On Thu, 28 Apr 2016 20:03:27 +0200
Benoit Gschwind wrote:
> Hello Pekka,
>
> The patch look good, but leave a leak in backend_init: the param.device
> is never freed.
That is because it is never malloc'd either, so this patch is fine.
> This fixed you get my Reviewed-by ;)
Pushed this and the
Hello Pekka,
The patch look good, but leave a leak in backend_init: the param.device
is never freed.
This fixed you get my Reviewed-by ;)
Best regards
Le 28/04/2016 16:08, Pekka Paalanen a écrit :
> From: Pekka Paalanen
>
> Ensuring that the pointer to the device path stays valid gets harder
On Thu, 28 Apr 2016 16:23:03 +0200
Quentin Glidic wrote:
> On 28/04/2016 16:08, Pekka Paalanen wrote:
> > From: Pekka Paalanen
> >
> > Ensuring that the pointer to the device path stays valid gets harder and
> > harder with migrating to the libweston-style config handling. Therefore,
> > make a
On 28/04/2016 16:08, Pekka Paalanen wrote:
From: Pekka Paalanen
Ensuring that the pointer to the device path stays valid gets harder and
harder with migrating to the libweston-style config handling. Therefore,
make a copy of the string, private to struct fbdev_output.
Now the pointer passed in
From: Pekka Paalanen
Ensuring that the pointer to the device path stays valid gets harder and
harder with migrating to the libweston-style config handling. Therefore,
make a copy of the string, private to struct fbdev_output.
Now the pointer passed in to fbdev_output_create() could be freed righ
