Re: [tip:x86/pti] x86/retpoline: Fill return stack buffer on vmexit

[Borislav Petkov]

On Thu, Jan 25, 2018 at 05:00:39PM +0000, David Woodhouse wrote:
> And the whole problem here is that patching it in with alternatives is
> painful on kernels < 4.1 because back then, we didn't cope with
> oldinstr and altinstr being different lengths.
> 
> And they don't want to fix *that* because kABI...

So if it were only because of the KABI, I would never sent a mail on the
list but would've done it in our tress and forgotten about it.

 [ And just to set one thing straight: I'm not the right person to
   complain to about KABI. ]

Now, I happen to think that those macros could be simplified regardless.
And I don't see *anything* wrong with that. Like making them more
readable, simpler, etc, etc. That's the only reason why I raised the
issue here. I mentioned the KABI because I didn't want to leave anything
out from the whole picture.

So forget the KABI angle and think: simpler, cleaner, more readable
macros.

Oh, and David, if while doing so I manage to add the alignment, then
*that* is even better.

Win-win-effing-win situation!

-- 
Regards/Gruss,
    Boris.

Good mailing practices for 400: avoid top-posting and trim the reply.