On Feb 5, 2013, at 7:23 PM, Alexander Hansen <alexanderk.han...@gmail.com>
wrote:
> On 2/5/13 2:23 PM, Alexander Hansen wrote:
>> On 2/5/13 1:52 PM, Eigil Krogh Sorensen wrote:
>>>
>>> Den 05/02/13 21.03 skrev "Alexander Hansen" <alexanderk.han...@gmail.com>:
>>>
>>>> On 2/5/13 10:35 AM, Eigil Krogh Sorensen wrote:
>>>>> I don't know if this is the right place to post this, but anyway...
>>>>>
>>>>> I have this small octave function, that runs fine on my Windows Octave,
>>>>> but crashes on my fink octave-atlas(3.6.3):
>>>>>
>>>>> "
>>>>> function simpleImageExample
>>>>> imageFile = "100-0039_IMG.JPG";
>>>>> info = imfinfo (imageFile)
>>>>> I = imread (imageFile);
>>>>> subplot(2,1,1);
>>>>> imshow(I);
>>>>> endfunction
>>>>> "
>>>>>
>>>>>
>>>>> If I run the function as shown above, octave crashes with this messages
>>>>> "
>>>>> warning: your version of GraphicsMagick limits images to 16 bits per
>>>>> pixel
>>>>> octave(18942,0x10b2db180) malloc: *** error for object 0x10b989570:
>>>>> pointer bei
>>>>> ng freed was not allocated
>>>>> *** set a breakpoint in malloc_error_break to debug
>>>>> panic: Abort trap: 6 -- stopping myself...
>>>>> attempting to save variables to `octave-core'...
>>>>> save to `octave-core' complete
>>>>> "
>>>>>
>>>>>
>>>>> The problem occurs in "imfinfo" because if I remove the line
>>>>> "info = imfinfo (imageFile)"
>>>>>
>>>>> then I only get the warning
>>>>> "
>>>>>>>> simpleImageExample
>>>>> warning: your version of GraphicsMagick limits images to 16 bits per
>>>>> pixel
>>>>>>>>
>>>>> "
>>>>>
>>>>> and the picture looks OK.
>>>>>
>>>>>
>>>>> --Eigil
>>>>>
>>>>>
>>>>>
>>>>
>>>> I can confirm this behavior.
>>>>
>>>> When I tried a local rebuild of graphicsmagick using
>>>> "--with-quantum-depth=32" I still got a "warning: your version of
>>>> GraphicsMagick limits images to 16 bits per pixel" message from Octave.
>>>>
>>>> The malloc() error goes away, but then the behavior gets really strange.
>>>> I made up identical test.jpg and test.jpeg files, so one would think
>>>> the error messages, if any, would be the same, but instead I get
>>>> (10.8/x86_64):
>>>>
>>>> octave:1> imfinfo ("test.jpeg")
>>>> warning: your version of GraphicsMagick limits images to 16 bits per pixel
>>>> error: Magick++ exception: Magick: Unable to open file (g) reported by
>>>> magick/constitute.c:1533 (ReadImage)
>>>> error: called from:
>>>> error: /sw/share/octave/3.6.3/m/image/imfinfo.m at line 147, column 10
>>>> octave:1> imfinfo ("test.jpg")
>>>> error: Magick++ exception: Magick: No decode delegate for this image
>>>> format () reported by magick/constitute.c:1522 (ReadImage)
>>>> error: called from:
>>>> error: /sw/share/octave/3.6.3/m/image/imfinfo.m at line 147, column 10
>>>>
>>>> When I rebuilt graphicsmagick using "--with-quantum-depth=8" the results
>>>> are still wrong, but at least more consistent:
>>>>
>>>> octave:1> imfinfo ("test.jpeg")
>>>> warning: your version of GraphicsMagick limits images to 16 bits per pixel
>>>> error: Magick++ exception: Magick: No decode delegate for this image
>>>> format () reported by magick/constitute.c:1522 (ReadImage)
>>>> error: called from:
>>>> error: /sw/share/octave/3.6.3/m/image/imfinfo.m at line 147, column 10
>>>> octave:1> imfinfo ("test.jpg")
>>>> error: Magick++ exception: Magick: No decode delegate for this image
>>>> format () reported by magick/constitute.c:1522 (ReadImage)
>>>> error: called from:
>>>> error: /sw/share/octave/3.6.3/m/image/imfinfo.m at line 147, column 10
>>>>
>>>>
>>>>
>>>> On 10.5/powerpc I get malloc() error messages but the image information
>>>> shows up (same test.jpg and test.jpeg):
>>>>
>>>> octave:2> imfinfo ("test.jpeg")
>>>> octave(47725) malloc: *** error for object 0xa08bf708: Non-aligned
>>>> pointer being freed
>>>> *** set a breakpoint in malloc_error_break to debug
>>>> octave(47725) malloc: *** error for object 0xa08bf708: Non-aligned
>>>> pointer being freed
>>>> *** set a breakpoint in malloc_error_break to debug
>>>> ans =
>>>>
>>>> scalar structure containing the fields:
>>>>
>>>> Filename = /Users/hansen/test.jpeg
>>>> FileModDate = 5-Feb-2013 12:21:54
>>>> FileSize = 26699
>>>> Height = 241
>>>> Width = 241
>>>> BitDepth = 8
>>>> Format = JPEG
>>>> LongFormat = Joint Photographic Experts Group JFIF format
>>>> XResolution = 0
>>>> YResolution = 0
>>>> TotalColors = 68
>>>> TileName =
>>>> AnimationDelay = 0
>>>> AnimationIterations = 0
>>>> ByteOrder = undefined
>>>> Gamma = 0
>>>> Matte = 0
>>>> ModulusDepth = 8
>>>> Quality = 75
>>>> QuantizeColors = 256
>>>> ResolutionUnits = undefined
>>>> ColorType = grayscale
>>>> View =
>>>>
>>>> octave:3> imfinfo ("test.jpg")
>>>> octave(47725) malloc: *** error for object 0xa08bf708: Non-aligned
>>>> pointer being freed
>>>> *** set a breakpoint in malloc_error_break to debug
>>>> octave(47725) malloc: *** error for object 0xa08bf708: Non-aligned
>>>> pointer being freed
>>>> *** set a breakpoint in malloc_error_break to debug
>>>> ans =
>>>>
>>>> scalar structure containing the fields:
>>>>
>>>> Filename = /Users/hansen/test.jpg
>>>> FileModDate = 5-Feb-2013 12:21:54
>>>> FileSize = 26699
>>>> Height = 241
>>>> Width = 241
>>>> BitDepth = 8
>>>> Format = JPEG
>>>> LongFormat = Joint Photographic Experts Group JFIF format
>>>> XResolution = 0
>>>> YResolution = 0
>>>> TotalColors = 68
>>>> TileName =
>>>> AnimationDelay = 0
>>>> AnimationIterations = 0
>>>> ByteOrder = undefined
>>>> Gamma = 0
>>>> Matte = 0
>>>> ModulusDepth = 8
>>>> Quality = 75
>>>> QuantizeColors = 256
>>>> ResolutionUnits = undefined
>>>> ColorType = grayscale
>>>> View =
>>>>
>>>> I'll investigate this further.
>>>>
>>>> --
>>>> Alexander Hansen, Ph.D.
>>>> Fink User Liaison
>>>> My package updates: http://finkakh.wordpress.com/
>>>
>>> FYI:
>>> Did try to rebuild graphicsMagick (and octave-atlas had to be build also)
>>> to 32 bits. On my system that din not make the performance noticeably
>>> worse.
>>>
>>> Now I'm back on latest fink versions of GraphicsMagick and octave-atlas
>>>
>>>
>>> --Eigil
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>
>> I wouldn't think that you'd need to rebuild octave-atlas. It links to
>> graphicsmagick dynamically rather than statically.
>>
>
> Ah, but I thought incorrectly. You're quite correct that you need to
> rebuild Octave when you change the GraphicsMagick quantum size.
>
> For further tests, I tried Fink's ImageMagick instead of GraphicsMagick,
> but I got the same error. And I tried different compilers and
> optimizations, but the error persists.
>
> I'm trying 10.6/i386 and 10.6/x86_64 now, to see if the problem is
> restricted to 64-bit builds.
I've updated graphicsmagick to the latest upstream snapshot, don't know if
that'll help. I've also added a graphicsmagick-q32 package with matching -dev
and -shlibs for those who wish to experiment. It uses --with-quantum-depth=32
and is installed so that it can coexist with the existing package.
Also, for those who have been experimenting with different builds of GM, the
quantum bit depth is recorded in the library's header files, so you do have to
rebuild octave when using a different GM or it is likely to die horribly.
I don't know why octave is complaining about the quantum depth being 8 bit
since no fink GM has been built that way. I suspect there's a problem in
octave's build. Alexander is trying different things to narrow it down.
Daniel
------------------------------------------------------------------------------
Free Next-Gen Firewall Hardware Offer
Buy your Sophos next-gen firewall before the end March 2013
and get the hardware for free! Learn more.
http://p.sf.net/sfu/sophos-d2d-feb
_______________________________________________
Fink-users mailing list
Fink-users@lists.sourceforge.net
List archive:
http://news.gmane.org/gmane.os.macosx.fink.user
Subscription management:
https://lists.sourceforge.net/lists/listinfo/fink-users
