Re: [Xen-devel] [PATCH v2] x86/PV: fix/generalize guest nul selector handling

["Jan Beulich" <JBeulich@xxxxxxxx>]

>>> On 29.09.17 at 17:17, <roger.pau@xxxxxxxxxx> wrote:
> On Thu, Sep 28, 2017 at 08:41:28AM +0000, Jan Beulich wrote:
>> --- a/xen/arch/x86/domain.c
>> +++ b/xen/arch/x86/domain.c
>> @@ -1237,6 +1237,18 @@ arch_do_vcpu_op(
>>      return rc;
>>  }
>>  
>> +/*
>> + * Loading a nul selector does not clear bases and limits on AMD CPUs. Be on
>> + * the safe side and re-initialize both to flat segment values before 
>> loading
>> + * a nul selector.
>> + */
>> +#define preload_segment(seg, value) do {              \
>> +    if ( !((value) & ~3) &&                           \
>> +         boot_cpu_data.x86_vendor == X86_VENDOR_AMD ) \
>> +        asm volatile ( "movl %k0, %%" #seg            \
> 
> Shouldn't this be a movw? Segment selectors are 16b, not 32b, but I
> might be missing something here.

Ideally it would be "mov" (i.e. without any suffix), but ...

> I see loadsegment is also using movl, so yes, I guess I'm missing
> something.

... my primary goal was to make the two match up in this regard.

Jan


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