Re: [PATCH 2/2] x86: Clear incorrectly forced X86_FEATURE_LAHF_LM flag

[Brian Gerst]

On Thu, Aug 13, 2009 at 10:54 AM, Kevin
Winchester<kjwinchester@xxxxxxxxx> wrote:
> 2009/8/13 Brian Gerst <brgerst@xxxxxxxxx>:
>> On Thu, Aug 13, 2009 at 8:31 AM, Borislav Petkov<borislav.petkov@xxxxxxx> wrote:
>>> From: Kevin Winchester <kjwinchester@xxxxxxxxx>
>>>
>>> Due to an erratum with certain AMD Athlon 64 processors, the BIOS may
>>> need to force enable the LAHF_LM capability. ÂUnfortunately, in at
>>> least one case, the BIOS does this even for processors that do not
>>> support the functionality.
>>>
>>> Add a specific check that will clear the feature bit for processors
>>> known not to support the LAHF/SAHF instructions.
>>>
>>> Borislav: turn off cpuid bit.
>>>
>>> Signed-off-by: Kevin Winchester <kjwinchester@xxxxxxxxx>
>>> Signed-off-by: Borislav Petkov <borislav.petkov@xxxxxxx>
>>> ---
>>> Âarch/x86/kernel/cpu/amd.c | Â 16 ++++++++++++++++
>>> Â1 files changed, 16 insertions(+), 0 deletions(-)
>>>
>>> diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c
>>> index e2485b0..9cd6fc7 100644
>>> --- a/arch/x86/kernel/cpu/amd.c
>>> +++ b/arch/x86/kernel/cpu/amd.c
>>> @@ -400,6 +400,22 @@ static void __cpuinit init_amd(struct cpuinfo_x86 *c)
>>> Â Â Â Â Â Â Â Âlevel = cpuid_eax(1);
>>> Â Â Â Â Â Â Â Âif((level >= 0x0f48 && level < 0x0f50) || level >= 0x0f58)
>>> Â Â Â Â Â Â Â Â Â Â Â Âset_cpu_cap(c, X86_FEATURE_REP_GOOD);
>>> +
>>> + Â Â Â Â Â Â Â /*
>>> + Â Â Â Â Â Â Â Â* Some BIOSes incorrectly force this feature, but only K8
>>> + Â Â Â Â Â Â Â Â* revision D (model = 0x14) and later actually support it.
>>> + Â Â Â Â Â Â Â Â*/
>>> + Â Â Â Â Â Â Â if (c->x86_model < 0x14) {
>>
>> Shouldn't you test that the flag is actually set before trying to clear it?
>>
>
> Possibly. ÂIf there were some concern that:
>
> - The extra instructions would cause a performance impact, and the
> test was significantly faster than the clear.

Testing a bit is cheap and MSR accesses are not.

> - The extra instructions might actually cause more problems if the
> flag is not set.

These MSRs don't exist on older cpus and will cause a fault, which is
handled at additional cost.

> Then we would certainly want to test it first. ÂIn my opinion, a few
> simple instructions to clear the flag and the CPUID bit will not
> affect performance, and clearing a flag that is already cleared should
> not cause any additional problems, so I would not bother testing the
> flag first. ÂThat results in fewer lines of code to change.

--
Brian Gerst
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/