linux-kernel - Re: [BUG mips llvm] MIPS: malformed R_MIPS

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]

Date:   Wed, 13 Jan 2021 10:16:41 +0000
From:   Alexander Lobakin <alobakin@...me>
To:     Nick Desaulniers <ndesaulniers@...gle.com>
Cc:     Alexander Lobakin <alobakin@...me>,
        Fangrui Song <maskray@...gle.com>,
        clang-built-linux <clang-built-linux@...glegroups.com>,
        linux-mips@...r.kernel.org,
        Thomas Bogendoerfer <tsbogend@...ha.franken.de>,
        Ralf Baechle <ralf@...ux-mips.org>,
        LKML <linux-kernel@...r.kernel.org>,
        linux-arch <linux-arch@...r.kernel.org>
Subject: Re: [BUG mips llvm] MIPS: malformed R_MIPS_{HI16,LO16} with LLVM

From: Nick Desaulniers <ndesaulniers@...gle.com>
Date: Tue, 12 Jan 2021 14:14:58 -0800

> On Mon, Jan 11, 2021 at 12:50 PM Alexander Lobakin <alobakin@...me> wrote:
>>
>>> The disassembly for me produces:
>>>     399c: 3c 03 00 00   lui     $3, 0 <phy_device_free>
>>>                         0000399c:  R_MIPS_HI16  .text
>>> ...
>>>     39a8: 24 63 3a 5c   addiu   $3, $3, 14940 <phy_probe>
>>>                         000039a8:  R_MIPS_LO16  .text
>>
>> So, in your case the values of the instructions that relocs refer are:
>>
>> 0x3c030000 R_MIPS_HI16
>> 0x24633a5c R_MIPS_LO16
>>
>> Mine were:
>>
>> 0x3c010000
>> 0x24339444
>>
>> Your second one doesn't have bit 15 set, so I think this pair won't
>> break the code.
>> Try to hunt for R_MIPS_LO16 that have this bit set, i.e. they have
>> '8', '9', 'a', 'b', 'c', 'd' or 'e' as their [15:12].
>
> I don't think any of my R_MIPS_LO16 in that file have that bit set.
> See attached.

Well, seems like you got lucky here.
You can try to check for other modules if any of them have
R_MIPS_LO16 with bit 15 set.

Also, you can try to enable:
CONFIG_LD_DEAD_CODE_DATA_ELIMINATION=y
CONFIG_TRIM_UNUSED_KSYMS=y

to alter the build process. These options didn't change anything
in terms of relocs for me though.

> --
> Thanks,
> ~Nick Desaulniers

Thanks,
Al