Re: [RFC PATCH] therm_throt: test bits as we build therm_intr_core_clear_mask
From: Kyle McMartin
Date: Thu Apr 04 2024 - 13:17:36 EST
On Wed, Apr 03, 2024 at 06:15:47PM -0700, srinivas pandruvada wrote:
> > On Broadwell and Broadwell-DE, the HWP flag is not set, but writing
> > these bits does not trap.
> >
> > On our Skylake-DE, Skylake, and Cooper Lake platforms, the HWP flag
> > is
> > set in CPUID, and writing 1 to these bits traps attempting to write
> > 0xAAA8 to MSR 0x19C (THERM_STATUS). Writing 0xAA8 from userspace
> > works
> > as expected to un-stick PROCHOT_LOG.
>
> I think this issue happens only on Skylake, Cascade Lake, Cooper Lake
> and not on any other systems.
>
> Please verify:
> GP# happens only when bit13 (Current Limit Log) or bit15 (Cross Domain
> Limit Log) is 1.
>
Yeah, if either of the bits are set, we'll trap and fail the WRMSRL.
> Basically writing 0x2000 or 0x8000 or A000 will cause this issue.
> Are you using the latest BIOS with microcode?
> Please confirm your microcode version, I can check internally.
>
On SkylakeDE, 6-85-4 we've got 0x2006e08 and 0x2006e05 as the most commonly
deployed microcodes. On Skylake, 6-85-4 we've got 0x2006e05 and 0x2000065.
Finally, on Cooper Lake, 6-85-11, we have 0x700001f and are in the process
of rolling out 0x7002503.
Rolling out new firmware is a pretty slow process... Since we're not
clearing those bits anywhere in the kernel we're deploying, I just
stubbed out setting BIT(13) and BIT(15) on those platforms for now while
we discuss a more durable fix.
Thanks for following up! --kyle
> Thanks,
> Srinivas
>
>
> >
> > On our Sapphire Rapids platforms, the HWP flag is set, and writing 1
> > to
> > these bits is successful.
> >
> > drivers/thermal/intel/therm_throt.c | 29 ++++++++++++++++++++++-----
> > --
> > 1 file changed, 22 insertions(+), 7 deletions(-)
> >
> > diff --git a/drivers/thermal/intel/therm_throt.c
> > b/drivers/thermal/intel/therm_throt.c
> > index e69868e868eb..3058d8fcfcef 100644
> > --- a/drivers/thermal/intel/therm_throt.c
> > +++ b/drivers/thermal/intel/therm_throt.c
> > @@ -196,8 +196,14 @@ static const struct attribute_group
> > thermal_attr_group = {
> > static u64 therm_intr_core_clear_mask;
> > static u64 therm_intr_pkg_clear_mask;
> >
> > +/* Probe each addition to the mask to ensure that our wrmsrl
> > + * won't fail to clear bits.
> > + */
> > static void thermal_intr_init_core_clear_mask(void)
> > {
> > + u64 bits = 0;
> > + u64 mask = 0;
> > +
> > if (therm_intr_core_clear_mask)
> > return;
> >
> > @@ -211,25 +217,34 @@ static void
> > thermal_intr_init_core_clear_mask(void)
> > * Bit 1, 3, 5: CPUID.01H:EDX[22] = 1. This driver will not
> > * enable interrupts, when 0 as it checks for
> > X86_FEATURE_ACPI.
> > */
> > - therm_intr_core_clear_mask = (BIT(1) | BIT(3) | BIT(5));
> > + mask = (BIT(1) | BIT(3) | BIT(5));
> >
> > /*
> > * Bit 7 and 9: Thermal Threshold #1 and #2 log
> > * If CPUID.01H:ECX[8] = 1
> > */
> > - if (boot_cpu_has(X86_FEATURE_TM2))
> > - therm_intr_core_clear_mask |= (BIT(7) | BIT(9));
> > + bits = BIT(7) | BIT(9);
> > + if (boot_cpu_has(X86_FEATURE_TM2) &&
> > + wrmsrl_safe(MSR_IA32_THERM_STATUS, mask | bits) >= 0)
> > + mask |= bits;
> > +
> >
> > /* Bit 11: Power Limitation log (R/WC0) If CPUID.06H:EAX[4] =
> > 1 */
> > - if (boot_cpu_has(X86_FEATURE_PLN))
> > - therm_intr_core_clear_mask |= BIT(11);
> > + bits = BIT(11);
> > + if (boot_cpu_has(X86_FEATURE_PLN) &&
> > + wrmsrl_safe(MSR_IA32_THERM_STATUS, mask | bits) >= 0)
> > + mask |= bits;
> >
> > /*
> > * Bit 13: Current Limit log (R/WC0) If CPUID.06H:EAX[7] = 1
> > * Bit 15: Cross Domain Limit log (R/WC0) If CPUID.06H:EAX[7]
> > = 1
> > */
> > - if (boot_cpu_has(X86_FEATURE_HWP))
> > - therm_intr_core_clear_mask |= (BIT(13) | BIT(15));
> > + bits = BIT(13) | BIT(15);
> > + if (boot_cpu_has(X86_FEATURE_HWP) &&
> > + wrmsrl_safe(MSR_IA32_THERM_STATUS, mask | bits) >= 0)
> > + mask |= bits;
> > +
> > + therm_intr_core_clear_mask = mask;
> > }
> >
> > static void thermal_intr_init_pkg_clear_mask(void)
>