The message posted by Pilar seems, as John pointed out, to be related to my
post from a couple of days ago. I too implemented the recent spm_spm.m fix,
and this did not eliminate our errors. I ran through batch versions of
several of our analyses, and some ran successfully, while others did not,
resulting in the same sort of errors we had been experiencing. We also have
been seeing file open errors like this:
****
Error in ==> /speechlab/software/SLT/startup/../spm2/spm_sample_vol.mexglx
Error in ==> /speechlab/software/SLT/startup/../spm2/spm_sample_vol.mexglx
Error in ==> /speechlab/software/SLT/startup/../spm2/spm_get_data.m
0);
Error in ==> /speechlab/software/SLT/startup/../spm2edits/spm_spm.m
On line 654 ==> Y(i,Cm) = spm_get_data(VY(i),xyz(:,Cm));
Error in ==> /speechlab/software/SLT/startup/../RAT2/roi_estimate.m
On line 16 ==> spm_spm(SPMmatfile);
Error in ==> /speechlab/software/SLT/startup/../RAT2/roi_randfx_subjects.m
On line 81 ==> roi_estimate('SPM.mat');
****
My initial thought was that this could also potentially involve a file
descriptor limit issue, so I probed this a bit further. In RedHat9, the
default number of open fd's allowed for a single process is 1024. You can
change this (but only for the root user) using ulimit -n. I changed this
limit to 2^15, and reran our scripts as root. I again saw the same errors
(in this case the write_analyze_plane error). I believe this is pretty
definitive evidence that the file descriptor limit is not the issue here.
If anyone has some additional insight I am all ears :)
Cheers,
Jay Bohland
--------
On Wed, 2 Feb 2005 13:54:15 +0000, John Ashburner <[log in to unmask]>
wrote:
>There are various problems with Matlab 6.5 under RedHat 9. These links may
>help:
>http://www.mathworks.com/support/solutions/data/1-1AY16.html?solution=1-1AY16
>http://www.mathworks.com/support/solutions/data/1-1B37U.html?solution=1-1B37U
>http://www.mathworks.com/support/solutions/data/1-1BOQ1.html?solution=1-1BOQ1
>http://www.mathworks.com/support/solutions/data/1-1BRZD.html?solution=1-1BRZD
>http://www.mathworks.com/support/solutions/data/1-1ATCE.html?solution=1-1ATCE
>http://www.mathworks.com/support/solutions/data/1-1BBE9.html?solution=1-1BBE9
>http://www.mathworks.com/support/sysreq/current_release/unix.html
>
>I thought the problem may have related to
>http://www.jiscmail.ac.uk/cgi-bin/wa.exe?A2=ind0501&L=spm&P=R55609&I=-1 , but
>as you have checked this out, then I guess it is not the cause. The problem
>may be related though. If spm or some other Matlab functions have crashed
>leaving too many open files, then I think you may see something similar.
>
>Do you still experience the same things if you start up a new Matlab?
>
>Best regards,
>-John
>
>On Wednesday 02 February 2005 11:24, SALGADO PINEDA Pilar wrote:
>> Hi everyone,
>>
>>
>> We have found a strange set of errors that seems to happen randomly when
>> SPM is reading or writing image files. We are running RedHat Linux 9
>> (2.4.20) with SPM2 on Matlab 6.5.
>>
>> In particular, we are frequently seeing errors when we permormed the smooth
>> step and this occurs in different scans, even with scans that we have
>> smoothed later without problem in the same computer, with the same spm and
>> matlab versions.
>>
>> Here is a sample error:
>>
>>
>> SPM2: spm_smooth_ui (v2.11) 12:50:09 - 02/02/2005
>>
>> ========================================================================
>>
>> * - SPM2: spm_write_plane -----------------------------------------
>>
>> Error writing:
>>
>>
>>
>> ..V7_2/fonc2/swavisageV7_2_fonc2_0082.img
>>
>>
>>
>> Check disk space / disk quota.
>>
>> ----------------------------------------- 12:50:28 - 02/02/2005
>>
>> ??? Error using ==> spm_write_plane (write_analyze_plane)
>>
>> Error writing
>> /extern/ParkEmotion3/analyse_SPM2/ldopa/visageV7_2/fonc2/swavisageV7_2_fonc
>>2_0082.img.
>>
>> Error in ==> /extern/ParkEmotion3/analyse_SPM2/spm_write_plane.m
>>
>> On line 20 ==> V = write_analyze_plane(V,A,p);
>>
>> Error in ==> /usr/local/SPM/spm2/spm_conv_vol.mexglx
>>
>> Error in ==> /usr/local/SPM/spm2/spm_smooth.m
>>
>> On line 64 ==> spm_conv_vol(P,Q,x,y,z,-[i,j,k]);
>>
>> Error in ==> /usr/local/SPM/spm2/spm_smooth_ui.m
>>
>> On line 45 ==> spm_smooth(Q,U,s);
>>
>> ??? Error while evaluating uicontrol Callback.
>>
>>
>>
>>
>>
>>
>> There is plenty of available disk space and permissions seem to be in
>> order.
>>
>> We have try the solutions suggested by John Ashburner in :
>> <http://www.jiscmail.ac.uk/cgi-bin/wa.exe?A2=ind03&L=spm&P=R258078&I=-1>
>> http://www.jiscmail.ac.uk/cgi-bin/wa.exe?A2=ind03&L=spm&P=R258078&I=-1; and
>> <http://www.jiscmail.ac.uk/cgi-bin/wa.exe?A2=ind0501&L=spm&P=R55609&I=-1>
>> http://www.jiscmail.ac.uk/cgi-bin/wa.exe?A2=ind0501&L=spm&P=R55609&I=-1.
>> But we have not resulted this problem....
>>
>> Anybody knows what is happening?
>> Thanks in advance
>>
>> Pilar
|