Messages in this thread | | | Date | Sun, 23 Nov 2008 17:32:40 +0100 | From | Ingo Molnar <> | Subject | Re: [Bug #12064] [regression] Measured 688 cycles TSC warp; marking TSC unstable |
| |
* Frans Pop <elendil@planet.nl> wrote:
> On Saturday 22 November 2008, Rafael J. Wysocki wrote: > > Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=12064 > > Date : 2008-11-16 19:27:43 (7 days old) > > References : http://marc.info/?l=linux-kernel&m=122686370308852&w=2 > > Looks to be fixed in -rc6: > > -Calibrating delay using timer specific routine.. 6198.05 BogoMIPS (lpj=12396107) > +Calibrating delay using timer specific routine.. 6399.82 BogoMIPS (lpj=12799647) > [...] > -checking TSC synchronization [CPU#0 -> CPU#1]: > -Measured 696 cycles TSC warp between CPUs, turning off TSC clock. > -Marking TSC unstable due to check_tsc_sync_source failed > +checking TSC synchronization [CPU#0 -> CPU#1]: passed. > > So, BogoMIPS value on CPU#1 is in line with CPU#0 again and TSC passes. > No idea what commit would be responsible.
this one from Venki fixed it:
| commit 93ce99e849433ede4ce8b410b749dc0cad1100b2 | Author: Venki Pallipadi <venkatesh.pallipadi@intel.com> | Date: Mon Nov 17 14:43:58 2008 -0800 | | x86: add rdtsc barrier to TSC sync check
Ingo
| |