linux-kernel - Re: [PATCH 1/3] LoongArch: tools: Add relocs tool support

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: <f0e77716-9533-724a-2ea9-86bc5b52066c@loongson.cn>
Date:   Sat, 3 Sep 2022 09:57:04 +0800
From:   Youling Tang <tangyouling@...ngson.cn>
To:     Xi Ruoyao <xry111@...111.site>, Huacai Chen <chenhuacai@...nel.org>
Cc:     linux-kernel@...r.kernel.org, loongarch@...ts.linux.dev,
        Xuerui Wang <kernel@...0n.name>,
        Jiaxun Yang <jiaxun.yang@...goat.com>,
        Jinyang He <hejinyang@...ngson.cn>
Subject: Re: [PATCH 1/3] LoongArch: tools: Add relocs tool support

Hi, Ruoyao

On 09/02/2022 07:27 PM, Xi Ruoyao wrote:
> On Fri, 2022-09-02 at 19:10 +0800, Xi Ruoyao wrote:
>> Hi Youling,
>>
>> On Fri, 2022-09-02 at 18:08 +0800, Youling Tang wrote:
>>> This tool is based on the arch/mips/boot/tools/relocs tool.
>>
>> I've dig some history of MIPS relocatable kernel
>> (https://lwn.net/Articles/682020/):
>>
>>    * Kernel is compiled & statically linked as normal, with no position
>>      independent code. MIPS before R6 only has limited relative jump
>>      instructions so the vast majority of jumps are absolute. To compile
>>      the kernel position independent would introduce a highly undesireable
>>      overhead. Relocating the static binary gives a small startup time
>>      penalty but the kernel otherwise perforns normally.
>>
>> Unlike (pre-r6) MIPS, LoongArch has a complete support for PIC, and
>> currently LoongArch toolchain always produces PIC (except, if -Wa,-mla-
>> {local,global}-with-abs or la.abs macros are used explicitly).
>>
>> So would it be easier to review and correct the uses of "la.abs" in the
>> code, and make the main kernel image a real PIE?  Then we can load it
>> everywhere w/o any need to do relocation at load time.

At the beginning I also wanted to make the main kernel image a real PIE
and tried it, some of the "la.abs" can be modified, but I encountered
difficulties in modifying the  exception handling code part, the kernel
will not  boot after modification :(, I will continue to work hard try.

I would be grateful if you all have better suggestions and ideas :).

>
> Arrgh, doing so will need to get rid of GOT (the GCC patch for this is
> still under review) or we still need to relocate the GOT at least :(.

Yes, now all global variable access is via got (la.got) in the new
toolchain, so we need to relocate the GOT in relocate.c. However there
is no need to relocate it in the old toolchain as it is accessed
via la.pcrel.

Thanks,
Youling

>
>