From: Andy Lutomirski <luto@amacapital.net>
To: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Herbert Xu <herbert@gondor.apana.org.au>,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
Linux Crypto Mailing List <linux-crypto@vger.kernel.org>,
Linus Torvalds <torvalds@linux-foundation.org>,
Ingo Molnar <mingo@kernel.org>,
Thomas Gleixner <tglx@linutronix.de>,
Andy Lutomirski <luto@kernel.org>
Subject: Re: x86-64: Maintain 16-byte stack alignment
Date: Tue, 10 Jan 2017 15:25:47 -0800 [thread overview]
Message-ID: <CALCETrW4DZeq=tcEnEyPc=H3Z35J-o6BGPwxEfcdKjg2tDqBGg@mail.gmail.com> (raw)
In-Reply-To: <CAKv+Gu_Te-D5V3eG-GHtpP2hdpb+aTCgxBhmvf1DVK-pKPT54g@mail.gmail.com>
On Tue, Jan 10, 2017 at 12:00 PM, Ard Biesheuvel
<ard.biesheuvel@linaro.org> wrote:
> On 10 January 2017 at 19:22, Andy Lutomirski <luto@amacapital.net> wrote:
>> On Tue, Jan 10, 2017 at 11:16 AM, Ard Biesheuvel
>> <ard.biesheuvel@linaro.org> wrote:
>>> On 10 January 2017 at 19:00, Andy Lutomirski <luto@amacapital.net> wrote:
>>>> On Tue, Jan 10, 2017 at 9:30 AM, Ard Biesheuvel
>>>> <ard.biesheuvel@linaro.org> wrote:
>>>>> On 10 January 2017 at 14:33, Herbert Xu <herbert@gondor.apana.org.au> wrote:
>>>>>> I recently applied the patch
>>>>>>
>>>>>> https://patchwork.kernel.org/patch/9468391/
>>>>>>
>>>>>> and ended up with a boot crash when it tried to run the x86 chacha20
>>>>>> code. It turned out that the patch changed a manually aligned
>>>>>> stack buffer to one that is aligned by gcc. What was happening was
>>>>>> that gcc can stack align to any value on x86-64 except 16. The
>>>>>> reason is that gcc assumes that the stack is always 16-byte aligned,
>>>>>> which is not actually the case in the kernel.
>>>>>>
>>>>>
>>>>> Apologies for introducing this breakage. It seemed like an obvious and
>>>>> simple cleanup, so I didn't even bother to mention it in the commit
>>>>> log, but if the kernel does not guarantee 16 byte alignment, I guess
>>>>> we should revert to the old method. If SSE instructions are the only
>>>>> ones that require this alignment, then I suppose not having a ABI
>>>>> conforming stack pointer should not be an issue in general.
>>>>
>>>> Here's what I think is really going on. This is partially from
>>>> memory, so I could be off base. The kernel is up against
>>>> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=53383, which means that,
>>>> on some GCC versions (like the bad one and maybe even current ones),
>>>> things compiled without -mno-sse can't have the stack alignment set
>>>> properly. IMO we should fix this in the affected code, not the entry
>>>> code. In fact, I think that fixing it in the entry code won't even
>>>> fully fix it because modern GCC will compile the rest of the kernel
>>>> with 8-byte alignment and the stack will get randomly unaligned (GCC
>>>> 4.8 and newer).
>>>>
>>>> Can we just add __attribute__((force_align_arg_pointer)) to the
>>>> affected functions? Maybe have:
>>>>
>>>> #define __USES_SSE __attribute__((force_align_arg_pointer))
>>>>
>>>> on affected gcc versions?
>>>>
>>>> ***HOWEVER***
>>>>
>>>> I think this is missing the tree for the supposed forest. The actual
>>>> affected code appears to be:
>>>>
>>>> static int chacha20_simd(struct blkcipher_desc *desc, struct scatterlist *dst,
>>>> struct scatterlist *src, unsigned int nbytes)
>>>> {
>>>> u32 *state, state_buf[16 + (CHACHA20_STATE_ALIGN / sizeof(u32)) - 1];
>>>>
>>>> ...
>>>>
>>>> state = (u32 *)roundup((uintptr_t)state_buf, CHACHA20_STATE_ALIGN);
>>>>
>>>> gcc presumably infers (incorrectly) that state_buf is 16-byte aligned
>>>> and optimizes out the roundup. How about just declaring an actual
>>>> __aligned(16) buffer, marking the function
>>>> __attribute__((force_align_arg_pointer)), and being done with it?
>>>> After all, we need that forcible alignment on *all* gcc versions.
>>>>
>>>
>>> Actually, the breakage is introduced by the patch Herbert refers to
>>>
>>> https://patchwork.kernel.org/patch/9468391/
>>>
>>> where the state is replaced by a simple
>>>
>>> u32 state[16] __aligned(CHACHA20_STATE_ALIGN);
>>>
>>> which seemed harmless enough to me. So the code above works fine.
>>
>> So how about just the one-line patch of adding the
>> force_align_arg_pointer? Would that solve the problem?
>
> If it does what it says on the tin, it should fix the issue, but after
> adding the attribute, I get the exact same object output, so there's
> something dodgy going on here.
Ugh, that's annoying. Maybe it needs noinline too?
--Andy
next prev parent reply other threads:[~2017-01-10 23:25 UTC|newest]
Thread overview: 54+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-01-10 14:33 x86-64: Maintain 16-byte stack alignment Herbert Xu
2017-01-10 14:39 ` Herbert Xu
2017-01-10 17:05 ` Linus Torvalds
2017-01-10 17:09 ` Andy Lutomirski
2017-01-11 3:11 ` Herbert Xu
2017-01-11 3:30 ` Linus Torvalds
2017-01-11 4:17 ` Linus Torvalds
2017-01-11 4:35 ` Herbert Xu
2017-01-11 6:01 ` Andy Lutomirski
2017-01-12 6:21 ` Andy Lutomirski
2017-01-12 7:40 ` Ingo Molnar
2017-01-12 14:02 ` Josh Poimboeuf
2017-01-12 19:51 ` Linus Torvalds
2017-01-12 20:08 ` Andy Lutomirski
2017-01-12 20:15 ` Josh Poimboeuf
2017-01-12 20:55 ` Josh Poimboeuf
2017-01-12 21:40 ` Linus Torvalds
2017-01-13 8:38 ` Herbert Xu
2017-01-13 1:46 ` Andy Lutomirski
2017-01-13 3:11 ` Josh Poimboeuf
2017-01-13 3:23 ` Andy Lutomirski
2017-01-13 4:27 ` Josh Poimboeuf
[not found] ` <CA+55aFzRrSwGxxfZk-RUEnsz=xhcSmOwE1CenfCPBWtsS9MwDw@mail.gmail.com>
2017-01-13 5:07 ` Josh Poimboeuf
2017-01-13 8:43 ` Herbert Xu
2017-01-13 8:42 ` Herbert Xu
2017-01-13 8:39 ` Herbert Xu
2017-01-13 8:36 ` Herbert Xu
2017-01-13 13:07 ` Josh Poimboeuf
[not found] ` <CA+55aFw+Z_ieo6DzTVB6_-TvQ0jj60s=T0mvXfqkBVFdKFPw_Q@mail.gmail.com>
2017-01-11 8:06 ` Ard Biesheuvel
2017-01-11 8:09 ` Herbert Xu
2017-01-11 18:20 ` Andy Lutomirski
2017-01-12 7:05 ` Herbert Xu
2017-01-12 7:46 ` Ingo Molnar
2017-01-12 14:49 ` Josh Poimboeuf
2017-01-12 7:51 ` Andy Lutomirski
2017-01-12 8:04 ` Herbert Xu
2017-01-12 8:18 ` Ingo Molnar
2017-01-12 15:03 ` Josh Poimboeuf
2017-01-12 15:06 ` Herbert Xu
2017-01-12 15:18 ` Josh Poimboeuf
2017-01-12 15:10 ` Josh Poimboeuf
2017-01-10 17:30 ` Ard Biesheuvel
2017-01-10 19:00 ` Andy Lutomirski
2017-01-10 19:16 ` Ard Biesheuvel
2017-01-10 19:22 ` Andy Lutomirski
2017-01-10 20:00 ` Ard Biesheuvel
2017-01-10 23:25 ` Andy Lutomirski [this message]
2017-01-11 3:26 ` Herbert Xu
2017-01-11 3:26 ` Herbert Xu
2017-01-11 3:16 ` Herbert Xu
2017-01-11 3:15 ` Herbert Xu
2017-01-12 6:12 ` Herbert Xu
2017-01-12 8:01 ` Ard Biesheuvel
2017-01-12 8:06 ` Herbert Xu
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to='CALCETrW4DZeq=tcEnEyPc=H3Z35J-o6BGPwxEfcdKjg2tDqBGg@mail.gmail.com' \
--to=luto@amacapital.net \
--cc=ard.biesheuvel@linaro.org \
--cc=herbert@gondor.apana.org.au \
--cc=linux-crypto@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=luto@kernel.org \
--cc=mingo@kernel.org \
--cc=tglx@linutronix.de \
--cc=torvalds@linux-foundation.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).