Re: Coverity: mt7915_mcu_get_chan_mib_info(): Memory - illegal accesses

[Kees Cook]

On Fri, Dec 02, 2022 at 11:42:36PM +0000, Ryder Lee wrote:
> On Fri, 2022-12-02 at 15:04 -0800, Kees Cook wrote:
> > > 
> > On Fri, Dec 02, 2022 at 10:56:19PM +0000, Ryder Lee wrote:
> > > On Fri, 2022-12-02 at 14:24 -0800, coverity-bot wrote:
> > > > Hello!
> > > > 
> > > > This is an experimental semi-automated report about issues
> > > > detected
> > > > by
> > > > Coverity from a scan of next-20221202 as part of the linux-next
> > > > scan
> > > > project:
> > > > 
> > > 
> > > 
> https://urldefense.com/v3/__https://scan.coverity.com/projects/linux-next-weekly-scan__;!!CTRNKA9wMg0ARbw!j7j_C0KpO4VD2yMOodvpeIexTGq4fhy2yq6nokNua9u4LToiUOLk4ou8JFFNrXkrh80d5BK2k44faRQstHE9$
> > >  
> > > >  
> > > > 
> > > > You're getting this email because you were associated with the
> > > > identified
> > > > lines of code (noted below) that were touched by commits:
> > > > 
> > > >   Thu Feb 3 13:57:56 2022 +0100
> > > >     417a4534d223 ("mt76: mt7915: update mt7915_chan_mib_offs for
> > > > mt7916")
> > > > 
> > > > Coverity reported the following:
> > > > 
> > > > *** CID 1527801:  Memory - illegal accesses  (OVERRUN)
> > > > drivers/net/wireless/mediatek/mt76/mt7915/mcu.c:3005 in
> > > > mt7915_mcu_get_chan_mib_info()
> > > > 2999     		start = 5;
> > > > 3000     		ofs = 0;
> > > > 3001     	}
> > > > 3002
> > > > 3003     	for (i = 0; i < 5; i++) {
> > > > 3004     		req[i].band = cpu_to_le32(phy->mt76->band_idx);
> > > > vvv     CID 1527801:  Memory - illegal accesses  (OVERRUN)
> > > > vvv     Overrunning array "offs" of 9 4-byte elements at element
> > > > index 9 (byte offset 39) using index "i + start" (which evaluates
> > > > to
> > > > 9).
> > > > 3005     		req[i].offs = cpu_to_le32(offs[i + start]);
> > > > 3006
> > > > 3007     		if (!is_mt7915(&dev->mt76) && i == 3)
> > > > 3008     			break;
> > > > 3009     	}
> > > > 3010
> > > > 
> > > > 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):
> > > > 
> > > 
> > > I think this is a false postive as the subsequent check 'if
> > > (!is_mt7915(&dev->mt76) && i == 3)' should break array "offs" of 8.
> > 
> > Ah, okay. What if is_mt7915(&dev->mt76) is always true?
> > 
> > -Kees
> 
> 	int start = 0;
> 
> 	if (!is_mt7915(&dev->mt76)) {
> 		start = 5;
> 		ofs = 0;
> 	}
> 
> 	for (i = 0; i < 5; i++) {
> 		req[i].band = cpu_to_le32(phy->band_idx);
> 		req[i].offs = cpu_to_le32(offs[i + start]);
> 
> 		if (!is_mt7915(&dev->mt76) && i == 3) //
> 			break;
> 	}
> 
> For 'is_mt7915' case, start:0 and i: 0 1 2 3 4, whereas !is_mt7915'
> case, start:5 and i: 0 1 2 3 (then break).
> 
> I know it's a bit tricky. This is used to differentiate chipset
> revision.

Ah-ha! Gotcha now. Thanks for the details and sorry for the noise! :)

-Kees

-- 
Kees Cook