Matthias, On that note, why are you using Clip? Clip on rectilinear grid will cause a memory bloat. If you're making axis aligned sections, you may want to use Extract Subset instead.
Utkarsh On Fri, Feb 10, 2012 at 12:12 PM, Yumin Yuan <yumin.y...@kitware.com> wrote: > So this is an out-of-memory issue after all. > > It crashes at this place: > > vtkFastGeomQuad* vtkDataSetSurfaceFilter::NewFastGeomQuad(int numPts) > { > ... > // Next: allocate a new array if necessary. > if (this->FastGeomQuadArrays[this->NextArrayIndex] == NULL) > { > this->FastGeomQuadArrays[this->NextArrayIndex] > = new unsigned char[this->FastGeomQuadArrayLength]; > } > ... > } > > Basically, the "new unsigned char[]" failed. We or someone else > familiar with this filter could improve its memory usage later, but > for this release, I don't think we will have a fix for it. > > Yumin > > > On Fri, Feb 10, 2012 at 10:49 AM, Yumin Yuan <yumin.y...@kitware.com> wrote: >> I believe I reproduced the crash on my 64 bit XP machine with 8G RAM. >> Looking into it right now. >> >> Yumin >> >> On Fri, Feb 10, 2012 at 10:46 AM, David E DeMarle >> <dave.dema...@kitware.com> wrote: >>> Yes I know that Utkarsh. >>> >>> I still would like Dr Zenker to reconfirm so that we can reproduce, >>> isolate and fix the issue. >>> >>> Please do your comparison runs on 3.12.rc2 and 3.12.0. The change >>> delta between them is smaller than to 3.14.rc2 which should help to >>> isolate the specific cause. >>> >>> David E DeMarle >>> Kitware, Inc. >>> R&D Engineer >>> 21 Corporate Drive >>> Clifton Park, NY 12065-8662 >>> Phone: 518-881-4909 >>> >>> >>> >>> On Fri, Feb 10, 2012 at 9:50 AM, Utkarsh Ayachit >>> <utkarsh.ayac...@kitware.com> wrote: >>>> Dave, >>>> >>>> We know Matthias previous emails that he had the same crash on clip on >>>> 3.12. >>>> >>>> Utkarsh >>>> >>>> On Fri, Feb 10, 2012 at 8:45 AM, David E DeMarle >>>> <dave.dema...@kitware.com> wrote: >>>>> Also what if any plugins do you have loaded, for both versions. >>>>> >>>>> David E DeMarle >>>>> Kitware, Inc. >>>>> R&D Engineer >>>>> 21 Corporate Drive >>>>> Clifton Park, NY 12065-8662 >>>>> Phone: 518-881-4909 >>>>> >>>>> >>>>> >>>>> On Fri, Feb 10, 2012 at 8:36 AM, David E DeMarle >>>>> <dave.dema...@kitware.com> wrote: >>>>>> Can you confirm for us that the same operation on the same machine >>>>>> with the same data did work on 3.12RC2? >>>>>> >>>>>> David E DeMarle >>>>>> Kitware, Inc. >>>>>> R&D Engineer >>>>>> 21 Corporate Drive >>>>>> Clifton Park, NY 12065-8662 >>>>>> Phone: 518-881-4909 >>>>>> >>>>>> >>>>>> >>>>>> On Fri, Feb 10, 2012 at 3:52 AM, Zenker, Dr. Matthias >>>>>> <matthias.zen...@erbe-med.com> wrote: >>>>>>> Hi Kyle, >>>>>>> >>>>>>> I have downloaded and tested the new release candidate with the file I >>>>>>> sent you. It loads OK, but crashes after a certain action with the clip >>>>>>> filter. I get a crash when I do the following: >>>>>>> >>>>>>> 1. Open the file, read as "Legacy vtk file" >>>>>>> 2. Apply >>>>>>> 3. Display -> Style -> Surface, Color by -> Value_of_SAR[mW/g]_0s >>>>>>> 4. Clip, Properties->Z Normal, Apply >>>>>>> 5. rotate around y axis to see the cut plane >>>>>>> 6. move the plane with the mouse towards the observer, i.e. in negative >>>>>>> z direction >>>>>>> 7. Apply => crash >>>>>>> >>>>>>> Since this is a release candidate, there are chances that this can be >>>>>>> fixed for the release... ;) >>>>>>> >>>>>>> Thank you, >>>>>>> >>>>>>> Matthias >>>>>>> >>>>>>> >>>>>>> -----Ursprüngliche Nachricht----- >>>>>>> Von: Kyle Lutz [mailto:kyle.l...@kitware.com] >>>>>>> Gesendet: Dienstag, 7. Februar 2012 19:05 >>>>>>> An: Zenker, Dr. Matthias >>>>>>> Cc: John Drescher; paraview@paraview.org >>>>>>> Betreff: Re: [Paraview] Crash on vtr file load >>>>>>> >>>>>>> I've tried and was able to reproduce the crash on WinXP-32bit with >>>>>>> ParaView-3.12. I then compiled ParaView from git-master (SHA: 818a8e9) >>>>>>> and the file could be opened succesfully. Could you try compiling >>>>>>> ParaView from the git? If not, ParaView 3.14 should be released this >>>>>>> month and should work. >>>>>>> >>>>>>> Cheers, >>>>>>> Kyle >>>>>>> >>>>>>> On Tue, Feb 7, 2012 at 11:10 AM, Zenker, Dr. Matthias >>>>>>> <matthias.zen...@erbe-med.com> wrote: >>>>>>>> Maybe - but why doesn't 3.12-RC2 crash at file load? >>>>>>>> With 3.12-RC2, I can open a 204 MB file (the file I have uploaded has >>>>>>>> 29 MB). Paraview will only crash when I start to play around with the >>>>>>>> clip filter, it then takes some 0.9 GB of memory. With the 29 MB file >>>>>>>> and one clip filter, I see 424 MB of memory usage in the task manager. >>>>>>>> When I move the clip plane, the memory usage grows to some 0.8 GB, and >>>>>>>> Paraview crashes. >>>>>>>> But 3.12.0 will crash immediately when I try to load any of the vtr >>>>>>>> files. >>>>>>>> >>>>>>>> So IMHO memory is at least not the only issue here. >>>>>>>> >>>>>>>> Matthias >>>>>>>> >>>>>>>> -----Ursprüngliche Nachricht----- >>>>>>>> Von: John Drescher [mailto:dresche...@gmail.com] >>>>>>>> Gesendet: Dienstag, 7. Februar 2012 16:53 >>>>>>>> An: Kyle Lutz >>>>>>>> Cc: Zenker, Dr. Matthias; paraview@paraview.org >>>>>>>> Betreff: Re: [Paraview] Crash on vtr file load >>>>>>>> >>>>>>>>> I've downloaded your vtr file and it opened it with my build of >>>>>>>>> paraview (Linux, 64bit, git-master). The file opens and the clip >>>>>>>>> filter works as expected. The memory usage for ParaView is 1-1.5GB so >>>>>>>>> I don't think lack of memory on your machine is an issue. >>>>>>>>> >>>>>>>>> I also tried it with the 3.12 binaries from paraview.org and >>>>>>>>> everything worked as expected. >>>>>>>>> >>>>>>>>> Perhaps this is a Windows issue? >>>>>>>>> >>>>>>>> >>>>>>>> 1-1.5GB is enough to get an out of memory on 32 bit windows if the >>>>>>>> memory is for the image needs to be allocated in a single buffer. This >>>>>>>> is a result of address space fragmentation and the default 2GB maximum >>>>>>>> address space limit per process. >>>>>>>> >>>>>>>> John >>>>>>>> >>>>>>>> >>>>>>>> _____________________________________________________________________ >>>>>>>> ERBE Elektromedizin GmbH >>>>>>>> Firmensitz: 72072 Tuebingen >>>>>>>> Geschaeftsfuehrer: Christian O. Erbe, Reiner Thede >>>>>>>> Registergericht: Stuttgart HRB 380137 >>>>>>>> >>>>>>> >>>>>>> >>>>>>> _____________________________________________________________________ >>>>>>> ERBE Elektromedizin GmbH >>>>>>> Firmensitz: 72072 Tuebingen >>>>>>> Geschaeftsfuehrer: Christian O. Erbe, Reiner Thede >>>>>>> Registergericht: Stuttgart HRB 380137 >>>>>>> >>>>>>> _______________________________________________ >>>>>>> Powered by www.kitware.com >>>>>>> >>>>>>> Visit other Kitware open-source projects at >>>>>>> http://www.kitware.com/opensource/opensource.html >>>>>>> >>>>>>> Please keep messages on-topic and check the ParaView Wiki at: >>>>>>> http://paraview.org/Wiki/ParaView >>>>>>> >>>>>>> Follow this link to subscribe/unsubscribe: >>>>>>> http://www.paraview.org/mailman/listinfo/paraview >>>>> _______________________________________________ >>>>> Powered by www.kitware.com >>>>> >>>>> Visit other Kitware open-source projects at >>>>> http://www.kitware.com/opensource/opensource.html >>>>> >>>>> Please keep messages on-topic and check the ParaView Wiki at: >>>>> http://paraview.org/Wiki/ParaView >>>>> >>>>> Follow this link to subscribe/unsubscribe: >>>>> http://www.paraview.org/mailman/listinfo/paraview >>> _______________________________________________ >>> Powered by www.kitware.com >>> >>> Visit other Kitware open-source projects at >>> http://www.kitware.com/opensource/opensource.html >>> >>> Please keep messages on-topic and check the ParaView Wiki at: >>> http://paraview.org/Wiki/ParaView >>> >>> Follow this link to subscribe/unsubscribe: >>> http://www.paraview.org/mailman/listinfo/paraview _______________________________________________ Powered by www.kitware.com Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html Please keep messages on-topic and check the ParaView Wiki at: http://paraview.org/Wiki/ParaView Follow this link to subscribe/unsubscribe: http://www.paraview.org/mailman/listinfo/paraview