Re: [Xen-devel] [PATCH v2 1/2] x86/msr: Carry on after a non-"safe" MSR access fails without !panic_on_oops

[Andy Lutomirski <luto@xxxxxxxxxxxxxx>]

On Mon, Sep 21, 2015 at 9:49 AM, Arjan van de Ven <arjan@xxxxxxxxxxxxxxx> wrote:
> On 9/21/2015 9:36 AM, Linus Torvalds wrote:
>>
>> On Mon, Sep 21, 2015 at 1:46 AM, Ingo Molnar <mingo@xxxxxxxxxx> wrote:
>>>
>>>
>>> Linus, what's your preference?
>>
>>
>> So quite frankly, is there any reason we don't just implement
>> native_read_msr() as just
>>
>>     unsigned long long native_read_msr(unsigned int msr)
>>     {
>>        int err;
>>        unsigned long long val;
>>
>>        val = native_read_msr_safe(msr, &err);
>>        WARN_ON_ONCE(err);
>>        return val;
>>     }
>>
>> Note: no inline, no nothing. Just put it in arch/x86/lib/msr.c, and be
>> done with it. I don't see the downside.
>>
>> How many msr reads are <i>so</i> critical that the function call
>> overhead would matter?
>
>
> if anything qualifies it'd be switch_to() and friends.

And maybe the KVM user return notifier.  Unfortunately, switch_to
might gain another two MSR accesses at some point if we decide to fix
the bugs in there.  Sigh.

>
> note that I'm not entirely happy about the notion of "safe" MSRs.
> They're safe as in "won't fault".
> Reading random MSRs isn't a generic safe operation though, but the name sort
> of gives people
> the impression that it is. Even with _safe variants, you still need to KNOW
> the MSR exists (by means
> of CPUID or similar) unfortunately.
>

I tend to agree.

Anyway, the fully out-of-line approach isn't obviously a bad idea, and
it simplifies the whole mess (we can drop most of the paravirt
patches, too).  I'll give it a try and see what happens.

--Andy

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel