linux-kernel - Re: [PATCH v3] net: don't call strlen on non-terminated string in dev_set

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: <CAG_fn=U0n4Sn4_44e63FjsYPTTDX4ivAbTcuL2s3erFc+g6n2g@mail.gmail.com>
Date:   Thu, 1 Jun 2017 15:50:33 +0200
From:   Alexander Potapenko <glider@...gle.com>
To:     Yury Norov <ynorov@...iumnetworks.com>
Cc:     Dmitriy Vyukov <dvyukov@...gle.com>,
        Kostya Serebryany <kcc@...gle.com>,
        Eric Dumazet <edumazet@...gle.com>,
        David Miller <davem@...emloft.net>, stephen@...workplumber.org,
        LKML <linux-kernel@...r.kernel.org>,
        Networking <netdev@...r.kernel.org>
Subject: Re: [PATCH v3] net: don't call strlen on non-terminated string in dev_set_alias()

On Thu, Jun 1, 2017 at 3:47 PM, Yury Norov <ynorov@...iumnetworks.com> wrote:
> On Thu, Jun 01, 2017 at 02:38:29PM +0200, Alexander Potapenko wrote:
>> KMSAN reported a use of uninitialized memory in dev_set_alias(),
>> which was caused by calling strlcpy() (which in turn called strlen())
>> on the user-supplied non-terminated string.
>>
>> Signed-off-by: Alexander Potapenko <glider@...gle.com>
>> ---
>> v3: removed the multi-line comment
>> v2: fixed an off-by-one error spotted by Dmitry Vyukov
>
> [...]
>
>> ---
>>  net/core/dev.c | 4 +++-
>>  1 file changed, 3 insertions(+), 1 deletion(-)
>>
>> diff --git a/net/core/dev.c b/net/core/dev.c
>> index fca407b4a6ea..3e3b29133cc9 100644
>> --- a/net/core/dev.c
>> +++ b/net/core/dev.c
>> @@ -1254,7 +1254,9 @@ int dev_set_alias(struct net_device *dev, const char *alias, size_t len)
>>               return -ENOMEM;
>>       dev->ifalias = new_ifalias;
>>
>> -     strlcpy(dev->ifalias, alias, len+1);
>> +     /* alias comes from the userspace and may not be zero-terminated. */
>
> So if the comment is correct, you'd use copy_from_user() instead.
Well, the contents of |alias| have been previously copied from the
userspace, but this is a pointer to a kernel buffer, as the function
prototype tells.
Do you think a confusion is possible here?
>> +     memcpy(dev->ifalias, alias, len);
>> +     dev->ifalias[len] = 0;
>>       return len;
>>  }
>>
>> --
>> 2.13.0.219.gdb65acc882-goog



-- 
Alexander Potapenko
Software Engineer

Google Germany GmbH
Erika-Mann-Straße, 33
80636 München

Geschäftsführer: Matthew Scott Sucherman, Paul Terence Manicle
Registergericht und -nummer: Hamburg, HRB 86891
Sitz der Gesellschaft: Hamburg