From aef1897cd36dcf5e296f1d2bae7e0d268561b685 Mon Sep 17 00:00:00 2001
From: Jianchao Wang <jianchao.w.wang@oracle.com>
Date: Tue, 12 Feb 2019 09:56:25 +0800
Subject: [PATCH 1/4] blk-mq: insert rq with DONTPREP to hctx dispatch list
when requeue
When requeue, if RQF_DONTPREP, rq has contained some driver
specific data, so insert it to hctx dispatch list to avoid any
merge. Take scsi as example, here is the trace event log (no
io scheduler, because RQF_STARTED would prevent merging),
kworker/0:1H-339 [000] ...1 2037.209289: block_rq_insert: 8,0 R 4096 () 32768 + 8 [kworker/0:1H]
scsi_inert_test-1987 [000] .... 2037.220465: block_bio_queue: 8,0 R 32776 + 8 [scsi_inert_test]
scsi_inert_test-1987 [000] ...2 2037.220466: block_bio_backmerge: 8,0 R 32776 + 8 [scsi_inert_test]
kworker/0:1H-339 [000] .... 2047.220913: block_rq_issue: 8,0 R 8192 () 32768 + 16 [kworker/0:1H]
scsi_inert_test-1996 [000] ..s1 2047.221007: block_rq_complete: 8,0 R () 32768 + 8 [0]
scsi_inert_test-1996 [000] .Ns1 2047.221045: block_rq_requeue: 8,0 R () 32776 + 8 [0]
kworker/0:1H-339 [000] ...1 2047.221054: block_rq_insert: 8,0 R 4096 () 32776 + 8 [kworker/0:1H]
kworker/0:1H-339 [000] ...1 2047.221056: block_rq_issue: 8,0 R 4096 () 32776 + 8 [kworker/0:1H]
scsi_inert_test-1986 [000] ..s1 2047.221119: block_rq_complete: 8,0 R () 32776 + 8 [0]
(32768 + 8) was requeued by scsi_queue_insert and had RQF_DONTPREP.
Then it was merged with (32776 + 8) and issued. Due to RQF_DONTPREP,
the sdb only contained the part of (32768 + 8), then only that part
was completed. The lucky thing was that scsi_io_completion detected
it and requeued the remaining part. So we didn't get corrupted data.
However, the requeue of (32776 + 8) is not expected.
Suggested-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Jianchao Wang <jianchao.w.wang@oracle.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
block/blk-mq.c | 12 ++++++++++--
1 file changed, 10 insertions(+), 2 deletions(-)
diff --git a/block/blk-mq.c b/block/blk-mq.c
index 8f5b533764ca..9437a5eb07cf 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -737,12 +737,20 @@ static void blk_mq_requeue_work(struct work_struct *work)
spin_unlock_irq(&q->requeue_lock);
list_for_each_entry_safe(rq, next, &rq_list, queuelist) {
- if (!(rq->rq_flags & RQF_SOFTBARRIER))
+ if (!(rq->rq_flags & (RQF_SOFTBARRIER | RQF_DONTPREP)))
continue;
rq->rq_flags &= ~RQF_SOFTBARRIER;
list_del_init(&rq->queuelist);
- blk_mq_sched_insert_request(rq, true, false, false);
+ /*
+ * If RQF_DONTPREP, rq has contained some driver specific
+ * data, so insert it to hctx dispatch list to avoid any
+ * merge.
+ */
+ if (rq->rq_flags & RQF_DONTPREP)
+ blk_mq_request_bypass_insert(rq, false);
+ else
+ blk_mq_sched_insert_request(rq, true, false, false);
}
while (!list_empty(&rq_list)) {
From 4726bcf30fad37cc555cd9dcd6c73f2b2668c879 Mon Sep 17 00:00:00 2001
From: Keith Busch <keith.busch@intel.com>
Date: Mon, 11 Feb 2019 09:23:50 -0700
Subject: [PATCH 2/4] nvme-pci: add missing unlock for reset error
The reset work holds a mutex to prevent races with removal modifying the
same resources, but was unlocking only on success. Unlock on failure
too.
Fixes: 5c959d73dba64 ("nvme-pci: fix rapid add remove sequence")
Signed-off-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
drivers/nvme/host/pci.c | 8 +++++---
1 file changed, 5 insertions(+), 3 deletions(-)
diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
index 022ea1ee63f8..7fee665ec45e 100644
--- a/drivers/nvme/host/pci.c
+++ b/drivers/nvme/host/pci.c
@@ -2560,15 +2560,15 @@ static void nvme_reset_work(struct work_struct *work)
mutex_lock(&dev->shutdown_lock);
result = nvme_pci_enable(dev);
if (result)
- goto out;
+ goto out_unlock;
result = nvme_pci_configure_admin_queue(dev);
if (result)
- goto out;
+ goto out_unlock;
result = nvme_alloc_admin_tags(dev);
if (result)
- goto out;
+ goto out_unlock;
/*
* Limit the max command size to prevent iod->sg allocations going
@@ -2651,6 +2651,8 @@ static void nvme_reset_work(struct work_struct *work)
nvme_start_ctrl(&dev->ctrl);
return;
+ out_unlock:
+ mutex_unlock(&dev->shutdown_lock);
out:
nvme_remove_dead_ctrl(dev, result);
}
From 96d7cb932e826219ec41ac02e5af037ffae6098c Mon Sep 17 00:00:00 2001
From: Yufen Yu <yuyufen@huawei.com>
Date: Tue, 29 Jan 2019 16:34:04 +0800
Subject: [PATCH 3/4] floppy: check_events callback should not return a
negative number
floppy_check_events() is supposed to return bit flags to say which
events occured. We should return zero to say that no event flags are
set. Only BIT(0) and BIT(1) are used in the caller. And .check_events
interface also expect to return an unsigned int value.
However, after commit a0c80efe5956, it may return -EINTR (-4u).
Here, both BIT(0) and BIT(1) are cleared. So this patch shouldn't
affect runtime, but it obviously is still worth fixing.
Reviewed-by: Dan Carpenter <dan.carpenter@oracle.com>
Fixes: a0c80efe5956 ("floppy: fix lock_fdc() signal handling")
Signed-off-by: Yufen Yu <yuyufen@huawei.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
drivers/block/floppy.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/block/floppy.c b/drivers/block/floppy.c
index 6f2856c6d0f2..55481b40df9a 100644
--- a/drivers/block/floppy.c
+++ b/drivers/block/floppy.c
@@ -4075,7 +4075,7 @@ static unsigned int floppy_check_events(struct gendisk *disk,
if (time_after(jiffies, UDRS->last_checked + UDP->checkfreq)) {
if (lock_fdc(drive))
- return -EINTR;
+ return 0;
poll_drive(false, 0);
process_fd_request();
}
From dfcc34c99f3ebc16b787b118763bf9cb6b1efc7a Mon Sep 17 00:00:00 2001
From: Nate Dailey <nate.dailey@stratus.com>
Date: Thu, 7 Feb 2019 14:19:01 -0500
Subject: [PATCH 4/4] md/raid1: don't clear bitmap bits on interrupted
recovery.
sync_request_write no longer submits writes to a Faulty device. This has
the unfortunate side effect that bitmap bits can be incorrectly cleared
if a recovery is interrupted (previously, end_sync_write would have
prevented this). This means the next recovery may not copy everything
it should, potentially corrupting data.
Add a function for doing the proper md_bitmap_end_sync, called from
end_sync_write and the Faulty case in sync_request_write.
backport note to 4.14: s/md_bitmap_end_sync/bitmap_end_sync
Cc: stable@vger.kernel.org 4.14+
Fixes: 0c9d5b127f69 ("md/raid1: avoid reusing a resync bio after error handling.")
Reviewed-by: Jack Wang <jinpu.wang@cloud.ionos.com>
Tested-by: Jack Wang <jinpu.wang@cloud.ionos.com>
Signed-off-by: Nate Dailey <nate.dailey@stratus.com>
Signed-off-by: Song Liu <songliubraving@fb.com>
---
drivers/md/raid1.c | 28 ++++++++++++++++++----------
1 file changed, 18 insertions(+), 10 deletions(-)
diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c
index 1d54109071cc..fa47249fa3e4 100644
--- a/drivers/md/raid1.c
+++ b/drivers/md/raid1.c
@@ -1863,6 +1863,20 @@ static void end_sync_read(struct bio *bio)
reschedule_retry(r1_bio);
}
+static void abort_sync_write(struct mddev *mddev, struct r1bio *r1_bio)
+{
+ sector_t sync_blocks = 0;
+ sector_t s = r1_bio->sector;
+ long sectors_to_go = r1_bio->sectors;
+
+ /* make sure these bits don't get cleared. */
+ do {
+ md_bitmap_end_sync(mddev->bitmap, s, &sync_blocks, 1);
+ s += sync_blocks;
+ sectors_to_go -= sync_blocks;
+ } while (sectors_to_go > 0);
+}
+
static void end_sync_write(struct bio *bio)
{
int uptodate = !bio->bi_status;
@@ -1874,15 +1888,7 @@ static void end_sync_write(struct bio *bio)
struct md_rdev *rdev = conf->mirrors[find_bio_disk(r1_bio, bio)].rdev;
if (!uptodate) {
- sector_t sync_blocks = 0;
- sector_t s = r1_bio->sector;
- long sectors_to_go = r1_bio->sectors;
- /* make sure these bits doesn't get cleared. */
- do {
- md_bitmap_end_sync(mddev->bitmap, s, &sync_blocks, 1);
- s += sync_blocks;
- sectors_to_go -= sync_blocks;
- } while (sectors_to_go > 0);
+ abort_sync_write(mddev, r1_bio);
set_bit(WriteErrorSeen, &rdev->flags);
if (!test_and_set_bit(WantReplacement, &rdev->flags))
set_bit(MD_RECOVERY_NEEDED, &
@@ -2172,8 +2178,10 @@ static void sync_request_write(struct mddev *mddev, struct r1bio *r1_bio)
(i == r1_bio->read_disk ||
!test_bit(MD_RECOVERY_SYNC, &mddev->recovery))))
continue;
- if (test_bit(Faulty, &conf->mirrors[i].rdev->flags))
+ if (test_bit(Faulty, &conf->mirrors[i].rdev->flags)) {
+ abort_sync_write(mddev, r1_bio);
continue;
+ }
bio_set_op_attrs(wbio, REQ_OP_WRITE, 0);
if (test_bit(FailFast, &conf->mirrors[i].rdev->flags))