Re: [PATCH 5/7] blk-mq: initialize request on allocation

From: Ming Lei
Date: Thu Apr 17 2014 - 10:54:44 EST


On Mon, Apr 14, 2014 at 4:30 PM, Christoph Hellwig <hch@xxxxxx> wrote:
> If we want to share tag and request allocation between queues we cannot
> initialize the request at init/free time, but need to initialize it
> at allocation time as it might get used for different queues over its
> lifetime.

Could you explain the use pattern? Looks you mean there are
still users of the tag/req even after it is freed, that looks a bit
weird since the tag/req can still be reallocated in another path
after it is freed.

>
> Signed-off-by: Christoph Hellwig <hch@xxxxxx>
> ---
> block/blk-mq.c | 4 +---
> 1 file changed, 1 insertion(+), 3 deletions(-)
>
> diff --git a/block/blk-mq.c b/block/blk-mq.c
> index ef7acc0..1cb7618 100644
> --- a/block/blk-mq.c
> +++ b/block/blk-mq.c
> @@ -82,6 +82,7 @@ static struct request *__blk_mq_alloc_request(struct blk_mq_hw_ctx *hctx,
> tag = blk_mq_get_tag(hctx->tags, gfp, reserved);
> if (tag != BLK_MQ_TAG_FAIL) {
> rq = hctx->rqs[tag];
> + blk_rq_init(hctx->queue, rq);
> rq->tag = tag;
>
> return rq;
> @@ -254,9 +255,7 @@ static void __blk_mq_free_request(struct blk_mq_hw_ctx *hctx,
> const int tag = rq->tag;
> struct request_queue *q = rq->q;
>
> - blk_rq_init(hctx->queue, rq);
> blk_mq_put_tag(hctx->tags, tag);
> -
> blk_mq_queue_exit(q);
> }
>
> @@ -1100,7 +1099,6 @@ static int blk_mq_init_rq_map(struct blk_mq_hw_ctx *hctx,
> left -= to_do * rq_size;
> for (j = 0; j < to_do; j++) {
> hctx->rqs[i] = p;
> - blk_rq_init(hctx->queue, hctx->rqs[i]);
> if (reg->ops->init_request) {
> error = reg->ops->init_request(driver_data,
> hctx, hctx->rqs[i], i);
> --
> 1.7.10.4
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@xxxxxxxxxxxxxxx
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/


Thanks,
--
Ming Lei
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/