Re: [PATCH] x86/PCI: Convert force_disable_hpet() to standard quirk

[Feng Tang]

On Sat, Oct 01, 2022 at 01:18:05PM +0800, Zhang Rui wrote:
[...]
> > > > Thanks Xiaofeng for the info.
> > > > 
> > > > Could you try the below patch? It is kinda extension of 
> > > > 
> > > > b50db7095fe0 ("x86/tsc: Disable clocksource watchdog for TSC on
> > > > qualified platorms") 
> > > > 
> > > > which I have run limited test on some 4 sockets Haswell and
> > > > Cascadelake
> > > > AP x86 servers.
> > > > 
> > > > 
> > > > Thanks,
> > > > Feng
> > > > ---
> > > > diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c
> > > > index cafacb2e58cc..b4ea79cb1d1a 100644
> > > > --- a/arch/x86/kernel/tsc.c
> > > > +++ b/arch/x86/kernel/tsc.c
> > > > @@ -1217,7 +1217,7 @@ static void __init
> > > > check_system_tsc_reliable(void)
> > > >  	if (boot_cpu_has(X86_FEATURE_CONSTANT_TSC) &&
> > > >  	    boot_cpu_has(X86_FEATURE_NONSTOP_TSC) &&
> > > >  	    boot_cpu_has(X86_FEATURE_TSC_ADJUST) &&
> > > > -	    nr_online_nodes <= 2)
> > > > +	    nr_online_nodes <= 8)
> > > >  		tsc_disable_clocksource_watchdog();
> > > >  }
> > > >  
> > > > 
> > > Hi Feng,
> > > 
> > > I tested this patch on a previous server and it fixes the issue.
> >  
> > Thanks for the testing, please do let us know if there is any TSC
> > problem after long time or stress running.
> > 
> > Plan to send the patch for merging.
> > 
> 
> Good to know.
> 
> This patch fixes the TSC issue for IA.
> while the patch at 
> https://lore.kernel.org/lkml/YzZDLBKbDTbNr45b@feng-clx/T/#m34094630193e8320c6d75e9c8aeabe7633e051d2
> impacts more Arches but I'm not sure if similar problem is observed on
> other Arches as well. So I will hold it for now.

I think your patch is more general and architecture independent. We've
seen quite some cases in boot phase on IA platforms, but I guess this
is also possible for other ARCHs.

Thanks,
Feng

> thanks,
> rui
>