Commit Graph

158 Commits

Author SHA1 Message Date
Bill(Shuzhou) Liu 79bd9c1d5f change sensor_type in amdsmi_dev_get_temp_metric() to enum
The sensor_type in amdsmi_dev_get_temp_metric() will be changed to
amdsmi_temperature_type_t

Change-Id: I72a7f271b0a55a025acc2ca523062a3d51cc036d
2023-01-04 13:01:04 -06:00
Dalibor Stanisavljevic e22e72d4c3 SWDEV-371492 - Added check that device_handle is valid
Change-Id: Ic1b593fd5f781650528c860c372fa9864624255d
Signed-off-by: Dalibor Stanisavljevic <Dalibor.Stanisavljevic@amd.com>
2022-12-22 12:57:42 +01:00
Dalibor Stanisavljevic 4c56e9e3d6 SWDEV-371199 - Return NOT_INIT when amdsmi initialization fails
Change-Id: Ifb40aef3a62885b08164e9aa944bf9b5c375ebfd
Signed-off-by: Dalibor Stanisavljevic <Dalibor.Stanisavljevic@amd.com>
2022-12-19 16:29:29 +01:00
Bill(Shuzhou) Liu 221d6fdc5c Make amdsmi function name consistent
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
2022-12-16 10:20:49 -06:00
Dalibor Stanisavljevic 238f885e14 SWDEV-371561 - Fixed vbios version string value
Change-Id: Ide06784200084741e6cde606492bf03a760b9601
Signed-off-by: Dalibor Stanisavljevic <Dalibor.Stanisavljevic@amd.com>
2022-12-09 15:19:12 +01:00
Dalibor Stanisavljevic a2a38a5aa2 SWDEV-371210 - Fixed pcie link speed
Change-Id: I736d8095c05ee0685db0c209ea0fdb5832e14744
Signed-off-by: Dalibor Stanisavljevic <Dalibor.Stanisavljevic@amd.com>
2022-12-08 12:03:50 -05:00
Dalibor Stanisavljevic b93baf686d SWDEV-371191 - Fixed amdsmi_get_bad_page_info
Change-Id: I97134f548164eff588d9caa9b9f31c4361c78804
Signed-off-by: Dalibor Stanisavljevic <Dalibor.Stanisavljevic@amd.com>
2022-12-08 12:03:34 -05:00
Dalibor Stanisavljevic b4b761d02f SWDEV-370223 - Change the name of the header to amdsmi.h
Change dev to device_handle throughout the file
Change the pcie_info pcie_speed field type to uint32_t
Add AMDSMI prefix before amdsmi_mm_ip enum

Change-Id: I242145389ddc3f2ad05dfd6ca371640f4d118fc4
Signed-off-by: Dalibor Stanisavljevic <Dalibor.Stanisavljevic@amd.com>
2022-12-08 13:34:34 +01:00
Jason Albert 3b1584915b Set status codes to fixed values
Assign fixed values to status codes to prevent enum auto assign
from changing them.

Signed-off-by: Jason Albert <jason.albert@amd.com>
Change-Id: I0ca1de7ba503ce8a75c56026f5a54e212204595b
2022-12-07 10:39:26 -05:00
Galantsev, Dmitrii aeb0bf5832 CMAKE: Repackage whole project for ROCm 5.5 release
Signed-off-by: Galantsev, Dmitrii <dmitrii.galantsev@amd.com>
Change-Id: I932b11a111c8e0db04bd8c5e0c3d1a470e5b2386
2022-11-29 17:04:32 -06:00
Dalibor Stanisavljevic 76f6cf7a9d SWDEV-366720 - Changed amdsmi_get_device_handle_from_bdf
Changed implementation and input parameters

Change-Id: Ifca3247132eb4033f99d74617a53f54ad076dad0
Signed-off-by: Dalibor Stanisavljevic <Dalibor.Stanisavljevic@amd.com>
2022-11-22 10:28:45 -05:00
Dalibor Stanisavljevic 9cad9e5216 SWDEV-361376 - Add README for python tool
- Add up to date README file for python tool

Change-Id: I7a02f79469e990870398b3741b033ea447998fdd
Signed-off-by: Dalibor Stanisavljevic <Dalibor.Stanisavljevic@amd.com>
2022-11-10 16:57:49 +01:00
Bill(Shuzhou) Liu b34b7451e8 Init the amdsmi using rocm_smi for libdrm
Init the ams_smi using the rocm-smi, which makes the GPU discovery
consistent with or without libdrm.

Change-Id: Ic714781f8ce791451b0c057621525926edb7f5ee
2022-11-07 11:09:09 -06:00
Bill(Shuzhou) Liu 9a92ea833f The device name and vbios version is incorrect
Get the device name from rocm-smi which is not displayed properly
in some cards. Set the vibos version using the rocm-smi.

Change-Id: I138f1760cde94007cb93cad02c6d8cccbb4afa28
2022-10-28 13:03:18 -05:00
Dejan Andjelkovic 6064f160a3 SWDEV-361376 - Add python wrapper
- Add generator for python wrapper
- Add interface, exception and init files
- Add CMake custom targets

Change-Id: I63c1d94fbb587387c22f559a3db79987eb214a2e
Signed-off-by: Dejan Andjelkovic <Dejan.Andjelkovic@amd.com>
2022-10-20 09:24:53 -05:00
Bill(Shuzhou) Liu 2b2d11c446 Change the get_socket_handles and get_device_handles APIs interface
Those two APIs are changed to let the user get the handles count,
allocate memory, and then return handles to the allocated memory.

Change-Id: Ibe28a89ad188c99da6af3af1740b2b25ff22ba06
2022-10-20 09:24:31 -05:00
Dalibor Stanisavljevic 3daf9c1063 SWDEV-353742 - Port smilib function to amdsmi
Change-Id: I99df249755a5c665a8dd1777fa82d046e139bd77
Signed-off-by: Dalibor Stanisavljevic <Dalibor.Stanisavljevic@amd.com>
2022-10-20 09:24:22 -05:00
Bill(Shuzhou) Liu 0c91ef919d Restructure the folder
Move rocm_smi related function to rocm_smi folder. Move amd_smi to
top level include/ and src/ folder. Remove obsolte oam folder.
Change the CMakeLists.txt to update folder locations.

Change-Id: I52e6be739e49f3b0545865f25364787f5985e9c3
2022-10-20 09:23:51 -05:00
Divya Shikre b23cfc0e82 Fix mem leaks observed while running rsmitst
1.  Memory allocated for handle was not deleted
when no variant, subvariant or supported function
was found
2. handle->func_id_iter address was set to 0
before delete[]

Signed-off-by: Divya Shikre <DivyaUday.Shikre@amd.com>
Change-Id: Iab50fdfbe03eec8e6fd0e84e03bd2c47e645b3d8
2022-05-18 14:31:44 -04:00
Divya Shikre afe996c2ed Update get_frequencies to handle failures.
Show an optional debug log (RSMI_DEBUG_BITFIELD=2) to
the user in the following scenarios:
1. If more than one current frequency is found
2. If frequencies are not read in increasing order of
   their value
If current frequency is not available, index for it is
set to -1, values will not have * next to it in the
output. This will also be handled in rocm_smi.py.

Signed-off-by: Divya Shikre <DivyaUday.Shikre@amd.com>
Change-Id: I477ec065f7513c8045d6392f12ef6cb835a6b8f6
2022-05-11 15:33:15 -04:00
Divya Shikre 99be3451d7 Add DEBUG_LOG macro
Add DEBUG_LOG that will optionally print error
message when RSMI_DEBUG_BITFIELD is set to 2.

Signed-off-by: Divya Shikre <DivyaUday.Shikre@amd.com>
Change-Id: I6017e92d8a9e5f9861ae29ece0488d4bc198f996
2022-05-11 11:03:24 -04:00
Divya Shikre c9b42bff57 Add RSMI_CLK_TYPE_PCIE to rsmi_clk_type_t
showclocks/showclkfrq does not display pp_dpm_pcie values
in sriov. This fix adds pcie clocks to rsmi_clk_type_t
where rest of the clocks are present.

Signed-off-by: Divya Shikre <DivyaUday.Shikre@amd.com>
Change-Id: I6d129ae412623b369c14456ae9781b2dbceb2139
2022-05-06 09:15:39 -04:00
Ori Messinger 9d6403bb17 ROCm SMI LIB: Add Missing GPU Blocks
This patch adds the following 4 missing GPU blocks to the SMI LIB:
-RSMI_GPU_BLOCK_MMHUB
-RSMI_GPU_BLOCK_PCIE_BIF
-RSMI_GPU_BLOCK_HDP
-RSMI_GPU_BLOCK_XGMI_WAFL

Signed-off-by: Ori Messinger <Ori.Messinger@amd.com>
Change-Id: Ia1ec6f53e195f4bf7b8f073d6bed4fdb6572e546
2022-05-05 00:44:16 -04:00
Bill(Shuzhou) Liu 7860de5107 Suppress "rsmi_init() failed" error message
When an application call the library in a system without amdgpu,
it may always print out "rsmi_init() failed". Suppress the error
message in the library.

Change-Id: Ice63dd3a764b221a6935536bff1bfa6aa3e51a46
2022-04-12 09:44:00 -04:00
Sreekant Somasekharan dbe3403bd3 make string variable 'tpath' an empty string.
string variable not being empty can lead to incorrect compilation
and corrupted output.

Change-Id: Ie66756c28aef7417759c29387500970a8b53e44c
2022-03-11 21:22:28 -05:00
Bill(Shuzhou) Liu 4b65b0307f Prevent stack buffer overflow
readlink() does not append a null byte to buffer. Initialize the
tpath to prevent stack buffer overflow.

Change-Id: I17895dc3576b080a0c35bd0528a5b83223ec1c1b
2022-03-03 15:43:53 -05:00
Laurent Morichetti 2804bf7c28 Don't use NDEBUG when the intent is !DEBUG
CMakeLists.txt does not set up the DEBUG macro correctly to mean
!NDEBUG, so, as a workaround, replace all uses of ifdef NDEBUG with
ifndef DEBUG in the library sources.

Change-Id: I408adb36d1a2310fb894a486574469662ebb27cd
(cherry picked from commit 9f87197d8d)
2022-01-27 11:08:48 -05:00
Divya Shikre ec71380e1c Add fix to check for vector size while reading pp_dpm_pcie
pop_back() was causing a seg fault when pp_dpm_pcie file is empty and returns whitespace.

Signed-off-by: Divya Shikre <DivyaUday.Shikre@amd.com>
Change-Id: I888f1f79751cd456e43751a5b96d08560a039677
2022-01-26 10:34:57 -05:00
Divya Shikre 432df20321 Add null ptr check for temperature read from all sensors.
The (temperature == nullptr) check happens only when HBM temperature is retrieved.
This check needs to apply in other cases as well, hence moving this outside the HBM condition.
This should return RSMI_STATUS_INVALID_ARGS consistently in all cases when nullptr is passed through rsmitst.

Signed-off-by: Divya Shikre <DivyaUday.Shikre@amd.com>
Change-Id: Iea3cec75312a0a669c7da27e15e9782e6a885c5f
2021-12-01 14:05:46 -05:00
Divya Shikre 7b1daaef96 Add fix to display correct GPU Memory Activity and GFX Activity value.
Driver mem fills in 0xFF for all for the metrices not supported for that ASIC.
So if 0xFF is detected, return RSMI_STATUS_NOT_SUPPORTED

Signed-off-by: Divya Shikre <DivyaUday.Shikre@amd.com>
Change-Id: I86a38148c7a288ea0db94893f685560eaac098ab
2021-11-25 14:28:06 -05:00
Divya Shikre f61cb1b41d Add fix for out of range temperature value for HBM.
Driver mem fills in 0xFF for all for the metrices not supported for that ASIC.
So if 0xFF is detected, return RSMI_STATUS_NOT_SUPPORTED

Signed-off-by: Divya Shikre <DivyaUday.Shikre@amd.com>
Change-Id: Iacb6474486e3732f2aa824ff447c17f8243b65cd
2021-11-23 15:37:41 -05:00
Elena Sakhnovitch 50ea68e694 [ROCm SMI LIB]: Add rsmi_minmax_bandwidth_get()
API provides min/max bandwidth values between nodes.
(Current implementation only supports directly (1 hop)
connected XGMI devices.

Signed-off-by: Elena Sakhnovitch
Change-Id: Ifc95da13845fbe7903c5386d320183ffd58c5b53
2021-10-28 17:00:41 -04:00
Ori Messinger ff02042c64 ROCm SMI LIB: Add rsmi_is_P2P_accessible() API
Implements rsmi_is_p2p_accessible API.
The function returns True if P2P is possible between two nodes.

Signed-off-by: Ori Messinger <Ori.Messinger@amd.com>
Change-Id: Ic7316eebcec4480175c7ad04c21a42b2e1a4c454
2021-10-13 22:01:33 -04:00
Elena Sakhnovitch 5e1bfcadd7 rocm_smi_lib: fix gpu_metrics_v1_3 support
Signed-off-by: Elena Sakhnovitch
Change-Id: Ia7a6b17eb0f317465613ba92ae7548a221c46ee3
2021-08-13 11:59:50 -04:00
Elena Sakhnovitch fee82af1fe rocm_smi_lib: add gpu_metrics_v1_3 support
Signed-off-by: Elena Sakhnovitch
Change-Id: I4a9dedc80b8fce60e12c5baf8651d54d16a6a41c
2021-08-13 09:23:35 -04:00
Harish Kasiviswanathan 7a8c3f3629 Fall back to pci-ids if FRU product_name is empty
rocm-smi --showproductname will not show "Card series" in its output if
product_name exported by Kernel is empty string. This has been raised a
regression by customer.

BUG: SWDEV-297228

Signed-off-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
Change-Id: I9aae24778e2d3a30aa661d8f338278c1666590fb
2021-08-04 10:53:55 -04:00
Bill(Shuzhou) Liu 8c60dbebaa AddressSanitizer report stack-use-after-scope
Fix the stack-use-after-scope error reported by the AddressSanitizer.

Bug: SWDEV-291913
Change-Id: I0ffd71af8679b8bff6c363096fafe75dffcf329e
2021-06-25 13:33:38 -04:00
Divya Shikre 462d4adc24 Return an error when user tries to set out of range clock values for setsrange functionality
Signed-off-by: Divya Shikre <DivyaUday.Shikre@amd.com>
Change-Id: Ibe1075c1d2b6c009332a52b81f4b41f7e93d0756
2021-05-11 12:32:19 -04:00
Harish Kasiviswanathan 6b10a7761b Add support to read gpu_metrics version 1.2
gpu_metrics version 1.2 provides atomic timestamp. Use this timestamp.

Signed-off-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
Change-Id: I7a1a675f53b93718f34b1f2979173e9064e0ef93
2021-05-05 12:31:10 -04:00
Harish Kasiviswanathan e83cf605c6 Change #define RSMI_GPU_METRICS_API_CONTENT_VER
Chnage to RSMI_GPU_METRICS_API_CONTENT_VER_1. In preparation for
supporting additional formats

Signed-off-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
Change-Id: I4367a2622a0fa41e6b05bc4436ecd24b8c4e30e2
2021-05-04 20:51:10 -04:00
Harish Kasiviswanathan c416726054 Move gpu_metrics functions to different file
No logic change. Only structural change

Signed-off-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
Change-Id: Id5e1a678c0888f04081ee06db4521c72b5eb9b16
2021-05-04 20:49:51 -04:00
Ori Messinger 83cd2fe4f1 ROCm SMI LIB: Add Default GPU Power Cap
Implement default GPU power cap functionality in the LIB.

Signed-off-by: Ori Messinger <Ori.Messinger@amd.com>
Change-Id: Ia6b3420beb0e4df5559c3e6d11d0667972590b53
2021-04-22 10:49:55 -04:00
Harish Kasiviswanathan 844acbc0d8 Add energy counter resolution to rsmi_dev_energy_count_get
Signed-off-by: Harish Kasiviswanathan <Harish.Kasiviswanathan@amd.com>
Change-Id: I03b70968257db7a45e21d7ba62542cdedd18eb85
2021-04-22 10:25:06 -04:00
Divya Shikre dc431506f5 Add support for mi200 clocks being continuous.
Signed-off-by: Divya Shikre <DivyaUday.Shikre@amd.com>
Change-Id: Ifb7570054572239b9f48eaefe51e879fb3569031
2021-04-20 13:12:27 -04:00
Divya Shikre 9f9a7aaf65 Add new setrange function in C++ lib
Signed-off-by: Divya Shikre <DivyaUday.Shikre@amd.com>
Change-Id: I670aaeb93827bf4b2cc08eb36d0f9756f00e4e4e
2021-04-19 22:38:59 -04:00
Elena a383dd23aa [rocm-smi-lib] add HBM temperature conversion factor
Change-Id: I45339c87c3d2a40670baf1b76ada60dceb650dc0
2021-04-19 16:41:48 -04:00
Bill(Shuzhou) Liu 8eec0a7d36 Add energy accumulator counter
The energy accumulator counter tracks all energy consumed.

Change-Id: I5b25f817b7802d81c477361447f0ecd7ec02fc61
2021-04-14 10:43:01 -04:00
Bill(Shuzhou) Liu 9bfb9ac297 Add coarse grain utilization counter
The coarse grain utilization counter includes GFX and Memory activity.

Change-Id: I5d09976792d3f4a1c1081651fa24ff857016d4c0
2021-04-14 10:40:19 -04:00
Divya Shikre aaf2120117 Update performance determinism api as per the modified sysfs interface.
Signed-off-by: Divya Shikre <DivyaUday.Shikre@amd.com>
Change-Id: Ib0ec5128819644a2ff6c916da9194a7fe1dad795
2021-04-07 16:38:48 -04:00
Bill(Shuzhou) Liu da480b4589 Add support for the HBM temperature
The rsmi_dev_temp_metric_get() can also support the HBM
temperatures which is retrieved from gpu_metrics.

Change-Id: I96b979296e90cf881523627b41b1a02849676416
2021-04-05 15:55:55 -04:00