Intel® C++ Compiler
Community support and assistance for creating C++ code that runs on platforms based on Intel® processors.

Compiler Options for Simulation

AbdelrahmanHussein
665 Views

Hi,

 

This is my first post here. I am a newbie with Intel compilers and trying to get started with them.

 

Basically, I am working on building SPEC OMP 2012 using Intel C/C++ and Fortran compilers. My end goal is to simulate that on the gem5 simulator. Below, are the compiler options that were used to build the benchmark:

 

 

 

CC	= icc 
FC	= ifort
CXX     = icpc

COPTIMIZE = -O3 -xSSE4.2  -ipo1 -qopenmp -ansi_alias -mcmodel=medium -shared-intel
CXXOPTIMIZE = -O3 -xSSE4.2 -ipo1  -qopenmp -ansi_alias -mcmodel=medium -shared-intel
FOPTIMIZE = -O3 -xSSE4.2 -ipo1  -qopenmp -mcmodel=medium -shared-intel

 

 

 

 

However, when I try to run the benchmarks on the targeted system for simulation, I encounter the following error:

Please verify that both the operating system and the processor support Intel(R) X87, CMOV, MMX, FXSAVE, SSE, SSE2, SSE3, SSSE3, SSE4_1, SSE4_2 and POPCNT instructions.

 

I am sure that my OS supports these instructions since I tested it on a different simulator and it worked just fine! 

Question #1: What should be the right compiler options to use to make the code compatible with the simulated system? I thought about some a few options but not sure if they are going to work:

  • Use -O2 instead of -O3 so that it does not generate vectorization (hopefully my understanding is correct!)
  • Remove the -xSSE4.2
  • Add -no-simd and -no-vec

Question #2: Do you recommend using -xHost? Why?

Question #3: Should I use  -no-intel-extensions? Or, it is not needed in my case?

Thanks.

0 Kudos
1 Solution
AbdelrahmanHussein
495 Views

Hi @SeshaP_Intel ,

 

I am sure I posted a very detailed reply to your questions on July 27. However, I cannot see it for some reason.

 

Anyway, I just would like to thank you so much. I think I solved the problem by just reducing the optimization options from using -xSSE4_2 to xSSE2, which is the highest SSE extension supported by the CPU implemented for this simulator.

 

Thanks again!

View solution in original post

0 Kudos
4 Replies
SeshaP_Intel
Moderator
633 Views

Hi,


Thank you for posting in Intel Communities.


Could you please provide the following details?

1. Is SPEC OMP 2012 a licensed or open-source benchmark? Please provide us with the link to download.

2. How to download and use the Gem5 Simulator? Also, provide the processor details. 

3. Operating system and ICC version you are using?

4. Are you using 32 bit or 64-bit machine?

5. CPU details

6. Which simulator you have used to run the benchmark successfully? 

7. Complete steps to reproduce the issue at our end?


Thanks and Regards,

Pendyala Sesha Srinivas


0 Kudos
SeshaP_Intel
Moderator
528 Views

Hi,


We haven't heard back from you. Could you please provide an update on your issue?


Thanks and Regards,

Pendyala Sesha Srinivas


0 Kudos
AbdelrahmanHussein
496 Views

Hi @SeshaP_Intel ,

 

I am sure I posted a very detailed reply to your questions on July 27. However, I cannot see it for some reason.

 

Anyway, I just would like to thank you so much. I think I solved the problem by just reducing the optimization options from using -xSSE4_2 to xSSE2, which is the highest SSE extension supported by the CPU implemented for this simulator.

 

Thanks again!

0 Kudos
SeshaP_Intel
Moderator
490 Views

Hi,


Glad to know that your issue is resolved. If you need any additional information, please post a new question as this thread will no longer be monitored by Intel.


Thanks and Regards,

Pendyala Sesha Srinivas


0 Kudos
Reply