Re: [i-g-t v6 1/4] lib/xe: Fix a comment error

["Dixit, Ashutosh" <ashutosh.dixit@intel.com>]

On Mon, 10 Mar 2025 17:03:03 -0700, Dixit, Ashutosh wrote:
>
> On Mon, 24 Feb 2025 09:40:07 -0800, Oak Zeng wrote:
> >
> > The timeout value of __xe_wait_ufence returns the
> > remaining time, not elapsed time. Fix it.
> >
> > Signed-off-by: Oak Zeng <oak.zeng@intel.com>
> > ---
> >  lib/xe/xe_ioctl.c | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/lib/xe/xe_ioctl.c b/lib/xe/xe_ioctl.c
> > index 6d8388918..028102aa2 100644
> > --- a/lib/xe/xe_ioctl.c
> > +++ b/lib/xe/xe_ioctl.c
> > @@ -496,7 +496,7 @@ void xe_exec_wait(int fd, uint32_t exec_queue, uint64_t addr)
> >   *
> >   * Function compares @value with memory pointed by @addr until they are equal.
> >   *
> > - * Returns (in @timeout), the elapsed time in nanoseconds if user fence was
> > + * Returns (in @timeout), the remaining time in nanoseconds if user fence was
>
> But the code in xe_wait_user_fence_ioctl() in the kernel is:
>
>	args->timeout -= ktime_to_ns(ktime_sub(ktime_get(), start));
>
> Which seems to be the elapsed, not the remaining time?

Sorry, I missed the '-=', so you are right, it is the remaing time. So
let's change the comment in xe_wait_ufence() too.

>
> Also, if this patch is correct, then let's change for xe_wait_ufence() too.
>
> >   * signalled. Returns 0 on success, -errno of ioctl on error.
> >   */
> >  int __xe_wait_ufence(int fd, uint64_t *addr, uint64_t value,
> > --
> > 2.26.3
>
> Need to update your git too, though not for this patch :/