Re: some small patches

I object to the sched_clock() change.  We've discussed this on threads_at_

On Thu, 17 Apr 2003, Julian Elischer wrote:

>
>
> On Thu, 17 Apr 2003, Andrew R. Reiter wrote:
>
> > On Thu, 17 Apr 2003, Julian Elischer wrote:
> >
> > :
> > :Here are two small patches they are pretty non-controversial in my
> > :opinion.
> > :
> > :they are part of a bigger patch, but I'd like to get them in separatly
> > :to simplify the bigger one.
> > :the first patch:
> > :http://www.freebsd.org/~julian/IDLETD.diff
> > :moves the flag that identifies a thread as being one of the idle threads
> > :from the KSE to the thread. The code that wants to know already
> > :has a thread pointer, but not the KSE pointer so this makes more sense.
> >
> > Is this correct?
> >
> > Index: sys/proc.h
> > ===================================================================
> > RCS file: /repos/projects/mirrored/freebsd/src/sys/sys/proc.h,v
> > retrieving revision 1.313
> > diff -u -r1.313 proc.h
> > --- sys/proc.h	2003/04/13 21:29:11	1.313
> > +++ sys/proc.h	2003/04/17 22:52:07
> > _at__at_ -349,6 +323,7 _at__at_
> >  #define	TDF_CAN_UNBIND	0x000004 /* Only temporarily bound. */
> >  #define	TDF_SINTR	0x000008 /* Sleep is interruptible. */
> >  #define	TDF_TIMEOUT	0x000010 /* Timing out during sleep. */
> > +#define	TDF_IDLETD	0x000040 /* This is an idle thread */
> >  #define	TDF_SELECT	0x000040 /* Selecting; wakeup/waiting
> > danger. */
> >  #define	TDF_CVWAITQ	0x000080 /* Thread is on a cv_waitq (not
> > slpq). */
> >  #define	TDF_UPCALLING	0x000100 /* This thread is doing an
> > upcall. */
> >
> >
> > Both TDF_IDLETD and TD_SELECT have the same value.
>
>
> uh, no, TDF_IDLETD should be 0x20
> I copied that line by hand from my test system because
> the diff from there is different (there are more changes frm which this
> was extracted) and flubbed it..  good catch.
> (That's why we do reviews right?) :-)
>
>
> >
> > Just curious.
> >
> > Cheers,
> > Andrew
> >
> >
> > :
> > :
> > :The second patch:
> > :http://www.freebsd.org/~julian/sched_clock.diff
> > :makes the sched_ API entrypoint sched_clock()
> > :take a thread argument instead of a KSE.
> > :Once again, the callers have the thread pointer and not the KSE pointer,
> > :and in fact they probably should not have the KSE pointer.
> > :
> > :
> > :anyone object to these patches?
> > :
> > :
> > :_______________________________________________
> > :freebsd-current_at_freebsd.org mailing list
> > :http://lists.freebsd.org/mailman/listinfo/freebsd-current
> > :To unsubscribe, send any mail to "freebsd-current-unsubscribe_at_freebsd.org"
> > :
> >
> > --
> > Andrew R. Reiter
> > arr_at_watson.org
> > arr_at_FreeBSD.org
> >
>