Skip to contentSkip to conversation

raid1: ensure bio doesn't have more than BIO_MAX_VECS sectors

PATCH In review Guoqing Jiang 5 participants 15 messages Aug 13, 2021 – Sep 25, 2021

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 raid1: ensure bio doesn't have more than BIO_MAX_VECS sectors

From: Guoqing Jiang <[email protected]>

We can't split bio with more than BIO_MAX_VECS sectors, otherwise the below call trace was triggered because we could allocate oversized write behind bio later.

[ 8.097936] bvec_alloc+0x90/0xc0
[ 8.098934] bio_alloc_bioset+0x1b3/0x260
[ 8.099959] raid1_make_request+0x9ce/0xc50 [raid1]
[ 8.100988] ? __bio_clone_fast+0xa8/0xe0
[ 8.102008] md_handle_request+0x158/0x1d0 [md_mod]
[ 8.103050] md_submit_bio+0xcd/0x110 [md_mod]
[ 8.104084] submit_bio_noacct+0x139/0x530
[ 8.105127] submit_bio+0x78/0x1d0
[ 8.106163] ext4_io_submit+0x48/0x60 [ext4]
[ 8.107242] ext4_writepages+0x652/0x1170 [ext4]
[ 8.108300] ? do_writepages+0x41/0x100
[ 8.109338] ? __ext4_mark_inode_dirty+0x240/0x240 [ext4]
[ 8.110406] do_writepages+0x41/0x100
[ 8.111450] __filemap_fdatawrite_range+0xc5/0x100
[ 8.112513] file_write_and_wait_range+0x61/0xb0
[ 8.113564] ext4_sync_file+0x73/0x370 [ext4]
[ 8.114607] __x64_sys_fsync+0x33/0x60
[ 8.115635] do_syscall_64+0x33/0x40
[ 8.116670] entry_SYSCALL_64_after_hwframe+0x44/0xae

[1]. https://bugs.archlinux.org/task/70992

---
Note: this depends on commit 018eca456c4b4dca56aaf1ec27f309c74d0fe246 in linux-block for-next branch.

 drivers/md/raid1.c | 1 +
 1 file changed, 1 insertion(+)
drivers/md/raid1.c +1 -0
index 3c44c4bb40fc..ab21abc056b8 100644
@@ -1454,6 +1454,7 @@ static void raid1_write_request(struct mddev *mddev, struct bio *bio,
 		goto retry_write;
 	}
 
+	max_sectors = min_t(int, max_sectors, BIO_MAX_VECS * PAGE_SECTORS);
 	if (max_sectors < bio_sectors(bio)) {
 		struct bio *split = bio_split(bio, max_sectors,
 					      GFP_NOIO, &conf->bio_split);
Signature · 1 line
2.25.1
#1 lore raw
Quoted 9 lines · Guoqing Jiang

From: Guoqing Jiang <[email protected]>

We can't split bio with more than BIO_MAX_VECS sectors, otherwise the below call trace was triggered because we could allocate oversized write behind bio later.

[ 8.097936] bvec_alloc+0x90/0xc0
[ 8.098934] bio_alloc_bioset+0x1b3/0x260
[ 8.099959] raid1_make_request+0x9ce/0xc50 [raid1]

Which bio_alloc_bioset is this? The one in alloc_behind_master_bio?

In which case I think you want to limit the reduction of max_sectors to just the write behind case, and clearly document what is going on.

In general the size of a bio only depends on the number of vectors, not the total I/O size. But alloc_behind_master_bio allocates new backing pages using order 0 allocations, so in this exceptional case the total size oes actually matter.

While we're at it: this huge memory allocation looks really deadlock prone.

#2 lore raw
Quoted 11 lines · Christoph Hellwig
Quoted 9 lines · Guoqing Jiang

From: Guoqing Jiang <[email protected]>

We can't split bio with more than BIO_MAX_VECS sectors, otherwise the below call trace was triggered because we could allocate oversized write behind bio later.

[ 8.097936] bvec_alloc+0x90/0xc0
[ 8.098934] bio_alloc_bioset+0x1b3/0x260
[ 8.099959] raid1_make_request+0x9ce/0xc50 [raid1]

Which bio_alloc_bioset is this? The one in alloc_behind_master_bio?

Yes, it should be the one since bio_clone_fast calls bio_alloc_bioset with 0 iovecs.

Quoted 2 lines

In which case I think you want to limit the reduction of max_sectors to just the write behind case, and clearly document what is going on.

Ok, thanks.

Quoted 7 lines

In general the size of a bio only depends on the number of vectors, not the total I/O size. But alloc_behind_master_bio allocates new backing pages using order 0 allocations, so in this exceptional case the total size oes actually matter.

While we're at it: this huge memory allocation looks really deadlock prone.

Hmm, let me think more about it, or could you share your thought? 😉

Thanks,
Guoqing

#3 lore raw
Quoted 8 lines · Guoqing Jiang

Ok, thanks.

Quoted 5 lines

In general the size of a bio only depends on the number of vectors, not the total I/O size. But alloc_behind_master_bio allocates new backing pages using order 0 allocations, so in this exceptional case the total size oes actually matter.

Quoted 4 lines
Quoted 2 lines

While we're at it: this huge memory allocation looks really deadlock prone.

Hmm, let me think more about it, or could you share your thought? ????

Well, you'd need a mempool which can fit the max payload of a bio, that is BIO_MAX_VECS pages.

FYI, this is what I'd do instead of this patch for now. We don't really need a vetor per sector, just per page. So this limits the I/O size a little less.

drivers/md/raid1.c +9 -0
index 3c44c4bb40fc..5b27d995302e 100644
@@ -1454,6 +1454,15 @@ static void raid1_write_request(struct mddev *mddev, struct bio *bio,
 		goto retry_write;
 	}
 
+	/*
+	 * When using a bitmap, we may call alloc_behind_master_bio below.
+	 * alloc_behind_master_bio allocates a copy of the data payload a page
+	 * at a time and thus needs a new bio that can fit the whole payload
+	 * this bio in page sized chunks.
+	 */
+	if (bitmap)
+		max_sectors = min_t(int, max_sectors, BIO_MAX_VECS * PAGE_SIZE);
+
 	if (max_sectors < bio_sectors(bio)) {
 		struct bio *split = bio_split(bio, max_sectors,
 					      GFP_NOIO, &conf->bio_split);
#4 lore raw
Quoted 27 lines · Christoph Hellwig
Quoted 8 lines · Guoqing Jiang

Ok, thanks.

Quoted 5 lines

In general the size of a bio only depends on the number of vectors, not the total I/O size. But alloc_behind_master_bio allocates new backing pages using order 0 allocations, so in this exceptional case the total size oes actually matter.

Quoted 4 lines
Quoted 2 lines

While we're at it: this huge memory allocation looks really deadlock prone.

Hmm, let me think more about it, or could you share your thought? ????

Well, you'd need a mempool which can fit the max payload of a bio, that is BIO_MAX_VECS pages.

FYI, this is what I'd do instead of this patch for now. We don't really need a vetor per sector, just per page. So this limits the I/O size a little less.

drivers/md/raid1.c +0 -0
index 3c44c4bb40fc..5b27d995302e 100644
@@ -1454,6 +1454,15 @@ static void raid1_write_request(struct mddev *mddev, struct bio *bio,
 		goto retry_write;

drivers/md/raid1.c static void raid1_write_request(struct mddev *mddev, struct bio *bio, · line 1464

Quoted 10 lines
drivers/md/raid1.c +8 -0
 	}
 
+	/*
+	 * When using a bitmap, we may call alloc_behind_master_bio below.
+	 * alloc_behind_master_bio allocates a copy of the data payload a page
+	 * at a time and thus needs a new bio that can fit the whole payload
+	 * this bio in page sized chunks.
+	 */
+	if (bitmap)
+		max_sectors = min_t(int, max_sectors, BIO_MAX_VECS * PAGE_SIZE);

s/PAGE_SIZE/PAGE_SECTORS

drivers/md/raid1.c static void raid1_write_request(struct mddev *mddev, struct bio *bio, · line 1468

Quoted 5 lines
drivers/md/raid1.c +1 -0
+
 	if (max_sectors < bio_sectors(bio)) {
 		struct bio *split = bio_split(bio, max_sectors,
 					      GFP_NOIO, &conf->bio_split);

Here the limit is max single-page vectors, and the above way may not work, such as:

0 ~ 254: each bvec's length is 512
255: bvec's length is 8192

the total length is just 512*255 + 8192 = 138752 bytes = 271 sectors, but it still may need 257 bvecs, which can't be allocated via bio_alloc_bioset().

One solution is to add queue limit of max_single_page_bvec, and let blk_queue_split() handle it.

Thanks,
Ming

#5 lore raw

Hi Ming and Christoph,

Quoted 9 lines · Ming Lei
Quoted 8 lines · Christoph Hellwig
Quoted 7 lines · Guoqing Jiang

Ok, thanks.

Quoted 5 lines

In general the size of a bio only depends on the number of vectors, not the total I/O size. But alloc_behind_master_bio allocates new backing pages using order 0 allocations, so in this exceptional case the total size oes actually matter.

Quoted 5 lines
Quoted 5 lines
Quoted 3 lines
Quoted 2 lines

While we're at it: this huge memory allocation looks really deadlock prone.

Hmm, let me think more about it, or could you share your thought? ????

Well, you'd need a mempool which can fit the max payload of a bio, that is BIO_MAX_VECS pages.

IIUC, the behind bio is allocated from bio_set (mddev->bio_set) which is allocated in md_run by call bioset_init, so the mempool (bvec_pool) of  this bio_set is created by biovec_init_pool which uses global biovec slabs. Do we really need another mempool? Or, there is no potential deadlock for this case.

Quoted 12 lines
Quoted 12 lines

FYI, this is what I'd do instead of this patch for now. We don't really need a vetor per sector, just per page. So this limits the I/O size a little less.

drivers/md/raid1.c +0 -0
index 3c44c4bb40fc..5b27d995302e 100644
@@ -1454,6 +1454,15 @@ static void raid1_write_request(struct mddev *mddev, struct bio *bio,
  		goto retry_write;
  	}
 
Quoted 6 lines

drivers/md/raid1.c static void raid1_write_request(struct mddev *mddev, struct bio *bio, · line 1462

Quoted 6 lines
drivers/md/raid1.c +6 -0
+	/*
+	 * When using a bitmap, we may call alloc_behind_master_bio below.
+	 * alloc_behind_master_bio allocates a copy of the data payload a page
+	 * at a time and thus needs a new bio that can fit the whole payload
+	 * this bio in page sized chunks.
+	 */

Thanks for the above, will copy it accordingly. I will check if WriteMostly is set before, then check both the flag and bitmap.

Quoted 3 lines

drivers/md/raid1.c static void raid1_write_request(struct mddev *mddev, struct bio *bio, · line 1464

Quoted 2 lines
drivers/md/raid1.c +2 -0
+	if (bitmap)
+		max_sectors = min_t(int, max_sectors, BIO_MAX_VECS * PAGE_SIZE);

s/PAGE_SIZE/PAGE_SECTORS

Agree.

Quoted 4 lines

drivers/md/raid1.c static void raid1_write_request(struct mddev *mddev, struct bio *bio, · line 1468

Quoted 4 lines
drivers/md/raid1.c +1 -0
+
  	if (max_sectors < bio_sectors(bio)) {
  		struct bio *split = bio_split(bio, max_sectors,
  					      GFP_NOIO, &conf->bio_split);
Quoted 8 lines

Here the limit is max single-page vectors, and the above way may not work, such as:ust splitted and not

0 ~ 254: each bvec's length is 512
255: bvec's length is 8192

the total length is just 512*255 + 8192 = 138752 bytes = 271 sectors, but it still may need 257 bvecs, which can't be allocated via bio_alloc_bioset().

Thanks for deeper looking! I guess it is because how vcnt is calculated.

Quoted 2 lines

One solution is to add queue limit of max_single_page_bvec, and let blk_queue_split() handle it.

The path (blk_queue_split -> blk_bio_segment_split -> bvec_split_segs) which respects max_segments of limit. Do you mean introduce max_single_page_bvec to limit? Then perform similar checking as for max_segment.

Thanks,
Guoqing

#6 lore raw
Quoted 69 lines · Guoqing Jiang

Hi Ming and Christoph,

Quoted 9 lines · Ming Lei
Quoted 8 lines · Christoph Hellwig
Quoted 7 lines · Guoqing Jiang

Ok, thanks.

Quoted 5 lines

In general the size of a bio only depends on the number of vectors, not the total I/O size. But alloc_behind_master_bio allocates new backing pages using order 0 allocations, so in this exceptional case the total size oes actually matter.

Quoted 5 lines
Quoted 5 lines
Quoted 3 lines
Quoted 2 lines

While we're at it: this huge memory allocation looks really deadlock prone.

Hmm, let me think more about it, or could you share your thought? ????

Well, you'd need a mempool which can fit the max payload of a bio, that is BIO_MAX_VECS pages.

IIUC, the behind bio is allocated from bio_set (mddev->bio_set) which is allocated in md_run by
call bioset_init, so the mempool (bvec_pool) of  this bio_set is created by biovec_init_pool which
uses global biovec slabs. Do we really need another mempool? Or, there is no potential deadlock
 for this case.

Quoted 4 lines
Quoted 4 lines

FYI, this is what I'd do instead of this patch for now. We don't really need a vetor per sector, just per page. So this limits the I/O size a little less.

Quoted 13 lines

drivers/md/raid1.c static void raid1_write_request(struct mddev *mddev, struct bio *bio, · line 1461

Quoted 13 lines
drivers/md/raid1.c +6 -0
index 3c44c4bb40fc..5b27d995302e 100644
@@ -1454,6 +1454,15 @@ static void raid1_write_request(struct mddev *mddev, struct bio *bio,
  		goto retry_write;
  	}
+	/*
+	 * When using a bitmap, we may call alloc_behind_master_bio below.
+	 * alloc_behind_master_bio allocates a copy of the data payload a page
+	 * at a time and thus needs a new bio that can fit the whole payload
+	 * this bio in page sized chunks.
+	 */

Thanks for the above, will copy it accordingly. I will check if WriteMostly is set before, then check both
the flag and bitmap.

Quoted 3 lines

drivers/md/raid1.c static void raid1_write_request(struct mddev *mddev, struct bio *bio, · line 1464

Quoted 2 lines
drivers/md/raid1.c +2 -0
+	if (bitmap)
+		max_sectors = min_t(int, max_sectors, BIO_MAX_VECS * PAGE_SIZE);

s/PAGE_SIZE/PAGE_SECTORS

Agree.

Quoted 4 lines

drivers/md/raid1.c static void raid1_write_request(struct mddev *mddev, struct bio *bio, · line 1468

Quoted 4 lines
drivers/md/raid1.c +1 -0
+
  	if (max_sectors < bio_sectors(bio)) {
  		struct bio *split = bio_split(bio, max_sectors,
  					      GFP_NOIO, &conf->bio_split);
Quoted 8 lines

Here the limit is max single-page vectors, and the above way may not work, such as:ust splitted and not

0 ~ 254: each bvec's length is 512
255: bvec's length is 8192

the total length is just 512*255 + 8192 = 138752 bytes = 271 sectors, but it still may need 257 bvecs, which can't be allocated via bio_alloc_bioset().

Thanks for deeper looking! I guess it is because how vcnt is calculated.

Quoted 8 lines
Quoted 2 lines

One solution is to add queue limit of max_single_page_bvec, and let blk_queue_split() handle it.

The path (blk_queue_split -> blk_bio_segment_split -> bvec_split_segs) which respects max_segments
of limit. Do you mean introduce max_single_page_bvec to limit? Then perform similar checking as for
 max_segment.

Yes, then the bio is guaranteed to not reach max single-page bvec limit, just like what __blk_queue_bounce() does.

thanks,
Ming

#7 lore raw
In reply to Ming Lei
Quoted 4 lines · Ming Lei

drivers/md/raid1.c static void raid1_write_request(struct mddev *mddev, struct bio *bio, · line 1464

Quoted 2 lines
drivers/md/raid1.c +2 -0
+	if (bitmap)
+		max_sectors = min_t(int, max_sectors, BIO_MAX_VECS * PAGE_SIZE);

s/PAGE_SIZE/PAGE_SECTORS

Yeah, max_sectors is in size units, I messed that up.

Quoted 7 lines

drivers/md/raid1.c static void raid1_write_request(struct mddev *mddev, struct bio *bio, · line 1468

Quoted 6 lines
drivers/md/raid1.c +1 -0
+
 	if (max_sectors < bio_sectors(bio)) {
 		struct bio *split = bio_split(bio, max_sectors,
 					      GFP_NOIO, &conf->bio_split);
Quoted 8 lines

Here the limit is max single-page vectors, and the above way may not work, such as:

0 ~ 254: each bvec's length is 512
255: bvec's length is 8192

the total length is just 512*255 + 8192 = 138752 bytes = 271 sectors, but it still may need 257 bvecs, which can't be allocated via bio_alloc_bioset().

Yes, we still need the rounding magic that alloc_behind_master_bio uses here.

Quoted 2 lines

One solution is to add queue limit of max_single_page_bvec, and let blk_queue_split() handle it.

I'd rather not bloat the core with this.

#8 lore raw
Quoted 18 lines · Christoph Hellwig
Quoted 4 lines · Ming Lei

drivers/md/raid1.c static void raid1_write_request(struct mddev *mddev, struct bio *bio, · line 1464

Quoted 2 lines
drivers/md/raid1.c +2 -0
+	if (bitmap)
+		max_sectors = min_t(int, max_sectors, BIO_MAX_VECS * PAGE_SIZE);

s/PAGE_SIZE/PAGE_SECTORS

Yeah, max_sectors is in size units, I messed that up.

Quoted 10 lines

drivers/md/raid1.c static void raid1_write_request(struct mddev *mddev, struct bio *bio, · line 1468

Quoted 6 lines
drivers/md/raid1.c +1 -0
+
 	if (max_sectors < bio_sectors(bio)) {
 		struct bio *split = bio_split(bio, max_sectors,
 					      GFP_NOIO, &conf->bio_split);

Here the limit is max single-page vectors, and the above way may not work, such as:

Quoted 8 lines
Quoted 5 lines

0 ~ 254: each bvec's length is 512
255: bvec's length is 8192

the total length is just 512*255 + 8192 = 138752 bytes = 271 sectors, but it still may need 257 bvecs, which can't be allocated via bio_alloc_bioset().

Yes, we still need the rounding magic that alloc_behind_master_bio uses here.

But it is wrong to use max sectors to limit number of bvecs(segments), isn't it?

Thanks,
Ming

#9 lore raw
Quoted 11 lines · Ming Lei
Quoted 9 lines
Quoted 6 lines

0 ~ 254: each bvec's length is 512
255: bvec's length is 8192

the total length is just 512*255 + 8192 = 138752 bytes = 271 sectors, but it still may need 257 bvecs, which can't be allocated via bio_alloc_bioset().

Yes, we still need the rounding magic that alloc_behind_master_bio uses here.

But it is wrong to use max sectors to limit number of bvecs(segments), isn't it?

The raid1 write behind code cares about the size ofa bio it can reach by adding order 0 pages to it. The bvecs are part of that and I think the calculation in the patch documents that a well.

#10 lore raw
Quoted 8 lines · Christoph Hellwig
Quoted 7 lines · Ming Lei
Quoted 7 lines
Quoted 6 lines

0 ~ 254: each bvec's length is 512
255: bvec's length is 8192

the total length is just 512*255 + 8192 = 138752 bytes = 271 sectors, but it still may need 257 bvecs, which can't be allocated via bio_alloc_bioset().

Quoted 8 lines
Quoted 4 lines
Quoted 2 lines

Yes, we still need the rounding magic that alloc_behind_master_bio uses here.

But it is wrong to use max sectors to limit number of bvecs(segments), isn't it?

The raid1 write behind code cares about the size ofa bio it can reach by adding order 0 pages to it. The bvecs are part of that and I think the calculation in the patch documents that a well.

Thinking of further, your and Guoqing's patch are correct & enough since bio_copy_data() just copies bytes(sectors) stream from fs bio to the write behind bio.

Thanks,
Ming

#11 lore raw

Hi all,

I just had the occasion to test the new patch as landed in arch linux 5.14.7. Unfortunately it does not work for me. Attached you can find a modification that works for me, though I am not really sure why write_behind seems not to be set to true on my configuration. If there is any more data I can provide to help you to investigate, please let me know.

Thanks for any clues,

Jens

My configuration:

[root@vdr jens]# mdadm --detail -v /dev/md0 /dev/md0:
           Version : 1.2
     Creation Time : Fri Dec 26 09:50:53 2014 Raid Level : raid1
        Array Size : 1953381440 (1862.89 GiB 2000.26 GB) Used Dev Size : 1953381440 (1862.89 GiB 2000.26 GB) Raid Devices : 2
     Total Devices : 2
       Persistence : Superblock is persistent

     Intent Bitmap : Internal

       Update Time : Fri Sep 24 17:30:51 2021
             State : active
    Active Devices : 2
   Working Devices : 2
    Failed Devices : 0
     Spare Devices : 0

Consistency Policy : bitmap

              Name : vdr:0  (local to host vdr) UUID : 5532ffda:ccbc790f:b50c4959:8f0fd43f Events : 32805

    Number   Major   Minor   RaidDevice State
       2       8       33        0      active sync   /dev/sdc1 3       8       17        1      active sync   /dev/sdb1

[root@vdr jens]# mdadm -X /dev/sdb1
        Filename : /dev/sdb1
           Magic : 6d746962
         Version : 4
            UUID : 5532ffda:ccbc790f:b50c4959:8f0fd43f Events : 32804
  Events Cleared : 32804
           State : OK
       Chunksize : 64 MB
          Daemon : 5s flush period
      Write Mode : Allow write behind, max 4096 Sync Size : 1953381440 (1862.89 GiB 2000.26 GB) Bitmap : 29807 bits (chunks), 3 dirty (0.0%)

[root@vdr jens]# mdadm -X /dev/sdc1
        Filename : /dev/sdc1
           Magic : 6d746962
         Version : 4
            UUID : 5532ffda:ccbc790f:b50c4959:8f0fd43f Events : 32804
  Events Cleared : 32804
           State : OK
       Chunksize : 64 MB
          Daemon : 5s flush period
      Write Mode : Allow write behind, max 4096 Sync Size : 1953381440 (1862.89 GiB 2000.26 GB) Bitmap : 29807 bits (chunks), 3 dirty (0.0%)

Am 17.08.21 um 14:32 schrieb Ming Lei:

Quoted 13 lines · Christoph Hellwig
Quoted 12 lines · Ming Lei
Quoted 8 lines
Quoted 7 lines
Quoted 5 lines

0 ~ 254: each bvec's length is 512
255: bvec's length is 8192

the total length is just 512*255 + 8192 = 138752 bytes = 271 sectors, but it still may need 257 bvecs, which can't be allocated via bio_alloc_bioset().

Yes, we still need the rounding magic that alloc_behind_master_bio uses here.

But it is wrong to use max sectors to limit number of bvecs(segments), isn't it?

The raid1 write behind code cares about the size ofa bio it can reach by adding order 0 pages to it. The bvecs are part of that and I think the calculation in the patch documents that a well.

Quoted 8 lines

Thinking of further, your and Guoqing's patch are correct & enough since bio_copy_data() just copies bytes(sectors) stream from fs bio to the write behind bio.

Thanks,
Ming

drivers/md/raid1.c +1 -1
diff --unified --recursive --text archlinux-linux/drivers/md/raid1.c archlinux-linux-diff/drivers/md/raid1.c
@@ -1501,7 +1501,7 @@
 			 * Not if there are too many, or cannot
 			 * allocate memory, or a reader on WriteMostly
 			 * is waiting for behind writes to flush */
-			if (bitmap &&
+			if (bitmap && write_behind &&
 			    (atomic_read(&bitmap->behind_writes)
 			     < mddev->bitmap_info.max_write_behind) &&
 			    !waitqueue_active(&bitmap->behind_wait)) {
drivers/md/raid1.c +2 -11
diff --unified --recursive --text archlinux-linux/drivers/md/raid1.c archlinux-linux-changed/drivers/md/raid1.c
@@ -1329,7 +1329,6 @@
 	struct raid1_plug_cb *plug = NULL;
 	int first_clone;
 	int max_sectors;
-	bool write_behind = false;
 
 	if (mddev_is_clustered(mddev) &&
 	     md_cluster_ops->area_resyncing(mddev, WRITE,
@@ -1383,14 +1382,6 @@
 	for (i = 0;  i < disks; i++) {
 		struct md_rdev *rdev = rcu_dereference(conf->mirrors[i].rdev);
 
-		/*
-		 * The write-behind io is only attempted on drives marked as
-		 * write-mostly, which means we could allocate write behind
-		 * bio later.
-		 */
-		if (rdev && test_bit(WriteMostly, &rdev->flags))
-			write_behind = true;
-
 		if (rdev && unlikely(test_bit(Blocked, &rdev->flags))) {
 			atomic_inc(&rdev->nr_pending);
 			blocked_rdev = rdev;
@@ -1470,7 +1461,7 @@
 	 * at a time and thus needs a new bio that can fit the whole payload
 	 * this bio in page sized chunks.
 	 */
-	if (write_behind && bitmap)
+	if (bitmap)
 		max_sectors = min_t(int, max_sectors,
 				    BIO_MAX_VECS * (PAGE_SIZE >> 9));
 	if (max_sectors < bio_sectors(bio)) {
@@ -1501,7 +1492,7 @@
 			 * Not if there are too many, or cannot
 			 * allocate memory, or a reader on WriteMostly
 			 * is waiting for behind writes to flush */
-			if (bitmap &&
+			if (bitmap && 
 			    (atomic_read(&bitmap->behind_writes)
 			     < mddev->bitmap_info.max_write_behind) &&
 			    !waitqueue_active(&bitmap->behind_wait)) {
#12 lore raw
Quoted 3 lines · Jens Stutte (Archiv)

Hi all,

I just had the occasion to test the new patch as landed in arch linux 5.14.7. Unfortunately it does not work for me. Attached you can find a modification that works for me, though I am not really sure why write_behind seems not to be set to true on my configuration. If there is any more data I can provide to help you to investigate, please let me know.

Thanks for the report!  As commented in bugzilla, this is because write-behind IO still happens even without write-mostly device. I will send new patch after you confirm it works.

1. https://bugzilla.kernel.org/show_bug.cgi?id=213181

Thanks,
Guoqing

#13 lore raw
In reply to Guoqing Jiang
Thank you for the patch! Yet something to improve:

Hi Guoqing,

Thank you for the patch! Yet something to improve:

[auto build test ERROR on song-md/md-next]
[also build test ERROR on v5.14-rc5 next-20210812]
[If your patch is applied to the wrong git tree, kindly drop us a note. And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]

url:    https://github.com/0day-ci/linux/commits/Guoqing-Jiang/raid1-ensure-bio-doesn-t-have-more-than-BIO_MAX_VECS-sectors/20210813-140810
base:   git://git.kernel.org/pub/scm/linux/kernel/git/song/md.git md-next
config: hexagon-randconfig-r001-20210813 (attached as .config)
compiler: clang version 12.0.0

reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross chmod +x ~/bin/make.cross
# https://github.com/0day-ci/linux/commit/29b7720a83de1deea0d8ecfafe0db46146636b15
git remote add linux-review https://github.com/0day-ci/linux
git fetch --no-tags linux-review Guoqing-Jiang/raid1-ensure-bio-doesn-t-have-more-than-BIO_MAX_VECS-sectors/20210813-140810 git checkout 29b7720a83de1deea0d8ecfafe0db46146636b15
# save the attached .config to linux build tree
mkdir build_dir
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross O=build_dir ARCH=hexagon SHELL=/bin/bash drivers/md/

If you fix the issue, kindly add following tag as appropriate

All errors (new ones prefixed by >>):

>> drivers/md/raid1.c:1459:55: error: use of undeclared identifier 'PAGE_SECTORS'
           max_sectors = min_t(int, max_sectors, BIO_MAX_VECS * PAGE_SECTORS);
                                                                ^
>> drivers/md/raid1.c:1459:55: error: use of undeclared identifier 'PAGE_SECTORS'

2 errors generated.

vim +/PAGE_SECTORS +1459 drivers/md/raid1.c

1320
1321 static void raid1_write_request(struct mddev *mddev, struct bio *bio,
1322 int max_write_sectors)
1323 {
1324 struct r1conf *conf = mddev->private;
1325 struct r1bio *r1_bio;
1326 int i, disks;
1327 struct bitmap *bitmap = mddev->bitmap;
1328 unsigned long flags;
1329 struct md_rdev *blocked_rdev;
1330 struct blk_plug_cb *cb;
1331 struct raid1_plug_cb *plug = NULL;
1332 int first_clone;
1333 int max_sectors;
1334
1335 if (mddev_is_clustered(mddev) &&
1336 md_cluster_ops->area_resyncing(mddev, WRITE,
1337 bio->bi_iter.bi_sector, bio_end_sector(bio))) {
1338
1339 DEFINE_WAIT(w);
1340 for (;;) {
1341 prepare_to_wait(&conf->wait_barrier,
1342 &w, TASK_IDLE);
1343 if (!md_cluster_ops->area_resyncing(mddev, WRITE,
1344 bio->bi_iter.bi_sector,
1345 bio_end_sector(bio)))
1346 break;
1347 schedule();
1348 }
1349 finish_wait(&conf->wait_barrier, &w);
1350 }
1351
1352 /*
1353 * Register the new request and wait if the reconstruction
1354 * thread has put up a bar for new requests.
1355 * Continue immediately if no resync is active currently.
1356 */
1357 wait_barrier(conf, bio->bi_iter.bi_sector);
1358
1359 r1_bio = alloc_r1bio(mddev, bio);
1360 r1_bio->sectors = max_write_sectors;
1361
1362 if (conf->pending_count >= max_queued_requests) {
1363 md_wakeup_thread(mddev->thread);
1364 raid1_log(mddev, "wait queued");
1365 wait_event(conf->wait_barrier,
1366 conf->pending_count < max_queued_requests);
1367 }
1368 /* first select target devices under rcu_lock and
1369 * inc refcount on their rdev. Record them by setting
1370 * bios[x] to bio
1371 * If there are known/acknowledged bad blocks on any device on
1372 * which we have seen a write error, we want to avoid writing those
1373 * blocks.
1374 * This potentially requires several writes to write around
1375 * the bad blocks. Each set of writes gets it's own r1bio
1376 * with a set of bios attached.
1377 */
1378
1379 disks = conf->raid_disks * 2;
1380 retry_write:
1381 blocked_rdev = NULL;
1382 rcu_read_lock();
1383 max_sectors = r1_bio->sectors;
1384 for (i = 0; i < disks; i++) {
1385 struct md_rdev *rdev = rcu_dereference(conf->mirrors[i].rdev);
1386 if (rdev && unlikely(test_bit(Blocked, &rdev->flags))) {
1387 atomic_inc(&rdev->nr_pending);
1388 blocked_rdev = rdev;
1389 break;
1390 }
1391 r1_bio->bios[i] = NULL;
1392 if (!rdev || test_bit(Faulty, &rdev->flags)) {
1393 if (i < conf->raid_disks)
1394 set_bit(R1BIO_Degraded, &r1_bio->state);
1395 continue;
1396 }
1397
1398 atomic_inc(&rdev->nr_pending);
1399 if (test_bit(WriteErrorSeen, &rdev->flags)) {
1400 sector_t first_bad;
1401 int bad_sectors;
1402 int is_bad;
1403
1404 is_bad = is_badblock(rdev, r1_bio->sector, max_sectors,
1405 &first_bad, &bad_sectors);
1406 if (is_bad < 0) {
1407 /* mustn't write here until the bad block is
1408 * acknowledged*/
1409 set_bit(BlockedBadBlocks, &rdev->flags);
1410 blocked_rdev = rdev;
1411 break;
1412 }
1413 if (is_bad && first_bad <= r1_bio->sector) {
1414 /* Cannot write here at all */
1415 bad_sectors -= (r1_bio->sector - first_bad);
1416 if (bad_sectors < max_sectors)
1417 /* mustn't write more than bad_sectors
1418 * to other devices yet
1419 */
1420 max_sectors = bad_sectors;
1421 rdev_dec_pending(rdev, mddev);
1422 /* We don't set R1BIO_Degraded as that
1423 * only applies if the disk is
1424 * missing, so it might be re-added,
1425 * and we want to know to recover this
1426 * chunk.
1427 * In this case the device is here,
1428 * and the fact that this chunk is not
1429 * in-sync is recorded in the bad
1430 * block log
1431 */
1432 continue;
1433 }
1434 if (is_bad) {
1435 int good_sectors = first_bad - r1_bio->sector;
1436 if (good_sectors < max_sectors)
1437 max_sectors = good_sectors;
1438 }
1439 }
1440 r1_bio->bios[i] = bio;
1441 }
1442 rcu_read_unlock();
1443
1444 if (unlikely(blocked_rdev)) {
1445 /* Wait for this device to become unblocked */
1446 int j;
1447
1448 for (j = 0; j < i; j++)
1449 if (r1_bio->bios[j])
1450 rdev_dec_pending(conf->mirrors[j].rdev, mddev);
1451 r1_bio->state = 0;
1452 allow_barrier(conf, bio->bi_iter.bi_sector);
1453 raid1_log(mddev, "wait rdev %d blocked", blocked_rdev->raid_disk);
1454 md_wait_for_blocked_rdev(blocked_rdev, mddev);
1455 wait_barrier(conf, bio->bi_iter.bi_sector);
1456 goto retry_write;
1457 }
1458

> 1459		max_sectors = min_t(int, max_sectors, BIO_MAX_VECS * PAGE_SECTORS);

1460 if (max_sectors < bio_sectors(bio)) {
1461 struct bio *split = bio_split(bio, max_sectors,
1462 GFP_NOIO, &conf->bio_split);
1463 bio_chain(split, bio);
1464 submit_bio_noacct(bio);
1465 bio = split;
1466 r1_bio->master_bio = bio;
1467 r1_bio->sectors = max_sectors;
1468 }
1469
1470 if (blk_queue_io_stat(bio->bi_bdev->bd_disk->queue))
1471 r1_bio->start_time = bio_start_io_acct(bio);
1472 atomic_set(&r1_bio->remaining, 1);
1473 atomic_set(&r1_bio->behind_remaining, 0);
1474
1475 first_clone = 1;
1476
1477 for (i = 0; i < disks; i++) {
1478 struct bio *mbio = NULL;
1479 struct md_rdev *rdev = conf->mirrors[i].rdev;
1480 if (!r1_bio->bios[i])
1481 continue;
1482
1483 if (first_clone) {
1484 /* do behind I/O ?
1485 * Not if there are too many, or cannot
1486 * allocate memory, or a reader on WriteMostly
1487 * is waiting for behind writes to flush */
1488 if (bitmap &&
1489 (atomic_read(&bitmap->behind_writes)
1490 < mddev->bitmap_info.max_write_behind) &&
1491 !waitqueue_active(&bitmap->behind_wait)) {
1492 alloc_behind_master_bio(r1_bio, bio);
1493 }
1494
1495 md_bitmap_startwrite(bitmap, r1_bio->sector, r1_bio->sectors,
1496 test_bit(R1BIO_BehindIO, &r1_bio->state));
1497 first_clone = 0;
1498 }
1499
1500 if (r1_bio->behind_master_bio)
1501 mbio = bio_clone_fast(r1_bio->behind_master_bio,
1502 GFP_NOIO, &mddev->bio_set);
1503 else
1504 mbio = bio_clone_fast(bio, GFP_NOIO, &mddev->bio_set);
1505
1506 if (r1_bio->behind_master_bio) {
1507 if (test_bit(CollisionCheck, &rdev->flags))
1508 wait_for_serialization(rdev, r1_bio);
1509 if (test_bit(WriteMostly, &rdev->flags))
1510 atomic_inc(&r1_bio->behind_remaining);
1511 } else if (mddev->serialize_policy)
1512 wait_for_serialization(rdev, r1_bio);
1513
1514 r1_bio->bios[i] = mbio;
1515
1516 mbio->bi_iter.bi_sector = (r1_bio->sector +
1517 conf->mirrors[i].rdev->data_offset);
1518 bio_set_dev(mbio, conf->mirrors[i].rdev->bdev);
1519 mbio->bi_end_io = raid1_end_write_request;
1520 mbio->bi_opf = bio_op(bio) | (bio->bi_opf & (REQ_SYNC | REQ_FUA));
1521 if (test_bit(FailFast, &conf->mirrors[i].rdev->flags) &&
1522 !test_bit(WriteMostly, &conf->mirrors[i].rdev->flags) &&
1523 conf->raid_disks - mddev->degraded > 1)
1524 mbio->bi_opf |= MD_FAILFAST;
1525 mbio->bi_private = r1_bio;
1526
1527 atomic_inc(&r1_bio->remaining);
1528
1529 if (mddev->gendisk)
1530 trace_block_bio_remap(mbio, disk_devt(mddev->gendisk),
1531 r1_bio->sector);
1532 /* flush_pending_writes() needs access to the rdev so...*/
1533 mbio->bi_bdev = (void *)conf->mirrors[i].rdev;
1534
1535 cb = blk_check_plugged(raid1_unplug, mddev, sizeof(*plug));
1536 if (cb)
1537 plug = container_of(cb, struct raid1_plug_cb, cb);
1538 else
1539 plug = NULL;
1540 if (plug) {
1541 bio_list_add(&plug->pending, mbio);
1542 plug->pending_cnt++;
1543 } else {
1544 spin_lock_irqsave(&conf->device_lock, flags);
1545 bio_list_add(&conf->pending_bio_list, mbio);
1546 conf->pending_count++;
1547 spin_unlock_irqrestore(&conf->device_lock, flags);
1548 md_wakeup_thread(mddev->thread);
1549 }
1550 }
1551
1552 r1_bio_write_done(r1_bio);
1553
1554 /* In case raid1d snuck in to freeze_array */
1555 wake_up(&conf->wait_barrier);
1556 }
1557

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all(a)lists.01.org

#14 lore raw
In reply to Guoqing Jiang
Thank you for the patch! Yet something to improve:

Hi Guoqing,

Thank you for the patch! Yet something to improve:

[auto build test ERROR on song-md/md-next]
[also build test ERROR on v5.14-rc5 next-20210812]
[If your patch is applied to the wrong git tree, kindly drop us a note. And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]

url:    https://github.com/0day-ci/linux/commits/Guoqing-Jiang/raid1-ensure-bio-doesn-t-have-more-than-BIO_MAX_VECS-sectors/20210813-140810
base:   git://git.kernel.org/pub/scm/linux/kernel/git/song/md.git md-next
config: nds32-randconfig-r035-20210813 (attached as .config)
compiler: nds32le-linux-gcc (GCC) 10.3.0

reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross chmod +x ~/bin/make.cross
# https://github.com/0day-ci/linux/commit/29b7720a83de1deea0d8ecfafe0db46146636b15
git remote add linux-review https://github.com/0day-ci/linux
git fetch --no-tags linux-review Guoqing-Jiang/raid1-ensure-bio-doesn-t-have-more-than-BIO_MAX_VECS-sectors/20210813-140810 git checkout 29b7720a83de1deea0d8ecfafe0db46146636b15
# save the attached .config to linux build tree
mkdir build_dir
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-10.3.0 make.cross O=build_dir ARCH=nds32 SHELL=/bin/bash drivers/md/

If you fix the issue, kindly add following tag as appropriate

All errors (new ones prefixed by >>):

In file included from include/linux/kernel.h:15,

                    from include/asm-generic/bug.h:20,
                    from ./arch/nds32/include/generated/asm/bug.h:1,
                    from include/linux/bug.h:5,
                    from include/linux/mmdebug.h:5,
                    from include/linux/gfp.h:5,
                    from include/linux/slab.h:15,
                    from drivers/md/raid1.c:26:
   drivers/md/raid1.c: In function 'raid1_write_request':
>> drivers/md/raid1.c:1459:55: error: 'PAGE_SECTORS' undeclared (first use in this function); did you mean 'PAGE_MEMORY'?
    1459 |  max_sectors = min_t(int, max_sectors, BIO_MAX_VECS * PAGE_SECTORS);
         |                                                       ^~~~~~~~~~~~
   include/linux/minmax.h:20:39: note: in definition of macro '__typecheck'
      20 |  (!!(sizeof((typeof(x) *)1 == (typeof(y) *)1)))
         |                                       ^
   include/linux/minmax.h:36:24: note: in expansion of macro '__safe_cmp'
      36 |  __builtin_choose_expr(__safe_cmp(x, y), \
         |                        ^~~~~~~~~~
   include/linux/minmax.h:104:27: note: in expansion of macro '__careful_cmp'
     104 | #define min_t(type, x, y) __careful_cmp((type)(x), (type)(y), <)
         |                           ^~~~~~~~~~~~~
   drivers/md/raid1.c:1459:16: note: in expansion of macro 'min_t'
    1459 |  max_sectors = min_t(int, max_sectors, BIO_MAX_VECS * PAGE_SECTORS);
         |                ^~~~~
   drivers/md/raid1.c:1459:55: note: each undeclared identifier is reported only once for each function it appears in
    1459 |  max_sectors = min_t(int, max_sectors, BIO_MAX_VECS * PAGE_SECTORS);
         |                                                       ^~~~~~~~~~~~
   include/linux/minmax.h:20:39: note: in definition of macro '__typecheck'
      20 |  (!!(sizeof((typeof(x) *)1 == (typeof(y) *)1)))
         |                                       ^
   include/linux/minmax.h:36:24: note: in expansion of macro '__safe_cmp'
      36 |  __builtin_choose_expr(__safe_cmp(x, y), \
         |                        ^~~~~~~~~~
   include/linux/minmax.h:104:27: note: in expansion of macro '__careful_cmp'
     104 | #define min_t(type, x, y) __careful_cmp((type)(x), (type)(y), <)
         |                           ^~~~~~~~~~~~~
   drivers/md/raid1.c:1459:16: note: in expansion of macro 'min_t'
    1459 |  max_sectors = min_t(int, max_sectors, BIO_MAX_VECS * PAGE_SECTORS);
         |                ^~~~~
>> include/linux/minmax.h:36:2: error: first argument to '__builtin_choose_expr' not a constant
      36 |  __builtin_choose_expr(__safe_cmp(x, y), \
         |  ^~~~~~~~~~~~~~~~~~~~~
   include/linux/minmax.h:104:27: note: in expansion of macro '__careful_cmp'
     104 | #define min_t(type, x, y) __careful_cmp((type)(x), (type)(y), <)
         |                           ^~~~~~~~~~~~~
   drivers/md/raid1.c:1459:16: note: in expansion of macro 'min_t'
    1459 |  max_sectors = min_t(int, max_sectors, BIO_MAX_VECS * PAGE_SECTORS);
         |                ^~~~~

vim +1459 drivers/md/raid1.c

1320
1321 static void raid1_write_request(struct mddev *mddev, struct bio *bio,
1322 int max_write_sectors)
1323 {
1324 struct r1conf *conf = mddev->private;
1325 struct r1bio *r1_bio;
1326 int i, disks;
1327 struct bitmap *bitmap = mddev->bitmap;
1328 unsigned long flags;
1329 struct md_rdev *blocked_rdev;
1330 struct blk_plug_cb *cb;
1331 struct raid1_plug_cb *plug = NULL;
1332 int first_clone;
1333 int max_sectors;
1334
1335 if (mddev_is_clustered(mddev) &&
1336 md_cluster_ops->area_resyncing(mddev, WRITE,
1337 bio->bi_iter.bi_sector, bio_end_sector(bio))) {
1338
1339 DEFINE_WAIT(w);
1340 for (;;) {
1341 prepare_to_wait(&conf->wait_barrier,
1342 &w, TASK_IDLE);
1343 if (!md_cluster_ops->area_resyncing(mddev, WRITE,
1344 bio->bi_iter.bi_sector,
1345 bio_end_sector(bio)))
1346 break;
1347 schedule();
1348 }
1349 finish_wait(&conf->wait_barrier, &w);
1350 }
1351
1352 /*
1353 * Register the new request and wait if the reconstruction
1354 * thread has put up a bar for new requests.
1355 * Continue immediately if no resync is active currently.
1356 */
1357 wait_barrier(conf, bio->bi_iter.bi_sector);
1358
1359 r1_bio = alloc_r1bio(mddev, bio);
1360 r1_bio->sectors = max_write_sectors;
1361
1362 if (conf->pending_count >= max_queued_requests) {
1363 md_wakeup_thread(mddev->thread);
1364 raid1_log(mddev, "wait queued");
1365 wait_event(conf->wait_barrier,
1366 conf->pending_count < max_queued_requests);
1367 }
1368 /* first select target devices under rcu_lock and
1369 * inc refcount on their rdev. Record them by setting
1370 * bios[x] to bio
1371 * If there are known/acknowledged bad blocks on any device on
1372 * which we have seen a write error, we want to avoid writing those
1373 * blocks.
1374 * This potentially requires several writes to write around
1375 * the bad blocks. Each set of writes gets it's own r1bio
1376 * with a set of bios attached.
1377 */
1378
1379 disks = conf->raid_disks * 2;
1380 retry_write:
1381 blocked_rdev = NULL;
1382 rcu_read_lock();
1383 max_sectors = r1_bio->sectors;
1384 for (i = 0; i < disks; i++) {
1385 struct md_rdev *rdev = rcu_dereference(conf->mirrors[i].rdev);
1386 if (rdev && unlikely(test_bit(Blocked, &rdev->flags))) {
1387 atomic_inc(&rdev->nr_pending);
1388 blocked_rdev = rdev;
1389 break;
1390 }
1391 r1_bio->bios[i] = NULL;
1392 if (!rdev || test_bit(Faulty, &rdev->flags)) {
1393 if (i < conf->raid_disks)
1394 set_bit(R1BIO_Degraded, &r1_bio->state);
1395 continue;
1396 }
1397
1398 atomic_inc(&rdev->nr_pending);
1399 if (test_bit(WriteErrorSeen, &rdev->flags)) {
1400 sector_t first_bad;
1401 int bad_sectors;
1402 int is_bad;
1403
1404 is_bad = is_badblock(rdev, r1_bio->sector, max_sectors,
1405 &first_bad, &bad_sectors);
1406 if (is_bad < 0) {
1407 /* mustn't write here until the bad block is
1408 * acknowledged*/
1409 set_bit(BlockedBadBlocks, &rdev->flags);
1410 blocked_rdev = rdev;
1411 break;
1412 }
1413 if (is_bad && first_bad <= r1_bio->sector) {
1414 /* Cannot write here at all */
1415 bad_sectors -= (r1_bio->sector - first_bad);
1416 if (bad_sectors < max_sectors)
1417 /* mustn't write more than bad_sectors
1418 * to other devices yet
1419 */
1420 max_sectors = bad_sectors;
1421 rdev_dec_pending(rdev, mddev);
1422 /* We don't set R1BIO_Degraded as that
1423 * only applies if the disk is
1424 * missing, so it might be re-added,
1425 * and we want to know to recover this
1426 * chunk.
1427 * In this case the device is here,
1428 * and the fact that this chunk is not
1429 * in-sync is recorded in the bad
1430 * block log
1431 */
1432 continue;
1433 }
1434 if (is_bad) {
1435 int good_sectors = first_bad - r1_bio->sector;
1436 if (good_sectors < max_sectors)
1437 max_sectors = good_sectors;
1438 }
1439 }
1440 r1_bio->bios[i] = bio;
1441 }
1442 rcu_read_unlock();
1443
1444 if (unlikely(blocked_rdev)) {
1445 /* Wait for this device to become unblocked */
1446 int j;
1447
1448 for (j = 0; j < i; j++)
1449 if (r1_bio->bios[j])
1450 rdev_dec_pending(conf->mirrors[j].rdev, mddev);
1451 r1_bio->state = 0;
1452 allow_barrier(conf, bio->bi_iter.bi_sector);
1453 raid1_log(mddev, "wait rdev %d blocked", blocked_rdev->raid_disk);
1454 md_wait_for_blocked_rdev(blocked_rdev, mddev);
1455 wait_barrier(conf, bio->bi_iter.bi_sector);
1456 goto retry_write;
1457 }
1458

> 1459		max_sectors = min_t(int, max_sectors, BIO_MAX_VECS * PAGE_SECTORS);

1460 if (max_sectors < bio_sectors(bio)) {
1461 struct bio *split = bio_split(bio, max_sectors,
1462 GFP_NOIO, &conf->bio_split);
1463 bio_chain(split, bio);
1464 submit_bio_noacct(bio);
1465 bio = split;
1466 r1_bio->master_bio = bio;
1467 r1_bio->sectors = max_sectors;
1468 }
1469
1470 if (blk_queue_io_stat(bio->bi_bdev->bd_disk->queue))
1471 r1_bio->start_time = bio_start_io_acct(bio);
1472 atomic_set(&r1_bio->remaining, 1);
1473 atomic_set(&r1_bio->behind_remaining, 0);
1474
1475 first_clone = 1;
1476
1477 for (i = 0; i < disks; i++) {
1478 struct bio *mbio = NULL;
1479 struct md_rdev *rdev = conf->mirrors[i].rdev;
1480 if (!r1_bio->bios[i])
1481 continue;
1482
1483 if (first_clone) {
1484 /* do behind I/O ?
1485 * Not if there are too many, or cannot
1486 * allocate memory, or a reader on WriteMostly
1487 * is waiting for behind writes to flush */
1488 if (bitmap &&
1489 (atomic_read(&bitmap->behind_writes)
1490 < mddev->bitmap_info.max_write_behind) &&
1491 !waitqueue_active(&bitmap->behind_wait)) {
1492 alloc_behind_master_bio(r1_bio, bio);
1493 }
1494
1495 md_bitmap_startwrite(bitmap, r1_bio->sector, r1_bio->sectors,
1496 test_bit(R1BIO_BehindIO, &r1_bio->state));
1497 first_clone = 0;
1498 }
1499
1500 if (r1_bio->behind_master_bio)
1501 mbio = bio_clone_fast(r1_bio->behind_master_bio,
1502 GFP_NOIO, &mddev->bio_set);
1503 else
1504 mbio = bio_clone_fast(bio, GFP_NOIO, &mddev->bio_set);
1505
1506 if (r1_bio->behind_master_bio) {
1507 if (test_bit(CollisionCheck, &rdev->flags))
1508 wait_for_serialization(rdev, r1_bio);
1509 if (test_bit(WriteMostly, &rdev->flags))
1510 atomic_inc(&r1_bio->behind_remaining);
1511 } else if (mddev->serialize_policy)
1512 wait_for_serialization(rdev, r1_bio);
1513
1514 r1_bio->bios[i] = mbio;
1515
1516 mbio->bi_iter.bi_sector = (r1_bio->sector +
1517 conf->mirrors[i].rdev->data_offset);
1518 bio_set_dev(mbio, conf->mirrors[i].rdev->bdev);
1519 mbio->bi_end_io = raid1_end_write_request;
1520 mbio->bi_opf = bio_op(bio) | (bio->bi_opf & (REQ_SYNC | REQ_FUA));
1521 if (test_bit(FailFast, &conf->mirrors[i].rdev->flags) &&
1522 !test_bit(WriteMostly, &conf->mirrors[i].rdev->flags) &&
1523 conf->raid_disks - mddev->degraded > 1)
1524 mbio->bi_opf |= MD_FAILFAST;
1525 mbio->bi_private = r1_bio;
1526
1527 atomic_inc(&r1_bio->remaining);
1528
1529 if (mddev->gendisk)
1530 trace_block_bio_remap(mbio, disk_devt(mddev->gendisk),
1531 r1_bio->sector);
1532 /* flush_pending_writes() needs access to the rdev so...*/
1533 mbio->bi_bdev = (void *)conf->mirrors[i].rdev;
1534
1535 cb = blk_check_plugged(raid1_unplug, mddev, sizeof(*plug));
1536 if (cb)
1537 plug = container_of(cb, struct raid1_plug_cb, cb);
1538 else
1539 plug = NULL;
1540 if (plug) {
1541 bio_list_add(&plug->pending, mbio);
1542 plug->pending_cnt++;
1543 } else {
1544 spin_lock_irqsave(&conf->device_lock, flags);
1545 bio_list_add(&conf->pending_bio_list, mbio);
1546 conf->pending_count++;
1547 spin_unlock_irqrestore(&conf->device_lock, flags);
1548 md_wakeup_thread(mddev->thread);
1549 }
1550 }
1551
1552 r1_bio_write_done(r1_bio);
1553
1554 /* In case raid1d snuck in to freeze_array */
1555 wake_up(&conf->wait_barrier);
1556 }
1557

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all(a)lists.01.org