netdev - Re: [PATCH net-next v6 3/5] dpll: zl3073x: Add firmware loading functionality

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: <4bd1847b-00b6-42a6-8391-aba08aeb3721@redhat.com>
Date: Mon, 15 Sep 2025 10:18:03 +0200
From: Ivan Vecera <ivecera@...hat.com>
To: Jakub Kicinski <kuba@...nel.org>
Cc: netdev@...r.kernel.org, Przemek Kitszel <przemyslaw.kitszel@...el.com>,
 Jiri Pirko <jiri@...nulli.us>, "David S. Miller" <davem@...emloft.net>,
 Eric Dumazet <edumazet@...gle.com>, Paolo Abeni <pabeni@...hat.com>,
 Simon Horman <horms@...nel.org>, Jonathan Corbet <corbet@....net>,
 Prathosh Satish <Prathosh.Satish@...rochip.com>, linux-doc@...r.kernel.org,
 linux-kernel@...r.kernel.org, Michal Schmidt <mschmidt@...hat.com>,
 Petr Oros <poros@...hat.com>
Subject: Re: [PATCH net-next v6 3/5] dpll: zl3073x: Add firmware loading
 functionality

On 14. 09. 25 11:45 odp., Jakub Kicinski wrote:
> On Tue,  9 Sep 2025 11:15:30 +0200 Ivan Vecera wrote:
>> +	/* Fetch image name and size from input */
>> +	strscpy(buf, *psrc, min(sizeof(buf), *psize));
>> +	rc = sscanf(buf, "%15s %u %n", name, &count, &pos);
>> +	if (!rc) {
>> +		/* No more data */
>> +		return 0;
>> +	} else if (rc == 1 || count > U32_MAX / sizeof(u32)) {
>> +		ZL3073X_FW_ERR_MSG(extack, "invalid component size");
>> +		return -EINVAL;
>> +	}
>> +	*psrc += pos;
>> +	*psize -= pos;
> 
> Still worried about pos not being bounds checked.
> Admin can crash the kernel with invalid FW file.
> 
> 	if (pos > *psize)
> 		/* error */

This cannot happen...

1) strscpy(buf, *psrc, min(sizeof(buf, *psize)) ensures that the string
    will be zero padded and strlen(buf) will be less than *psize

2) sscanf(buf, "%15s %u %n", name, &count, &pos) scans for string with
    max length of 15, one or more whitespace(s), number and one or more
    whitespaces(s). And reports number of parsed arguments. Note that
    the %n does not increase the count returned.

So... if:
1) buf is empty then sscanf returns 0 and /* No more data */ code path
    is executed
2) buf contains only string (1st argument) OR string and non-numeric
    2nd argument then the sscanf returns 1 and 'invalid component size'
    is executed
3) buf contains string (1st arg) and numeric 2nd arg then the sscanf
    returns 2 and the 2nd arg is stored into 'count' and number of
    consumed characters into 'pos'
4) if 'count' > (U32_MAX / 4) then 'invalid component size' path is
    executed

> Also what if sscanf() return 2? pos is uninitialized?

If the sscanf() returns 2 then pos is initialized, in other words
'pos' is initialized and less then *psize if the 'count' is correctly
parsed as the numeric value.

Thanks,
Ivan