Btrfs: force unplugs when switching from high to regular priority bios

Btrfs does bio submissions from a worker thread, and each device
has a list of high priority bios and regular priority bios.

Synchronous writes go to the high priority thread while async writes
go to regular list.  This commit brings back an explicit unplug
any time we switch from high to regular priority, which makes it
easier for the block layer to give us low latencies.

Signed-off-by: Chris Mason <chris.mason@oracle.com>

diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index 53875ae73ad4f634548e8d1d315056822c9cbd4b..3c5f2fcd82c1c0e9216a46964f73b3fd1fb97cef 100644 (file)
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -142,6 +142,7 @@ static noinline int run_scheduled_bios(struct btrfs_device *device)
        unsigned long limit;
        unsigned long last_waited = 0;
        int force_reg = 0;
+       int sync_pending;
        struct blk_plug plug;
 
        /*
@@ -229,6 +230,22 @@ loop_lock:
 
                BUG_ON(atomic_read(&cur->bi_cnt) == 0);
 
+               /*
+                * if we're doing the sync list, record that our
+                * plug has some sync requests on it
+                *
+                * If we're doing the regular list and there are
+                * sync requests sitting around, unplug before
+                * we add more
+                */
+               if (pending_bios == &device->pending_sync_bios) {
+                       sync_pending = 1;
+               } else if (sync_pending) {
+                       blk_finish_plug(&plug);
+                       blk_start_plug(&plug);
+                       sync_pending = 0;
+               }
+
                submit_bio(cur->bi_rw, cur);
                num_run++;
                batch_run++;