| From: |
| David Woodhouse <dwmw2-AT-infradead.org> |
| To: |
| Thomas Gleixner <tglx-AT-linutronix.de>, Jon Masters <jcm-AT-jonmasters.org> |
| Subject: |
| Re: Improve retpoline for Skylake |
| Date: |
| Mon, 15 Jan 2018 11:20:32 +0100 |
| Message-ID: |
| <1516011632.22147.137.camel@infradead.org> |
| Cc: |
| Henrique de Moraes Holschuh <hmh-AT-hmh.eng.br>, Andi Kleen <andi-AT-firstfloor.org>, x86-AT-kernel.org, linux-kernel-AT-vger.kernel.org, pjt-AT-google.com, torvalds-AT-linux-foundation.org, gregkh-AT-linux-foundation.org, peterz-AT-infradead.org, luto-AT-amacapital.net, thomas.lendacky-AT-amd.com, arjan.van.de.ven-AT-intel.com |
On Mon, 2018-01-15 at 11:03 +0100, Thomas Gleixner wrote:
>
> > Our numbers on Skylake weren't bad, and there seem to be all kinds of
> > corner cases, so again, it seems as if IBRS is the safest choice.
>
> Talk is cheap. Show numbers comparing the full retpoline/RBS mitigation
> compared to IBRS.
Right now my intent is to add IBRS support and use it by default on
Skylake instead of retpoline. Users will be able to add
'spectre_v2=retpoline' to use retpoline on Skylake instead, if they
want to. But we'll default to the safest option.
Once that is done, we can do proper comparisons, and also look at
Andi's patches to attempt to make retpoline completely safe on Skylake.
