getting sdk to run on suse linux machine missing files necessary to compile sdk package...

I got a new Intel box running Suse 11 (Linux 2.6.27.23-0.1-default x86_64) to test the capabilities of
cuda programming with my models which are similar to the OceanFFT model.
It has the geforce 9600 gt, which should be better than the geforce 8600 gt that is in my MacbookPro.

The linux cuda download I got is giving an error upon make:

----CUT HERE----
/usr/include/gnu/stubs.h:7:27: error: gnu/stubs-32.h: No such file or directory
In file included from src/bank_checker.cpp:33:
./…/common/inc/bank_checker.h:52: warning: type qualifiers ignored on function return type
./…/common/inc/bank_checker.h:238: warning: type qualifiers ignored on function return type
src/bank_checker.cpp:75: warning: type qualifiers ignored on function return type
src/bank_checker.cpp:277: warning: type qualifiers ignored on function return type
make[1]: *** [obj/release/bank_checker.cpp.o] Error 1
make[1]: Leaving directory `/home/Mike/NVIDIA_CUDA_SDK/common’
make: *** [lib/libcutil.so] Error 2
—END CUT----

I find /usr/include/gnu/stubs-64.h , but not the -32.h file. Did I grab the wrong cuda download package,
or do I need to modify the Makefile somehow?

Try installing SDK 2.2.1 or install ia32-libs

N.

From the cuda download page, after entering my OS, the option to download SDK2.2.1 vanishes.

I’ll search to install the ia32-libs. (ia ?)

I believe ia32 stands for Intel Architecture 32 bit.

gnu/stubs-32.h is also in libc6-dev-i386 which is included by ia32-libs

N.

Hmmm, seems odd that 2.2 offers SDKs for 64 bit, but we should load 32 bit as well. How is this being compiled? Is it being compiled for 32 bit hence the need for those libs? I suppose 2.1 may work, but feels like an odd solution.

I am a newbie so sorry if I am totally off base here.

From the SDK 2.2.1 release notes:

N.