[PATCH v2 0/2] fix wrong value when user modify nr_request by sysfs

[PATCH v2 0/2] fix wrong value when user modify nr_request by sysfs

[weiping zhang]

Hi Jens,

This is v2 of fixing nr_request.

v1 -> v2:

blk-mq: fix nr_requests wrong value when modify it from sysfs
this patch return -EINVAL when user write a value that's too large.

blk-sysfs: return EINVAL when user modify nr_request less than
  BLKDEV_MIN_RQ
In order to keep same behavior with former patch, also return EINVAL 
when user write a value less than BLKDEV_MIN_RQ

weiping zhang (2):
  blk-mq: fix nr_requests wrong value when modify it from sysfs
  blk-sysfs: return EINVAL when user modify nr_request less than
    BLKDEV_MIN_RQ

 block/blk-mq.c    | 8 ++++++--
 block/blk-sysfs.c | 2 +-
 2 files changed, 7 insertions(+), 3 deletions(-)

-- 
2.9.4

[PATCH v2 1/2] blk-mq: fix nr_requests wrong value when modify it from sysfs

[weiping zhang]

if blk-mq use "none" io scheduler, nr_request get a wrong value when
input a number > tag_set->queue_depth. blk_mq_tag_update_depth will get
the smaller one min(nr, set->queue_depth), and then q->nr_request get a
wrong value.

Reproduce:

echo none > /sys/block/nvme0n1/queue/ioscheduler
echo 1000000 > /sys/block/nvme0n1/queue/nr_requests
cat /sys/block/nvme0n1/queue/nr_requests
1000000

Signed-off-by: weiping zhang <zhangweiping@didichuxing.com>
---
 block/blk-mq.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/block/blk-mq.c b/block/blk-mq.c
index 98a1860..479c35a 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -2642,8 +2642,12 @@ int blk_mq_update_nr_requests(struct request_queue *q, unsigned int nr)
 		 * queue depth. This is similar to what the old code would do.
 		 */
 		if (!hctx->sched_tags) {
-			ret = blk_mq_tag_update_depth(hctx, &hctx->tags,
-							min(nr, set->queue_depth),
+			if (nr > set->queue_depth) {
+				ret = -EINVAL;
+				break;
+			}
+
+			ret = blk_mq_tag_update_depth(hctx, &hctx->tags, nr,
 							false);
 		} else {
 			ret = blk_mq_tag_update_depth(hctx, &hctx->sched_tags,
-- 
2.9.4

[PATCH v2 2/2] blk-sysfs: return EINVAL when user modify nr_request less than BLKDEV_MIN_RQ

[weiping zhang]

Signed-off-by: weiping zhang <zhangweiping@didichuxing.com>
---
 block/blk-sysfs.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/block/blk-sysfs.c b/block/blk-sysfs.c
index b8362c0..03a6e19 100644
--- a/block/blk-sysfs.c
+++ b/block/blk-sysfs.c
@@ -75,7 +75,7 @@ queue_requests_store(struct request_queue *q, const char *page, size_t count)
 		return ret;
 
 	if (nr < BLKDEV_MIN_RQ)
-		nr = BLKDEV_MIN_RQ;
+		return -EINVAL;
 
 	if (q->request_fn)
 		err = blk_update_nr_requests(q, nr);
-- 
2.9.4

Re: [PATCH v2 1/2] blk-mq: fix nr_requests wrong value when modify it from sysfs

[Jens Axboe]

On 09/21/2017 09:17 AM, weiping zhang wrote:
> if blk-mq use "none" io scheduler, nr_request get a wrong value when
> input a number > tag_set->queue_depth. blk_mq_tag_update_depth will get
> the smaller one min(nr, set->queue_depth), and then q->nr_request get a
> wrong value.
> 
> Reproduce:
> 
> echo none > /sys/block/nvme0n1/queue/ioscheduler
> echo 1000000 > /sys/block/nvme0n1/queue/nr_requests
> cat /sys/block/nvme0n1/queue/nr_requests
> 1000000
> 
> Signed-off-by: weiping zhang <zhangweiping@didichuxing.com>
> ---
>  block/blk-mq.c | 8 ++++++--
>  1 file changed, 6 insertions(+), 2 deletions(-)
> 
> diff --git a/block/blk-mq.c b/block/blk-mq.c
> index 98a1860..479c35a 100644
> --- a/block/blk-mq.c
> +++ b/block/blk-mq.c
> @@ -2642,8 +2642,12 @@ int blk_mq_update_nr_requests(struct request_queue *q, unsigned int nr)
>  		 * queue depth. This is similar to what the old code would do.
>  		 */
>  		if (!hctx->sched_tags) {
> -			ret = blk_mq_tag_update_depth(hctx, &hctx->tags,
> -							min(nr, set->queue_depth),
> +			if (nr > set->queue_depth) {
> +				ret = -EINVAL;
> +				break;
> +			}
> +
> +			ret = blk_mq_tag_update_depth(hctx, &hctx->tags, nr,
>  							false);

What am I missing here? blk_mq_tag_update_depth() should already return
-EINVAL for the case where we can't grow the tags. Looks like this patch
should simply remove the min(nr, set->queue_depth) and just pass in 'nr'.
Should not need the duplicated check for depth.

-- 
Jens Axboe

Re: [PATCH v2 2/2] blk-sysfs: return EINVAL when user modify nr_request less than BLKDEV_MIN_RQ

[Jens Axboe]

On 09/21/2017 09:17 AM, weiping zhang wrote:
> Signed-off-by: weiping zhang <zhangweiping@didichuxing.com>
> ---
>  block/blk-sysfs.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/block/blk-sysfs.c b/block/blk-sysfs.c
> index b8362c0..03a6e19 100644
> --- a/block/blk-sysfs.c
> +++ b/block/blk-sysfs.c
> @@ -75,7 +75,7 @@ queue_requests_store(struct request_queue *q, const char *page, size_t count)
>  		return ret;
>  
>  	if (nr < BLKDEV_MIN_RQ)
> -		nr = BLKDEV_MIN_RQ;
> +		return -EINVAL;

This is potentially breaking existing scripts.

-- 
Jens Axboe

Re: [PATCH v2 1/2] blk-mq: fix nr_requests wrong value when modify it from sysfs

[weiping zhang]

On Thu, Sep 21, 2017 at 10:37:48AM -0600, Jens Axboe wrote:
> On 09/21/2017 09:17 AM, weiping zhang wrote:
> > if blk-mq use "none" io scheduler, nr_request get a wrong value when
> > input a number > tag_set->queue_depth. blk_mq_tag_update_depth will get
> > the smaller one min(nr, set->queue_depth), and then q->nr_request get a
> > wrong value.
> > 
> > Reproduce:
> > 
> > echo none > /sys/block/nvme0n1/queue/ioscheduler
> > echo 1000000 > /sys/block/nvme0n1/queue/nr_requests
> > cat /sys/block/nvme0n1/queue/nr_requests
> > 1000000
> > 
> > Signed-off-by: weiping zhang <zhangweiping@didichuxing.com>
> > ---
> >  block/blk-mq.c | 8 ++++++--
> >  1 file changed, 6 insertions(+), 2 deletions(-)
> > 
> > diff --git a/block/blk-mq.c b/block/blk-mq.c
> > index 98a1860..479c35a 100644
> > --- a/block/blk-mq.c
> > +++ b/block/blk-mq.c
> > @@ -2642,8 +2642,12 @@ int blk_mq_update_nr_requests(struct request_queue *q, unsigned int nr)
> >  		 * queue depth. This is similar to what the old code would do.
> >  		 */
> >  		if (!hctx->sched_tags) {
> > -			ret = blk_mq_tag_update_depth(hctx, &hctx->tags,
> > -							min(nr, set->queue_depth),
> > +			if (nr > set->queue_depth) {
> > +				ret = -EINVAL;
> > +				break;
> > +			}
> > +
> > +			ret = blk_mq_tag_update_depth(hctx, &hctx->tags, nr,
> >  							false);
> 
> What am I missing here? blk_mq_tag_update_depth() should already return
> -EINVAL for the case where we can't grow the tags. Looks like this patch
> should simply remove the min(nr, set->queue_depth) and just pass in 'nr'.
> Should not need the duplicated check for depth.
> 
Ya, you are right, I will send V3 later.

Thanks

Re: [PATCH v2 2/2] blk-sysfs: return EINVAL when user modify nr_request less than BLKDEV_MIN_RQ

[weiping zhang]

On Thu, Sep 21, 2017 at 10:38:22AM -0600, Jens Axboe wrote:
> On 09/21/2017 09:17 AM, weiping zhang wrote:
> > Signed-off-by: weiping zhang <zhangweiping@didichuxing.com>
> > ---
> >  block/blk-sysfs.c | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> > 
> > diff --git a/block/blk-sysfs.c b/block/blk-sysfs.c
> > index b8362c0..03a6e19 100644
> > --- a/block/blk-sysfs.c
> > +++ b/block/blk-sysfs.c
> > @@ -75,7 +75,7 @@ queue_requests_store(struct request_queue *q, const char *page, size_t count)
> >  		return ret;
> >  
> >  	if (nr < BLKDEV_MIN_RQ)
> > -		nr = BLKDEV_MIN_RQ;
> > +		return -EINVAL;
> 
> This is potentially breaking existing scripts.
> 
I just want this keep same behavior with the too larger case, If this
may make other side effect, I drop this patch.

Thanks
weiping