2007/12/16, Jeff Roberson <[EMAIL PROTECTED]>:
> On Sat, 15 Dec 2007, Jeff Roberson wrote:
>
> > jeff 2007-12-15 23:13:32 UTC
> >
> > FreeBSD src repository
> >
> > Modified files:
> > sys/conf options
> > sys/kern kern_mutex.c kern_rwlock.c kern_sx.c
> > kern_thread.c sched_4bsd.c sched_ule.c
> > subr_lock.c
> > sys/sys _lock.h lock_profile.h proc.h sx.h
>
> Some of you may complain about the inclusion of lock_profile.h in
> sys/proc.h. I hope you will forgive this since lock profiling now no
> longer disturbs the ABI and will work with binary modules as well.
Hello,
+void
+lock_profile_obtain_lock_success(struct lock_object *lo, int contested,
+ uint64_t waittime, const char *file, int line)
+{
+ static int lock_prof_count;
+ struct lock_profile_object *l;
+ int spin;
+
+ /* don't reset the timer when/if recursing */
+ if (!lock_prof_enable || (lo->lo_flags & LO_NOPROFILE))
+ return;
+ if (lock_prof_skipcount &&
+ (++lock_prof_count % lock_prof_skipcount) == 0)
+ return;
+ spin = LOCK_CLASS(lo) == &lock_class_mtx_spin;
+ if (spin && lock_prof_skipspin == 1)
+ return;
I would really like to get ride of this kind of checks against
lock_class_* objects.
The correct thing to do here is to check for the LC_SPINLOCK in the
class flags (this is exactly why they have been introduced I think).
I have a question though.
Does this new implementation track adaptive spinning in locks?
I would have expected a different function for something like that as
"lock_profile_obtain_lock_spinning()" or similar in order to track
both spinning and sleeping times about the same locks, but looking at
function prototypes, at consumers and looking at members of struct
lock_prof I don't see any obvious evidence of it.
Thanks for this work, I really like that our ABI is no longer broken :)
Attilio
--
Peace can only be achieved by understanding - A. Einstein
_______________________________________________
cvs-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/cvs-all
To unsubscribe, send any mail to "[EMAIL PROTECTED]"
