[dpdk-stable] [PATCH] eal/x86: fix some CPU extended features

David Marchand david.marchand at redhat.com
Fri Oct 8 13:08:54 CEST 2021


On Fri, Oct 8, 2021 at 12:15 PM Bruce Richardson
<bruce.richardson at intel.com> wrote:
>
> On Wed, Oct 06, 2021 at 09:20:08PM +0200, David Marchand wrote:
> > Caught while checking CPUID related stuff in OVS.
> >
> > According to [1], for Structured Extended Feature Flags Enumeration Leaf
> > (EAX = 0x07H, ECX = 0):
> >
> > - BMI1 is associated to EBX, bit 3 (was incorrectly 2),
> > - SMEP is associated to EBX, bit 7 (was incorrectly 6),
> > - BMI2 is associated to EBX, bit 8 (was incorrectly 7),
> > - ERMS is associated to EBX, bit 9 (was incorrectly 8),
> >
> > This patch then sorts the rest of the extended features (leaf 0) for
> > readability.
> >
> > 1: https://software.intel.com/sites/default/files/managed/c5/15/architecture-instruction-set-extensions-programming-reference.pdf
> >
> > Fixes: af75078fece3 ("first public release")
>
> Wow, I'm surprised at it being incorrect this long, but checking the
> original commit of code, it does indeed seem true that this is the correct
> fixes line!
>
> > Cc: stable at dpdk.org
> >
> > Signed-off-by: David Marchand <david.marchand at redhat.com>
>
> Acked-by: Bruce Richardson <bruce.richardson at intel.com>
>
> Ideally, I think the fixes should go in a separate patch from the
> reordering, but I don't feel strongly about it so this is ok as either one
> or two patches IMHO.

I was reading the datasheet and comparing the defines one by one, so
easier with sorted defines.
Now that I identified the wrong entries, I am fine with splitting
since it eases backport on 19.11.

I'll send v2 with your ack on both.


-- 
David Marchand



More information about the stable mailing list