[Surefit-users] error messages
Dear list, I am encountering some problems in Running SureFit 4.38 on a Red Hat 7.3 Linux platform. Oddly, the program used to work all right about 8 months ago, when I first installed it. I have not been using it for a while and just resumed using it now. The first error is in trying to run the Analyze2Minc program: a .mnc is generated, however the file is empty (size 0 bytes). This is the report I get: -- $ Analyze2Minc nE15512S3_A.hdr Traceback (innermost last): File "./UTILITIES/Analyze2Minc.py", line 62, in ? File "./volume.py", line 250, in WriteNetCDFFile File "/usr/local/lib/python1.5/site-packages/Scientific/IO/NetCDF.py", line 143, in NetCDFFile IOError: netcdf: Operation not permitted Starting SureFitOps... dims: 79 95 68 bitpix 16 reads as typecode s Image will be byteswapped. Read netCDF file nE15512S3_A.img Read RAW file nE15512S3_A.img Reading short Volume Dimensions 79 95 68 Reading complete nE15512S3_A.img s s max, min 135 -9 Scale, size of array 510340 3; strides 15010 158 2 Min -9.00; max 135.00 scale 1.89 min 0.00 at 95, max 135.00 at 3894 I have also tried doing the same thins as a superuser, but that's even worse: - bash-2.05a# Analyze2Minc nE15512S3_A.hdr python: error while loading shared libraries: libtk8.0.so: cannot open shared object file: No such file or directory - I then converted the file to .mnc format on a Sun Ultrasparc station where the same SureFit version (4.38) works without problems, and then went back to the Linux platform to run the segmentation. I get an other error message when starting the segmentation by clicking on 'Run SureFit'. The following error message appears in a separate windows and the process terminates: Error: 1 IOError Exception in Tk callback Function: at 8672910> (type: ) Args: () Traceback (innermost last): File "/usr/local/lib/python1.5/site-packages/Pmw/Pmw_0_8_3/lib/PmwBase.py", line 1653, in __call__ None File "./RunSureFit.py", line 277, in None File "./RunSureFit.py", line 114, in DoShellScript None File "./RunSureFit.py", line 151, in DoShellScriptNow None File "./Scripts.py", line 656, in SegmentCerebralWhiteMatter None File "./Scripts.py", line 4664, in WriteNetCDFFile None File "/usr/local/lib/python1.5/site-packages/Scientific/IO/NetCDF.py", line 143, in NetCDFFile None IOError: netcdf: Operation not permitted This is the report I get in the shell from which SureFit was launched: - $ 'VTK Version vtk version 3.2.0, vtk source $Revision: 1.788 $, $Date: 2001/02/23 01:04:24 $ (GMT) VTK Version vtk version 3.2.0, vtk source $Revision: 1.788 $, $Date: 2001/02/23 01:04:24 $ (GMT) Checking pad parameters ACx, ACy, ACz: 0 0 0 Starting SureFitOps... ['/usr/local/bin/SUREfit/lib/python1.5/site-packages/SureFit.pyc'] Read netCDF file /mnt/data/work/somatotopy/subj_03/tmp/nE15512S3_A.L.full.sMRI.mnc CDF file 83 189 136 1 126 -128 NETCDF Reading complete /mnt/data/work/somatotopy/subj_03/tmp/nE15512S3_A.L.full.sMRI.mnc b b 254 0 Reloading parameters Paramsfile found: /mnt/data/work/somatotopy/subj_03/tmp/nE15512S3_A.L.full.sMRI.params WMThresh = 68.0 Checking pad parameters ACx, ACy, ACz: 78 115 44 Initiate SliceVis SwitchVolumeToDisplay 0 Set Horizontal Plane ClickedOnTab Volume Information datatype = MRI Resolution set to 1.0 Species = human Volume already cropped -- not a whole volume. Hem = 0 oriented = yes Checking pad parameters Set Sagittal Plane ClickedOnTab List Parameters WMThresh = 68.0 Checking pad parameters WMThresh = 68.0 Running SureFit: self.segment extract_cwm self.fillvents yes self.correcterrors yes self.prepflatten no self.labelsulci yes Segment Hindbrain 83 189 136 mkdir /mnt/data/work/somatotopy/subj_03/tmp/INTERMEDIATE_VOLUMES Hem1: 0 Hem2: 1 HemDbl: 0 Hem3: -1 xAC_1: 77 xAC_8: 70 xAC_10: 68 xAC_15: 63 xAC_20: 58 xAC_30: 48 xAC_40: 38 xAC_45: 33 xAC_50: 28 xAC_1_low: 77 xAC_1_high: 78 xAC_20_low: 58 xAC_20_high: 78 xAC_45_low: 33 xAC_45_high: 78 xAC_10_40_low: 38 xAC_10_40_high: 68 xAC_15_low: 63 xAC_15_high: 78 xAC_15_40_low: 38 xAC_15_40_high: 63 xAC_40_low: 38 xAC_40_high: 78 xAC_20_30_low: 48 xAC_20_30_high: 58 xMedLimit_20_low: 58 xMedLimit_20_high: 83 xMedLimit_40_low: 38 xMedLimit_40_high: 83 xMed2AClow: 78 xMed2AChigh: 83 xAClow: 0 xAChigh: 78 xMedLimit_low: 78 xMedLimit_high: 83 WMThresh = 68.0 Threshold 68.00 ThresholdVolume 68 Thresholded 511162 voxels 23.96... --------
Re: [Surefit-users] error messages
Dear Donna, thanks a lot for your reply. I have lot of free space (more than 12 GB), so I don't think that this can be the problem. The recommended environment variables in the .bashrc file were already set, so this cannot be the problem neither. Nevertheless, I have tried upgrading to version 4.45. The error messages do change a bit, but I still get 0 bytes .mnc images after conversion with Analyze2Minc and I still get an error message when clicking on 'Run SureFit'. I report the error messages below for reference. Any help is greatly appreciated, Best wishes, Marco - $ Analyze2Minc nE15512S3_A.hdr Starting SureFitOps... dims: 79 95 68 bitpix 16 reads as typecode s Image will be byteswapped. Read minc file nE15512S3_A.img Read RAW file nE15512S3_A.img Reading short Volume Dimensions 79 95 68 Reading complete nE15512S3_A.img s s max, min 135 -9 Scale, size of array 510340 3; strides 15010 158 2 Min -9.00; max 135.00 scale 1.89 min 0.00 at 95, max 135.00 at 3894 Traceback (most recent call last): File "", line 1, in ? File "/usr/local/bin/SUREfit/lib/SureFit/Analyze2Minc.py", line 85, in ? vol.WriteNetCDFFile (mncfname, output) File "/usr/local/bin/SUREfit/lib/SureFit/volume.py", line 319, in WriteNetCDFFile fd = NetCDFFile (fname, 'w') File "./python2.1/site-packages/Scientific/IO/NetCDF.py", line 143, in NetCDFFile IOError: netcdf: Operation not permitted --- -- Error: 1 IOError Exception in Tk callback Function: at 0x8808a4c> (type: ) Args: () Traceback (innermost last): File "./python2.1/site-packages/Pmw/Pmw_0_8_5/lib/PmwBase.py", line 1690, in __call__ None File "./SureFit/RunSureFit.py", line 309, in command=lambda s=self: s.DoShellScript (8)) File "./SureFit/RunSureFit.py", line 131, in DoShellScript self.DoShellScriptNow (shell_index) File "./SureFit/RunSureFit.py", line 172, in DoShellScriptNow Scripts.SegmentCerebralWhiteMatter (SureSliceFcn.Volumes [0], keepfilesflag, refreshslice) File "/space/SureFitSrc/Linux/SUREfit/lib/SureFit/Scripts.py", line 741, in SegmentCerebralWhiteMatter WriteNetCDFFile (wmthreshfname, data, xdim, ydim, zdim) File "/space/SureFitSrc/Linux/SUREfit/lib/SureFit/Scripts.py", line 5050, in WriteNetCDFFile fd = NetCDFFile (fname, 'w') File "./python2.1/site-packages/Scientific/IO/NetCDF.py", line 143, in NetCDFFile None IOError: netcdf: Operation not permitted -- > Hi Marco, > > First, I recommend you upgrade to the latest version of SureFit. > Several bugs have been fixed since SureFit 4,38. I don't recall any > bugs with Analyze2Minc, but I do recall a bug fix affecting > Minc2Analyze. This is more of a general recommendation. > > Regarding your specific errors, they suggest a problem writing to your > local file system. The fact that you had an empty .mnc suggests you > have write permission, but whenever I've seen errors like yours, I've > had a full file system or a maxed quota. > > You can do "df ." to see if your local file system is full, or try just > copying the .img file (e.g., cp myanalyze.img newanalyze.img) and see if > that command succeeds or fails. > > I don't mind admitting that this has happened to me more than once. ;-) > > The reason your bash session fails is that SureFit requires some > environment variables (e.g., SFHOME, and a number of others in that old > release, including LD_LIBRARY_PATH). While your .cshrc file probably > has these variables set, they are evidently missing from your .bashrc or > equivalent environment file. but if I'm right about the true cause of > your problem, it doesn't matter what shell you use. > > Donna Hanlon > -- Marco Tettamanti Scientific Institute San Raffaele c/o L.I.T.A. - room 24/5 Via Fratelli Cervi 93 I-20090 Segrate (MI) Italy Tel. +39-02-21717548 Fax +39-02-21717558 email: [EMAIL PROTECTED]
Re: [Surefit-users] error messages
Dear Donna, I certanly do not have any permission problems, and in fact the same errors appear if I run SureFit as a superuser. The cp nE15512S3_A.img testimg.img command worked all right and diff nE15512S3_A.img testimg.img does not report any differences. I can remove files without any problems. Looking forward for a solution, Best wishes, Marco > Hi Marco, > > For whatever reason, SureFit can't write .mnc files in this directory. > Both of these operations are failing when trying to write the .mnc file, > with the complaint "IOError: netcdf: Operation not permitted". > > In the directory where you run Analyze2Minc, what happens when you > execute this command: > > cp nE15512S3_A.img testimg.img > > If it doesn't complain, then do "diff nE15512S3_A.img testimg.img" to > make sure the copy matches nE15512S3_A.img. If diff doesn't report a > difference, then we'll have to look harder for reasons why the file > system won't let SureFit write .mnc files. > > First, get rid of the previous empty .mnc file, in case it's read-only: > > rm nE15512S3_A.mnc > > Use 'ls -l' to make sure it is really gone. If it won't remove, we > could have a permissions problem (i.e., SureFit can't overwrite an > existing file of the same name). > > Please let me know what you find trying these commands. > > Donna -- Marco Tettamanti Scientific Institute San Raffaele c/o L.I.T.A. - room 24/5 Via Fratelli Cervi 93 I-20090 Segrate (MI) Italy Tel. +39-02-21717548 Fax +39-02-21717558 email: [EMAIL PROTECTED]
Re: [Surefit-users] error messages
Dear Donna, sorry to cause you so many troubles. I have uploaded the nE15512S3_A.img and the .hdr files to the website you suggested. Thanks a lot in advance Marco > Hi Marco, > > You have stumped me. It seems the usual suspects for write trouble > (full file system, lack of write permission) are not the culprits here. > > Given the nature of the problem, I doubt that looking at your volume > will shed much light on the issue. But since I am at a loss for better > ideas, I will give it a try. Please upload your .img and .hdr here: > > http://pulvinar.wustl.edu/cgi-bin/upload.cgi > > I'll give Analyze2Minc a shot and hope for more ideas. > > Donna -- Marco Tettamanti Scientific Institute San Raffaele c/o L.I.T.A. - room 24/5 Via Fratelli Cervi 93 I-20090 Segrate (MI) Italy Tel. +39-02-21717548 Fax +39-02-21717558 email: [EMAIL PROTECTED]
Re: [Surefit-users] error messages
Hi Donna, no problem, I hope you had nice vacations! Thanks for the hint on normalization: in fact I normally work with normalized files with a cubic voxel size of 1 mm; for some reasons, the file I have uploaded to your ftp site is 2x2x2, but it wasn't in my intentions. Sorry for that. I have now tried to run Analyze2Minc and the Run SureFit routine with a 1x1x1 mm normalized files: I get the same errors. I typed the ulimit -a command and this is the output: core file size(blocks, -c) 0 data seg size (kbytes, -d) unlimited file size (blocks, -f) unlimited max locked memory (kbytes, -l) unlimited max memory size (kbytes, -m) unlimited open files(-n) 1024 pipe size (512 bytes, -p) 8 stack size(kbytes, -s) unlimited cpu time (seconds, -t) unlimited max user processes(-u) 6139 virtual memory(kbytes, -v) unlimited I do not understand very much of this, but as far as I can see it looks quite similar to yours. I have tried setting core file size to unlimited (which seemed to me as one of the only relevant difference), but that doesn't solve things. I can write .params files with SureFit. No error message is reported and the parameters are properly stored. Thanks a lot for any further help Marco > Hi Marco, > > SureFit that is causing the trouble here -- not you. I am sorry I > wasn't able to solve this problem before leaving on vacation. > > I'm afraid I don't have many more ideas than I did a week ago. I was > able to Analyze2Minc your volume with no problem. One thing I did > notice was that the input volume appeared to be spatially normalized SPM > default grid -- perfectly fine, except you want to change your voxel > dimensions to cubic 1mm, which is what SureFit wants for humans. You > can do this by changing SPM's defaults for writing normalized; you can > use the .mat file SPM already generated during normalization. > > Back to the real problem: Python can't open .mnc files for writing > (occurs not only during Analyze2Minc, but also during first write .mnc > step in Run SureFit routine), complaining like this: > > File "/usr/local/bin/SUREfit/lib/SureFit/volume.py", line 319, in > WriteNetCDFFile > fd = NetCDFFile (fname, 'w') >File "./python2.1/site-packages/Scientific/IO/NetCDF.py", line 143, > in NetCDFFile > IOError: netcdf: Operation not permitted > > You have ruled out permissions problems and file space issues. I > considered low limit values: If your shell is csh or tcsh, enter > "limit" at the command line. My output looks like this: > > cputime unlimited > filesizeunlimited > datasizeunlimited > stacksize 8192 kbytes > coredumpsizeunlimited > memoryuse unlimited > descriptors 1024 > memorylockedunlimited > maxproc 7168 > openfiles 1024 > > If you're a bash user, use "ulimit -a" instead of limit. But if low > limits were the culprit, then I'd expect you to have trouble with the cp > command, which you were able to execute successfully. But if you have > lines like "limit filesize unlimited" in your .cshrc file (or just "cd > $HOME ; grep limit .*rc"), then this could explain why the shell could > do something python couldn't. Anyway, I doubt this is the problem. > > One other test might give us a bit more insight: When you load your > intensity volume in SureFit, are you able to save your settings under > the SureFit: Volume Preparation tab (i.e., write your .params file)? > I'm curious as to whether python can write *any* files -- .mnc or > .params. What happens when you press the Save Settings button? > > Donna > > marco tettamanti wrote: > > Dear Donna, > > sorry to cause you so many troubles. I have uploaded the nE15512S3_A.img > > and the .hdr files to the website you suggested. > > Thanks a lot in advance > > Marco > > > >>Hi Marco, > >> > >>You have stumped me. It seems the usual suspects for write trouble > >>(full file system, lack of write permission) are not the culprits here. > >> > >>Given the nature of the problem, I doubt that looking at your volume > >>will shed much light on the issue. But since I am at a loss for better > >>ideas, I will give it a try. Please upload your .img and .hdr here: > >> > >>http://pulvinar.wustl.edu/cgi-bin/upload.cgi > >> > >>I'll give Analyze2Minc a shot and hope for more ideas. > >> > >>Donna > > ___ > surefit-users mailing list > [EMAIL PROTECTED] > http://pulvinar.wustl.edu/mailman/listinfo/surefit-users -- Marco Tettamanti Scientific Institute San Raffaele c/o L.I.T.A. - room 24/5 Via Fratelli Cervi 93 I-20090 Segrate (MI) Italy Tel. +39-02-21717548 Fax +39-02-21717558 email: [EMAIL PROTECTED]
Re: [Surefit-users] error messages
Hi Donna, I do not know whether my file system is a FAT32 (I am not too familiar here). When I type: df -T I get: FilesystemType 1k-blocks Used Available Use% Mounted on /dev/hda6 ext3 7061696 4236480 2466496 64% / none tmpfs 386168 0386168 0% /dev/shm /dev/hda2 vfat20470016 8826176 11643840 44% /mnt/data /dev/hda6 is the partition where SureFit is located (/usr/local) and it doesn't look as a FAT32 system /dev/hda2 is the partition where I have my datasets; as far as I understand, this is a FAT32 system (?). Do you suggest to follow the hints given in the website you suggested (http://www.unidata.ucar.edu/packages/netcdf/known_problems.html#fat32-linux)? Which practical steps should I follow? 1) delete the following line in src/libsrc/ncconfig.h #define HAVE_FTRUNCATE 1 2) rebuild the library Which library should I rebuild? How do I do that? :-( Marco > Hi Marco, > > The limits look fine, which doesn't surprise me; it was a very long > shot, but easy enough to rule out. The core file size isn't the issue, > and I very much doubt the open files or max user processes limits are. > > I *am* surprised that you can write params files, though. Most > interesting. This is the first report of a problem specifically with > the netcdf-python stuff. I'd like to rule out a problem described here: > > http://www.unidata.ucar.edu/packages/netcdf/known_problems.html#fat32-linux > > Do you know whether your file system is a FAT32 file system? > > I'm not finding anything else promising on the unidata site, but the > python interface to netcdf is part of the ScientificPython module > (http://starship.python.net/~hinsen/ScientificPython). I'll look futher > for similar problems with this module. > > Also, I will contact you off-list about debugging alternatives. > > Donna -- Marco Tettamanti Scientific Institute San Raffaele c/o L.I.T.A. - room 24/5 Via Fratelli Cervi 93 I-20090 Segrate (MI) Italy Tel. +39-02-21717548 Fax +39-02-21717558 email: [EMAIL PROTECTED]
Re: [Surefit-users] error messages
Hi Donna, Wonderful! You hit the right point. I tried running Analyze2Minc in the /usr directory and it works perfectly. I have also tried to run the Run SureFit routine and it seems to work properly (the process is now in progress, I'll let you know about the outcome): the error message did not show up this time. Just a question: the Analyze2Minc routine worked properly and the .mnc looks fine, but the process generated a surefit.err file, which contains a number or error reports (I list report only some below, as they all look of the same type). A similar file is generated by the Run SureFit routine. This seems a general feature of version 4.45 (as opposed to versione 4.38, which didn't generate a surefit.err file). Should I worry about these error messages? And also, just a suggestion for future releases: a file like surefit.err where error reports are stored is definetely very useful, but it would help a lot if the error report was accompanied by the date and time when it occurred; it is otherwise sometimes hard to reconstruct to which process an error message was associated. Thanks a lot, Marco > Hi Marco, > > The file system to which SureFit is trying to write is a vfat file > system, which quite possibly won't permit the POSIX ftrucate function -- > the same problem the unidata site below describes for FAT32. > > I will try rebuilding the netcdf library with HAVE_FTRUNCATE undefined, > and I'll send it to you for testing. I don't expect you to build it, > but we may well be on to something here. > > Meanwhile, it wouldn't hurt for you to copy your .img and .hdr file to > /tmp, and try running Analyze2Minc there. Assuming /tmp is under /, an > ext3 file system, you might see success there, and it would support the > vfat-as-culprit hypothesis. -- Marco Tettamanti Scientific Institute San Raffaele c/o L.I.T.A. - room 24/5 Via Fratelli Cervi 93 I-20090 Segrate (MI) Italy Tel. +39-02-21717548 Fax +39-02-21717558 email: [EMAIL PROTECTED]
Re: [Surefit-users] error messages
ups! I forgot to include the content of the surefit.err file. Here are some of the error reports: ERROR: In /space/SureFitSrc/Linux.3.2.2.nothreads/VTK/Common/vtkDataObject.cxx, line 544 vtkImageData (0x8678c08): Update extent does not lie within whole extent ERROR: In /space/SureFitSrc/Linux.3.2.2.nothreads/VTK/Common/vtkDataObject.cxx, line 551 vtkImageData (0x8678c08): Update extent is: 0, 156, 0, 188, 0, 0 ERROR: In /space/SureFitSrc/Linux.3.2.2.nothreads/VTK/Common/vtkDataObject.cxx, line 558 vtkImageData (0x8678c08): Whole extent is: 0, 156, 0, 135, 0, 188 ERROR: In /space/SureFitSrc/Linux.3.2.2.nothreads/VTK/Common/vtkDataObject.cxx, line 544 vtkImageData (0x8679b28): Update extent does not lie within whole extent ERROR: In /space/SureFitSrc/Linux.3.2.2.nothreads/VTK/Common/vtkDataObject.cxx, line 551 vtkImageData (0x8679b28): Update extent is: 0, 156, 0, 188, 0, 0 ERROR: In /space/SureFitSrc/Linux.3.2.2.nothreads/VTK/Common/vtkDataObject.cxx, line 558 vtkImageData (0x8679b28): Whole extent is: 0, 156, 0, 135, 0, 188 ERROR: In /space/SureFitSrc/Linux.3.2.2.nothreads/VTK/Common/vtkDataObject.cxx, line 544 vtkImageData (0x8679b28): Update extent does not lie within whole extent ERROR: In /space/SureFitSrc/Linux.3.2.2.nothreads/VTK/Common/vtkDataObject.cxx, line 551 vtkImageData (0x8679b28): Update extent is: 0, 156, 0, 188, 0, 0 ERROR: In /space/SureFitSrc/Linux.3.2.2.nothreads/VTK/Common/vtkDataObject.cxx, line 558 vtkImageData (0x8679b28): Whole extent is: 0, 156, 0, 135, 0, 188 ERROR: In /space/SureFitSrc/Linux.3.2.2.nothreads/VTK/Common/vtkDataObject.cxx, line 544 vtkImageData (0x8679b28): Update extent does not lie within whole extent ERROR: In /space/SureFitSrc/Linux.3.2.2.nothreads/VTK/Common/vtkDataObject.cxx, line 551 vtkImageData (0x8679b28): Update extent is: 0, 156, 0, 188, 0, 0 Cheers, Marco > Hi Donna, > Wonderful! You hit the right point. I tried running Analyze2Minc in the > /usr directory and it works perfectly. I have also tried to run the Run > SureFit routine and it seems to work properly (the process is now in > progress, I'll let you know about the outcome): the error message did not > show up this time. > > Just a question: the Analyze2Minc routine worked properly and the .mnc > looks fine, but the process generated a surefit.err file, which contains a > number or error reports (I list report only some below, as they all look of > the same type). A similar file is generated by the Run SureFit routine. > This seems a general feature of version 4.45 (as opposed to versione 4.38, > which didn't generate a surefit.err file). > Should I worry about these error messages? > And also, just a suggestion for future releases: a file like surefit.err > where error reports are stored is definetely very useful, but it would help > a lot if the error report was accompanied by the date and time when it > occurred; it is otherwise sometimes hard to reconstruct to which process an > error message was associated. > > Thanks a lot, > Marco -- Marco Tettamanti Scientific Institute San Raffaele c/o L.I.T.A. - room 24/5 Via Fratelli Cervi 93 I-20090 Segrate (MI) Italy Tel. +39-02-21717548 Fax +39-02-21717558 email: [EMAIL PROTECTED]
Re: [Surefit-users] error messages (vfat and netcdf don't get along)
Hi Donna, great! I have dowloaded the library without HAVE_FTRUNCATE defined, and SureFit is now writing in the vfat filesystem without any problem (I have only tested Analyze2Minc for the moment). I have never encountered any problems with Caret, yet (hope I never will!). Thanks a lot! Marco > Hi Marco, > > Ignore error messages and surefit.err lines like those listed here: > > http://brainmap.wustl.edu/SureFit/RELEASE_NOTES.html#exitmessages > > I'll add your recommendation about adding the date to the surefit.err > file to my SureFit wish list (way down below the MacOSX port ;-). > > Now that I've solved some problems on my Linux build host, I've rebuilt > what I hope is the culprit library without HAVE_FTRUNCATE defined. > You'll need write permission to $SFHOME to install the new library. > > cd $SFHOME/lib/python2.1/site-packages/Scientific/linux2/ > mv Scientific_netcdf.so Scientific_netcdf.so.HAVE_FTRUNCATE > wget http://brainmap.wustl.edu/donna/SureFit/Linux/Scientific_netcdf.so > > If you don't have wget, you can download the file via browser into that > directory. Then, give Analyze2Minc another try on the vfat file system. > > Assuming you also will be using Caret and/or its fMRI mapper, your vfat > file system might also complain about functions using the netcdf > libraries, because they almost certainly were built with HAVE_FTRUNCATE > defined. But Caret uses netcdf *far* less than does SureFit. > > Donna -- Marco Tettamanti Scientific Institute San Raffaele c/o L.I.T.A. - room 24/5 Via Fratelli Cervi 93 I-20090 Segrate (MI) Italy Tel. +39-02-21717548 Fax +39-02-21717558 email: [EMAIL PROTECTED]
Re: [Surefit-users] error messages (vfat and netcdf don't get along)
Hi Donna, I have now gone through the whole Run SureFit routine and everything seems to work properly (writing in the /usr directory). All the files are properly written, the surfaces look fine, and the *spec file is properly recognized by caret. I will now test the Run SureFit routine, using the library without HAVE_FTRUNCATE defined. I' ll let you know. Thanks again, Marco -- Marco Tettamanti Scientific Institute San Raffaele c/o L.I.T.A. - room 24/5 Via Fratelli Cervi 93 I-20090 Segrate (MI) Italy Tel. +39-02-21717548 Fax +39-02-21717558 email: [EMAIL PROTECTED]
[Surefit-users] another error message
Hi Donna, bad news... I have tried running the Run Surefit routine in the vfat filesystem, as announced yesterday. While the routine worked properly in the /usr directory (ext3 filesystem), in the vfat filesystem an error occurred. The error message sounds: --- nccreate: filename "/mnt/data/work/somatotopy/surefit_mean17/tmp/ERROR_CORRECTION_INTERMEDIATES/NearCrossovers.mnc": Operation not permitted - It is generated after completion of the following step: InflateSurface done SetFileName /mnt/data/work/somatotopy/surefit_mean17/tmp/SURFACES/mean17s.L.full.segment_vent.ellipsoid.57640.vtk Starting to do normals... normals complete... Write surface data to /mnt/data/work/somatotopy/surefit_mean17/tmp/SURFACES/mean17s.L.full.segment_vent.ellipsoid.57640.coord Number of Point 57640 Removed /mnt/data/work/somatotopy/surefit_mean17/tmp/SURFACES/LowSmooth.vtk Removed /mnt/data/work/somatotopy/surefit_mean17/tmp/SURFACES/HighSmooth.vtk Removed /mnt/data/work/somatotopy/surefit_mean17/tmp/ERROR_CORRECTION_INTERMEDIATES/average_cosi.metric Removed /mnt/data/work/somatotopy/surefit_mean17/tmp/ERROR_CORRECTION_INTERMEDIATES/compressed.metric Removed /mnt/data/work/somatotopy/surefit_mean17/tmp/ERROR_CORRECTION_INTERMEDIATES/gaussian_neg.metric Removed /mnt/data/work/somatotopy/surefit_mean17/tmp/ERROR_CORRECTION_INTERMEDIATES/smoothed_surface.vtk Removed /mnt/data/work/somatotopy/surefit_mean17/tmp/ERROR_CORRECTION_INTERMEDIATES/stretched.metric GenerateEllipsoid done Center of points: -0.620868 -0.764846 -0.654845 Iter: 0Points: 57640 with crossovers: 56736 Iter: 1Points: 57640 with crossovers: 2339 There are 57640 points. There are 2970 crossovers. --- May this be a similar problem related to the vfat filesystem? Thanks a lot for your help, Marco -- Marco Tettamanti Scientific Institute San Raffaele c/o L.I.T.A. - room 24/5 Via Fratelli Cervi 93 I-20090 Segrate (MI) Italy Tel. +39-02-21717548 Fax +39-02-21717558 email: [EMAIL PROTECTED]
Re: [Surefit-users] another error message
Hi Donna, thanks a lot. I am not really in a hurry and, as you said, I can always use a non vfat filesystem. Best wishes, Marco > Hi Marco, > > Yes -- this error is the same problem encountered before, but creeping > up in a different shared library. The SureFit build process makes a > single libnetcdf.a static library to which at least five shared > libraries are linked. One of them is the Scientific_netcdf.so library I > rebuilt -- the one that allowed python to write .mnc files. I knew that > SureFitOps and some of its supporting libraries ($SFHOME/lib/lib*Caret*) > linked to the netcdf library, but I didn't think it would be a problem, > because I didn't think they involved writing volume files. But I forgot > about SureFitOps.metric_file_to_volume, which is used during automatic > error correction and preparation for flattening. It writes a netcdf > volume, and I'm afraid we need it. > > I need to rebuild the other four libraries, linking to the libnetcdf.a > library I generated without the offending define. It might finish > today, but if I run into problems, I won't be able to get them to you > until late next week. Tomorrow is a U.S. holiday, and I'll be at > Dartmouth's fMRI workshop Monday-Wednesday of next week. > > Meanwhile, you'll need to run SureFit on a non-vfat/fat32 file system. > > Donna -- Marco Tettamanti Scientific Institute San Raffaele c/o L.I.T.A. - room 24/5 Via Fratelli Cervi 93 I-20090 Segrate (MI) Italy Tel. +39-02-21717548 Fax +39-02-21717558 email: [EMAIL PROTECTED]
Re: [Surefit-users] another error message
Hi Donna, great, now it works fine! I have tested your new libraries and they seem to solve all problems: I could run a Run SureFit routine on the vfat filesystem and no error messages were generated. Thank you 1'000'000 times! Marco > Hi Marco, > > I'm on my way out the door and won't be back in the office until > Thursday, but the build did just finish and I managed to copy the > libraries to our web site. > > If you want to give them a try: > > login in as a user with write permission to $SFHOME/lib > cd $SFHOME/lib > mkdir BAK > mv libCaretCommon.so libCaretSurface.so \ > libCaretUtilities.so libVtkCaretClasses.so \ > SureFitOps.so BAK > wget http://brainmap.wustl.edu/donna/SureFit/Linux/libCaretCommon.so > wget http://brainmap.wustl.edu/donna/SureFit/Linux/libCaretSurface.so > wget http://brainmap.wustl.edu/donna/SureFit/Linux/libCaretUtilities.so > wget http://brainmap.wustl.edu/donna/SureFit/Linux/libVtkCaretClasses.so > wget http://brainmap.wustl.edu/donna/SureFit/Linux/SureFitOps.so > > Give them a try on the vfat file system. I haven't had a chance to try > them on our system, so I don't know what will happen. But if they fail > badly, you can do "cd $SFHOME/lib/BAK ; mv * .." to get back to the old > ones, which work on non-vfat file systems. > > Donna Marco Tettamanti Scientific Institute San Raffaele c/o L.I.T.A. - room 24/5 Via Fratelli Cervi 93 I-20090 Segrate (MI) Italy Tel. +39-02-21717548 Fax +39-02-21717558 email: [EMAIL PROTECTED]
[caret-users] Re: Show Raised Foci
Hello John and Caret users, I have posted this message a while ago, but didn't receive any reply. I try again, in case something has got lost. Thank you in advance for your help and sorry for bothering you! Sincerely, Marco -- Concerning the issue about the 'Raised Foci', I am still not sure that I understand. Considering what you have said, I now understand that the cortical sheet of a flat map is associated with a z = 0. Is that right? Is the z=0 level set at the level of the external surface of the cortical sheet? Or somewhere deep in the cortical sheet? Now, you said that when the 'Show Raised Foci' option is enabled, all foci are displayed at the level z = 0.5. That is, slighlty above the cortical sheet. Is that correct? If I disable the 'Show Raised Foci' option, two things happen: i) a few foci disappear completely ii) a few foci shrink in size I assume that the foci that shrink in size lie within the cortical sheet, but below the external surface (if the external surface is set to z=0, these foci are e.g. at a z=-1 or z=-2 level). Is that correct? Concerning those foci that completely disppear, I still do not understand where they do lie. Maybe still in the cortical sheet, but the farthest apart from the external surface? Why are you sure that they not subcortical? If I play with the 'Layers: Foci: Adjust Focus Distance' option, instead than with the 'Show Raised Foci' option, I see that those foci that completely disappear when disabling the 'Show Raised Foci' option, with the 'Layers: Foci: Adjust Focus Distance' option they disappear when the distance is set to < 10 mm. Given that the cortex is about 3-4 mm thick, I thought it would be safe to conclude that these foci are subcortical. Or is that wrong? Thanks a lot for any clarifications! Marco - Original Message - From: John Harwell <[EMAIL PROTECTED]> Date: Thursday, July 3, 2003 11:39 pm Subject: Re: Show Raised Foci > Hello Marco, > > On Thursday, July 3, 2003, at 04:06 PM, <[EMAIL PROTECTED]> > wrote: > > Dear John, > > thank you very much for your reply. I am still a bit uncertain, > though:> does "foci inside the surface" for you mean that these > foci are > > subcortical? > > No, they are not subcortical. -- Marco Tettamanti Scientific Institute San Raffaele c/o L.I.T.A. - room 24/5 Via Fratelli Cervi 93 I-20090 Segrate (MI) Italy Tel. +39-02-21717548 Fax +39-02-21717558 email: [EMAIL PROTECTED]
[caret-users] Re: Show Raised Foci
Hallo John, thank you very much for your clarifications. May I still ask you a couple of questions? > If you used SureFit to generate your surface, the surface represents > layer 4 of the cortex. So, for the flat surface, "z = 0" represents > layer 4. I did use SureFit. However, I did something which may be somewhat unorthodox. I generated surfaces from the an average image, which was obtained by averaging the structural images of 17 subjects (all individual images had been previously registered to the MNI stereotactic space). Am I safe to assume that the reconstructed surface represents the 'average" cortical layer 4 in my population sample? > Foci that disappear when "Show Raised Foci" is disabled are beneath the > flat cortical surface with a negative Z coordinate. A Focus that > shrinks in size is most likely due to several foci in close proximity, > some of which are beneath the flat surface. This doesn't seem to me to be the case. There are no other foci in proximity of those ones that shrink. Moreover, while the size of the foci that shrinked is the same as that of all other foci (i.e. the specified size) when 'Show Raised Foci' is enabled, when it is disabled, the size of the "shrinked foci" is smaller (to a variable extent, depending on the particular "shrinked" focus). If it were an overlap of distinct foci, I would rather expect that the 'initial size' is larger than normal, and the 'final size', in the minimal case, is as normal, but not smaller. If you wish, I can send you a .jpg picture of what I see. A marginal problem: there seem to be a small incompatability between Caret 4.6 and my linux platform. Whenever I attempt to Map a Talairach Focus, the focus is correctly mapped, but the color scales are unavailable (the foci are consequently all shown in black. After clicking on 'Apply' in the 'Map Talairach Focus' window, an error message appears, saying: PROGRAM ERROR: Unable to create GLX visual I am using a Red Hat Linux release 7.3 (Valhalla). Thanks again for your kind help, Marco -- Marco Tettamanti Scientific Institute San Raffaele c/o L.I.T.A. - room 24/5 Via Fratelli Cervi 93 I-20090 Segrate (MI) Italy Tel. +39-02-21717548 Fax +39-02-21717558 email: [EMAIL PROTECTED]
[caret-users] Map fMRI to Right Hemisphere
Dear Caret authors, I am sorry to come up with new questions. I am experiencing some problems in Mapping fMRI data onto RIGHT hemispheric surfaces (not on the atlas, but on my own generated average dataset). In the Map fMRI to Surface dialogue window, I select my own fMRI dataset and specify Analyze volume type and Left-Posterior-Inferior orientation. I then select the Map to Individual option. Choosing Negative X - Left (Anatomical) (i.e. the orientation of average structural image) in the Individual parameters dialogue window, produces some unexpected results: the left hemispheric activations, rather than the right hemispheric ones, are mapped onto the right hemispheric surface. Why is that? Appearently, choosing Negative X - Right (Radiological) in the Individual parameters dialogue window, solves the problem, i.e. the right hemispheric activations are mapped onto the right hemispheric surface. However, only some of the expected foci light up (I was expecting - e.g. based on the SPM maps - to see a focus in the inferior frontal gyrus, but it is not there; and other foci are also not there). Using the SAME fMRI data files for Mapping onto the left hemispheric surface (derived from the SAME average structural image) does not produce such problems. Choosing Negative X - Left (Anatomical) in the Individual parameters dialogue window, has the expected effect of mapping left hemispheric foci onto the left hemisphere. Just as a control, I have then mapped the right hemispheric activations onto the left hemisphere (by choosing Negative X - Right (Radiological) in the Individual parameters dialogue window). When I do this, all expected foci (included the one in the inferior frontal gyrus) light up. I am aware that this doesn't mean much; the difference may be due to an asymmetry between the right and the left hemisphere. But I do not see such a big asymmetry, and I just wanted to be sure that there is not something wrong with the way I am using the mapping procedure. Thanks for your help as always! Marco -- Marco Tettamanti Scientific Institute San Raffaele c/o L.I.T.A. - room 24/5 Via Fratelli Cervi 93 I-20090 Segrate (MI) Italy Tel. +39-02-21717548 Fax +39-02-21717558 email: [EMAIL PROTECTED]
Re: [caret-users] Map fMRI to Right Hemisphere
Thanks a lot, Donna! > * Load the appropriate .params file for the right hemisphere (i.e., the > one that is named closest to your surface name, or rather the intensity > volume from which it was generated); its filename should appear in your > spec file, but the mapper does not load it automatically. That's the part I was missing! Loading the .params file tells Caret how to crop the fMRI data files based on the origin on the x-axis. Is this correct? Anyway, the maps now look fine! Thanks a lot for your help, Marco
[caret-users] caret 5.2: Segmentation Error
Dear Caret authors, I have downloaded caret 5.2 and have attempted to use the integrated SureFit functions, but I am experiencing some serious problems and cannot procede with segmentation. After I have completed the preparation step and hit the segmentation button, I receive the following error message: ERROR IN EYE FAT SEGMENTATION I have therefore tried to uncheck "Disconnect Eye and Skull" from the Segmentation Operations Dialog, procede again with segmentation, but again, another error message is displayed: ERROR IN HINDBRAIN SEGMENTATION Trying to uncheck this operation as well leads to a segmentation fault and to caret termination. Apart of this, the other caret functions seem to work correctly, and I can import segmented surfaces from SureFit 4.45. I am working on a Debian linux platform. I have anectoctical informations from one colleague in the lab, who is working on a different linux distribution (Red Hat I believe) and is experiencing the same problems (I can gether more precise infos if you like). Any hints on what is happening? Thank you a lot for your help. Best wishes, Marco -- Marco Tettamanti, Ph.D. Department of Neuroscience Scientific Institute San Raffaele c/o L.I.T.A. - room 25/5 Via Fratelli Cervi 93 I-20090 Segrate (MI) Italy Tel. +39-02-21717552 Fax +39-02-21717558 email: [EMAIL PROTECTED]
Re: [caret-users] caret 5.2: Segmentation Error
Hi Donna, I remember you too! You have provided me a valuable help quite many times. Thank you for your quick reply. I have uploaded the following files: Human.lvs7b.L.params_file_19.2005-06-09.params wmean_anat3D_AC_L+orig.BRIK wmean_anat3D_AC_L+orig.HEAD Processing the same images with SureFit 5.2 is not a problem, so I do not think there is a problem with the data; but I hope it turns out I a wrong! All the best, Marco > Hi Marco, > > I remember you! You helped educate me about the differences between > talairach and MNI space a couple of years ago. > > Is there any chance you're trying to segment both hemispheres at once? > If not, then can you upload your cropped .HEAD and .BRIK structural > volumes along with your params file (or just tell me your GM/WM peaks) > here: > > http://pulvinar.wustl.edu/cgi-bin/upload.cgi > > Other potential problems that can cause such problems: > > * AC not set correctly > * Bad peaks > > The best way for me to help is to look at your volume. > > Donna -- Marco Tettamanti, Ph.D. Department of Neuroscience Scientific Institute San Raffaele c/o L.I.T.A. - room 25/5 Via Fratelli Cervi 93 I-20090 Segrate (MI) Italy Tel. +39-02-21717552 Fax +39-02-21717558 email: [EMAIL PROTECTED]
Re: [caret-users] caret 5.2: Segmentation Error
Hi Donna, yes! I have converted my original data from 16 to 8-bit format and segmentation now seems to run fine. Why does Caret only read unsigned byte volumes? SureFit read 16-bit volumes without any problems (or are they converted to 8-bit in backgound?). Isn't the loss of intensity associated with converting to unsigned format going to be a problem for surface image quality? Thanks a lot, Marco > Hi Marco, > > Very nice structural volume. My Caret gives a very nice segmentation > with no complaints, but I had to scale it from float to byte (not sure > what your original datum type was). We ported the SureFit algorithm > almost completely "as is," and its "SureFitOps" utilities work only on > unsigned byte (0-255) volumes. Shortly after we released Caret 5.2, > John added some features to rescale the volume in the Edit Volume > Attributes menu. If your Caret doesn't have a Data tab on the Volume > Attributes Editor with a button that says "Rescale Voxels...," then you > can get an updated version here: > > http://brainmap.wustl.edu/pub/john/mcw/ > login pub > password download > > Alternatively, you can write your original full brain volume in unsigned > byte form and input it to Caret. There should be no need to rescale if > the input is byte. > > I think this probably is the problem (relatively easy fix). > > Donna > > On 06/09/2005 02:14 PM, marco tettamanti wrote: > >Hi Donna, > >I remember you too! You have provided me a valuable help quite many times. > >Thank you for your quick reply. > >I have uploaded the following files: > > > >Human.lvs7b.L.params_file_19.2005-06-09.params > >wmean_anat3D_AC_L+orig.BRIK > >wmean_anat3D_AC_L+orig.HEAD > > > >Processing the same images with SureFit 5.2 is not a problem, so I do not > >think there is a problem with the data; but I hope it turns out I a wrong! > >All the best, > >Marco > > > >>Hi Marco, > >> > >>I remember you! You helped educate me about the differences between > >>talairach and MNI space a couple of years ago. > >> > >>Is there any chance you're trying to segment both hemispheres at once? > >>If not, then can you upload your cropped .HEAD and .BRIK structural > >>volumes along with your params file (or just tell me your GM/WM peaks) > >>here: > >> > >>http://pulvinar.wustl.edu/cgi-bin/upload.cgi > >> > >>Other potential problems that can cause such problems: > >> > >>* AC not set correctly > >>* Bad peaks > >> > >>The best way for me to help is to look at your volume. > >> > >>Donna > > ___ > caret-users mailing list > caret-users@brainvis.wustl.edu > http://pulvinar.wustl.edu/mailman/listinfo/caret-users -- Marco Tettamanti, Ph.D. Department of Neuroscience Scientific Institute San Raffaele c/o L.I.T.A. - room 25/5 Via Fratelli Cervi 93 I-20090 Segrate (MI) Italy Tel. +39-02-21717552 Fax +39-02-21717558 email: [EMAIL PROTECTED]
[caret-users] problems with Map Volumes
Dear authors, I am having problems with the "Map Volumes to Surface" utility. I am following the very same procedure that used to work for me until release Caret5.2 (included). Now I have dowloaded the release 5.3 and this procedure no longer seems to work. I am trying to map a metric (functional) image to a Fiducial coord file. The algorithm doesn't report any errors and outputs: Map to Spec File: lvs7b.L.spec Metric File: map_data_0_10_Jun_2005_13_09_06.metric Algorithm: Average Voxel but in the bash shell I have got the followint report: Unknown Spec File tag moved to header: When I then try to open the .metric file that has been created, I receive an error message: Error while accessing X.metric: contains a different number of nodes than Y.coord This is strange, because the Y.coord file is exactly the same as the one entered in the mapping algorithm. Thank you for your advice! Marco -- Marco Tettamanti, Ph.D. Department of Neuroscience Scientific Institute San Raffaele c/o L.I.T.A. - room 25/5 Via Fratelli Cervi 93 I-20090 Segrate (MI) Italy Tel. +39-02-21717552 Fax +39-02-21717558 email: [EMAIL PROTECTED]
Re: [caret-users] problems with Map Volumes
Hi Donna! I am back and was finally able to check. The metric file, as you suspected, only contains the header. I have uploaded the metric file, the functional volume (.hdr and .img in Analyze format) and the fiducial surface. I am working on a Debian Linux platform. Thank you for your help, Marco > Hi Marco, > > The Unknown spec file tag can be safely ignored. > > My hunch is that the metric file is empty (except perhaps for the > header). How big is the file? If >1K, then upload your metric, volume, > and fiducial surface here: > > http://pulvinar.wustl.edu/cgi-bin/upload.cgi > > You were running Debian Linux, right? > > Donna > > On 06/10/2005 12:16 PM, marco tettamanti wrote: > >Dear authors, > >I am having problems with the "Map Volumes to Surface" utility. I am > > following the very same procedure that used to work for me until release > > Caret5.2 (included). Now I have dowloaded the release 5.3 and this > > procedure no longer seems to work. > >I am trying to map a metric (functional) image to a Fiducial coord file. > > The algorithm doesn't report any errors and outputs: > > > >Map to Spec File: lvs7b.L.spec > >Metric File: map_data_0_10_Jun_2005_13_09_06.metric > >Algorithm: Average Voxel > > > >but in the bash shell I have got the followint report: > > > >Unknown Spec File tag moved to header: > > > >When I then try to open the .metric file that has been created, I receive > > an error message: > > > >Error while accessing X.metric: contains a different number of nodes than > >Y.coord > > > >This is strange, because the Y.coord file is exactly the same as the one > >entered in the mapping algorithm. > > > >Thank you for your advice! > >Marco -- Marco Tettamanti, Ph.D. Department of Neuroscience Scientific Institute San Raffaele c/o L.I.T.A. - room 25/5 Via Fratelli Cervi 93 I-20090 Segrate (MI) Italy Tel. +39-02-21717552 Fax +39-02-21717558 email: [EMAIL PROTECTED]
[caret-users] [SPAM] caret on AMD64
Dear Donna and John,
first of all I would like to express my congratulations for the
improvements you made in Caret! I haven't been using the software for
some times, but now that I have some new data to work on and I have
started to use the latest release (5.4), I find that many options have
been made more intuitive and easy to use, especially for the manual
correction of segmentation errors.
I am trying to install caret on an an AMD64 linux architecture (CentOS
4.3, kernel 2.6.9, 2 x AMD Opteron dual-core).
I have tried installing from the linux binaries, but I receive an error
message that some libraries cannot be opened ("error while loading
shared libraries: libpng12.so.0: cannot open shared object file: No such
file or directory"), even if I create symbolic links to e.g.
/usr/lib64/libpng12.so.0.
Is there any way I can get the caret binaries to work on AMD64, without
having to build from the source files?
Thank you a lot for your help,
Marco
P.S.: I am using this alternative email address because but I have tried
several times to send an email to caret-users@brainvis.wustl.edu from my
address [EMAIL PROTECTED], but for some reasons my emails are rejected with the
following error message:
The Symantec Mail Security program
: host acuda.wustl.edu[128.252.248.242] said:
550 <[EMAIL PROTECTED]>: Sender address rejected: Blocked (in reply
to RCPT TO command)
--
Marco Tettamanti, Ph.D.
San Raffaele Scientific Institute
Facoltà di Psicologia
Via Olgettina 58
I-20132 Milano, Italy
Tel. ++39-02-26434888
Fax ++39-02-26434892
Email: [EMAIL PROTECTED]
-
How low will we go? Check out Yahoo! Messengers low PC-to-Phone call rates.
[SPAM] Re: Re: [caret-users] [SPAM] caret on AMD64
Hi Donna and John,
thank you for your answers! Unfortunately I couldn't solve the problem.
I have tried exporting /usr/lib64 to LD_LIBRARY_PATH, but I still get the same
error message:
"error while loading shared libraries: libpng12.so.0: cannot open shared
object file: No such file or directory"
The output of ldd caret5 is:
[EMAIL PROTECTED] bin]# ldd caret5
linux-gate.so.1 => (0xe000)
libpng12.so.0 => not found
libSM.so.6 => not found
libICE.so.6 => not found
libXi.so.6 => not found
libXrender.so.1 => not found
libXrandr.so.2 => not found
libXcursor.so.1 => not found
libfreetype.so.6 => not found
libfontconfig.so.1 => not found
libXext.so.6 => not found
libX11.so.6 => not found
libz.so.1 => not found
libpthread.so.0 => /lib/tls/libpthread.so.0 (0xf7fdb000)
libdl.so.2 => /lib/libdl.so.2 (0xf7fd7000)
libGLU.so.1 => not found
libGL.so.1 => /usr/lib/libGL.so.1 (0xf7f51000)
libm.so.6 => /lib/tls/libm.so.6 (0xf7f2e000)
libc.so.6 => /lib/tls/libc.so.6 (0x00669000)
/lib/ld-linux.so.2 (0x0065)
libGLcore.so.1 => /usr/lib/libGLcore.so.1 (0xf776c000)
libnvidia-tls.so.1 => /usr/lib/tls/libnvidia-tls.so.1 (0xf776a000)
libXext.so.6 => not found
libX11.so.6 => not found
I have tried creating a symolic link to the existing /usr/lib64/libpng12.so.0
both in the /lib and /usr/lib directories, but I still get the same error
message.
Permissions and file integrity should all be ok.
Do you have any other suggestions?
Thank you a lot!
Marco
Marco Tettamanti wrote:
*/Donna Dierker <[EMAIL PROTECTED]>/* wrote:
Date: Tue, 29 Aug 2006 08:47:13 -0500
From: Donna Dierker <[EMAIL PROTECTED]>
To: "Caret, SureFit, and SuMS software users"
Subject: Re: [caret-users] [SPAM] caret on AMD64
Try making sure /usr/lib64 is in your LD_LIBRARY_PATH variable, e.g.:
setenv LD_LIBRARY_PATH /usr/lib64:$LD_LIBRARY_PATH
... or if your shell is bash:
export LD_LIBRARY_PATH=/usr/lib64:$LD_LIBRARY_PATH
Thanks for reporting the issue with the mailing list. It sounds like I
need to check the spam settings for [EMAIL PROTECTED]
On 08/29/2006 03:01 AM, Marco Tettamanti wrote:
> Dear Donna and John,
> first of all I would like to express my congratulations for the
> improvements you made in Caret! I haven't been using the software for
> some times, but now that I have some new data to work on and I have
> started to use the latest release (5.4), I find that many options
have
> been made more intuitive and easy to use, especially for the manual
> correction of segmentation errors.
>
> I am trying to install caret on an an AMD64 linux architecture
(CentOS
> 4.3, kernel 2.6.9, 2 x AMD Opteron dual-core).
> I have tried installing from the linux binaries, but I receive an
error
> message that some libraries cannot be opened ("error while loading
> shared libraries: libpng12.so.0: cannot open shared object file:
No such
> file or directory"), even if I create symbolic links to e.g.
> /usr/lib64/libpng12.so.0.
> Is there any way I can get the caret binaries to work on AMD64,
without
> having to build from the source files?
>
> Thank you a lot for your help,
> Marco
>
> P.S.: I am using this alternative email address because but I have
> tried several times to send an email to
caret-users@brainvis.wustl.edu
> from my address [EMAIL PROTECTED], but for some reasons my
> emails are rejected with the following error message:
>
> The Symantec Mail Security program
>
> : host
> acuda.wustl.edu[128.252.248.242] said:
> 550 : Sender address rejected: Blocked (in reply
> to RCPT TO command)
--
Marco Tettamanti, Ph.D.
San Raffaele Scientific Institute
Facoltà di Psicologia
Via Olgettina 58
I-20132 Milano, Italy
Tel. ++39-02-26434888
Fax ++39-02-26434892
Email: [EMAIL PROTECTED]
-
How low will we go? Check out Yahoo! Messengers low PC-to-Phone call rates.
Re: [SPAM] Re: Re: [caret-users] [SPAM] caret on AMD64
Hi Donna!
Upon a fresh login, the output of 'echo $LD_LIBRARY_PATH' is usually empty.
After giving the command 'export LD_LIBRARY_PATH=/usr/lib64:$LD_LIBRARY_PATH',
the output is:
echo $LD_LIBRARY_PATH
/usr/lib64:
The output of the 'find' command is:
find / -name libX11.so.6
/usr/X11R6/lib64/libX11.so.6
Thank you again,
Marco
Donna Dierker <[EMAIL PROTECTED]> wrote: Hi Marco,
John may have more concrete suggestions, but I want to see the output of
two commands:
echo $LD_LIBRARY_PATH
and the other is:
find / -name libX11.so.6
Based on the output below, I still suspect an issue with your
LD_LIBRARY_PATH. It could also be that there are too many shared library
differences between our Linux build platform and your runtime platform.
But let's clarify before reaching that conclusion.
On 08/30/2006 07:06 AM, Marco Tettamanti wrote:
> Hi Donna and John,
> thank you for your answers! Unfortunately I couldn't solve the problem.
>
> I have tried exporting /usr/lib64 to LD_LIBRARY_PATH, but I still get
> the same error message:
> "error while loading shared libraries: libpng12.so.0: cannot open
> shared object file: No such file or directory"
>
> The output of ldd caret5 is:
>
> [EMAIL PROTECTED] bin]# ldd caret5
> linux-gate.so.1 => (0xe000)
> libpng12.so.0 => not found
> libSM.so.6 => not found
> libICE.so.6 => not found
> libXi.so.6 => not found
> libXrender.so.1 => not found
> libXrandr.so.2 => not found
> libXcursor.so.1 => not found
> libfreetype.so.6 => not found
> libfontconfig.so.1 => not found
> libXext.so.6 => not found
> libX11.so.6 => not found
> libz.so.1 => not found
> libpthread.so.0 => /lib/tls/libpthread.so.0 (0xf7fdb000)
> libdl.so.2 => /lib/libdl.so.2 (0xf7fd7000)
> libGLU.so.1 => not found
> libGL.so.1 => /usr/lib/libGL.so.1 (0xf7f51000)
> libm.so.6 => /lib/tls/libm.so.6 (0xf7f2e000)
> libc.so.6 => /lib/tls/libc.so.6 (0x00669000)
> /lib/ld-linux.so.2 (0x0065)
> libGLcore.so.1 => /usr/lib/libGLcore.so.1 (0xf776c000)
> libnvidia-tls.so.1 => /usr/lib/tls/libnvidia-tls.so.1 (0xf776a000)
> libXext.so.6 => not found
> libX11.so.6 => not found
>
> I have tried creating a symolic link to the existing
> /usr/lib64/libpng12.so.0 both in the /lib and /usr/lib directories,
> but I still get the same error message.
> Permissions and file integrity should all be ok.
>
> Do you have any other suggestions?
> Thank you a lot!
> Marco
>
>
>
> Marco Tettamanti wrote:
>>
>>
>> */Donna Dierker /* wrote:
>>
>> Date: Tue, 29 Aug 2006 08:47:13 -0500
>> From: Donna Dierker
>> To: "Caret, SureFit, and SuMS software users"
>>
>> Subject: Re: [caret-users] [SPAM] caret on AMD64
>>
>> Try making sure /usr/lib64 is in your LD_LIBRARY_PATH variable, e.g.:
>>
>> setenv LD_LIBRARY_PATH /usr/lib64:$LD_LIBRARY_PATH
>>
>> ... or if your shell is bash:
>>
>> export LD_LIBRARY_PATH=/usr/lib64:$LD_LIBRARY_PATH
>>
>> Thanks for reporting the issue with the mailing list. It sounds like I
>> need to check the spam settings for [EMAIL PROTECTED]
>>
>> On 08/29/2006 03:01 AM, Marco Tettamanti wrote:
>> > Dear Donna and John,
>> > first of all I would like to express my congratulations for the
>> > improvements you made in Caret! I haven't been using the software for
>> > some times, but now that I have some new data to work on and I have
>> > started to use the latest release (5.4), I find that many options
>> have
>> > been made more intuitive and easy to use, especially for the manual
>> > correction of segmentation errors.
>> >
>> > I am trying to install caret on an an AMD64 linux architecture
>> (CentOS
>> > 4.3, kernel 2.6.9, 2 x AMD Opteron dual-core).
>> > I have tried installing from the linux binaries, but I receive an
>> error
>> > message that some libraries cannot be opened ("error while loading
>> > shared libraries: libpng12.so.0: cannot open shared object file:
>> No such
>> > file or directory"), even if I create symbolic links to e.g.
>> > /usr/lib64/libpng12.so.0.
>> > Is there any way I can get the caret binaries to work on AMD64,
>> without
>> > having to build from the source files?
>> >
>> > Thank you a lot for your help,
>> > Marco
>> >
>> > P.S.: I am using this alternative email address because but I have
>> > tried several times to send an email to
>> caret-users@brainvis.wustl.edu
>>
Re: [SPAM] Re: Re: [caret-users] [SPAM] caret on AMD64
Hi Donna, I have tried to add a few directories: echo $LD_LIBRARY_PATH /lib64:/usr/X11R6/lib:/usr/lib:/lib:/usr/X11R6/lib64:/usr/lib64: The output of ldd caret5 doesn't seem to be affected at all... ldd caret5 linux-gate.so.1 => (0xe000) libpng12.so.0 => not found libSM.so.6 => not found libICE.so.6 => not found libXi.so.6 => not found libXrender.so.1 => not found libXrandr.so.2 => not found libXcursor.so.1 => not found libfreetype.so.6 => not found libfontconfig.so.1 => not found libXext.so.6 => not found libX11.so.6 => not found libz.so.1 => not found libpthread.so.0 => /lib/tls/libpthread.so.0 (0xf7fda000) libdl.so.2 => /lib/libdl.so.2 (0xf7fd6000) libGLU.so.1 => not found libGL.so.1 => /usr/lib/libGL.so.1 (0xf7f5) libm.so.6 => /lib/tls/libm.so.6 (0xf7f2d000) libc.so.6 => /lib/tls/libc.so.6 (0x00669000) /lib/ld-linux.so.2 (0x0065) libGLcore.so.1 => /usr/lib/libGLcore.so.1 (0xf776b000) libnvidia-tls.so.1 => /usr/lib/tls/libnvidia-tls.so.1 (0xf7769000) libXext.so.6 => not found libX11.so.6 => not found Marco Donna Dierker <[EMAIL PROTECTED]> wrote: Hi Marco, It doesn't seem like this should be necessary, but let's see how your ldd caret5 output is affected by this: export LD_LIBRARY_PATH=/usr/lib64:/usr/lib:/usr/X11R6/lib64/lib Add any other library directories to the path above (delimited by :) that you think might help. On 08/30/2006 08:14 AM, Marco Tettamanti wrote: > Hi Donna! > Upon a fresh login, the output of 'echo $LD_LIBRARY_PATH' is usually > empty. > After giving the command 'export > LD_LIBRARY_PATH=/usr/lib64:$LD_LIBRARY_PATH', the output is: > > echo $LD_LIBRARY_PATH > /usr/lib64: > > The output of the 'find' command is: > find / -name libX11.so.6 > /usr/X11R6/lib64/libX11.so.6 > > Thank you again, > Marco > > > */Donna Dierker /* wrote: > > Hi Marco, > > John may have more concrete suggestions, but I want to see the > output of > two commands: > > echo $LD_LIBRARY_PATH > > and the other is: > > find / -name libX11.so.6 > > Based on the output below, I still suspect an issue with your > LD_LIBRARY_PATH. It could also be that there are too many shared > library > differences between our Linux build platform and your runtime > platform. > But let's clarify before reaching that conclusion. > > On 08/30/2006 07:06 AM, Marco Tettamanti wrote: > > Hi Donna and John, > > thank you for your answers! Unfortunately I couldn't solve the > problem. > > > > I have tried exporting /usr/lib64 to LD_LIBRARY_PATH, but I > still get > > the same error message: > > "error while loading shared libraries: libpng12.so.0: cannot open > > shared object file: No such file or directory" > > > > The output of ldd caret5 is: > > > > [EMAIL PROTECTED] bin]# ldd caret5 > > linux-gate.so.1 => (0xe000) > > libpng12.so.0 => not found > > libSM.so.6 => not found > > libICE.so.6 => not found > > libXi.so.6 => not found > > libXrender.so.1 => not found > > libXrandr.so.2 => not found > > libXcursor.so.1 => not found > > libfreetype.so.6 => not found > > libfontconfig.so.1 => not found > > libXext.so.6 => not found > > libX11.so.6 => not found > > libz.so.1 => not found > > libpthread.so.0 => /lib/tls/libpthread.so.0 (0xf7fdb000) > > libdl.so.2 => /lib/libdl.so.2 (0xf7fd7000) > > libGLU.so.1 => not found > > libGL.so.1 => /usr/lib/libGL.so.1 (0xf7f51000) > > libm.so.6 => /lib/tls/libm.so.6 (0xf7f2e000) > > libc.so.6 => /lib/tls/libc.so.6 (0x00669000) > > /lib/ld-linux.so.2 (0x0065) > > libGLcore.so.1 => /usr/lib/libGLcore.so.1 (0xf776c000) > > libnvidia-tls.so.1 => /usr/lib/tls/libnvidia-tls.so.1 (0xf776a000) > > libXext.so.6 => not found > > libX11.so.6 => not found > > > > I have tried creating a symolic link to the existing > > /usr/lib64/libpng12.so.0 both in the /lib and /usr/lib directories, > > but I still get the same error message. > > Permissions and file integrity should all be ok. > > > > Do you have any other suggestions? > > Thank you a lot! > > Marco
Re: [SPAM] Re: Re: [caret-users] [SPAM] caret on AMD64
Thank you Volkmar and John! Volkmar is right, I have both /usr/lib and /usr/X11R6/lib directories on my machine. But I have almost exclusively 64bit libraries installed, no 32 bit libraries. This was actually my original question: whether there was a mean to install Caret on a full 64bit machine, without having to compile from the source files... Given that this is not the case, I may go on and install the full set of 32 bit libraries. Does anyone know the name of the packages that need to be installed in CentOS, e.g. by using the command 'yum', in order to obtain all the libraries required by Caret? Thank you again, Marco John Harwell <[EMAIL PROTECTED]> wrote: Marco, As Volkmar points out, Caret is built on a 32-bit machine. Is there a "/usr/lib32" or "/usr/X11R6/lib32" directory on your computer? When we used Silicon Graphics computers (running their version of UNIX called IRIX), it contained both /usr/lib32 and /usr/ lib64 so that both 32 and 64 bit executables could be used. I don't know if this is how Linux is setup. You may want to try the website for your version of Linux to inquire about obtaining 32-bit libraries. RedHat has info at "http:// www.redhat.com/magazine/009jul05/features/multilib/". -- John Harwell [EMAIL PROTECTED] 314-362-3467 Department of Anatomy and Neurobiology Washington University School of Medicine 660 S. Euclid Ave.Box 8108 St. Louis, MO 63110 USA On Aug 30, 2006, at 8:44 AM, Volkmar Glauche wrote: > Dear Marco & others, > > it appears that your caret5 is a 32-bit executable. At least it is > for my > binary 5.3 distribution. > > file caret5 > > should tell you what your system thinks of the binary format, and > the ldd > output suggests that only 32-bit libraries are used. To use this > binary > as it is, you will have to install the whole bunch of 32-bit shared > libraries it needs. Or go and recompile, but then you will need to > install > the development files for these libraries. > > Volkmar > > On Wed, 30 Aug 2006, Marco Tettamanti wrote: > >> Hi Donna and John, >> thank you for your answers! Unfortunately I couldn't solve the >> problem. >> >> I have tried exporting /usr/lib64 to LD_LIBRARY_PATH, but I still >> get the same error message: >> "error while loading shared libraries: libpng12.so.0: cannot open >> shared object file: No such file or directory" >> >> The output of ldd caret5 is: >> >> [EMAIL PROTECTED] bin]# ldd caret5 >> linux-gate.so.1 => (0xe000) >> libpng12.so.0 => not found >> libSM.so.6 => not found >> libICE.so.6 => not found >> libXi.so.6 => not found >> libXrender.so.1 => not found >> libXrandr.so.2 => not found >> libXcursor.so.1 => not found >> libfreetype.so.6 => not found >> libfontconfig.so.1 => not found >> libXext.so.6 => not found >> libX11.so.6 => not found >> libz.so.1 => not found >> libpthread.so.0 => /lib/tls/libpthread.so.0 (0xf7fdb000) >> libdl.so.2 => /lib/libdl.so.2 (0xf7fd7000) >> libGLU.so.1 => not found >> libGL.so.1 => /usr/lib/libGL.so.1 (0xf7f51000) >> libm.so.6 => /lib/tls/libm.so.6 (0xf7f2e000) >> libc.so.6 => /lib/tls/libc.so.6 (0x00669000) >> /lib/ld-linux.so.2 (0x0065) >> libGLcore.so.1 => /usr/lib/libGLcore.so.1 (0xf776c000) >> libnvidia-tls.so.1 => /usr/lib/tls/libnvidia-tls.so.1 >> (0xf776a000) >> libXext.so.6 => not found >> libX11.so.6 => not found >> >> I have tried creating a symolic link to the existing /usr/lib64/ >> libpng12.so.0 both in the /lib and /usr/lib directories, but I >> still get the same error message. >> Permissions and file integrity should all be ok. >> >> Do you have any other suggestions? >> Thank you a lot! >> Marco >> >> >> >> Marco Tettamanti wrote: >> >> >> */Donna Dierker /* wrote: >> >> Date: Tue, 29 Aug 2006 08:47:13 -0500 >> From: Donna Dierker >> To: "Caret, SureFit, and SuMS software users" >> >> Subject: Re: [caret-users] [SPAM] caret on AMD64 >> >> Try making sure /usr/lib64 is in your LD_LIBRARY_PATH >> variable, e.g.: >> >> setenv LD_LIBRARY_PATH /usr/lib64:$LD_LIBRARY_PATH >> >> ... or if your shell is bash: >> >> ex
Re: [caret-users] caret on AMD64
Dear Volkmar, John and Donna,
this is an old thread, but I just wanted to say that I eventually found
some time to get back to my problem of running caret on an AMD64
architecture and solved it: the solution proposed by Volkmar did the
trick! This basically amounted to run ldd caret5 and install the missing
32-bit libraries with the following commands:
yum install libpng.i386
yum install xorg-x11-libs.i386
yum install xorg-x11-Mesa-libGLU.i386
Now everything's running properly. All the best!
Marco
Marco Tettamanti wrote:
*/Volkmar Glauche <[EMAIL PROTECTED]>/* wrote:
Dear Marco & others,
it appears that your caret5 is a 32-bit executable. At least it is
for my
binary 5.3 distribution.
file caret5
should tell you what your system thinks of the binary format, and
the ldd
output suggests that only 32-bit libraries are used. To use this binary
as it is, you will have to install the whole bunch of 32-bit shared
libraries it needs. Or go and recompile, but then you will need to
install
the development files for these libraries.
Volkmar
On Wed, 30 Aug 2006, Marco Tettamanti wrote:
> Hi Donna and John,
> thank you for your answers! Unfortunately I couldn't solve the
problem.
>
> I have tried exporting /usr/lib64 to LD_LIBRARY_PATH, but I still
get the same error message:
> "error while loading shared libraries: libpng12.so.0: cannot open
shared object file: No such file or directory"
>
> The output of ldd caret5 is:
>
> [EMAIL PROTECTED] bin]# ldd caret5
> linux-gate.so.1 => (0xe000)
> libpng12.so.0 => not found
> libSM.so.6 => not found
> libICE.so.6 => not found
> libXi.so.6 => not found
> libXrender.so.1 => not found
> libXrandr.so.2 => not found
> libXcursor.so.1 => not found
> libfreetype.so.6 => not found
> libfontconfig.so.1 => not found
> libXext.so.6 => not found
> libX11.so.6 => not found
> libz.so.1 => not found
> libpthread.so.0 => /lib/tls/libpthread.so.0 (0xf7fdb000)
> libdl.so.2 => /lib/libdl.so.2 (0xf7fd7000)
> libGLU.so.1 => not found
> libGL.so.1 => /usr/lib/libGL.so.1 (0xf7f51000)
> libm.so.6 => /lib/tls/libm.so.6 (0xf7f2e000)
> libc.so.6 => /lib/tls/libc.so.6 (0x00669000)
> /lib/ld-linux.so.2 (0x0065)
> libGLcore.so.1 => /usr/lib/libGLcore.so.1 (0xf776c000)
> libnvidia-tls.so.1 => /usr/lib/tls/libnvidia-tls.so.1 (0xf776a000)
> libXext.so.6 => not found
> libX11.so.6 => not found
>
> I have tried creating a symolic link to the existing
/usr/lib64/libpng12.so.0 both in the /lib and /usr/lib directories,
but I still get the same error message.
> Permissions and file integrity should all be ok.
>
> Do you have any other suggestions?
> Thank you a lot!
> Marco
>
>
>
> Marco Tettamanti wrote:
>
>
> */Donna Dierker /* wrote:
>
> Date: Tue, 29 Aug 2006 08:47:13 -0500
> From: Donna Dierker
> To: "Caret, SureFit, and SuMS software users"
>
> Subject: Re: [caret-users] [SPAM] caret on AMD64
>
> Try making sure /usr/lib64 is in your LD_LIBRARY_PATH variable,
e.g.:
>
> setenv LD_LIBRARY_PATH /usr/lib64:$LD_LIBRARY_PATH
>
> ... or if your shell is bash:
>
> export LD_LIBRARY_PATH=/usr/lib64:$LD_LIBRARY_PATH
>
> Thanks for reporting the issue with the mailing list. It sounds
like I
> need to check the spam settings for [EMAIL PROTECTED]
>
> On 08/29/2006 03:01 AM, Marco Tettamanti wrote:
> > Dear Donna and John,
> > first of all I would like to express my congratulations for the
> > improvements you made in Caret! I haven't been using the
software for
> > some times, but now that I have some new data to work on and I
have
> > started to use the latest release (5.4), I find that many options
> have
> > been made more intuitive and easy to use, especially for the
manual
> > correction of segmentation errors.
> >
> > I am trying to install caret on an an AMD64 linux architecture
> (CentOS
> > 4.3, kernel 2.6.9, 2 x AMD Opteron dual-core).
> > I have tried installing from the linux binaries, but I receive an
> error
> > message that some libraries cannot be opened ("error while loading
> > shared libraries: libpng12.so.0: cannot open shared object file:
> No such
> > file
[caret-users] crash of X KDE session
Hi Donna and Hohn, I am experiencing a severe problem with my caret installation in linux: whenever I try to import an Analyze Volume file, my entire X session immediately crashes, including all applications and I have to login in KDE from scratch. I couldn't tell whether the problem is limited to the Import function only, because the system crashes so badly that I thought it would be better to obtain some feedback from you before doing some debugging. But all the other display functions of caret that I have tried out (rendering surfaces, overlaying metric maps and so on) seem to work fine. I first encountered the problem a few days ago. I completely ignore what could be the origin of the problem. I had used the same installation without any problems before, including the Import function. I have also tried to do a fresh install, of both versions 5.4 and 5.5, but the problem persists. I did some updates of KDE and xorg recently, though, and I suspect that the problem may have arosen since then. All other applications run fine, though. Also, I have installed the same .zip caret tar ball on another machine and there it runs fine... This is the kernel of my machine: Linux bll4 2.6.11-kanotix-11 #1 Sun May 29 22:32:10 CEST 2005 i686 GNU/Linux Any help would be greatly appreciated! Best, Marco -- Marco Tettamanti, Ph.D. San Raffaele Scientific Institute Facoltà di Psicologia Via Olgettina 58 I-20132 Milano, Italy Tel. ++39-02-26434888 Fax ++39-02-26434892 Email: [EMAIL PROTECTED]
Re: [caret-users] crash of X KDE session
Hi John and Donna, I tried both of your methods, but unfortunately none solved my problem... I have tried to rename the file .fonts.conf, alone and, following the thread between Roland and you, together with .fonts.cache-1 and .fontconfig/ I also tried running Donna's script sh ./caret_prefs_stripper.sh. Helas, my entire X session always crashes when I import an Analyze image file. My problem seem to be different form Roland's and Veronica's ones, in that the problem is not "limited" to Caret, but it affects the entire X environment. Best, Marco John Harwell wrote: Marco, Another user had a problem similar to this. I believe it has to do with the FreeType library which gets used to create text characters that are drawn on the volume slices to show the voxel and its stereotaxic coordinate. See if there is a file in your home directory named ".fonts.conf" (note that the file begins with a period so use the command "ls -a" to see the file). If ".fonts.conf" exists, rename it, start Caret, and see if you can import a volume. For more info: http://www.mail-archive.com/caret-users@brainvis.wustl.edu/msg00892.html -- John Harwell [EMAIL PROTECTED] 314-362-3467 Department of Anatomy and Neurobiology Washington University School of Medicine 660 S. Euclid Ave.Box 8108 St. Louis, MO 63110 USA On Nov 29, 2006, at 10:42 AM, marco tettamanti wrote: Hi Donna and Hohn, I am experiencing a severe problem with my caret installation in linux: whenever I try to import an Analyze Volume file, my entire X session immediately crashes, including all applications and I have to login in KDE from scratch. I couldn't tell whether the problem is limited to the Import function only, because the system crashes so badly that I thought it would be better to obtain some feedback from you before doing some debugging. But all the other display functions of caret that I have tried out (rendering surfaces, overlaying metric maps and so on) seem to work fine. I first encountered the problem a few days ago. I completely ignore what could be the origin of the problem. I had used the same installation without any problems before, including the Import function. I have also tried to do a fresh install, of both versions 5.4 and 5.5, but the problem persists. I did some updates of KDE and xorg recently, though, and I suspect that the problem may have arosen since then. All other applications run fine, though. Also, I have installed the same .zip caret tar ball on another machine and there it runs fine... This is the kernel of my machine: Linux bll4 2.6.11-kanotix-11 #1 Sun May 29 22:32:10 CEST 2005 i686 GNU/Linux Any help would be greatly appreciated! Best, Marco --Marco Tettamanti, Ph.D. San Raffaele Scientific Institute Facoltà di Psicologia Via Olgettina 58 I-20132 Milano, Italy Tel. ++39-02-26434888 Fax ++39-02-26434892 Email: [EMAIL PROTECTED] ___ caret-users mailing list caret-users@brainvis.wustl.edu http://pulvinar.wustl.edu/mailman/listinfo/caret-users ___ caret-users mailing list caret-users@brainvis.wustl.edu http://pulvinar.wustl.edu/mailman/listinfo/caret-users -- Marco Tettamanti, Ph.D. San Raffaele Scientific Institute Facoltà di Psicologia Via Olgettina 58 I-20132 Milano, Italy Tel. ++39-02-26434888 Fax ++39-02-26434892 Email: [EMAIL PROTECTED]
Re: [caret-users] crash of X KDE session
Hi Donna, I have tried to use File: Open Volume Anatomy and Open Volume Functional on: image.hdr image.HEAD image.nii as you suggested, but unfortunately they all cause the X session to crash. Thank you for you help! Best, Marco Donna Dierker wrote: Marco, Does the problem occur when you use File: Open Volume Anatomy (or functional) File to open the same file? If so, does it occur if you convert the file to NIfTI and/or AFNI? I'm guessing it will (e.g., suggesting a display issue, rather than a file I/O issue), but it helps to be sure. If you have trouble converting to another file format, then upload the file here, and I'll provide you the converted version: http://pulvinar.wustl.edu/cgi-bin/upload.cgi On 11/30/2006 04:08 AM, marco tettamanti wrote: Hi John and Donna, I tried both of your methods, but unfortunately none solved my problem... I have tried to rename the file .fonts.conf, alone and, following the thread between Roland and you, together with .fonts.cache-1 and .fontconfig/ I also tried running Donna's script sh ./caret_prefs_stripper.sh. Helas, my entire X session always crashes when I import an Analyze image file. My problem seem to be different form Roland's and Veronica's ones, in that the problem is not "limited" to Caret, but it affects the entire X environment. Best, Marco John Harwell wrote: Marco, Another user had a problem similar to this. I believe it has to do with the FreeType library which gets used to create text characters that are drawn on the volume slices to show the voxel and its stereotaxic coordinate. See if there is a file in your home directory named ".fonts.conf" (note that the file begins with a period so use the command "ls -a" to see the file). If ".fonts.conf" exists, rename it, start Caret, and see if you can import a volume. For more info: http://www.mail-archive.com/caret-users@brainvis.wustl.edu/msg00892.html -- John Harwell [EMAIL PROTECTED] 314-362-3467 Department of Anatomy and Neurobiology Washington University School of Medicine 660 S. Euclid Ave.Box 8108 St. Louis, MO 63110 USA On Nov 29, 2006, at 10:42 AM, marco tettamanti wrote: Hi Donna and Hohn, I am experiencing a severe problem with my caret installation in linux: whenever I try to import an Analyze Volume file, my entire X session immediately crashes, including all applications and I have to login in KDE from scratch. I couldn't tell whether the problem is limited to the Import function only, because the system crashes so badly that I thought it would be better to obtain some feedback from you before doing some debugging. But all the other display functions of caret that I have tried out (rendering surfaces, overlaying metric maps and so on) seem to work fine. I first encountered the problem a few days ago. I completely ignore what could be the origin of the problem. I had used the same installation without any problems before, including the Import function. I have also tried to do a fresh install, of both versions 5.4 and 5.5, but the problem persists. I did some updates of KDE and xorg recently, though, and I suspect that the problem may have arosen since then. All other applications run fine, though. Also, I have installed the same .zip caret tar ball on another machine and there it runs fine... This is the kernel of my machine: Linux bll4 2.6.11-kanotix-11 #1 Sun May 29 22:32:10 CEST 2005 i686 GNU/Linux Any help would be greatly appreciated! Best, Marco --Marco Tettamanti, Ph.D. San Raffaele Scientific Institute Facoltà di Psicologia Via Olgettina 58 I-20132 Milano, Italy Tel. ++39-02-26434888 Fax ++39-02-26434892 Email: [EMAIL PROTECTED] ___ caret-users mailing list caret-users@brainvis.wustl.edu http://pulvinar.wustl.edu/mailman/listinfo/caret-users ___ caret-users mailing list caret-users@brainvis.wustl.edu http://pulvinar.wustl.edu/mailman/listinfo/caret-users -- Marco Tettamanti, Ph.D. San Raffaele Scientific Institute Facoltà di Psicologia Via Olgettina 58 I-20132 Milano, Italy Tel. ++39-02-26434888 Fax ++39-02-26434892 Email: [EMAIL PROTECTED]
Re: [caret-users] crash of X KDE session
Hi John and Donna,
thank you for your suggestions! I will no try to dowload the debug
release and follow your instructions.
In the meantime, I ran Caret with the 'strace' linux diagnostic command,
hoping it may help you to trace the problem. I have attached the file
with the whole text output.
I am a quite naive linux user, so I do not know how to read the output
of the trace command, but as some point toward the end of the file I
have found:
open("/mnt/hda5/work/negazione/neg_01b/anat3D/mean_anat3D.hdr",
O_RDONLY|O_LARGEFILE) = 12
and at the end of the text output:
write(7, "\221\5\4\0\17\0\300\3\16\0\300\3\1\0\0\0", 16) = 16
read(7, 0xbfffcbbc, 32) = -1 EAGAIN (Resource
temporarily unavailable)
poll([{fd=7, events=POLLIN, revents=POLLIN|POLLERR|POLLHUP}], 1, -1) = 1
read(7, 0xbfffcbbc, 32) = -1 ECONNRESET (Connection
reset by peer)
write(2, "caret5: Fatal IO error: client k"..., 38) = 38
munmap(0xb7723000, 163840) = 0
write(6, "@", 1)= 1
close(6)= 0
close(5)= 0
close(3)= 0
close(4)= 0
munmap(0xb7955000, 4096)= 0
exit_group(1) = ?
I will let you know about using debug_caret!
Best,
Marco
John Harwell wrote:
Hi Marco,
We had a problem similar to this that had to do with the "freetype"
library that is used to render the stereotaxic coordinates on the volume
display window. I have built a new version of Caret5 for linux that has
an option, "-notext", that will prevent Caret5 from trying to draw any
text characters.
* Go to "http://brainmap.wustl.edu/pub/john/marco/";. Use the username
"pub" and the password "download" to access the website.
* Download the file caret5_exe_linux_debug.zip. This is a big file
since it contains debugging information. If it takes too long to
download, download caret5_exe_linux.zip instead.
* Unzip this file which will produce the "caret5" executable. In the
directory where you place the caret5 executable, start it with the
command "./caret5 -notext". Load a volume into Caret and see if Caret
crashes. If Caret does not crash after loading a volume, it probably
means there is a problem with the freetype library on your computer
(perhaps google freetype server crash.
* If Caret still crashes with the "-notext" option and you downloaded
the "debug" zip file do the following: Run the command "gdb ./caret5".
After about 15-30 seconds there will be a gdb prompt and enter the
command "run". Load a volume into Caret. When Caret crashes enter the
command "where" at the gdb and send me the output. If your X session
crashes, log back in, and, in the same directory, see if there is a file
named "core". If there is a file named core, enter the command "gdb
./caret5 core". At the gdb prompt, enter the command "where" and send
me the output.
Good Luck.
--
John Harwell
[EMAIL PROTECTED]
314-362-3467
Department of Anatomy and Neurobiology
Washington University School of Medicine
660 S. Euclid Ave.Box 8108
St. Louis, MO 63110 USA
On Dec 1, 2006, at 8:42 AM, marco tettamanti wrote:
Hi Donna,
I have tried to use File: Open Volume Anatomy and Open Volume
Functional on:
image.hdr
image.HEAD
image.nii
as you suggested, but unfortunately they all cause the X session to
crash.
Thank you for you help!
Best,
Marco
Donna Dierker wrote:
Marco,
Does the problem occur when you use File: Open Volume Anatomy (or
functional) File to open the same file?
If so, does it occur if you convert the file to NIfTI and/or AFNI?
I'm guessing it will (e.g., suggesting a display issue, rather than a
file I/O issue), but it helps to be sure. If you have trouble
converting to another file format, then upload the file here, and
I'll provide you the converted version:
http://pulvinar.wustl.edu/cgi-bin/upload.cgi
On 11/30/2006 04:08 AM, marco tettamanti wrote:
Hi John and Donna,
I tried both of your methods, but unfortunately none solved my
problem...
I have tried to rename the file .fonts.conf, alone and, following
the thread between Roland and you, together with .fonts.cache-1 and
.fontconfig/
I also tried running Donna's script sh ./caret_prefs_stripper.sh.
Helas, my entire X session always crashes when I import an Analyze
image file.
My problem seem to be different form Roland's and Veronica's ones,
in that the problem is not "limited" to Caret, but it affects the
entire X environment.
Best,
Marco
John Harwell wrote:
Marco,
Another user had a problem similar to this. I believe it has to do
with the
Re: [caret-users] crash of X KDE session
Hi John, I have downloaded caret5_exe_linux_debug and ran it, but strange things happened... :-| 1) First, I ran the debug executable as caret5 -notext as you suggested: I selected File Import, and, wow! caret opened the file. I did the same with File Open and it worked. Solved, I thought: must be the freetype library. So I clicked on the upper right corner to quit Caret and, whoops, the X session crashed. 2) Second, I started X again, and wanted to replicate this behavior. I launched the debug caret5 -notext, choose File Import... Well, this time the X session crashed at this point already. 3) Third, I followed your suggestion and typed gdb ./caret5 (the debug version), choose File Import and the X session crashed. Unfortunately, at rebooting there was no core file in the directory from where I launched caret. 4) I have tried to re-download caret5_exe_linux_debug to check whether something had gone corrupted during 1), but Caret continued to crash right after a File Import command. Thank you again! Marco John Harwell wrote: Hi Marco, We had a problem similar to this that had to do with the "freetype" library that is used to render the stereotaxic coordinates on the volume display window. I have built a new version of Caret5 for linux that has an option, "-notext", that will prevent Caret5 from trying to draw any text characters. * Go to "http://brainmap.wustl.edu/pub/john/marco/";. Use the username "pub" and the password "download" to access the website. * Download the file caret5_exe_linux_debug.zip. This is a big file since it contains debugging information. If it takes too long to download, download caret5_exe_linux.zip instead. * Unzip this file which will produce the "caret5" executable. In the directory where you place the caret5 executable, start it with the command "./caret5 -notext". Load a volume into Caret and see if Caret crashes. If Caret does not crash after loading a volume, it probably means there is a problem with the freetype library on your computer (perhaps google freetype server crash. * If Caret still crashes with the "-notext" option and you downloaded the "debug" zip file do the following: Run the command "gdb ./caret5". After about 15-30 seconds there will be a gdb prompt and enter the command "run". Load a volume into Caret. When Caret crashes enter the command "where" at the gdb and send me the output. If your X session crashes, log back in, and, in the same directory, see if there is a file named "core". If there is a file named core, enter the command "gdb ./caret5 core". At the gdb prompt, enter the command "where" and send me the output. Good Luck. -- John Harwell [EMAIL PROTECTED] 314-362-3467 Department of Anatomy and Neurobiology Washington University School of Medicine 660 S. Euclid Ave.Box 8108 St. Louis, MO 63110 USA On Dec 1, 2006, at 8:42 AM, marco tettamanti wrote: Hi Donna, I have tried to use File: Open Volume Anatomy and Open Volume Functional on: image.hdr image.HEAD image.nii as you suggested, but unfortunately they all cause the X session to crash. Thank you for you help! Best, Marco Donna Dierker wrote: Marco, Does the problem occur when you use File: Open Volume Anatomy (or functional) File to open the same file? If so, does it occur if you convert the file to NIfTI and/or AFNI? I'm guessing it will (e.g., suggesting a display issue, rather than a file I/O issue), but it helps to be sure. If you have trouble converting to another file format, then upload the file here, and I'll provide you the converted version: http://pulvinar.wustl.edu/cgi-bin/upload.cgi On 11/30/2006 04:08 AM, marco tettamanti wrote: Hi John and Donna, I tried both of your methods, but unfortunately none solved my problem... I have tried to rename the file .fonts.conf, alone and, following the thread between Roland and you, together with .fonts.cache-1 and .fontconfig/ I also tried running Donna's script sh ./caret_prefs_stripper.sh. Helas, my entire X session always crashes when I import an Analyze image file. My problem seem to be different form Roland's and Veronica's ones, in that the problem is not "limited" to Caret, but it affects the entire X environment. Best, Marco John Harwell wrote: Marco, Another user had a problem similar to this. I believe it has to do with the FreeType library which gets used to create text characters that are drawn on the volume slices to show the voxel and its stereotaxic coordinate. See if there is a file in your home directory named ".fonts.conf" (note that the file begins with a period so use the command "ls -a" to see the file). If ".fonts.conf" exists, rename it, start Car
Re: [caret-users] crash of X KDE session
Hi John, I have tried unlimiting the coredumpsize from a tcsh shell (as I was indeed using a bash shell and the coredumpsize was limited to 0 kbytes) but, for some reasons, even if caret crashes no core file gets written out. All the display functions that I have tried out (rendering surfaces, overlaying metric maps and so on) seem to work fine. But I couldn't tell whether the problem is really limited to opening volumes, as I haven't tried out many other things. I will try to upgrade the graphic card drivers! Thanks! Marco John Harwell wrote: Marco, The creation of a core file may be inhibited. So, make sure you are running one of the c-shell variances (enter the command "/bin/csh" or "/bin/tcsh". Enter the command "limit". If "coredumpsize" is anything but unlimited, do the following: * Enter the command "unlimit coredumpsize". * Enter "limit" to verify that "coredumpsize" is unlimited. * Run caret by typing "./caret5" but do not use gdb. * Import the volume to get Caret to crash. * Relogin. * Enter the command "gdb ./caret5 core". * Type "where" at the gdb prompt and send me the output. Are your able to load a surface into Caret5? Have you tried updating your graphics card drivers? John On Dec 5, 2006, at 10:24 AM, marco tettamanti wrote: Hi John, I have downloaded caret5_exe_linux_debug and ran it, but strange things happened... :-| 1) First, I ran the debug executable as caret5 -notext as you suggested: I selected File Import, and, wow! caret opened the file. I did the same with File Open and it worked. Solved, I thought: must be the freetype library. So I clicked on the upper right corner to quit Caret and, whoops, the X session crashed. 2) Second, I started X again, and wanted to replicate this behavior. I launched the debug caret5 -notext, choose File Import... Well, this time the X session crashed at this point already. 3) Third, I followed your suggestion and typed gdb ./caret5 (the debug version), choose File Import and the X session crashed. Unfortunately, at rebooting there was no core file in the directory from where I launched caret. 4) I have tried to re-download caret5_exe_linux_debug to check whether something had gone corrupted during 1), but Caret continued to crash right after a File Import command. Thank you again! Marco John Harwell wrote: Hi Marco, We had a problem similar to this that had to do with the "freetype" library that is used to render the stereotaxic coordinates on the volume display window. I have built a new version of Caret5 for linux that has an option, "-notext", that will prevent Caret5 from trying to draw any text characters. * Go to "http://brainmap.wustl.edu/pub/john/marco/";. Use the username "pub" and the password "download" to access the website. * Download the file caret5_exe_linux_debug.zip. This is a big file since it contains debugging information. If it takes too long to download, download caret5_exe_linux.zip instead. * Unzip this file which will produce the "caret5" executable. In the directory where you place the caret5 executable, start it with the command "./caret5 -notext". Load a volume into Caret and see if Caret crashes. If Caret does not crash after loading a volume, it probably means there is a problem with the freetype library on your computer (perhaps google freetype server crash. * If Caret still crashes with the "-notext" option and you downloaded the "debug" zip file do the following: Run the command "gdb ./caret5". After about 15-30 seconds there will be a gdb prompt and enter the command "run". Load a volume into Caret. When Caret crashes enter the command "where" at the gdb and send me the output. If your X session crashes, log back in, and, in the same directory, see if there is a file named "core". If there is a file named core, enter the command "gdb ./caret5 core". At the gdb prompt, enter the command "where" and send me the output. Good Luck. ------ John Harwell [EMAIL PROTECTED] 314-362-3467 Department of Anatomy and Neurobiology Washington University School of Medicine 660 S. Euclid Ave.Box 8108 St. Louis, MO 63110 USA On Dec 1, 2006, at 8:42 AM, marco tettamanti wrote: Hi Donna, I have tried to use File: Open Volume Anatomy and Open Volume Functional on: image.hdr image.HEAD image.nii as you suggested, but unfortunately they all cause the X session to crash. Thank you for you help! Best, Marco Donna Dierker wrote: Marco, Does the problem occur when you use File: Open Volume Anatomy (or functional) File to open the same file? If so, does it occur if you convert the file to NIfTI and/or AFNI? I'm guessing it will (e.g., suggesting
