blkcg: implement REQ_CGROUP_PUNT

When a shared kthread needs to issue a bio for a cgroup, doing so
synchronously can lead to priority inversions as the kthread can be
trapped waiting for that cgroup.  This patch implements
REQ_CGROUP_PUNT flag which makes submit_bio() punt the actual issuing
to a dedicated per-blkcg work item to avoid such priority inversions.

This will be used to fix priority inversions in btrfs compression and
should be generally useful as we grow filesystem support for
comprehensive IO control.

Cc: Chris Mason <clm@fb.com>
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

1 files changed, 10 insertions, 3 deletions

diff --git a/include/linux/writeback.h b/include/linux/writeback.h
index e056a22075cf..8945aac31392 100644
--- a/include/linux/writeback.h
+++ b/include/linux/writeback.h
@@ -78,6 +78,8 @@ struct writeback_control {
          */
         unsigned no_cgroup_owner:1;
 
+        unsigned punt_to_cgroup:1;      /* cgrp punting, see __REQ_CGROUP_PUNT */
+
 #ifdef CONFIG_CGROUP_WRITEBACK
         struct bdi_writeback *wb;       /* wb this writeback is issued under */
         struct inode *inode;            /* inode being written out */
@@ -94,12 +96,17 @@ struct writeback_control {
 
 static inline int wbc_to_write_flags(struct writeback_control *wbc)
 {
+        int flags = 0;
+
+        if (wbc->punt_to_cgroup)
+                flags = REQ_CGROUP_PUNT;
+
         if (wbc->sync_mode == WB_SYNC_ALL)
-                return REQ_SYNC;
+                flags |= REQ_SYNC;
         else if (wbc->for_kupdate || wbc->for_background)
-                return REQ_BACKGROUND;
+                flags |= REQ_BACKGROUND;
 
-        return 0;
+        return flags;
 }
 
 static inline struct cgroup_subsys_state *