[Pkg-gmagick-im-team] Bug#773980: libmagickcore-6.q16-2: 8:6.8.9.9-4 breaks images in GNU Emacs 24 (displayed as single colour rectangles)

[Bastien ROUCARIES]

On Sat, Dec 27, 2014 at 8:33 PM, Bastien ROUCARIES
<roucaries.bastien at gmail.com> wrote:
> control: tags -1 + confirmed
>
> Hi,
>
> I found it
>
> In order to reproduce run
> convert rose: rose.xpm
> display rose.xpm
>
> it should show strange stuff.
>
> The problematic commit is http://trac.imagemagick.org/changeset/17297
aka
fcf54f0b0f1a277f2e480ac37d731c6a0e5eb473
> and it is needed from a security point of view :S
>
> Will see we upstream
>
> On Sat, Dec 27, 2014 at 8:01 PM, Bastien ROUCARIES
> <roucaries.bastien at gmail.com> wrote:
>> On Sat, Dec 27, 2014 at 3:11 PM, Adam Sjøgren <asjo at koldfront.dk> wrote:
>>> Bastien writes:
>>>
>>>> On Fri, Dec 26, 2014 at 10:24 PM, Adam Sjøgren <asjo at koldfront.dk> wrote:
>>>>> Bastien writes:
>>>
>>>>>>> When displaying an image in GNU Emacs 24 (package emacs24), after
>>>>>>> upgrading from ImageMagick 8:6.8.9.9-3 to 8:6.8.9.9-4, images with
>>>>>>> :type 'imagemagick are displayed as single colour rectangles.
>>>
>>>>>> Thanks could you get the exact command running ?
>>>
>>>>> I don't understand your question.
>>>
>>>> I suppose emacs execute imagmagick. Thus that is the command (shell)
>>>> and parameter that emacs use ?
>>>
>>> I don't think Emacs executes any ImageMagick commands - Emacs is linked
>>> to the ImageMagick libraries, which is why I reported the bug against
>>> libmagickcore-6, and not on the package with the binaries.
>>>
>>> I.e.:
>>>
>>>   $ ldd /usr/bin/emacs24 | grep -i magick
>>>           libMagickWand-6.Q16.so.2 => /usr/lib/x86_64-linux-gnu/libMagickWand-6.Q16.so.2 (0x00007f7d415b8000)
>>>           libMagickCore-6.Q16.so.2 => /usr/lib/x86_64-linux-gnu/libMagickCore-6.Q16.so.2 (0x00007f7d410f8000)
>>>   $
>>>
>>> And this is also why the recipe for reproducing the problem I wrote
>>> includes running Emacs.
>>
>>
>> It seems it is only xpm file. Could you try with other format of input
>> image ? I have applied some patch to xpm coder in order to solve
>> security problem. Will try to reproduce.
>>
>> Staying in -3 is not a solution due to the security problem.
>>
>> Bastien
>>
>>>
>>> I hope this clears it up.
>>>
>>>
>>>   Best regards,
>>>
>>>     Adam
>>>
>>> --
>>>  "May the force be...                                         Adam Sjøgren
>>>   ... equal to mass · acceleration"                      asjo at koldfront.dk