On Mon, Jul 28, 2025 at 1:09 AM Yu Kuai <yukuai@xxxxxxxxxx> wrote:
>
> Hi,
>
> 在 2025/7/28 0:47, Tang Yizhou 写道:
> > From: Tang Yizhou <yizhou.tang@xxxxxxxxxx>
> >
> > The symbol wb_window_usec cannot be found. Update the doc to reflect the
> > latest implementation, in other words, the cur_win_nsec member of struct
> > rq_wb.
> >
> > Signed-off-by: Tang Yizhou <yizhou.tang@xxxxxxxxxx>
> > ---
> > Documentation/ABI/stable/sysfs-block | 2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/Documentation/ABI/stable/sysfs-block b/Documentation/ABI/stable/sysfs-block
> > index 4ba771b56b3b..277d89815edd 100644
> > --- a/Documentation/ABI/stable/sysfs-block
> > +++ b/Documentation/ABI/stable/sysfs-block
> > @@ -731,7 +731,7 @@ Contact: linux-block@xxxxxxxxxxxxxxx
> > Description:
> > [RW] If the device is registered for writeback throttling, then
> > this file shows the target minimum read latency. If this latency
> > - is exceeded in a given window of time (see wb_window_usec), then
> > + is exceeded in a given window of time (see cur_win_nsec), then
> Is this a typo? Jan suggested curr_win_nsec from v1.
>
> BTW, I don't mind rename rwb->cur_win_nsec to curr_win_nsec as well.
Sorry, that was indeed a typo.
I checked the code, and now both cur_win_nsec and curr_win_nsec are
used. The latter was introduced by Ming in commit d19afebca476
("blk-wbt: export internal state via debugfs"). In the Linux kernel,
both 'cur' and 'curr' are commonly used as abbreviations for
'current'. If we were to unify the naming, I suspect it could spark
some debate. For now, I won’t pursue unified naming — in the next
version of the patch, I’ll change it to curr_win_nsec.
Thanks,
Yi
>
> Thanks,
> Kuai
>
> > the writeback throttling will start scaling back writes. Writing
> > a value of '0' to this file disables the feature. Writing a
> > value of '-1' to this file resets the value to the default
>
