[PATCH] Fix current_io_context() vs set_task_ioprio() race

I know nothing about io scheduler, but I suspect set_task_ioprio() is not safe.

current_io_context() initializes "struct io_context", then sets ->io_context.
set_task_ioprio() running on another cpu may see the changes out of order, so
->set_ioprio(ioc) may use io_context which was not initialized properly.

Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Signed-off-by: Jens Axboe <axboe@suse.de>

2 files changed, 5 insertions, 0 deletions

diff --git a/block/ll_rw_blk.c b/block/ll_rw_blk.c
index 61d6b3c65b66..ddd9253f9d55 100644
--- a/block/ll_rw_blk.c
+++ b/block/ll_rw_blk.c
@@ -3628,6 +3628,8 @@ struct io_context *current_io_context(gfp_t gfp_flags)
                 ret->nr_batch_requests = 0; /* because this is 0 */
                 ret->aic = NULL;
                 ret->cic_root.rb_node = NULL;
+                /* make sure set_task_ioprio() sees the settings above */
+                smp_wmb();
                 tsk->io_context = ret;
         }
 
diff --git a/fs/ioprio.c b/fs/ioprio.c
index 3db31038e9ab..06578311c63f 100644
--- a/fs/ioprio.c
+++ b/fs/ioprio.c
@@ -44,6 +44,9 @@ static int set_task_ioprio(struct task_struct *task, int ioprio)
         task->ioprio = ioprio;
 
         ioc = task->io_context;
+        /* see wmb() in current_io_context() */
+        smp_read_barrier_depends();
+
         if (ioc && ioc->set_ioprio)
                 ioc->set_ioprio(ioc, ioprio);