linux-kernel - Re: [RFC] iio: Release irq if set_trigger

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 for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <5728988D.1080502@intel.com>
Date:	Tue, 3 May 2016 15:24:45 +0300
From:	Crestez Dan Leonard <leonard.crestez@...el.com>
To:	Jonathan Cameron <jic23@...nel.org>, linux-iio@...r.kernel.org
Cc:	linux-kernel@...r.kernel.org, Hartmut Knaack <knaack.h@....de>,
	Lars-Peter Clausen <lars@...afoo.de>,
	Peter Meerwald-Stadler <pmeerw@...erw.net>,
	Daniel Baluta <daniel.baluta@...el.com>
Subject: Re: [RFC] iio: Release irq if set_trigger_state fails

On 05/01/2016 10:11 PM, Jonathan Cameron wrote:
> On 28/04/16 17:40, Crestez Dan Leonard wrote:
>> When attaching a pollfunc iio_trigger_attach_poll_func will allocate a
>> virtual irq and call the driver's set_trigger_state function. Fix error
>> handling to release the irq if set_trigger_state fails.
>>
>> Otherwise when using triggered buffers and the driver's
>> set_trigger_state fails once then the buffer becomes unusable.
>>
>> It is not possible to handle this sort of error by calling
>> iio_trigger_detach_poll_func externally somehow. That function should
>> only be called if attach is successful.
>>
>> Signed-off-by: Crestez Dan Leonard <leonard.crestez@...el.com>
> I'm embarrassed :(  good find! Not sure why you made such an obvious
> bug fix an RFC though! 
> 
> Slight issue in the new error handling though I think..
> 
>> +
>> +out_put_irq:
>> +	iio_trigger_put_irq(trig, pf->irq);
>> +	free_irq(pf->irq, pf);
>> +out_put_module:
> I think the iio_trigger_put_irq should be here as it will have been gotten before the 
> request_threaded_irq call and hence should always be unwound on error not just in the
> case above.
>
Yes. I sent V2 which should properly handle errors from
iio_trigger_get_irq and request_threaded_irq separately.

-- 
Regards,
Leonard