Re: [PATCH] x86/entry/64/compat: Preserve r8-r11 in int $0x80

[Borislav Petkov <bp@alien8.de>]

On Tue, Apr 17, 2018 at 07:36:36AM -0700, Andy Lutomirski wrote:
> 32-bit user code that uses int $80 doesn't care about r8-r11.  There is,
> however, some 64-bit user code that intentionally uses int $0x80 to
> invoke 32-bit system calls.  From what I've seen, basically all such
> code assumes that r8-r15 are all preserved, but the kernel clobbers
> r8-r11.  Since I doubt that there's any code that depends on int $0x80
> zeroing r8-r11, change the kernel to preserve them.
> 
> I suspect that very little user code is broken by the old clobber,
> since r8-r11 are only rarely allocated by gcc, and they're clobbered
> by function calls, so they only way we'd see a problem is if the
> same function that invokes int $0x80 also spills something important
> to one of these registers.
> 
> The current behavior seems to date back to the historical commit
> "[PATCH] x86-64 merge for 2.6.4".  Before that, all regs were
> preserved.  I can't find any explanation of why this change was made.

Probably because r8-r11 are callee-clobbered, according to ABI so
someone decided to whack them so that code which doesn't adhere to the
ABI would fall on its face...

Also, looking at PUSH_AND_CLEAR_REGS and how we call it on the 64-bit
entry path, we probably should keep clearing those regs to avoid
speculation crap.

Methinks.

-- 
Regards/Gruss,
    Boris.

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