Server Processors

[yilongl]

Hi,

I am trying to use AMDuProfPcm to collect metrics for my application. I was able to install the AMDuProf successfully on my Ubuntu 22.04 system following the instructions in the user guide. However, I cannot launch any application using AMDuProfPcm. There are two types of errors: "Fail to execute with execvp" and "Error: Invalid Launch application path" (two examples and some info about my system are attached below). Unfortunately, I cannot find any relevant information in the user guide. Has anyone seen these error messages before? Thank you!

Example 1:

root@node0:/opt/AMDuProf_4.0-341# bin/AMDuProfPcm -m ipc -- sleep 1
Error: Invalid Launch application path : sleep

Example 2:

root@node0:/opt/AMDuProf_4.0-341# bin/AMDuProfPcm -m ipc -- Examples/AMDTClassicMatMul/bin/AMDTClassicMatMul-bin

AMDuProfPcm Report

System:,Family(0x17) Model(0x31) Stepping(0)
Number of Sockets :,1
Number of CCDs :,4
Number of Core Complexes(CCX) :,8
Number of Cores :,16
Number of Threads :,32
SMT Enabled in HW:,True
SMT Enabled by OS :,True
Number of Threads sharing L3:,4

CPU Topology:
Socket, CCD, CCX, Core(s)
0,0,0, 0 1 16 17
0,0,1, 2 3 18 19
0,1,2, 4 5 20 21
0,1,3, 6 7 22 23
0,2,4, 8 9 24 25
0,2,5, 10 11 26 27
0,3,6, 12 13 28 29
0,3,7, 14 15 30 31

Core P0 state frequency (MHz):,3000.000000
Measured DF Frequency (MHz):,231.992432
Multiplex interval (ms):,1000
Sample interval (ms):,1000
Version:,4.0.341

Hypervisor Info:
Hypervisor Enabled:,0

Abbreviation :
IPC    : Instructions Per CPU Cycle
CPI    : CPU Cycles Per Instructions
pti    : Per Thousand Instructions
ptc    : Per Thousand CPU Cycles

Profile Time: 2023/02/14 01:47:36:927
Core-0,,,,,
Utilization (%),Eff Freq,IPC (Sys + User),CPI (Sys + User),Branch Misprediction Ratio,
Fail to execute with execvp. 

AMDuProfPcm Report

System:,Family(0x17) Model(0x31) Stepping(0)
Number of Sockets :,1
Number of CCDs :,4
Number of Core Complexes(CCX) :,8
Number of Cores :,16
Number of Threads :,32
SMT Enabled in HW:,True
SMT Enabled by OS :,True
Number of Threads sharing L3:,4

CPU Topology:
Socket, CCD, CCX, Core(s)
0,0,0, 0 1 16 17
0,0,1, 2 3 18 19
0,1,2, 4 5 20 21
0,1,3, 6 7 22 23
0,2,4, 8 9 24 25
0,2,5, 10 11 26 27
0,3,6, 12 13 28 29
0,3,7, 14 15 30 31

Core P0 state frequency (MHz):,3000.000000
Measured DF Frequency (MHz):,231.992432
Multiplex interval (ms):,1000
Sample interval (ms):,1000
Version:,4.0.341

Hypervisor Info:
Hypervisor Enabled:,0

Abbreviation :
IPC    : Instructions Per CPU Cycle
CPI    : CPU Cycles Per Instructions
pti    : Per Thousand Instructions
ptc    : Per Thousand CPU Cycles

Profile Time: 2023/02/14 01:47:36:928
Core-0,,,,,
Utilization (%),Eff Freq,IPC (Sys + User),CPI (Sys + User),Branch Misprediction Ratio,
0.00,2341.66,0.00,0.00,0.00,
0.00,1788.32,0.00,0.00,0.00,
^C0.02,1817.47,0.00,0.00,0.00,
0.02,1731.41,0.00,0.00,0.00,
root@node0:/opt/AMDuProf_4.0-341# 

CPU:

Architecture:                    x86_64
CPU op-mode(s):                  32-bit, 64-bit
Address sizes:                   43 bits physical, 48 bits virtual
Byte Order:                      Little Endian
CPU(s):                          32
On-line CPU(s) list:             0-31
Vendor ID:                       AuthenticAMD
Model name:                      AMD EPYC 7302P 16-Core Processor
CPU family:                      23
Model:                           49
Thread(s) per core:              2
Core(s) per socket:              16
Socket(s):                       1
Stepping:                        0
BogoMIPS:                        5988.90
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 nopl nonstop_tsc cpuid extd_apicid aperfmperf rapl pni pclmulqdq monitor ssse3 fma cx16 sse4_1 sse4_2 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 ibrs ibpb stibp vmmcall fsgsbase bmi1 avx2 smep bmi2 cqm rdt_a rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local clzero irperf xsaveerptr rdpru wbnoinvd amd_ppin arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold avic v_vmsave_vmload vgif v_spec_ctrl umip rdpid overflow_recov succor smca sme sev sev_es
Virtualization:                  AMD-V
L1d cache:                       512 KiB (16 instances)
L1i cache:                       512 KiB (16 instances)
L2 cache:                        8 MiB (16 instances)
L3 cache:                        128 MiB (8 instances)
NUMA node(s):                    1
NUMA node0 CPU(s):               0-31
Vulnerability Itlb multihit:     Not affected
Vulnerability L1tf:              Not affected
Vulnerability Mds:               Not affected
Vulnerability Meltdown:          Not affected
Vulnerability Mmio stale data:   Not affected
Vulnerability Retbleed:          Mitigation; untrained return thunk; SMT enabled with STIBP protection
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl and seccomp
Vulnerability Spectre v1:        Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2:        Mitigation; Retpolines, IBPB conditional, STIBP always-on, RSB filling, PBRSB-eIBRS Not affected
Vulnerability Srbds:             Not affected
Vulnerability Tsx async abort:   Not affected

OS:

Distributor ID:	Ubuntu
Description:	Ubuntu 22.04.1 LTS
Release:	22.04
Codename:	jammy

[santosh_zanjurne]

Hello, for both issues you reported, can you please try using absolute path for the executable you are trying to profile? 

AMDuProfPcm -m ipc -- `which sleep` 1

bin/AMDuProfPcm -m ipc -- $PWD/Examples/AMDTClassicMatMul/bin/AMDTClassicMatMul-bin

Regards,

Santosh

[yilongl]

Hi Santosh,

Sorry for the late reply. I didn't get a notification of your message.

I can confirm that absolute paths work properly in both cases.

Thanks,

Yilong

[yilongl]

So I assume this is a bug?

[santosh_zanjurne]

Hi yilong,

This is not a bug really - but i have filed a bug report / feature request to either allow users to specify relative path or document the requirement of absolute path in the user guide.

[santosh_zanjurne]

Hi, this issue has been fixed in our latest release of uProf 4.1.  Please visit AMD μProf | AMD for details.