Intel® C++ Compiler
Community support and assistance for creating C++ code that runs on platforms based on Intel® processors.
7954 Discussions

icpc: command line warning #10121: overriding '-march=core-avx2' with '-xHOST'

samfux84
New Contributor I
3,490 Views

Hi,

 

I am trying to compile SuiteSparse 5.10.1 with the "Intel parallel studio xe 2020 r0".

 

[spackapps@eu-a6-008-07 bin]$ icpc --version
icpc (ICC) 19.1.0.166 20191121
Copyright (C) 1985-2019 Intel Corporation. All rights reserved.

[spackapps@eu-a6-008-07 bin]$

 

I explicitly set -march=core-avx2 and I see in the logs the warning:

icpc: command line warning #10121: overriding '-march=core-avx2' with '-xHOST'

 

How can I achieve that the Intel compiler respects my setting of march and does not replace it? If I want to optimize for the current host, then I would set -march=native. But I explicitly want to optimize for -march=core-avx2 and not for anything else.

 

Any help is appreciated.

 

Thank you and best regards

Sam

0 Kudos
1 Solution
samfux84
New Contributor I
3,399 Views

Hi Rahul,

 

Thank you for your feedback. I encountered this problem when compiling SuiteSparse 5.10.1 with Intel 19.1.0. I don't have any other example to reproduce this. Unfortunately there is a bug in the Intel 19.1.0 such that SuiteSparse 5.10.1 cannot be compiled successfully:

 

https://community.intel.com/t5/Intel-C-Compiler/error-type-name-is-not-allowed/m-p/1295999/emcs_t/S2h8ZW1haWx8dG9waWNfc3Vic2NyaXB0aW9ufEtRUU1NTjlXWUpZUFQxfDEyOTU5OTl8U1VCU0NSSVBUSU9OU3xoSw#M39069

 

Please feel free to close this thread. It does not make sense to spend more time on this. I will switch to GCC,  where compilation of most packages that I need works.

View solution in original post

0 Kudos
10 Replies
RahulV_intel
Moderator
3,464 Views

Hi,


Could you provide your OS environment details?

Also, check whether AVX2 instructions are supported on your machine.


Thanks,

Rahul


0 Kudos
samfux84
New Contributor I
3,410 Views

Hi Rahul,

 

Thank you for your reply.

 

Please find below the information regarding the operating system and kernel:

 

[sfux@eu-login-05 ~]$ cat /etc/redhat-release
CentOS Linux release 7.9.2009 (Core)
[sfux@eu-login-05 ~]$ uname -a
Linux eu-login-05 3.10.0-1160.24.1.el7.x86_64 #1 SMP Thu Apr 8 19:51:47 UTC 2021 x86_64 GNU/Linux
[sfux@eu-login-05 ~]$

 

All hardware in our cluster all supports avx2 (I have compiled 100's of software packages with march=core-avx2 already):

 

Please find below the information about the CPU from the oldest generation of our hardware (all other nodes are have newer CPUs):

 

[sfux@eu-login-05 ~]$ cat /proc/cpuinfo | head -n 25
processor : 0
vendor_id : GenuineIntel
cpu family : 6
model : 71
model name : Intel(R) Xeon(R) CPU E3-1284L v4 @ 2.90GHz
stepping : 1
microcode : 0x21
cpu MHz : 922.180
cache size : 6144 KB
physical id : 0
siblings : 4
core id : 0
cpu cores : 4
apicid : 0
initial apicid : 0
fpu : yes
fpu_exception : yes
cpuid level : 20
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc aperfmperf eagerfpu pni pclmulqdq dtes64 monitor ds_cpl vmx smx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm 3dnowprefetch epb invpcid_single intel_pt ssbd ibrs ibpb stibp tpr_shadow vnmi flexpriority ept vpid fsgsbase tsc_adjust bmi1 hle avx2 smep bmi2 erms invpcid rtm rdseed adx smap xsaveopt dtherm ida arat pln pts md_clear spec_ctrl intel_stibp flush_l1d
bogomips : 5799.40
clflush size : 64
cache_alignment : 64
address sizes : 39 bits physical, 48 bits virtual
power management:
[sfux@eu-login-05 ~]$

 

Best regards

 

Sam

 

 

0 Kudos
samfux84
New Contributor I
3,406 Views

Dear Rahul,

 

Thank you for your reply.

 

[sfux@eu-login-15 ~]$ cat /etc/redhat-release
CentOS Linux release 7.9.2009 (Core)
[sfux@eu-login-15 ~]$ uname -a
Linux eu-login-15 3.10.0-1160.24.1.el7.x86_64 #1 SMP Thu Apr 8 19:51:47 UTC 2021 x86_64 GNU/Linux
[sfux@eu-login-15 ~]$ cat /proc/cpuinfo | head -n 25
processor : 0
vendor_id : GenuineIntel
cpu family : 6
model : 71
model name : Intel(R) Xeon(R) CPU E3-1284L v4 @ 2.90GHz
stepping : 1
microcode : 0x21
cpu MHz : 1012.097
cache size : 6144 KB
physical id : 0
siblings : 4
core id : 0
cpu cores : 4
apicid : 0
initial apicid : 0
fpu : yes
fpu_exception : yes
cpuid level : 20
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc aperfmperf eagerfpu pni pclmulqdq dtes64 monitor ds_cpl vmx smx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm 3dnowprefetch epb invpcid_single intel_pt ssbd ibrs ibpb stibp tpr_shadow vnmi flexpriority ept vpid fsgsbase tsc_adjust bmi1 hle avx2 smep bmi2 erms invpcid rtm rdseed adx smap xsaveopt dtherm ida arat pln pts md_clear spec_ctrl intel_stibp flush_l1d
bogomips : 5799.77
clflush size : 64
cache_alignment : 64
address sizes : 39 bits physical, 48 bits virtual
power management:
[sfux@eu-login-15 ~]$

 

I don't think that it is related to the OS or the CPU model.

 

If I specify an option to the compiler, it should not overwritten.

 

Best regards

 

Sam

0 Kudos
RahulV_intel
Moderator
3,430 Views

Hi,


Just a quick reminder to provide the required details.


Thanks,

Rahul


0 Kudos
samfux84
New Contributor I
3,406 Views

Dear Rahul,

 

Thank you for the reminder. I posted my answer already twice and I don't know why this is not added to this thread. Trying for the third time:

[sfux@eu-login-28 ~]$ cat /etc/redhat-release
CentOS Linux release 7.9.2009 (Core)
[sfux@eu-login-28 ~]$ uname -a
Linux eu-login-28 3.10.0-1160.31.1.el7.x86_64 #1 SMP Thu Jun 10 13:32:12 UTC 2021 x86_64 GNU/Linux
[sfux@eu-login-28 ~]$ cat /proc/cpuinfo | head -n 25
processor : 0
vendor_id : GenuineIntel
cpu family : 6
model : 71
model name : Intel(R) Xeon(R) CPU E3-1284L v4 @ 2.90GHz
stepping : 1
microcode : 0x21
cpu MHz : 947.314
cache size : 6144 KB
physical id : 0
siblings : 4
core id : 0
cpu cores : 4
apicid : 0
initial apicid : 0
fpu : yes
fpu_exception : yes
cpuid level : 20
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc aperfmperf eagerfpu pni pclmulqdq dtes64 monitor ds_cpl vmx smx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm 3dnowprefetch epb invpcid_single intel_pt ssbd ibrs ibpb stibp tpr_shadow vnmi flexpriority ept vpid fsgsbase tsc_adjust bmi1 hle avx2 smep bmi2 erms invpcid rtm rdseed adx smap xsaveopt dtherm ida arat pln pts md_clear spec_ctrl intel_stibp flush_l1d
bogomips : 5800.03
clflush size : 64
cache_alignment : 64
address sizes : 39 bits physical, 48 bits virtual
power management:
[sfux@eu-login-28 ~]$

 

As you can see, the hardware supports AVX2.

 

Best regards

 

Sam

0 Kudos
samfux84
New Contributor I
3,424 Views

I sent the required details already 3 times, but it does not get added to this thread ...

0 Kudos
samfux84
New Contributor I
3,406 Views

[sfux@eu-login-29 ~]$ cat /etc/redhat-release
CentOS Linux release 7.9.2009 (Core)
[sfux@eu-login-29 ~]$ uname -a
Linux eu-login-29 3.10.0-1160.31.1.el7.x86_64 #1 SMP Thu Jun 10 13:32:12 UTC 2021 x86_64 x86_64 x86_64 GNU/Linux
[sfux@eu-login-29 ~]$ cat /proc/cpuinfo | head -n 25
processor : 0
vendor_id : GenuineIntel
cpu family : 6
model : 71
model name : Intel(R) Xeon(R) CPU E3-1284L v4 @ 2.90GHz
stepping : 1
microcode : 0x21
cpu MHz : 1299.902
cache size : 6144 KB
physical id : 0
siblings : 4
core id : 0
cpu cores : 4
apicid : 0
initial apicid : 0
fpu : yes
fpu_exception : yes
cpuid level : 20
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc aperfmperf eagerfpu pni pclmulqdq dtes64 monitor ds_cpl vmx smx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm 3dnowprefetch epb invpcid_single intel_pt ssbd ibrs ibpb stibp tpr_shadow vnmi flexpriority ept vpid fsgsbase tsc_adjust bmi1 hle avx2 smep bmi2 erms invpcid rtm rdseed adx smap xsaveopt dtherm ida arat pln pts md_clear spec_ctrl intel_stibp flush_l1d
bogomips : 5799.63
clflush size : 64
cache_alignment : 64
address sizes : 39 bits physical, 48 bits virtual
power management:
[sfux@eu-login-29 ~]$
0 Kudos
RahulV_intel
Moderator
3,402 Views

Hi,


Thanks for providing the required details. All your previous posts have reappeared now. Sorry for the inconvenience caused.


Do you have a minimal reproducer code that I can try out at my end?

Also, is this issue specific to a particular source code or generic?


Thanks,

Rahul


0 Kudos
samfux84
New Contributor I
3,400 Views

Hi Rahul,

 

Thank you for your feedback. I encountered this problem when compiling SuiteSparse 5.10.1 with Intel 19.1.0. I don't have any other example to reproduce this. Unfortunately there is a bug in the Intel 19.1.0 such that SuiteSparse 5.10.1 cannot be compiled successfully:

 

https://community.intel.com/t5/Intel-C-Compiler/error-type-name-is-not-allowed/m-p/1295999/emcs_t/S2h8ZW1haWx8dG9waWNfc3Vic2NyaXB0aW9ufEtRUU1NTjlXWUpZUFQxfDEyOTU5OTl8U1VCU0NSSVBUSU9OU3xoSw#M39069

 

Please feel free to close this thread. It does not make sense to spend more time on this. I will switch to GCC,  where compilation of most packages that I need works.

0 Kudos
RahulV_intel
Moderator
3,313 Views

Hi Samuel,


As discussed via email, we have taken note of your suggestions. We would discontinue monitoring this issue. Please raise a new thread if you have further issues.


Thanks,

Rahul


0 Kudos
Reply