Re: [PATCH] x86: entry_32.S: change ESPFIX test to not touch PT_OLDSS(%esp)

From: Denys Vlasenko <vda.linux@googlemail.com>
To: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Denys Vlasenko <dvlasenk@redhat.com>,
	Andy Lutomirski <luto@amacapital.net>,
	Steven Rostedt <rostedt@goodmis.org>,
	Ingo Molnar <mingo@kernel.org>, Borislav Petkov <bp@alien8.de>,
	"H. Peter Anvin" <hpa@zytor.com>, Oleg Nesterov <oleg@redhat.com>,
	Frederic Weisbecker <fweisbec@gmail.com>,
	Alexei Starovoitov <ast@plumgrid.com>,
	Will Drewry <wad@chromium.org>, Kees Cook <keescook@chromium.org>,
	"the arch/x86 maintainers" <x86@kernel.org>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH] x86: entry_32.S: change ESPFIX test to not touch PT_OLDSS(%esp)
Date: Mon, 9 Mar 2015 17:28:26 +0100	[thread overview]
Message-ID: <CAK1hOcNmfy3RS20t-+Pmn+KFtoYM+RqyCx=LgfcHaaXnzF1M5w@mail.gmail.com> (raw)
In-Reply-To: <CA+55aFzck8nuV06AZhhFXgUQEsujZCtufw0jY4iswJUggh6ruA@mail.gmail.com>

On Mon, Mar 9, 2015 at 5:08 PM, Linus Torvalds
<torvalds@linux-foundation.org> wrote:
> On Mon, Mar 9, 2015 at 7:05 AM, Denys Vlasenko <dvlasenk@redhat.com> wrote:
>>     New:
>>      1e6:   0f ba 64 24 38 11       btl    $0x11,0x38(%esp)
>
> btl? Really?
>
> Why isn't that just
>
>     testb $2,0x3a(%esp)
>
> which is both smaller and quite a bit faster on older machines.

Ok. (I thought people won't be happy about this.)

> Or just at least *partially* do what we used to do, and make it all be
>
>     movb  PT_EFLAGS+2(%esp),%al
>     andb $2,%al
>     orb PT_CS(%esp),%al
>     testb $3,%al
>     je restore_nocheck

These insns must run serially. 4 cycles.

        test
        branch
        test
        branch

can execute both test/branch'es out-of-order in parallel.
2 cycles.