Welcome to iraf.net Thursday, May 02 2024 @ 05:44 AM GMT


 Forum Index > Help Desk > General IRAF New Topic Post Reply
 Memory error in iraf imfort libraries?
   
Anonymous: Guest
 08/28/2006 08:13PM (Read 3596 times)  



To whom it may concern:We have come across a possible bug/incompatibility than may exist in the
iraf imfort libraries. The bug concerns the imclos function. We have
noticed that whenever this function is called in any of our many fortran
programs it exits with the error:PANIC in `': Memory has been corrupted.We are currently running the Redhat version of NOAO PC-IRAF Revision
2.12.2a-EXPORT released on Wed Jul 14 20:45:34 MST 2004 on a machine
running the Mandriva Linux release 2006.0 (Official) for i586. Our system
uses the following glibc versionglibc-2.3.5-5mdk, and
glibc-devel-2.3.5-5mdk.The command to create the executable is as follows:f77 -O2 -Wall -pedantic strcat.f -c
f77 -O2 -Wall -pedantic specindex.f -c
f77 -O2 -Wall -pedantic specindex.o strcat.o /iraf/iraf/unix/bin.redhat/libcompat.a /iraf/iraf/bin.redhat/libimfort.a /iraf/iraf/bin.redhat/libsys.a /iraf/iraf/bin.redhat/libvops.a /iraf/iraf/unix/bin.redhat/libos.a -L/usr/local/lib -lplotsub_d -ldevices_d -lutils_d -L/usr/local/lib -lcfitsio -lcdl -L/usr/X11R6/lib -lX11 -lm -o ../../bin/specindexYou may notice the method used to link with the IRAF libraries. Linking
with the iraf libraries in the manner is the only successful of creating
the executables that we have found. If the libraries are included on the
command line in the normal fashion -l<library> the linking process fails.
If this problem is linked to the memory corruption problem mentioned
above, I am not sure. Any insight as to why this might be would be
helpful.We have also noticed that the problem does not exist for an older version
of Redhat linux. Independent of the machine where the binary is compiled,
the program executes in normal fashion. The machine where the executable
functions properly runs Red Hat Linux release 9 (Shrike) with
glibc-2.3.2-27.9.7. Any assistance you can give us in resolving this
problem will be greatly appreciated. If any further information is need
to diagnose this problem, do not hesitate to contact me at the information
provided below.Sincerely,Robert C. Berrington
--
Robert C. Berrington Phone #: (307) 766-4366 (W)
University of Wyoming Cell #: (703) 963-8052
Dept of Physics & Astronomy Fax #: (307) 766-2652
P. O. Box 3905 E-Mail: rberring@uwyo.edu (W)
1000 E. University Avenue http://physics.uwyo.edu/~rberring
Laramie, WY 82071God made the fly, but forgot to tell us why. --Ogden Nash

 
 Quote
fitz
 08/28/2006 08:13PM  
AAAAA
Admin

Status: offline


Registered: 09/30/2005
Posts: 4040
For Linux 2.6 kernels and IMFORT tasks you need to specify that the 'stacksize' is unlimited in order to work around changes in the memio system. For example[code:1:7bda7db531]
linux% limit stacksize unlimited [for c-shell]
linux% ulimit -s unlimited [for bash]
[/code:1:7bda7db531]Put this in your environment file or on the commandline before running the task. The normal CL startup command does this for you automatically, however imfort tasks are normally declared as foreign commands and so the above should go in your .cshrc/.bashrc file.Cheers,
-Mike

 
Profile Email
 Quote
   
Content generated in: 0.04 seconds
New Topic Post Reply

Normal Topic Normal Topic
Sticky Topic Sticky Topic
Locked Topic Locked Topic
New Post New Post
Sticky Topic W/ New Post Sticky Topic W/ New Post
Locked Topic W/ New Post Locked Topic W/ New Post
View Anonymous Posts 
Anonymous users can post 
Filtered HTML Allowed 
Censored Content 
dog allergies remedies cialis 20 mg chilblain remedies


Privacy Policy
Terms of Use

User Functions

Login