[vtkusers] Crash with FixedPointVolumeRayCastMapper

[Antoine Rosset]

Program received signal EXC_BAD_ACCESS, Could not access memory.
Reason: KERN_INVALID_ADDRESS at address: 0xce9f6917
[Switching to process 8764 thread 0x88fb]
0x0063e200 in  
vtkFixedPointCompositeShadeHelperGenerateImageOneTrilin<float> ()

#0  0x0063e200 in  
vtkFixedPointCompositeShadeHelperGenerateImageOneTrilin<float> ()
#1  0x0028e4c4 in  
vtkFixedPointVolumeRayCastCompositeShadeHelper::GenerateImage ()
#2  0x001bd8b4 in FixedPointVolumeRayCastMapper_CastRays ()
#3  0x9002c3b4 in _pthread_body ()

With : reader->SetDataSpacing( 4.6875, 4.6875, 15)

Sincerely,

Antoine Rosset

<> <> <> <> <> <> <> <> <> <> <> <> <> <> <> <> <><> <>
<> OSIRIX: Next Generation DICOM Viewer
<> http://homepage.mac.com/rossetantoine/osirix
<>
<> Antoine ROSSET, M.D. (RossetAntoine at bluewin.ch)
<> Blog : http://rossetantoine.blogspot.com/
<> Geneva University Hospital - Radiology Department
<> 20 Micheli-Du-Crest, CH-1205 Geneva
<> Switzerland, Europe
<> <> <> <> <> <> <> <> <> <> <> <> <> <> <> <> <><> <>


On May 16, 2005, at 3:36 PM, Mathieu Malaterre wrote:

> Antoine,
>
>     Your feedback definitely narrows down the parameters, but it  
> would be even more usefull if you could send us the backtrace of  
> the crash. I'll assume you are running osirix compiled in Debug  
> mode for that. Simply do:
>
>
> $ gdb osirix
> (gdb) run
>
> ... reload your usual volume ...
> when your program hangs then simply do:
>
> (gdb) bt
>
> Thanks a lot
> Mathieu
>
>
> Antoine Rosset wrote:
>
>> Hi,
>> Yes, this crash occurs repeatedly when I rotate the object:  
>> suddenly  it crashes at a specific angle. It seems the function  
>> try to access  something outside the data memory zone.
>> Here are the parameters:
>> - float and unsigned char data (data is 512 x 512 x 51)
>> - 1 or 4 components
>> - with or without shading
>> - with or without linear opacity function
>> The crashes are not influenced by these parameters.
>> If I replace the vtkFixedPointVolumeRayCastMapper function by   
>> vtkVolumeRayCastMapper function, the crashes completely disappear.
>> BUT: if I change the reader->SetDataSpacing( 0.46875, 0.46875,  
>> 1.5)  function to reader->SetDataSpacing( 1, 1, 1); the crashes  
>> DISAPPEAR.
>> The library is compiled under MacOS X 10.4, dual-G5, 2 GB of RAM.
>> BTW: This new function accelerate renderings by a factor from 4 to  
>> 8  on dual-PowerPC systems compared to vtkVolumeRayCastMapper.  
>> And  thanks for the new RGB support!
>> Sincerely,
>> Antoine Rosset
>> <> <> <> <> <> <> <> <> <> <> <> <> <> <> <> <> <><> <>
>> <> OSIRIX: Next Generation DICOM Viewer
>> <> http://homepage.mac.com/rossetantoine/osirix
>> <>
>> <> Antoine ROSSET, M.D. (RossetAntoine at bluewin.ch)
>> <> Blog : http://rossetantoine.blogspot.com/
>> <> Geneva University Hospital - Radiology Department
>> <> 20 Micheli-Du-Crest, CH-1205 Geneva
>> <> Switzerland, Europe
>> <> <> <> <> <> <> <> <> <> <> <> <> <> <> <> <> <><> <>
>> On May 16, 2005, at 6:09 AM, Lisa Avila wrote:
>>
>>> Hi Antoine,
>>>
>>> If you have some way of repeatedly causing the crash, I would  
>>> love  to hear about it. Also, can you describe the type of data  
>>> you are  rendering (data type and number of components, along  
>>> with the  options you've selected such as MIP vs. composite, was  
>>> shading on?  Is the gradient opacity function not a constant 1?),  
>>> and your system?
>>>
>>> Thanks,
>>>
>>> Lisa
>>>
>> _______________________________________________
>> This is the private VTK discussion list. Please keep messages on- 
>> topic. Check the FAQ at: http://www.vtk.org/Wiki/VTK_FAQ
>> Follow this link to subscribe/unsubscribe:
>> http://www.vtk.org/mailman/listinfo/vtkusers
>>
>
>