[Mvapich-discuss] CPU Affinity is undefined error on newer AMD CPU

[Shineman, Nat]

Hi Blaise,

MVAPICH2 2.3.7 is several years old and is no longer being actively developed, as a result issues like this can arise when new CPUs become available. I recommend migrating to the latest release version MVAPICH-Plus 4.0. This version should not have this issue. Currently only our closed source MVAPICH-Plus library is available, but an open source MVAPICH 4.0 version will be released soon.

Thanks,
Nat
________________________________
From: Mvapich-discuss <mvapich-discuss-bounces at lists.osu.edu> on behalf of Blaise Bourdin via Mvapich-discuss <mvapich-discuss at lists.osu.edu>
Sent: Friday, January 17, 2025 10:07
To: mvapich-discuss at lists.osu.edu <mvapich-discuss at lists.osu.edu>
Subject: [Mvapich-discuss] CPU Affinity is undefined error on newer AMD CPU

Hi, We are adding new nodes (AMD EPYC 9754) to an already deployed cluster (all AMD AMD EPYC 7543 running Rocky Linux 9. 5) and are encountering MPI issues with them: Here is the output of the standard cpi. c example (we get the same error message
ZjQcmQRYFpfptBannerStart
This Message Is From an External Sender
This message came from outside your organization.
<https://us-phishalarm-ewt.proofpoint.com/EWT/v1/KGKeukY!s0QdHUjgDYcOBBqRf64qtH2ENAKEcUjYPLdGbKDpmLa3clc87E2bJdzlPrvx_jAvJbd72NmtKT7vErVGZsAuq880lfpZPLfXvQhhlnEeJGXpvw$>
Report Suspicious

ZjQcmQRYFpfptBannerEnd
Hi,

We are adding new nodes (AMD EPYC 9754) to an already deployed cluster (all AMD AMD EPYC 7543 running Rocky Linux 9.5)  and are encountering MPI issues with them:
Here is the output of the standard cpi.c example (we get the same error message on all MPI jobs when running on more that a few cores)

[bourdinb at bbserv2 MPI]$ mpicc -o cpi cpi.c
[bourdinb at bbserv2 MPI]$ srun -n 13 -p nk cpi
Warning! : Core id 32578 does not exist on this architecture!  CPU Affinity is undefined  Fatal error in MPI_Init:
Other MPI error, error stack:
MPIR_Init_thread(493)........:  MPID_Init(400)...............:  MPIDI_CH3I_set_affinity(3594):  smpi_setaffinity(2758).......: CPU Affinity is undefined.

srun: Job step aborted: Waiting up to 32 seconds for job step to finish.
slurmstepd: error: *** STEP 20384.0 ON bb13 CANCELLED AT 2025-01-17T09:58:17 ***
srun: error: bb13: tasks 0-10,12: Killed
srun: error: bb13: task 11: Exited with exit code 1

For reference, here are the details of the processor type

[bourdinb at bbserv2 MPI]$ ssh bb13 lscpu
Architecture:                         x86_64
CPU op-mode(s):                       32-bit, 64-bit
Address sizes:                        52 bits physical, 57 bits virtual
Byte Order:                           Little Endian
CPU(s):                               512
On-line CPU(s) list:                  0-511
Vendor ID:                            AuthenticAMD
Model name:                           AMD EPYC 9754 128-Core Processor
CPU family:                           25
Model:                                160
Thread(s) per core:                   2
Core(s) per socket:                   128
Socket(s):                            2
Stepping:                             2
BogoMIPS:                             4500.11
Flags:                                fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good amd_lbr_v2 nopl nonstop_tsc cpuid extd_apicid aperfmperf rapl pni pclmulqdq monitor ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs skinit wdt tce topoext perfctr_core perfctr_nb bpext perfctr_llc mwaitx cpb cat_l3 cdp_l3 hw_pstate ssbd mba perfmon_v2 ibrs ibpb stibp ibrs_enhanced vmmcall fsgsbase bmi1 avx2 smep bmi2 erms invpcid cqm rdt_a avx512f avx512dq rdseed adx smap avx512ifma clflushopt clwb avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local avx512_bf16 clzero irperf xsaveerptr rdpru wbnoinvd amd_ppin cppc arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold avic v_vmsave_vmload vgif x2avic v_spec_ctrl vnmi avx512vbmi umip pku ospke avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg avx512_vpopcntdq la57 rdpid overflow_recov succor smca fsrm flush_l1d debug_swap
Virtualization:                       AMD-V
L1d cache:                            8 MiB (256 instances)
L1i cache:                            8 MiB (256 instances)
L2 cache:                             256 MiB (256 instances)
L3 cache:                             512 MiB (32 instances)
NUMA node(s):                         2
NUMA node0 CPU(s):                    0-127,256-383
NUMA node1 CPU(s):                    128-255,384-511
Vulnerability Gather data sampling:   Not affected
Vulnerability Itlb multihit:          Not affected
Vulnerability L1tf:                   Not affected
Vulnerability Mds:                    Not affected
Vulnerability Meltdown:               Not affected
Vulnerability Mmio stale data:        Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed:               Not affected
Vulnerability Spec rstack overflow:   Mitigation; Safe RET
Vulnerability Spec store bypass:      Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1:             Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2:             Mitigation; Enhanced / Automatic IBRS; IBPB conditional; STIBP always-on; RSB filling; PBRSB-eIBRS Not affected; BHI Not affected
Vulnerability Srbds:                  Not affected
Vulnerability Tsx async abort:        Not affected

And details about the (standard) mvapich2 configuration

[bourdinb at bbserv2 mvapich2-2.3.7-1-fgy2vo7eexx3booddriqbkn6hbd5puf7]$ cat lib/pkgconfig/mvapich2.pc
# this gives access to the mvapich2 header files
prefix=/2/sw/spack/opt/spack/linux-rocky9-x86_64_v3/gcc-13.2.0/mvapich2-2.3.7-1-fgy2vo7eexx3booddriqbkn6hbd5puf7
exec_prefix=${prefix}
libdir=${exec_prefix}/lib
includedir=${prefix}/include

Name: mvapich2
Description: High Performance and portable MPI
Version: 2.3.7
URL: http://mvapich.cse.ohio-state.edu
Requires:
Libs: -Wl,-rpath -Wl,${exec_prefix}/lib  -L${libdir} -lmpi  -lpmi2   -lpmi2 -lpthread
Cflags:   -I${includedir}

# pkg-config does not understand Cxxflags, etc. So we allow users to
# query them using the --variable option

cxxflags=  -I${includedir}
fflags= -I${includedir}
fcflags= -I${includedir}

For reference, we get the same error message with mvapich3.0 but don’t have any problems with openmpi-4.1.7.

Does anybody have an idea on how to get mvapich running on the new nodes?

Regards,
Blaise


—
Canada Research Chair in Mathematical and Computational Aspects of Solid Mechanics (Tier 1)
Professor, Department of Mathematics & Statistics
Hamilton Hall room 409A, McMaster University
1280 Main Street West, Hamilton, Ontario L8S 4K1, Canada
https://www.math.mcmaster.ca/bourdin<https://urldefense.com/v3/__https://www.math.mcmaster.ca/bourdin__;!!KGKeukY!xNacGdMGaPECzNbB3tBpV9J_3zV3k9Mz08dLYfUqVxo6f2H7xxoU-SoOSFGn7Wk-m2WD5MTm6SHmx-vSmQOq3FkDCQ_RHg$> | +1 (905) 525 9140 ext. 27243

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osu.edu/pipermail/mvapich-discuss/attachments/20250117/b7366226/attachment-0002.html>