linux-kernel - Re: sparse warnings in overflow.h

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: <4d2aaa98-1655-3ab8-8ce8-bf4b9c7d2dbc@rasmusvillemoes.dk>
Date:   Thu, 7 Jun 2018 21:25:09 +0200
From:   Rasmus Villemoes <linux@...musvillemoes.dk>
To:     Kees Cook <keescook@...omium.org>, Miroslav Benes <mbenes@...e.cz>
Cc:     LKML <linux-kernel@...r.kernel.org>
Subject: Re: sparse warnings in overflow.h

On 2018-06-07 20:35, Kees Cook wrote:
> On Thu, Jun 7, 2018 at 4:01 AM, Miroslav Benes <mbenes@...e.cz> wrote:
>>
>> Hi Kees,
>>
>> sparse (make C=1) gives me this warnings today...
>>
>> ./include/linux/overflow.h:254:13: error: undefined identifier '__builtin_mul_overflow'
>> ./include/linux/overflow.h:254:13: error: incorrect type in conditional
>> ./include/linux/overflow.h:254:13:    got void
>> ./include/linux/overflow.h:256:13: error: undefined identifier '__builtin_add_overflow'
>> ./include/linux/overflow.h:256:13: error: incorrect type in conditional
>> ./include/linux/overflow.h:256:13:    got void
>>
>> This hunk obviously fixes it, but I'm really not sure if it is a proper
>> fix.
>>
>> diff --git a/include/linux/compiler-gcc.h b/include/linux/compiler-gcc.h
>> index f1a7492a5cc8..15e55b89e952 100644
>> --- a/include/linux/compiler-gcc.h
>> +++ b/include/linux/compiler-gcc.h
>> @@ -344,6 +344,6 @@
>>   */
>>  #define uninitialized_var(x) x = x
>>
>> -#if GCC_VERSION >= 50100
>> +#if GCC_VERSION >= 50100 && !defined(__CHECKER__)
>>  #define COMPILER_HAS_GENERIC_BUILTIN_OVERFLOW 1
>>  #endif
>>
>> Any idea?
> 
> Hi! Yeah, that's probably the correct approach. If the checker doesn't
> have it, the builtins need to be disabled. :)

IIRC, the problem is that sparse pretends to be the gcc sparse itself
was built with, which is obviously entirely unrelated to the C dialect
that that particular sparse version groks. Sigh. Ack to the fix above.

Rasmus