Profile mode:
Fix roofline plots for datatypes that have peakVALU only. Check for highest roofline to plot the bandwidth lines to proper height, don't rely on existence of peakMFMA for every datatype.
Analyze mode:
Add roofline-data-type option for viewing pdfs in standalone gui. Default is same as profile mode, FP32.
---------
Signed-off-by: Carrie Fallows <Carrie.Fallows@amd.com>
* Removed unnecessary rocm-smi files
* Moved the update wrapper script into the tools folder
---------
Signed-off-by: Maisam Arif <Maisam.Arif@amd.com>
* Removed unnecessary rocm-smi files
* Moved the update wrapper script into the tools folder
---------
Signed-off-by: Maisam Arif <Maisam.Arif@amd.com>
[ROCm/amdsmi commit: 50d7d5287f]
Unit_hipMalloc_AllocateMoreThanTotalRAM is an invalid test. hipMalloc is able to allocate more than RAM size if VRam is larger, unless the test is meant to only allocate on host device.
Previously this test may have passed because of an incorrect check on hipMalloc limit: we had incorrectly limited allocation to physical ram size. This is no longer correct. We could always allocate Vram limit or sometimes system ram limit if Vram is full, fixed in https://github.com/AMD-ROCm-Internal/clr/pull/122
Co-authored-by: Jimbo Xie <jiabaxie@amd.com>
[ROCm/hip-tests commit: e25680e4d1]
Unit_hipMalloc_AllocateMoreThanTotalRAM is an invalid test. hipMalloc is able to allocate more than RAM size if VRam is larger, unless the test is meant to only allocate on host device.
Previously this test may have passed because of an incorrect check on hipMalloc limit: we had incorrectly limited allocation to physical ram size. This is no longer correct. We could always allocate Vram limit or sometimes system ram limit if Vram is full, fixed in https://github.com/AMD-ROCm-Internal/clr/pull/122
Co-authored-by: Jimbo Xie <jiabaxie@amd.com>
Add VmHeapArray class to reduce the pressure on VA reservation, since
multiple memory pools can be active at the same time.
[ROCm/clr commit: e974f7fde1]
Use spaces consistently to format the trap handler code. This patch
does not introduce any change in the trap handler. Using `git show -w`
on this patch shows an empty diff.
Change-Id: Ic0244dd203347146ffde65460cd87ecbcc43732a
Use spaces consistently to format the trap handler code. This patch
does not introduce any change in the trap handler. Using `git show -w`
on this patch shows an empty diff.
Change-Id: Ic0244dd203347146ffde65460cd87ecbcc43732a
[ROCm/ROCR-Runtime commit: e0359e5d35]
* Add MI 350 hardware information
* Refactor MI GPU YAML file and corresponding interface
* Add SoC file for gfx950 architecture
* Add analysis report configs for MI 350 containing existing metrics
* Add placeholder None valued metrics for previous architectures to make
baseline comparison work
* Enable testing on MI 350
* Analysis config metric changes
- SPI changes
- Update metric formula for default SPI pipe counter
- Use efficiently collected pipe wise SPI counters
- Add SPI Wave Occupancy
- Add Scheduler-Pipe Wave Utilization
- Update formula for VGPR Writes
- Add Scheduler-Pipe FIFO Full Rate
- CPC changes
- Add CPC SYNC FIFO Full Rate
- Add CPC CANE Stall Rate
- Add CPC ADC Utilization
- SQ changes
- Add VALU co-issue efficiency
- Add F6F4 datatype metrics
- Update formula for total FLOPs by adding F6F4 counters
- Add LDS STORE / LOAD / ATOMIC metrics
- Add LDS STORE / LOAD / ATOMIC bandwidth
- Add LDS FIFO and TA ADDR / CMD / DATA FIFO full rates
* Collect TCP_TCP_LATENCY_sum only for gfx950 (MI 350)
* Do not inject SQ_ACCUM_PREV_HIRES unnecesarily
* Do not hardcode memory and shader clock speeds
* Write num_hbm_channels to sysinfo.csv instead of hbm_bw while profiling
* Move generate sysinfo.csv to pre processing step of profiling
* Add warnings to use --specs-correction for missing sysinfo.csv values during analysis phase
* Update CHANGELOG
* Analysis phase warning to use --specs-correction when needed
[ROCm/rocprofiler-compute commit: f9aa7be97c]
cfallows-amd
Choudhary, Rahul
Betigeri, Sourabh
Pham, Gabriel
Arif, Maisam
Galantsev, Dmitrii
Yiannis Papadopoulos
Justin Williams
Sajina PK
Cheruvally, Aravindan
Xie, Jiabao(Jimbo)
Andryeyev, German
dependabot[bot]
Rawat, Swati
Arandjelovic, Marko
Lancelot SIX
vedithal-amd