linux-kernel - Re: [PATCH] proc: fix proc-self-map-files selftest for arm

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: <52504bd5-b4ed-518b-f748-c634172c6c68@linaro.org>
Date:   Mon, 15 Oct 2018 13:55:14 -0300
From:   Rafael David Tinoco <rafael.tinoco@...aro.org>
To:     Cyrill Gorcunov <gorcunov@...il.com>,
        Alexey Dobriyan <adobriyan@...il.com>
Cc:     linux-kselftest@...r.kernel.org, linux-fsdevel@...r.kernel.org,
        linux-kernel@...r.kernel.org, akpm@...ux-foundation.org,
        shuah@...nel.org
Subject: Re: [PATCH] proc: fix proc-self-map-files selftest for arm

On 10/11/18 7:00 PM, Cyrill Gorcunov wrote:
> On Fri, Oct 12, 2018 at 12:30:06AM +0300, Alexey Dobriyan wrote:
>> On Fri, Oct 12, 2018 at 12:02:56AM +0300, Cyrill Gorcunov wrote:
>>> On Thu, Oct 11, 2018 at 11:56:01PM +0300, Alexey Dobriyan wrote:
>>>>
>>>> As the comment in the beginning says this test is specifically for addresss 0.
>>>> Maybe it should be ifdeffed with __arm__ then.
>>>
>>> Is there some other reason than allocating non-mergable VMA?
>>
>> IIRC the reason is to test address 0 as it is effectively banned
>> for userspace so if it will be broken, it will be broken silently
>> for a long time.
> 
> This is rather a side effect of the test because the primary reason
> was to check procfs numbers conversion, right? Don't get me wrong,
> I don't mind about __arm__ define or similar, this is fine for
> one architecture, but if there comes more we will get a number
> of #ifdefs which is unrelated to procfs numeric routines at all.

That is what I also had in mind, thus the patch. I just realized we had 
another issue on LKFT (our functional tests tool) for 
proc-self-map-files-001.c. Test 001 does pretty much the same as 002, 
but without the MAP_FIXED mmap flag.

Is it okay to consolidate both tests into just 1, and focus in checking 
procfs numbers conversion only, rather than if mapping 0 is allowed or 
not ? Can I send a v2 with that in mind ?

> 
>> As for "unmergeable" libc here doesn't map /dev/zero. I know how to
>> avoid even theoretical breakage by creating binaries by hand but it
>> will be probably too much.
> 
> Sure.
>