linux-kernel - Re: [PATCH] leds: ti-lmu-common: Fix coccinelle issue in TI LMU

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]

Message-ID: <f36eac2f-6643-c4f3-2b73-45f147b019a7@gmail.com>
Date:   Tue, 27 Aug 2019 23:02:17 +0200
From:   Jacek Anaszewski <jacek.anaszewski@...il.com>
To:     Dan Murphy <dmurphy@...com>, pavel@....cz
Cc:     linux-leds@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] leds: ti-lmu-common: Fix coccinelle issue in TI LMU

Dan,

On 8/27/19 3:37 PM, Dan Murphy wrote:
> Jacek
> 
> On 8/26/19 2:34 PM, Jacek Anaszewski wrote:
>> Dan,
>>
>> On 8/26/19 4:53 PM, Dan Murphy wrote:
>>> Jacek
>>>
>>> On 8/24/19 10:18 AM, Jacek Anaszewski wrote:
>>>> Hi Dan,
>>>>
>>>> Thank you for the patch.
>>>>
>>>> On 8/23/19 9:55 PM, Dan Murphy wrote:
>>>>> Fix the coccinelle issues found in the TI LMU common code
>>>>>
>>>>> drivers/leds/leds-ti-lmu-common.c:97:20-29: WARNING: Unsigned
>>>>> expression compared with zero: ramp_down < 0
>>>>> drivers/leds/leds-ti-lmu-common.c:97:5-12: WARNING: Unsigned
>>>>> expression compared with zero: ramp_up < 0
>>>> Wouldn't it make more sense to remove those pointless checks?
>>>> Clearly a correct index of an array cannot be negative.
>>>> Looking at the code I would make more int -> unsigned int conversions:
>>>>
>>>> - ramp_table should be unsigned int
>>>> - ti_lmu_common_convert_ramp_to_index should return unsigned int
>>>>
>>> Yeah I was going to just remove the code but when I was writing the
>>> original code my intent was
>>>
>>> to extend the ramp call to allow other TI LMU driver to pass in the
>>> device specific ramp table.
>>>
>>> But since I don't currently have any devices on my plate that require
>>> that I can just remove the code as well
>> You don't need to remove, just do the conversions I proposed.
>> Unless it introduces some other problems I am currently not aware of.
>>
> Well just converting those two would/did not fix the issue.

I implicitly assumed that you'd just drop the check since it
would make no sense to check unsigned int for being lower than 0.

And I propose to not return any error code from
ti_lmu_common_convert_ramp_to_index(), just make sure inside it
you return sane value. Ramp should, well, ramp.

> 
> But actually there is only 1 possibility that could happen if the
> convert function returns -EINVAL
> 
> So the check should be
> 
> if (ramp_up == -EINVAL || ramp_down == -EINVAL)
> 
> Because ramp_up/down should never be less then zero otherwise.
> 
> Dan
> 
> 
> 

-- 
Best regards,
Jacek Anaszewski