Intel® C++ Compiler

omp_set_num_threads

I tried the following code section on one of the Dell computers equipped with dual P4 CPUs running linux. It resulted in "segmentation fault" after executing the function "omp_set_num_threads(t)", no matter what value t assumes. Other omp operations are fine. The compiler is intel C++ 9.x for linux. Is this a known bug?

{
int nthreads, tid;
int nCPU;

nCPU = omp_get_num_procs();
printf("number of CPUs is %d
",nCPU);
omp_set_num_threads(nCPU);
printf("number of threads is%d
", omp_get_num_threads());

IPO doesn't work!

I have installed ICC v9.0.21 on my FC4 system, but IPO doesn't work. The mcpcom thread will use all the cpu resources and compile progress stop there when using -ipo command. My GCC version is 4.0.2 and glibc is 2.3.5 and all the other rpms are updated to latest release. Could somebody help me?

movntdq instruction

I have an application that I believe the intel compiler generated the "movntdq" instruction to clear a memory region to zero. The compiler seems to generate the instruction twice like the following,

pxor xmm0, xmm0
labe1: movntdq XMMWORD PTR [ebx+eax*4+04h] xmm0 24,224
movntdq XMMWORD PTR[ebx+eax*014h] xmm0 2,013
add eax 0x8h 380
cmp eax, edi
jb label1

I dont understand
1. why movntdq was generated twice and
2. why the first one spent more time than the second one from the Vtune analyzer? and
3. what is XMMWORD?

GNU gsl test failed

Hi all,

yesterday I've start playing with GNU gsl library a I've got suspicious results.

Test PC #1: P4 2.8HT, Shuttle AB60,1.5 GB RAM, CentOS 4.2 i386, icc 9.0.030, gsl-1.7

#source /opt/intel/cc/9.0/bin/iccvars.sh

#./configure CC=icc CFLAGS='-O2 -no-ipo'
#make clean;make;make check
Test OK

#./configure CC=icc CFLAGS='-O3 -no-ipo'
#make clean;make;make check
Test OK

#./configure CC=icc CFLAGS='-O3 -xN -no-ipo'
#make clean;make;make check
Test FAIL

Intel 9.0 and mpich

Hi all,
I was wondering has anybody try to get intel icc and mpich 1.2.6 or higher build correctly?
platform fedora core 4 (em64t)
gcc 4.0(I know 4.0 is not currently supported)
checked the obvious already:
sourcing the iccvars.sh

I'm getting undefined reference _intel_fast_memcpy
eventhough it's defined in libirc.a
nm libirc.a |grep intel_fast_memcpy
0000000000000020 T _intel_fast_memcpy
0000000000000000 T _intel_fast_memcpy.A
0000000000000018 T _intel_fast_memcpy.J

Subscribe to Intel® C++ Compiler