linux-kernel - Re: [PATCH 13/15] x86/static_call: Add inline static call implementation for x86-64

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <F9768052-29D8-48EF-86E9-B4C3C5672A67@vmware.com>
Date:   Mon, 10 Jun 2019 19:20:31 +0000
From:   Nadav Amit <namit@...are.com>
To:     Josh Poimboeuf <jpoimboe@...hat.com>
CC:     Peter Zijlstra <peterz@...radead.org>,
        the arch/x86 maintainers <x86@...nel.org>,
        LKML <linux-kernel@...r.kernel.org>,
        Ard Biesheuvel <ard.biesheuvel@...aro.org>,
        Andy Lutomirski <luto@...nel.org>,
        Steven Rostedt <rostedt@...dmis.org>,
        Ingo Molnar <mingo@...nel.org>,
        Thomas Gleixner <tglx@...utronix.de>,
        Linus Torvalds <torvalds@...ux-foundation.org>,
        Masami Hiramatsu <mhiramat@...nel.org>,
        Jason Baron <jbaron@...mai.com>, Jiri Kosina <jkosina@...e.cz>,
        David Laight <David.Laight@...LAB.COM>,
        Borislav Petkov <bp@...en8.de>,
        Julia Cartwright <julia@...com>, Jessica Yu <jeyu@...nel.org>,
        "H. Peter Anvin" <hpa@...or.com>,
        Rasmus Villemoes <linux@...musvillemoes.dk>,
        Edward Cree <ecree@...arflare.com>,
        Daniel Bristot de Oliveira <bristot@...hat.com>
Subject: Re: [PATCH 13/15] x86/static_call: Add inline static call
 implementation for x86-64

> On Jun 10, 2019, at 11:55 AM, Josh Poimboeuf <jpoimboe@...hat.com> wrote:
> 
> On Mon, Jun 10, 2019 at 06:45:52PM +0000, Nadav Amit wrote:
>>> On Jun 10, 2019, at 11:33 AM, Josh Poimboeuf <jpoimboe@...hat.com> wrote:
>>> 
>>> On Wed, Jun 05, 2019 at 03:08:06PM +0200, Peter Zijlstra wrote:
>>>> --- a/arch/x86/include/asm/static_call.h
>>>> +++ b/arch/x86/include/asm/static_call.h
>>>> @@ -2,6 +2,20 @@
>>>> #ifndef _ASM_STATIC_CALL_H
>>>> #define _ASM_STATIC_CALL_H
>>>> 
>>>> +#include <asm/asm-offsets.h>
>>>> +
>>>> +#ifdef CONFIG_HAVE_STATIC_CALL_INLINE
>>>> +
>>>> +/*
>>>> + * This trampoline is only used during boot / module init, so it's safe to use
>>>> + * the indirect branch without a retpoline.
>>>> + */
>>>> +#define __ARCH_STATIC_CALL_TRAMP_JMP(key, func)				\
>>>> +	ANNOTATE_RETPOLINE_SAFE						\
>>>> +	"jmpq *" __stringify(key) "+" __stringify(SC_KEY_func) "(%rip) \n"
>>>> +
>>>> +#else /* !CONFIG_HAVE_STATIC_CALL_INLINE */
>>> 
>>> I wonder if we can simplify this (and drop the indirect branch) by
>>> getting rid of the above cruft, and instead just use the out-of-line
>>> trampoline as the default for inline as well.
>>> 
>>> Then the inline case could fall back to the out-of-line implementation
>>> (by patching the trampoline's jmp dest) before static_call_initialized
>>> is set.
>> 
>> I must be missing some context - but what guarantees that this indirect
>> branch would be exactly 5 bytes long? Isn’t there an assumption that this
>> would be the case? Shouldn’t there be some handling of the padding?
> 
> We don't patch the indirect branch.  It's just part of a temporary
> trampoline which is called by the call site, and which does "jmp
> key->func" during boot until static call initialization is done.
> 
> (Though I'm suggesting removing that.)

Oh... I see.

On another note - even if this branch is only executed during module
initialization, it does seem safer to use a retpoline instead of an indirect
branch (consider a branch that is run many times on one hardware thread on
SMT, when STIBP is not set, and attacker code is run on the second thread).

I guess you don’t simply call the retpoline code so since you don’t have a
clobbered register to hold the target. But it still seems possible to use
a retpoline. Anyhow, it might be a moot discussion if this code is removed.