Updates:
* Added rsmi_dev_nps_mode_set and rsmi_dev_nps_mode_get
* Added ability to set multiple SYSFS files in debug build
* Added ability to see user's env variables set for debug build
* Added tests for rsmi_dev_nps_mode_set and rsmi_dev_nps_mode_get
* Added ability to restart AMD GPU driver, used in nps_mode_set
* Updated ROCm_SMI_Manual.pdf to include new APIs
* Added progress bar for long running python_smi_tools, used
in setting nps_mode if runs longer than .1 seconds
Change-Id: I6d61bedd28d7cba6aff432ad2d127ba741b7d15a
Signed-off-by: Charis Poag <Charis.Poag@amd.com>
[ROCm/amdsmi commit: 9ef376cd61]
Fallback to other methods if the device name in sysfs is not printable.
Change-Id: I20b22950399d4a515d2688b955248a3de3c61d05
[ROCm/amdsmi commit: 00a6c78a51]
This patch fixes a --showproductname bug, which is related to the
device's SKU. If a device with a VBIOS value that cannot be decoded
is used, that device's SKU cannot be parsed out of the VBIOS string.
Now, when the VBIOS value cannot be decoded, an error will be
printed instead of crashing with an 'UnboundLocalError' message.
Signed-off-by: Ori Messinger <Ori.Messinger@amd.com>
Change-Id: I680a182e94107e782235b8a2477ab165988f7703
[ROCm/amdsmi commit: 56f9d6bfc0]
On APU's vbios_version string might not be exposed. Relying on vendor ID
to detect AMDGPU is sufficient
Signed-off-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
Change-Id: I08ef4b4bc7491a40f318791803aeaf261f7fac25
[ROCm/amdsmi commit: 142dcfa8f4]
The format of the fdinfo file has changed
Change-Id: Iad2e26487e75f3e614e364456e929aa1f6f949a4
Signed-off-by: Dalibor Stanisavljevic <Dalibor.Stanisavljevic@amd.com>
[ROCm/amdsmi commit: 411ef54087]
This reverts commit 64b1e4e65b.
Reason for revert: Branch is still WIP
Change-Id: I75eec813b3d81049f033fe0a534251bd69eeca0e
[ROCm/amdsmi commit: 6aa91da74c]
The tag values largely were not used and were causing doxygen
generation issues.
In the few cases where the tags were being referenced, clean up
those compile issues.
Signed-off-by: Jason Albert <jason.albert@amd.com>
Change-Id: I7b32eac742fb5af560400c13dda2721705d882bc
Signed-off-by: Dalibor Stanisavljevic <Dalibor.Stanisavljevic@amd.com>
[ROCm/amdsmi commit: 86de0f441f]
Original updates:
* Added .gitignore to help with future commits
* Updated/added copyrights on modified or added files
* Updated rocm_smi.h/.cc
- Added 3 new SMI API functions:
rsmi_dev_compute_partition_set &
rsmi_dev_compute_partition_get
- Added helpful maps/enums used in
new get/set compute_partition API calls
* Updated rocm_smi.py
- Added --showcomputepartition
- Added --setcomputepartition
- Fixed a few mistypes
* Updated rsmiBindings.py - added helpful class/dict/list
* Updated rocm_smi_example.cc
- Added helpful MACRO to detect if api is not supported.
- Added current_compute_partition set/get rocm lib calls
- Added helpful macro to discover future RSMI errors
- Commented out test_set_freq, was having permission issues
on a Navi21
* Updated rocm_smi_main.cc
- Added helpful map to debug API calls, left in for future use
- Added comment to better understand a non-class function returns
* Added computepartition_read_write.cc/.h
- Added get/set compute partition API test calls
- Confirmed on devices that do not support the API calls, tests pass
* Updated rocm_smi_test/main.cc
- Calls new compute partition gtests
Added following updates from review feedback:
* Updated rocm_smi.h/cc
- Removed C++ API calls, adding support for both C/C++
API calls could cause confusion and adds extra work for us
- rsmi_dev_compute_partition_get -> Fixed an edge case where
user gives a small buffer length size (smaller than data
received), but does not receive the partial buffer back.
google Tests are updated to reflect this find.
* Updated rocm_smi_example.cc
- Fixed test_set_freq, issue was that file was not writable.
We now indicate this warning, so prior errors make sense.
- General test code cleanup. Removed extra code,
by creating loops for tests.
* Updated rocm_smi_main.cc
- Moved and got rid of an external reference to a map used
for debugging RSMI enums, now is a const public reference.
* Updated rocm_smi.py
- Updated python code to identify NOT_SUPPORTED due to
(currently) only a few GPU support the feature
Change-Id: I4a567acbb59d6771fb64df08d19175fe3604fd1b
[ROCm/amdsmi commit: 4d7f3f2bc7]
When it fails to open sysfs file, it may crash. Modify the condition
to check the file descriptor after open the file.
Change-Id: I2acdc55f8194a2d734db20d16e1660a20ba09574
[ROCm/amdsmi commit: f19da1bb2c]
The amdsmi_dev_get_temp_metric() will cover both function:
amdsmi_get_temperature_measure() using AMDSMI_TEMP_CURRENT
and
amdsmi_get_temperature_limit() using AMDSMI_TEMP_CRITICAL
Remove those two function.
It also merge the amdsmi_get_power_limit() into
amdsmi_get_power_measure()
Change-Id: I40d4afeb2ec0ac7b64832729f36adfaae120c990
[ROCm/amdsmi commit: ec48312c61]
This patch fixes a couple of --showproductname bugs, both of which
are related to the device's SKU.
Previously if a device with a non-standard VBIOS name was used,
fetching that device's SKU wasn't working correctly.
A standard VBIOS name should follow the following pattern:
AAA-BBBBBB-CCC
Where the middle section "BBBBBB" between the hypens is the SKU.
Now, SKU can be correctly fetched even with a non-standard VBIOS
name, and return 'unkown' if SKU does not exist.
Signed-off-by: Ori Messinger <Ori.Messinger@amd.com>
Change-Id: I5899a859c6131c6048bb31a4305ddacbac3075a9
[ROCm/amdsmi commit: 5c478e9eb9]
The sensor_type in amdsmi_dev_get_temp_metric() will be changed to
amdsmi_temperature_type_t
Change-Id: I72a7f271b0a55a025acc2ca523062a3d51cc036d
[ROCm/amdsmi commit: 79bd9c1d5f]
Tests overwritten the linker flags resulting in failed build with clang.
This change improves ASAN linker flag assignment and fixes test issue.
Change-Id: I88f38360d46b20f6cc7298ad0d1fd09ff6ce47d6
Signed-off-by: Galantsev, Dmitrii <dmitrii.galantsev@amd.com>
[ROCm/amdsmi commit: 2184d0c3d7]
Use ctypes to import types instead of via amdsmi_wrapper
Change-Id: I217e90b74aafdd39eaab5f50edfda80e0bf91cce
Signed-off-by: Dalibor Stanisavljevic <Dalibor.Stanisavljevic@amd.com>
[ROCm/amdsmi commit: 0f7c440d95]
Some of the amdsmi function have the verb (set or get) at the
end of the function. Move it to the middle to be consistent with
other APIs.
Change-Id: I8053d16f46af951c25aaaf8febf2896a33633fa1
[ROCm/amdsmi commit: 221d6fdc5c]
The purpose of this patch is to add a new feature to the smi cli.
Use ./rocm-smi --showtempgraph to print a persistant bar graph for
each GPU's temperature.
The bar graphs refresh continuously to show current temps, and the
graphs change in a color gradient depending on the temperature.
Signed-off-by: Ori Messinger <Ori.Messinger@amd.com>
Change-Id: I98902b76c42cc7281420759f5ebe8c78f7785e66
[ROCm/amdsmi commit: 932feb6e49]
amdsmitst was failing and not finding libgtest and libamd_smi.
This change resolves the issue by
1. Installing gtest into tests directory
2. Modifying RUNPATH variable to point to libamd_smi.so
Change-Id: I126d01c88116d37c5f2b55b9ecb2c9f1313f26fe
Signed-off-by: Galantsev, Dmitrii <dmitrii.galantsev@amd.com>
[ROCm/amdsmi commit: 0c52236abd]
We append CE then UE, but in the table right after, it goes UE then CE.
Fix the order of the table, and add capitals for consistency
Change-Id: I208f37685508ab1e2ff83d3456620bbbf3a16268
[ROCm/amdsmi commit: 248c6f79f4]
Galantsev, Dmitrii
Charis Poag
Bill(Shuzhou) Liu
Ranjith Ramakrishnan
Ori Messinger
Harish Kasiviswanathan
Dalibor Stanisavljevic
Maisam Arif
Elena Sakhnovitch
Jason Albert
David Ceranic
kent.russell@amd.com