Updated get_cache_info to use level, type, cache_size_kb,
and num_cu_shared as key to increment num_cache_instance.
Signed-off-by: Juan Castillo <juan.castillo@amd.com>
Change-Id: I54c88263bce5a49713f15e2687ae05212423b186
[ROCm/amdsmi commit: 8c9e77bb7b]
Most sysfs reads do not require cross-process level mutex, and writes to sysfs
should be protected by the kernel already.
Users can still switch to the old behavior by setting the environment variable AMDSMI_MUTEX_CROSS_PROCESS=1.
[ROCm/amdsmi commit: 8e74024b11]
After some update to amdclang a tag '--no-allow-shlib-undefine' began
being enforced. This caused compile failure when building with:
CC=amdclang CXX=amdclang++ cmake -B build
DRM_AMDGPU_LIB is never used.
DRM_LIBRARIES is the one set by pkg_check_modules(DRM...)
Change-Id: I031aa1325b7a09a027aba8ca5cc7cc44fe86d3be
Signed-off-by: Galantsev, Dmitrii <dmitrii.galantsev@amd.com>
[ROCm/amdsmi commit: c3864bf617]
Blacklisted TestVoltCurvRead for devices with gfx_target_version
90400, 90401 and 90402 as it is not supported on these systems.
Signed-off-by: Bindhiya Kanangot Balakrishnan <Bindhiya.KanangotBalakrishnan@amd.com>
[ROCm/amdsmi commit: a7283196a7]
1) When `clang` is used as system compiler, libraries were built without respecting LDFLAGS. For example, this affected LTO flags, if any (and it only affected clang, not gcc).
2) Linker flags are registered as CXX flags, which produces warnings during compilation:
```
clang++: warning: -Wl,-z,noexecstack: 'linker' input unused [-Wunused-command-line-argument]
clang++: warning: -Wl,-znoexecheap: 'linker' input unused [-Wunused-command-line-argument]
clang++: warning: -Wl,-z,relro: 'linker' input unused [-Wunused-command-line-argument]
clang++: warning: -Wl,-z,now: 'linker' input unused [-Wunused-command-line-argument]
```
3) Clang does not support `-Wtrampolines` flag:
```
warning: unknown warning option '-Wtrampolines' [-Wunknown-warning-option]
```
4) No linkers support `noexecheap` anymore. `noexecheap` linker flag was a part of PaX patches to GNU ld, (which were dropped in 2017)[https://www.gentoo.org/support/news-items/2017-08-19-hardened-sources-removal.html]. Now ld/ld.lld/ld.gold don't support it and protection of heap is managed by NX bit. Therefore every compiler produces this warning:
```
ld.lld: warning: unknown -z value: noexecheap
```
Change-Id: I2334a4d4c745df2abc12d543616ca179f85c3575
Signed-off-by: Galantsev, Dmitrii <dmitrii.galantsev@amd.com>
Co-authored-by: Sv. Lockal <lockalsash@gmail.com>
[ROCm/amdsmi commit: b256c1b6de]
Full output:
$ amd-smi metric:
AttributeError: 'AMDSMILogger' object has no attribute 'clear_multiple_devices_output'. Did you mean: 'clear_multiple_devices_ouput'?
Changes:
* Changed CLI function definition clear_multiple_devices_ouput(self) ->
clear_multiple_devices_output(self)
* Updated all references to clear_multiple_devices_ouput() to use
clear_multiple_devices_ouput()
Change-Id: Ibd4e210ea30c9dd51fba17981a524b823f2db054
[ROCm/amdsmi commit: 1d2272490e]
BM specific help text contained functions that required the driver to be loaded.
this was causing amd-smi not supported error on Linux guests.
Fixed this by wrapping the help text in the proper checks
Signed-off-by: Bindhiya Kanangot Balakrishnan <Bindhiya.KanangotBalakrishnan@amd.com>
[ROCm/amdsmi commit: 51c705fd43]
* amdsmi: Adding Support to get hsmp Driver version
Adding Support to fetch hsmp driver version from ESmi Interfaces.
Adding Support to fetch memory bandwidth per socket.
Signed-off-by: muthusamy <muthusamy.ramalingam@amd.com>
[ROCm/amdsmi commit: ced110dbb6]
The count field was not pythonic nor needed
Signed-off-by: Maisam Arif <Maisam.Arif@amd.com>
Change-Id: I212f43dc11f2f2c7eddd39900e6e3aaec03f3f8f
[ROCm/amdsmi commit: 803b18fe95]
Logging specific clocks in csv format was causing TypeError as the levels were int.
Fixed this by appending Level string at the beginning.
Signed-off-by: Bindhiya Kanangot Balakrishnan <Bindhiya.KanangotBalakrishnan@amd.com>
[ROCm/amdsmi commit: e3e11835e4]
The target_graphics_version was not formatted properly and was
showing incorrect Target Name. Corrected this by fomatting
major, minor and revision numbers.
Signed-off-by: Bindhiya Kanangot Balakrishnan <Bindhiya.KanangotBalakrishnan@amd.com>
[ROCm/amdsmi commit: 6fa991c39c]
* [SWDEV-508173] Updates include:
- Updating py-interface to import amdsmi_get_gpu_reg_table_info and amdsmi_get_gpu_pm_metrics_info.
- Updating the ctypes from byref to pointer.
Signed-off-by: Castillo, Juan <Juan.Castillo@amd.com>
[ROCm/amdsmi commit: 9cc5c303a2]
Multi-threaded application rsmi_dev_gpu_metrics_info_get() causes crash
Code changes related to the following:
* API implementation changes
Change-Id: I1f1fb39c1125569ec5d534b37fd6f68c8829eef7
Signed-off-by: Oliveira, Daniel <daniel.oliveira@amd.com>
Authored-by: Oliveira, Daniel <daniel.oliveira@amd.com>
[ROCm/amdsmi commit: 03a2368655]
Corrected clean_local_data error in Linux VM's while doing
amd-smi set without args.
Signed-off-by: Bindhiya Kanangot Balakrishnan <Bindhiya.KanangotBalakrishnan@amd.com>
[ROCm/amdsmi commit: 834993e1c3]