Re: [PATCH AUTOSEL 5.4 016/187] x86/intel: Disable HPET on Intel Ice Lake platforms

From: Len Brown
Date: Thu Jan 30 2020 - 14:51:14 EST


Can somebody point to an ICL system that is running production-level
firmware that exhibits this issue?

The hardware people are telling me that zero such systems should exist.

(The reported symptom that I've seen is that the TSC vs HPET
comparison fails, and Linux keeps the (bad) HPET and disables the
(good) TSC)

thanks,
-Len

On Fri, Dec 27, 2019 at 12:41 PM Sasha Levin <sashal@xxxxxxxxxx> wrote:
>
> From: Kai-Heng Feng <kai.heng.feng@xxxxxxxxxxxxx>
>
> [ Upstream commit e0748539e3d594dd26f0d27a270f14720b22a406 ]
>
> Like CFL and CFL-H, ICL SoC has skewed HPET timer once it hits PC10.
> So let's disable HPET on ICL.
>
> Signed-off-by: Kai-Heng Feng <kai.heng.feng@xxxxxxxxxxxxx>
> Cc: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>
> Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
> Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
> Cc: bp@xxxxxxxxx
> Cc: feng.tang@xxxxxxxxx
> Cc: harry.pan@xxxxxxxxx
> Cc: hpa@xxxxxxxxx
> Link: https://lkml.kernel.org/r/20191129062303.18982-2-kai.heng.feng@xxxxxxxxxxxxx
> Signed-off-by: Ingo Molnar <mingo@xxxxxxxxxx>
> Signed-off-by: Sasha Levin <sashal@xxxxxxxxxx>
> ---
> arch/x86/kernel/early-quirks.c | 2 ++
> 1 file changed, 2 insertions(+)
>
> diff --git a/arch/x86/kernel/early-quirks.c b/arch/x86/kernel/early-quirks.c
> index 606711f5ebf8..2f9ec14be3b1 100644
> --- a/arch/x86/kernel/early-quirks.c
> +++ b/arch/x86/kernel/early-quirks.c
> @@ -714,6 +714,8 @@ static struct chipset early_qrk[] __initdata = {
> PCI_CLASS_BRIDGE_HOST, PCI_ANY_ID, 0, force_disable_hpet},
> { PCI_VENDOR_ID_INTEL, 0x3ec4,
> PCI_CLASS_BRIDGE_HOST, PCI_ANY_ID, 0, force_disable_hpet},
> + { PCI_VENDOR_ID_INTEL, 0x8a12,
> + PCI_CLASS_BRIDGE_HOST, PCI_ANY_ID, 0, force_disable_hpet},
> { PCI_VENDOR_ID_BROADCOM, 0x4331,
> PCI_CLASS_NETWORK_OTHER, PCI_ANY_ID, 0, apple_airport_reset},
> {}
> --
> 2.20.1
>


--
Len Brown, Intel Open Source Technology Center