netdev - AW: Where is this patch buried?

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: <1612776251858.30522@mentor.com>
Date:   Mon, 8 Feb 2021 09:24:11 +0000
From:   "Schmid, Carsten" <Carsten_Schmid@...tor.com>
To:     Greg KH <gregkh@...uxfoundation.org>
CC:     "netdev@...r.kernel.org" <netdev@...r.kernel.org>,
        "weiwan@...gle.com" <weiwan@...gle.com>,
        "idosch@...sch.org" <idosch@...sch.org>,
        "jesse@...ki-mvuki.org" <jesse@...ki-mvuki.org>,
        "kafai@...com" <kafai@...com>,
        "dsahern@...il.com" <dsahern@...il.com>,
        "davem@...emloft.net" <davem@...emloft.net>
Subject: AW: Where is this patch buried?

>> Hi Greg,
>>
>> in kernel 4.14 i have seen a NULL pointer deref in
>> [65064.613465] RIP: 0010:ip_route_output_key_hash_rcu+0x755/0x850
>> (i have a core dump and detailed analysis)
>>
>> That looks like this patch could have prevented it:
>>
>> https://www.spinics.net/lists/stable-commits/msg133055.html
>>
>> this one was queued for 4.14, but i can't find it in git tree?
>> Any idea who/what buried this one?
>>
>> In 4.19 it is present.
>>
>> Because our customer uses 4.14 (going to 4.14.212 in a few days) i kindly want to
>> ask why this patch hasn't entered 4.14.
>
> Because it breaks the build?  Try it yourself and see what happens :)

yep. rt_add_uncached_list is implemented _after_ the call :-(

>
> I will gladly take a working backport if you can submit it.
>
Please find it attached - i needed to move rt_add_uncached_list before
the rt_cache_route, nothing else changed.
This is for 4.14 only, as other kernels do have this patch.

I can't reproduce the crash at will, but at least i could compile and flash it on my target.
And the target comes up .. hopefully the testing in other/newer kernels
done by the developers of the patch is also valid for 4.14.

> thanks,
>
> greg k-h

Thanks,
Carsten
-----------------
Mentor Graphics (Deutschland) GmbH, Arnulfstrasse 201, 80634 München Registergericht München HRB 106955, Geschäftsführer: Thomas Heurung, Frank Thürauf

View attachment "0001-ipv4-fix-race-condition-between-route-lookup-and-inv.patch" of type "text/x-patch" (4140 bytes)