I have installed CUDA 4.1 on a RHEL 5.5 Linux workstation.
nvcc works fine, but cuda-gdb fails.
I see the following error:
$ cuda-gdb hello
'import site' failed; use -v for traceback
Traceback (most recent call last):
File "<string>", line 32, in ?
ImportError: No module named os.path
NVIDIA (R) CUDA Debugger
4.1 release
I have used
nvcc -g -G -pg -D_DEBUG hello_gpu.cu -o hello
to generate the executable.
Any suggestions?
Thanks in advance.
PS: cuda-gdb was working fine with the previous version (CUDA 4.0)
There is not enough information posted to make a diagnosis. I checked with the debugger team and they would need at least the following additonal information / data:
(1) Host GCC version (gcc –v)
(2) System/kernel info (uname –a)
(3) The binary for the test program (files can be attached to forum posts)
(4) Does starting cuda-gdb without any arguments result in the same behavior?
$ gcc -v
Using built-in specs.
Target: x86_64-unknown-linux-gnu
Configured with: ../gcc-4.4.0/configure --prefix=/gcc/4.4.0/x86_64 --with-gmp=/gmp/4.3.1/x86_64 --with-mpfr=/mpfr/2.4.1/x86_64
Thread model: posix
gcc version 4.4.0 (GCC)
$ uname -a
Linux 2.6.18-194.el5 #1 SMP Tue Mar 16 21:52:39 EDT 2010 x86_64 x86_64 x86_64 GNU/Linux
$ gdb
GNU gdb (GDB) Red Hat Enterprise Linux (7.0.1-23.el5)
Copyright (C) 2009 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law. Type "show copying"
and "show warranty" for details.
This GDB was configured as "x86_64-redhat-linux-gnu".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>.
$ python
Python 2.7.2 (default, Aug 8 2011, 10:58:53)
[GCC 4.1.2 20080704 (Red Hat 4.1.2-50)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>>