[PATCH] block, bfq: Init saved_wr_start_at_switch_to_srt in unlikely case

[PATCH] block, bfq: Init saved_wr_start_at_switch_to_srt in unlikely case

[Douglas Anderson]

Some debug code suggested by Paolo was tripping when I did reboot
stress tests.  Specifically in bfq_bfqq_resume_state()
"bic->saved_wr_start_at_switch_to_srt" was later than the current
value of "jiffies".  A bit of debugging showed that
"bic->saved_wr_start_at_switch_to_srt" was actually 0 and a bit more
debugging showed that was because we had run through the "unlikely"
case in the bfq_bfqq_save_state() function.

Let's init "saved_wr_start_at_switch_to_srt" in the unlikely case to
something sane.

NOTE: this fixes no known real-world errors.

Signed-off-by: Douglas Anderson <dianders@chromium.org>
Reviewed-by: Paolo Valente <paolo.valente@linaro.org>
---
Paolo said to add his Reviewed-by in https://crrev.com/c/1678756.

 block/bfq-iosched.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/block/bfq-iosched.c b/block/bfq-iosched.c
index 72840ebf953e..008c93d6b8d7 100644
--- a/block/bfq-iosched.c
+++ b/block/bfq-iosched.c
@@ -2678,6 +2678,7 @@ static void bfq_bfqq_save_state(struct bfq_queue *bfqq)
 		 * to enjoy weight raising if split soon.
 		 */
 		bic->saved_wr_coeff = bfqq->bfqd->bfq_wr_coeff;
+		bic->saved_wr_start_at_switch_to_srt = bfq_smallest_from_now();
 		bic->saved_wr_cur_max_time = bfq_wr_duration(bfqq->bfqd);
 		bic->saved_last_wr_start_finish = jiffies;
 	} else {
-- 
2.22.0.410.gd8fdbe21b5-goog

Re: [PATCH] block, bfq: Init saved_wr_start_at_switch_to_srt in unlikely case

[Guenter Roeck]

On Wed, Jun 26, 2019 at 1:00 PM Douglas Anderson <dianders@chromium.org> wrote:
>
> Some debug code suggested by Paolo was tripping when I did reboot
> stress tests.  Specifically in bfq_bfqq_resume_state()
> "bic->saved_wr_start_at_switch_to_srt" was later than the current
> value of "jiffies".  A bit of debugging showed that
> "bic->saved_wr_start_at_switch_to_srt" was actually 0 and a bit more
> debugging showed that was because we had run through the "unlikely"
> case in the bfq_bfqq_save_state() function.
>
> Let's init "saved_wr_start_at_switch_to_srt" in the unlikely case to
> something sane.
>
> NOTE: this fixes no known real-world errors.
>
> Signed-off-by: Douglas Anderson <dianders@chromium.org>
> Reviewed-by: Paolo Valente <paolo.valente@linaro.org>

Reviewed-by: Guenter Roeck <groeck@chromium.org>

>
> ---
> Paolo said to add his Reviewed-by in https://crrev.com/c/1678756.
>
>  block/bfq-iosched.c | 1 +
>  1 file changed, 1 insertion(+)
>
> diff --git a/block/bfq-iosched.c b/block/bfq-iosched.c
> index 72840ebf953e..008c93d6b8d7 100644
> --- a/block/bfq-iosched.c
> +++ b/block/bfq-iosched.c
> @@ -2678,6 +2678,7 @@ static void bfq_bfqq_save_state(struct bfq_queue *bfqq)
>                  * to enjoy weight raising if split soon.
>                  */
>                 bic->saved_wr_coeff = bfqq->bfqd->bfq_wr_coeff;
> +               bic->saved_wr_start_at_switch_to_srt = bfq_smallest_from_now();
>                 bic->saved_wr_cur_max_time = bfq_wr_duration(bfqq->bfqd);
>                 bic->saved_last_wr_start_finish = jiffies;
>         } else {
> --
> 2.22.0.410.gd8fdbe21b5-goog
>

Re: [PATCH] block, bfq: Init saved_wr_start_at_switch_to_srt in unlikely case

[Jens Axboe]

On 6/26/19 1:59 PM, Douglas Anderson wrote:
> Some debug code suggested by Paolo was tripping when I did reboot
> stress tests.  Specifically in bfq_bfqq_resume_state()
> "bic->saved_wr_start_at_switch_to_srt" was later than the current
> value of "jiffies".  A bit of debugging showed that
> "bic->saved_wr_start_at_switch_to_srt" was actually 0 and a bit more
> debugging showed that was because we had run through the "unlikely"
> case in the bfq_bfqq_save_state() function.
> 
> Let's init "saved_wr_start_at_switch_to_srt" in the unlikely case to
> something sane.
> 
> NOTE: this fixes no known real-world errors.

Applied, thanks.

-- 
Jens Axboe