linux-kernel - Re: [PATCH 1/2] iio: generic_buffer: Cleanup when receiving signals

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: <481f5a02-6683-d8a1-b31d-2ea2011810f7@kernel.org>
Date:	Sun, 29 May 2016 20:11:28 +0100
From:	Jonathan Cameron <jic23@...nel.org>
To:	Crestez Dan Leonard <leonard.crestez@...el.com>,
	Peter Meerwald-Stadler <pmeerw@...erw.net>
Cc:	linux-iio@...r.kernel.org, linux-kernel@...r.kernel.org,
	Hartmut Knaack <knaack.h@....de>,
	Lars-Peter Clausen <lars@...afoo.de>,
	Daniel Baluta <daniel.baluta@...el.com>
Subject: Re: [PATCH 1/2] iio: generic_buffer: Cleanup when receiving signals

On 23/05/16 17:10, Crestez Dan Leonard wrote:
> On 05/21/2016 07:28 PM, Jonathan Cameron wrote:
>> On 20/05/16 16:55, Peter Meerwald-Stadler wrote:
>>>
>>>> This also drops all the code freeing string buffers at the end of main.
>>>> Memory is freed when the process exits anyway so there's no point in
>>>> cluttering the code with all those gotos.
>>>
>>> well, it helps to see that all memory has been released when looking for 
>>> leaks :)
>>> e.g. valgrind becomes much less useful when the program exits with tons of 
>>> memory still allocated
>> Beyond that we are looking at code here that will get cut and paste into other
>> peoples applications - they might not pick up that it doesn't clean up properly
>> after itself.
>>
>> I'd much prefer to keep these explicit frees in place.
> 
> I think this would make more sense for a library (like libiio). But
> isn't the code in tools/iio merely an a test tool?
Absolutely - but as we all know test tool code gets copied when one is an hurry!
> 
> I submitted v2 which keeps the frees. It still simplifies them by
> relying on stuff like free(NULL) being allowed.
That's a nicer approach.

Jonathan
>