/raid5: use sysfs_notify_dirent_safe to avoid NULL pointer NeilBrown
2011-01-11 22:12 ` [md PATCH 07/13] md: Be more careful about clearing flags bit in ->recovery NeilBrown
2011-01-11 22:12 ` [md PATCH 04/13] md: Ensure no IO request to get md device before it is properly initialised NeilBrown
2011-01-11 22:12 ` [md PATCH 01/13] md: fix regression with re-adding devices to arrays with no metadata NeilBrown
2011-01-11 22:12 ` [md PATCH 06/13] md: md_stop_writes requires mddev_lock NeilBrown
2011-01-11 22:12 ` [md PATCH 02/13] md: fix regression resulting in delays in clearing bits in a bitmap NeilBrown
2011-01-11 22:12 ` [md PATCH 03/13] md: Fix single printks with multiple KERN_<level>s NeilBrown
2011-01-11 22:12 ` [md PATCH 13/13] md: fix sync_completed reporting for very large drives (>2TB) NeilBrown
2011-01-11 22:12 ` [md PATCH 09/13] md-new-param-to_sync_page_io NeilBrown
2011-01-11 22:12 ` [md PATCH 12/13] md: allow suspend_lo and suspend_hi to decrease as well as increase NeilBrown
2011-01-11 22:12 ` [md PATCH 11/13] md: Don't let implementation detail of curr_resync leak out through sysfs NeilBrown
2011-01-11 22:12 ` NeilBrown [this message]
2011-01-11 22:12 ` [md PATCH 08/13] md-new-param-to-calc_dev_sboffset NeilBrown

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=20110111221247.12732.87882.stgit@notabene.brown \
    --to=neilb@suse.de \
    --cc=jbrassow@redhat.com \
    --cc=linux-raid@vger.kernel.org \
    /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).