Re: [PATCH] x86/retpoline/entry: Disable the entire SYSCALL64 fast path with retpolines on

[Brian Gerst]

In Thu, Jan 25, 2018 at 2:16 PM, Linus Torvalds
<torvalds@xxxxxxxxxxxxxxxxxxxx> wrote:
> On Thu, Jan 25, 2018 at 10:48 AM, Linus Torvalds
> <torvalds@xxxxxxxxxxxxxxxxxxxx> wrote:
>>
>> So the biggest impact of this is the extra register saves
>
> Actually, the other noticeable part is the reloading of the argument
> registers from ptregs. Together with just the extra level of
> 'call/ret' and the stack setup, I'm guessing we're talking maybe 20
> cycles or so.
>
> So there's the extra register saves, and simply the fact that the
> fastpath had a flatter calling structure.
>
> It still feels worth it. And if we do decide that we want to do the
> register clearing on kernel entry for some paranoid mode, we'd pretty
> much have to do this anyway.
>
>                  Linus

Another extra step the slow path does is checking to see if ptregs is
safe for SYSRET.  I think that can be mitigated by moving the check to
the places that do modify ptregs (ptrace, sigreturn, and exec) which
would set a flag to force return with IRET if the modified regs do not
satisfy the criteria for SYSRET.

--
Brian Gerst