blk-mq: Fix preempt count imbalance

Avoid that the following kernel bug gets triggered:

BUG: sleeping function called from invalid context at ./include/linux/buffer_head.h:349
in_atomic(): 1, irqs_disabled(): 0, pid: 8019, name: find
CPU: 10 PID: 8019 Comm: find Tainted: G        W I     4.11.0-rc4-dbg+ #2
Call Trace:
 dump_stack+0x68/0x93
 ___might_sleep+0x16e/0x230
 __might_sleep+0x4a/0x80
 __ext4_get_inode_loc+0x1e0/0x4e0
 ext4_iget+0x70/0xbc0
 ext4_iget_normal+0x2f/0x40
 ext4_lookup+0xb6/0x1f0
 lookup_slow+0x104/0x1e0
 walk_component+0x19a/0x330
 path_lookupat+0x4b/0x100
 filename_lookup+0x9a/0x110
 user_path_at_empty+0x36/0x40
 vfs_statx+0x67/0xc0
 SYSC_newfstatat+0x20/0x40
 SyS_newfstatat+0xe/0x10
 entry_SYSCALL_64_fastpath+0x18/0xad

This happens since the big if/else in blk_mq_make_request() doesn't
have final else section that also drops the ctx. Add that.

Fixes: b00c53e ("blk-mq: fix schedule-while-atomic with scheduler attached")
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Cc: Omar Sandoval <osandov@fb.com>

Added a bit more to the commit log.

Signed-off-by: Jens Axboe <axboe@fb.com>

block/blk-mq.c

@@ -1639,7 +1639,8 @@ static blk_qc_t blk_mq_make_request(struct request_queue *q, struct bio *bio)
 	} else if (!blk_mq_merge_queue_io(data.hctx, data.ctx, rq, bio)) {
 		blk_mq_put_ctx(data.ctx);
 		blk_mq_run_hw_queue(data.hctx, true);
-	}
+	} else
+		blk_mq_put_ctx(data.ctx);
 
 	return cookie;
 }