Re: [PATCH 5/8] x86: Make old K8 swapgs workaround conditional

2015-04-11 Thread Borislav Petkov
On Sat, Apr 11, 2015 at 01:10:13AM +0200, Andi Kleen wrote: > I don't know which one, but I don't think it really matters. We're not punishing innocent K8 machines just because you don't remember. -- Regards/Gruss, Boris. ECO tip #101: Trim your mails when you reply. -- -- To unsubscribe fr

Re: [PATCH 5/8] x86: Make old K8 swapgs workaround conditional

2015-04-10 Thread Andi Kleen
On Sat, Apr 11, 2015 at 12:01:10AM +0200, Borislav Petkov wrote: > > diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c > > index a220239..e7f5667 100644 > > --- a/arch/x86/kernel/cpu/amd.c > > +++ b/arch/x86/kernel/cpu/amd.c > > @@ -551,6 +551,9 @@ static void init_amd_k8(struct cp

Re: [PATCH 5/8] x86: Make old K8 swapgs workaround conditional

2015-04-10 Thread Borislav Petkov
On Fri, Apr 10, 2015 at 08:50:30AM -0700, Andi Kleen wrote: > From: Andi Kleen > > Every gs selector/index reload always paid an extra MFENCE > between the two SWAPGS. This was to work around an old > bug in early K8 steppings. All other CPUs don't need the extra > mfence. Patch the extra MFENCE

Re: [PATCH 5/8] x86: Make old K8 swapgs workaround conditional

2015-04-10 Thread Andy Lutomirski
On Fri, Apr 10, 2015 at 8:50 AM, Andi Kleen wrote: > From: Andi Kleen > > Every gs selector/index reload always paid an extra MFENCE > between the two SWAPGS. This was to work around an old > bug in early K8 steppings. All other CPUs don't need the extra > mfence. Patch the extra MFENCE only in

Re: [PATCH 5/8] x86: Make old K8 swapgs workaround conditional

2014-11-11 Thread Andy Lutomirski
On 11/10/2014 03:55 PM, Andi Kleen wrote: > From: Andi Kleen > > Every gs selector/index reload always paid an extra MFENCE > between the two SWAPGS. This was to work around an old > bug in early K8 steppings. All other CPUs don't need the extra > mfence. Patch the extra MFENCE only in for K8. >

[PATCH 5/8] x86: Make old K8 swapgs workaround conditional

2014-11-10 Thread Andi Kleen
From: Andi Kleen Every gs selector/index reload always paid an extra MFENCE between the two SWAPGS. This was to work around an old bug in early K8 steppings. All other CPUs don't need the extra mfence. Patch the extra MFENCE only in for K8. Signed-off-by: Andi Kleen --- arch/x86/include/asm/c

[PATCH 5/8] x86: Make old K8 swapgs workaround conditional

2014-10-14 Thread Andi Kleen
From: Andi Kleen Every gs selector/index reload always paid an extra MFENCE between the two SWAPGS. This was to work around an old bug in early K8 steppings. All other CPUs don't need the extra mfence. Patch the extra MFENCE only in for K8. Signed-off-by: Andi Kleen --- arch/x86/include/asm/c