On Thu, Feb 13, 2025 at 10:56 AM Yu Kuai <yukuai1@xxxxxxxxxxxxxxx> wrote:
>
> Hi,
>
> 在 2025/02/12 11:00, Tang Yizhou 写道:
> > From: Tang Yizhou <yizhou.tang@xxxxxxxxxx>
> >
> > wbt_wait() no longer uses a spinlock as a parameter. Update the
> > function comments accordingly.
> >
> > Additionally, revise other comments to ensure they align with the
> > actual implementation.
> >
> > Signed-off-by: Tang Yizhou <yizhou.tang@xxxxxxxxxx>
> > ---
> > block/blk-wbt.c | 17 +++++++----------
> > 1 file changed, 7 insertions(+), 10 deletions(-)
> >
> > diff --git a/block/blk-wbt.c b/block/blk-wbt.c
> > index 6dfc659d22e2..f1754d07f7e0 100644
> > --- a/block/blk-wbt.c
> > +++ b/block/blk-wbt.c
> > @@ -136,8 +136,9 @@ enum {
> > RWB_MIN_WRITE_SAMPLES = 3,
> >
> > /*
> > - * If we have this number of consecutive windows with not enough
> > - * information to scale up or down, scale up.
> > + * If we have this number of consecutive windows without enough
> > + * information to scale up or down, slowly return to center state
> > + * (step == 0).
> > */
> > RWB_UNKNOWN_BUMP = 5,
> > };
> > @@ -446,9 +447,9 @@ static void wb_timer_fn(struct blk_stat_callback *cb)
> > break;
> > case LAT_UNKNOWN_WRITES:
> > /*
> > - * We started a the center step, but don't have a valid
> > - * read/write sample, but we do have writes going on.
> > - * Allow step to go negative, to increase write perf.
> > + * We don't have a valid read/write sample, but we do have
> > + * writes going on. Allow step to go negative, to increase
> > + * write performance.
>
> Other than this clean up, the others are actually fix. Can you remove
> this one and change the title to "Fix some comments"?
>
> Thanks,
> Kuai
>
Will do.
Thanks
Yi
> > */
> > scale_up(rwb);
> > break;
> > @@ -638,11 +639,7 @@ static void wbt_cleanup(struct rq_qos *rqos, struct bio *bio)
> > __wbt_done(rqos, flags);
> > }
> >
> > -/*
> > - * May sleep, if we have exceeded the writeback limits. Caller can pass
> > - * in an irq held spinlock, if it holds one when calling this function.
> > - * If we do sleep, we'll release and re-grab it.
> > - */
> > +/* May sleep, if we have exceeded the writeback limits. */
> > static void wbt_wait(struct rq_qos *rqos, struct bio *bio)
> > {
> > struct rq_wb *rwb = RQWB(rqos);
> >
>
>
