Skip to contentSkip to conversation

get rid of unused request_queue field queue_wait

RFC PATCH In review Lou Langholtz 2 participants 3 messages Jul 28, 2003 Block layer

lore
Share & tools
lore
#0 lore raw

RFC PATCH 1/1 get rid of unused request_queue field queue_wait

Are we going to use the queue_wait field of struct request_queue someday? As of 2.6.0-test2, I don't see any use of it. If it's not needed anymore, the following patch gets rid of it. Tested this patch by compiling for i386 and also doing a grep through all .h and .c files to see if it's used somewhere else (admittedly weak).

drivers/block/ll_rw_blk.c +0 -1
diff -urN linux-2.6.0-test2/drivers/block/ll_rw_blk.c linux-2.6.0-test2-no_queue_wait/drivers/block/ll_rw_blk.c
@@ -225,7 +225,6 @@
 	 */
 	blk_queue_bounce_limit(q, BLK_BOUNCE_HIGH);
 
-	init_waitqueue_head(&q->queue_wait);
 	INIT_LIST_HEAD(&q->plug_list);
 }
 
include/linux/blkdev.h +0 -2
diff -urN linux-2.6.0-test2/include/linux/blkdev.h linux-2.6.0-test2-no_queue_wait/include/linux/blkdev.h
@@ -337,8 +337,6 @@
 	unsigned long		seg_boundary_mask;
 	unsigned int		dma_alignment;
 
-	wait_queue_head_t	queue_wait;
-
 	struct blk_queue_tag	*queue_tags;
 
 	/*
#1 lore raw
Quoted 5 lines · Lou Langholtz

Are we going to use the queue_wait field of struct request_queue someday? As of 2.6.0-test2, I don't see any use of it. If it's not needed anymore, the following patch gets rid of it. Tested this patch by compiling for i386 and also doing a grep through all .h and .c files to see if it's used somewhere else (admittedly weak).

It's a relic from before dynamic request allocation, for now it can definitely go.

Signature · 1 line
Jens Axboe
#2 lore raw
Quoted 9 lines · Jens Axboe
Quoted 5 lines · Lou Langholtz

Are we going to use the queue_wait field of struct request_queue someday? As of 2.6.0-test2, I don't see any use of it. If it's not needed anymore, the following patch gets rid of it. Tested this patch by compiling for i386 and also doing a grep through all .h and .c files to see if it's used somewhere else (admittedly weak).

It's a relic from before dynamic request allocation, for now it can definitely go.

Eh sorry, still very much jet lagged. It has nothing to do with request allocation. The point still stands though, it can be removed.

Signature · 1 line
Jens Axboe