Re: Avoid speculative indirect calls in kernel

[Thomas Gleixner]

On Thu, 4 Jan 2018, Alan Cox wrote:
> On Wed, 3 Jan 2018 16:15:01 -0800
> Andi Kleen <andi@xxxxxxxxxxxxxx> wrote:
> 
> > > It should be a CPU_BUG bit as we have for the other mess. And that can be
> > > used for patching.  
> > 
> > It has to be done at compile time because it requires a compiler option.
> > 
> > Most of the indirect calls are in C code.
> > 
> > So it cannot just patched in, only partially out.
> 
> You can replace the pushl ; jmp  with an alternatives section (although
> there might be a lot of them). Even if gcc isn't smart enough to do that
> perl is.

So you say, that we finally need a perl interpreter in the kernel to do
alternative patching?

Thanks,

	tglx