Messages in this thread | | | Date | Wed, 1 Dec 2021 13:19:35 +0800 | From | Feng Tang <> | Subject | Re: [PATCH v3 2/2] x86/tsc: skip tsc watchdog checking for qualified platforms |
| |
On Tue, Nov 30, 2021 at 11:45:32PM -0500, Luming Yu wrote: > On Tue, Nov 16, 2021 at 11:18 PM Feng Tang <feng.tang@intel.com> wrote: > > > > There are cases that tsc clocksources are wrongly judged as unstable by > > clocksource watchdogs like hpet, acpi_pm or 'refined-jiffies'. While > > there is hardly a general reliable way to check the validity of a > > watchdog, and to protect the innocent tsc, Thomas Gleixner proposed [1]: > > > > "I'm inclined to lift that requirement when the CPU has: > > > > 1) X86_FEATURE_CONSTANT_TSC > > 2) X86_FEATURE_NONSTOP_TSC > > 3) X86_FEATURE_NONSTOP_TSC_S3 > > 4) X86_FEATURE_TSC_ADJUST > > 5) At max. 4 sockets > > > > Hi Feng, > > We do need to decouple tsc from HPET as the current HPET as a > clocksource watchdog for tsc > is only useful to find HPET read skews in some circumstances and the > variations of HPET read come from many different sources. But > none of which really came from the tsc quality, AFAICT. > > so this patch is in line with my understanding of the problem. > So , please use reviewed-by : luming.yu@intel.com , if it can help > the merge of the patch. : -)
Thanks for the review and sharing the real world cases you've met!
- Feng
| |