Re: Coverity: mcp251xfd_dump_rx_ring(): Memory - illegal accesses

[Kees Cook <keescook@chromium.org>]

On Thu, Apr 01, 2021 at 09:49:04AM +0200, Marc Kleine-Budde wrote:
> On 31.03.2021 14:59:44, coverity-bot wrote:
> > This is an experimental semi-automated report about issues detected by
> > Coverity from a scan of next-20210331 as part of the linux-next scan project:
> > https://scan.coverity.com/projects/linux-next-weekly-scan
> > 
> > You're getting this email because you were associated with the identified
> > lines of code (noted below) that were touched by commits:
> > 
> >   None
> >     e0ab3dd5f98f ("can: mcp251xfd: add dev coredump support")
> > 
> > Coverity reported the following:
> > 
> > *** CID 1503585:  Memory - illegal accesses  (OVERRUN)
> > /drivers/net/can/spi/mcp251xfd/mcp251xfd-dump.c: 190 in mcp251xfd_dump_rx_ring()
> > 184     static void mcp251xfd_dump_rx_ring(const struct mcp251xfd_priv *priv,
> > 185     				   struct mcp251xfd_dump_iter *iter)
> > 186     {
> > 187     	struct mcp251xfd_rx_ring *rx_ring;
> > 188     	unsigned int i;
> > 189
> > vvv     CID 1503585:  Memory - illegal accesses  (OVERRUN)
> > vvv     Overrunning array of 1 8-byte elements at element index 1 (byte offset 15) by dereferencing pointer "priv->rx + i".
> > 190     	mcp251xfd_for_each_rx_ring(priv, rx_ring, i)
> > 191     		mcp251xfd_dump_rx_ring_one(priv, iter, rx_ring);
> 
> mcp251xfd_for_each_rx_ring is a macro that iterates over all RX rings in
> the struct mcp251xfd_priv. It's not pretty (as it uses its arguments
> more than once), but it works:

Ah yes; thanks! This is another "for each" macro that that confuses
Coverity. I'll try to silence these...

Thanks for the details and taking a look at it!

-Kees

> 
> | #define mcp251xfd_for_each_rx_ring(priv, ring, n) \
> | 	for ((n) = 0, (ring) = *((priv)->rx + (n)); \
> | 	     (n) < (priv)->rx_ring_num; \
> | 	     (n)++, (ring) = *((priv)->rx + (n)))
> 
> For now there is only one rx ring...
> 
> | struct mcp251xfd_priv {
> |       [...]
> | 	struct mcp251xfd_rx_ring *rx[1];
> | 	u8 rx_ring_num;
> |       [...]
> | }
> 
> ...and rx_ring_num is initialized as "1".
> 
> | 	for (i = 0;
> | 	     i < ARRAY_SIZE(priv->rx) && ram_free >= rx_obj_size;
> | 	     i++) {
> |               [...]
> | 	}
> | 	priv->rx_ring_num = i;
> 
> 
> > 192     }
> > 193
> > 194     static void mcp251xfd_dump_tx_ring(const struct mcp251xfd_priv *priv,
> > 195     				   struct mcp251xfd_dump_iter *iter)
> > 
> > If this is a false positive, please let us know so we can mark it as
> > such, or teach the Coverity rules to be smarter. If not, please make
> > sure fixes get into linux-next. :) For patches fixing this, please
> > include these lines (but double-check the "Fixes" first):
> 
> This looks indeed like a false positive to me.
> 
> > Reported-by: coverity-bot <keescook+coverity-bot@chromium.org>
> > Addresses-Coverity-ID: 1503585 ("Memory - illegal accesses")
> > Fixes: e0ab3dd5f98f ("can: mcp251xfd: add dev coredump support")
> 
> regards,
> Marc
> 
> -- 
> Pengutronix e.K.                 | Marc Kleine-Budde           |
> Embedded Linux                   | https://www.pengutronix.de  |
> Vertretung West/Dortmund         | Phone: +49-231-2826-924     |
> Amtsgericht Hildesheim, HRA 2686 | Fax:   +49-5121-206917-5555 |



-- 
Kees Cook