linux-kernel - Re: WARNING in refcount

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] [day] [month] [year] [list]

Date:   Thu, 1 Feb 2018 17:24:14 +0100
From:   Dmitry Vyukov <dvyukov@...gle.com>
To:     Andrey Konovalov <andreyknvl@...gle.com>
Cc:     Eric Biggers <ebiggers3@...il.com>,
        syzbot 
        <bot+944a064dbc165044476317df06ebfb218415e951@...kaller.appspotmail.com>,
        Andrew Morton <akpm@...ux-foundation.org>,
        "David S. Miller" <davem@...emloft.net>,
        Deepa Dinamani <deepa.kernel@...il.com>,
        Giuseppe Scrivano <gscrivan@...hat.com>,
        Herbert Xu <herbert@...dor.apana.org.au>,
        LKML <linux-kernel@...r.kernel.org>, luc.vanoostenryck@...il.com,
        Ingo Molnar <mingo@...nel.org>,
        netdev <netdev@...r.kernel.org>,
        Stephen Rothwell <sfr@...b.auug.org.au>,
        Steffen Klassert <steffen.klassert@...unet.com>,
        syzkaller-bugs@...glegroups.com,
        Alexander Viro <viro@...iv.linux.org.uk>,
        Cong Wang <xiyou.wangcong@...il.com>
Subject: Re: WARNING in refcount_inc (2)

On Thu, Feb 1, 2018 at 3:34 PM, 'Andrey Konovalov' via syzkaller-bugs
<syzkaller-bugs@...glegroups.com> wrote:
> On Wed, Jan 31, 2018 at 8:27 AM, Eric Biggers <ebiggers3@...il.com> wrote:
>>
>> Also Dmitry, syzbot seems to be grouping together unrelated bugs under the
>> refcount_t WARNINGs; maybe those should be on a blacklist?
>
> Not a blacklist, we need a proper way of extracting the offending
> caller like it's done for reports from __this_cpu_* [1].
>
> [1] https://github.com/google/syzkaller/blob/master/pkg/report/linux.go#L579

Thanks, I've added this as a test case:
https://github.com/google/syzkaller/commit/e525e980eaed440e278614b9e887270ca67d2dde

We mishandle __this_cpu_* as well:
https://github.com/google/syzkaller/blob/master/pkg/report/testdata/linux/report/136
https://github.com/google/syzkaller/blob/master/pkg/report/testdata/linux/report/137
It does not seem that regexps can sustain this anymore (we also
mishandle rcu stalls and some other cases), I guess we need manual
parsing of stack traces that can handle all these cases.