MKL bug mailing list

[jaewonj]

Is there any mailing list where I can send a bug report?

[TimP]

Quoting - jaewonj

Is there any mailing list where I can send a bug report?

If you haven't created a support account on https://premier.intel.com, you can do so by registering your serial number at https://registrationcenter.intel.com. Then you can submit the bug report on premier.

[Gennady_F_Intel]

Quoting - jaewonj

Gennady,

I think this is a threading bug. In case I missed something, please let me know. Thanks.

When performing dense sparse level 3 operation (mkl_dcsrmm), I got inconsistent result depending on the number of cores used. Please see the attached c file (dcsrmm.c).

B (1001-by-1001) dense matrix with only 1 nonzero, B(0,0) = 1.0.
A (1001-by-1001) CSR sparse matrix also with only 1 nz, A(0,0) = 1.0

For C = B * A, I should get C(0,0) = 1.0 and this happens when I use 1 core or more than 2 cores. However, if I set OMP_NUM_THREADS=2, I get C(0,0) = 0.0 which is incorrect.

It seems this can be reproducible on all platforms with more than 2 cores.

I'm using MKL 10.2.2.


Jaewon

Jaewon,
Thanks for the test.
It's easy to check the is it threading bug or not to link the test with sequential libraries.
Can you check it on your side?
--Gennady

[jaewonj]

Quoting - Gennady Fedorov (Intel)

Jaewon,
Thanks for the test.
It's easy to check the is it threading bug or not to link the test with sequential libraries.
Can you check it on your side?
--Gennady

Gennady,

Unable to reproduce with sequential libraries.

Jaewon

[Gennady_F_Intel]

Quoting - jaewonj

Gennady,

Unable to reproduce with sequential libraries.

Jaewon

Jaewon,
yes i've got the same result. We will investigate the problem and will back if any news asap.
--Gennady

[jaewonj]

Quoting - Gennady Fedorov (Intel)

Jaewon,
yes i've got the same result. We will investigate the problem and will back if any news asap.
--Gennady

Gennady,

Any update on this issue?

Jaewon

[ali_m_2]

I have tried the following simple program to generate random numbers for parallel threads using Cilk. 

#include "mkl.h"
#include <cilk/cilk.h> 

#define NUM_THREADS 2
VSLStreamStatePtr stream[NUM_THREADS];

void cal(int sid) {
    
    double buff[1024];
    vdRngUniform(VSL_METHOD_DUNIFORM_STD, stream[sid], 1024, buff, -1.0, 1.0);
}

int main() {

    for (int i = 0; i < NUM_THREADS; i++) {
        vslNewStream(&stream, VSL_BRNG_WH, 1);
    }

    for (int i = 0; i < NUM_THREADS; i++) {
        cilk_spawn cal(i);
    }
    cilk_sync;

    for (int i = 0; i < NUM_THREADS; i++)
        vslDeleteStream(&stream);
    return 0;
}

The I compile it with icpc compiler icpc (ICC) 14.0.1 20131008.

icpc -std=c++11 -fopenmp -mkl -I/opt/intel/include/ -g -o pi_mont pi_mont.cpp

If I run it inside cilkscreen, it shows the race condition error.

$ cilkscreen ./pi_mont

Cilkscreen Race Detector V2.0.0, Build 4225

Race condition on location 0x7fc000ee8d00
  write access at 0x7fc000b877e3: (vdRngUniform+0xf3)
  read access at 0x7fc000b87761: (vdRngUniform+0x71)
    called by 0x40172c: (/localstore/theorie/amirsol/paralleluct/pi_mont.cpp:10, cal+0x60)
    called by 0x401a3a: (/localstore/theorie/amirsol/paralleluct/pi_mont.cpp:20, main+0x2fc)
1 error found by Cilkscreen
Cilkscreen suppressed 1 duplicate error messages

Ali

[ali_m_2]

I have tried the following simple program to generate random numbers for parallel threads using Cilk. 

#include "mkl.h"
#include <cilk/cilk.h> 

#define NUM_THREADS 2
VSLStreamStatePtr stream[NUM_THREADS];

void cal(int sid) {
    
    double buff[1024];
    vdRngUniform(VSL_METHOD_DUNIFORM_STD, stream[sid], 1024, buff, -1.0, 1.0);
}

int main() {

    for (int i = 0; i < NUM_THREADS; i++) {
        vslNewStream(&stream, VSL_BRNG_WH, 1);
    }

    for (int i = 0; i < NUM_THREADS; i++) {
        cilk_spawn cal(i);
    }
    cilk_sync;

    for (int i = 0; i < NUM_THREADS; i++)
        vslDeleteStream(&stream);
    return 0;
}

The I compile it with icpc compiler icpc (ICC) 14.0.1 20131008.

icpc -std=c++11 -fopenmp -mkl -I/opt/intel/include/ -g -o pi_mont pi_mont.cpp

If I run it inside cilkscreen, it shows the race condition error.

$ cilkscreen ./pi_mont

Cilkscreen Race Detector V2.0.0, Build 4225

Race condition on location 0x7fc000ee8d00
  write access at 0x7fc000b877e3: (vdRngUniform+0xf3)
  read access at 0x7fc000b87761: (vdRngUniform+0x71)
    called by 0x40172c: (/localstore/theorie/amirsol/paralleluct/pi_mont.cpp:10, cal+0x60)
    called by 0x401a3a: (/localstore/theorie/amirsol/paralleluct/pi_mont.cpp:20, main+0x2fc)
1 error found by Cilkscreen
Cilkscreen suppressed 1 duplicate error messages

Ali

[Andrey_N_Intel]

The topic is discussed in another thread, https://software.intel.com/en-us/forums/intel-math-kernel-library/topic/283349#comment-1846156