AI/rocm-systems
2
0
Fork 0
程式碼 問題 合併請求 Actions 套件 專案 版本發布 Wiki 動態
1,236 提交 3 分支 0 標籤
f8c19dce67351aef9649e247aa513ce1598e25d7
提交線圖

52 次程式碼提交

作者 SHA1 備註 日期
Maisam Arif 11c72946eb Revert "SWDEV-458102 - Deprecated Voltage Curve API" 
This reverts commit 1423fb632e.

Change-Id: I8a3eaf0a9f28200e09fb35d5260fbc070fe8a4a9
 2024-05-02 15:27:16 -05:00
Maisam Arif 1423fb632e SWDEV-458102 - Deprecated Voltage Curve API 
Signed-off-by: Maisam Arif <maisarif@amd.com>
Change-Id: I111c3ce26d2ab66d5e755432f4b8a9bfa631f805
 2024-05-02 02:53:29 -04:00
Bill(Shuzhou) Liu a0d0210761 Process isolation sysfs format change 
The process isolation sysfs format is changed. This fix will
adapt to the new sysfs format.

Change-Id: Id6fd7eeb3e25525047dccab248fd9cfb206cbf62
 2024-04-25 08:11:05 -05:00
Bill(Shuzhou) Liu 7d2ab7970d Process isolation and clean shader 
A few APIs and command line options are added to support process
isolation and clean shader.

Change-Id: I98ad3fc9fc7429799a21798b7fca1c307de7f403
 2024-04-24 13:22:20 -04:00
Oliveira, Daniel 1ae3a5b6cb fix: [SWDEV-458102] [rocm/amd_smi_lib] 
Drops checks that are invalid with the new pp_od_clk_voltage format

Code changes related to the following:
  * get_od_clk_volt_info()
  * get_od_clk_volt_curve_regions()

Change-Id: I534c920e00fa3dacdb980f431db5eef260ac93f5
Signed-off-by: Oliveira, Daniel <daniel.oliveira@amd.com>
 2024-04-23 18:23:39 -05:00
Oliveira, Daniel 08e2e21bab fix: [SWDEV-442525] [rocm/amd_smi_lib] 
Fixes gpu_process_list

Code changes related to the following:
  * amdsmi_get_gpu_process_list()
  * CLI
  * Examples
  * Unit tests
  * Changelog
  * Readme
  * rocm_smi_lib commit: 677433b367

Change-Id: I9210fbca7a5da92d0a8b472b72ca82597c8e4fb5
Signed-off-by: Oliveira, Daniel <daniel.oliveira@amd.com>
 2024-03-27 16:48:24 -05:00
Bill(Shuzhou) Liu e4085c6414 Get and set the XGMI PLPD 
Update the API and CLI to support XGMI Per-Link Power Down Policy.

Change-Id: Iaf04a771eb8bb0829a5b3088d803a7355a8dfd0b
 2024-03-26 01:48:14 -05:00
Bill(Shuzhou) Liu 108e6d4ae6 Set and get DPM policy for GPU device 
Add new APIs to set and get dpm policy for the GPU device.

Change-Id: I26fa49cd17d0ce66bda3446c38945a6cf35717ff
 2024-03-12 10:32:31 -04:00
Galantsev, Dmitrii 44c189b9f5 Fix misc memory leaks 
Change-Id: I3dbf56e98d8c1312f9081956ed590962b2bdace3
Signed-off-by: Galantsev, Dmitrii <dmitrii.galantsev@amd.com>
 2024-03-07 04:56:16 -06:00
Galantsev, Dmitrii 0155219389 Fix memory leak created by hanging opendir 
Change-Id: I01e372c6a6b427f21e89cb5e4217f876346a35be
Signed-off-by: Galantsev, Dmitrii <dmitrii.galantsev@amd.com>
 2024-03-07 04:35:30 -06:00
Bill(Shuzhou) Liu c489cb8f3f Add support for deferred RAS errors in API 
The API will support the deferred errors

Change-Id: I221a146f09fefde1fc31e5f746d0870e07c93561
 2024-03-04 22:46:44 -05:00
Oliveira, Daniel 475424525e fix: [rocm/amd_smi_lib] TestFrequenciesRead & TestPciReadWrite test cases failed 
Fixes asserts in unit tests, and 'pp_dpm_pcie' condition

Code changes related to the following:
  * rsmi_dev_pci_bandwidth_set()
  * Functional tests

Change-Id: Id5e6851393fa3b51bb8cad87daca1efaf500a7e0
Signed-off-by: Oliveira, Daniel <daniel.oliveira@amd.com>
 2024-02-22 03:40:50 -05:00
Oliveira, Daniel 78074d7d77 fix: [rocm/amd_smi_lib] amdsmi_get_gpu_activity gfx/memory activity does not update 
Checks and forces rereading gpu metrics unconditionally

Code changes related to the following:
  * Device::dev_log_gpu_metrics()
  * amdsmi_get_gpu_metrics_header_info()
    Removed unintentionally during work on 'header cleanup Remove non-unified headers'
  * Examples
  * Unit tests

Change-Id: I83710e173c0f7102d0b7f865c18474c979a95cd8
Signed-off-by: Oliveira, Daniel <daniel.oliveira@amd.com>
 2024-02-13 10:15:17 -06:00
Bill(Shuzhou) Liu 9683fc26ca UMC ECC count error 
Remove another place which will check the size of the file.

Change-Id: If681a84253a8594c9f6d44d84dbf5d04a5e35612
 2024-02-05 11:07:29 -06:00
Bill(Shuzhou) Liu b125f3ee8f UMC ECC count return not supported 
The current code assume err_count sysfs only have 2 lines, which is
changed for umc_err_count by adding extra line for defer errors.
The code is changed to relax such check.

Change-Id: I7129bfcf9ef47346bca3c5a9535bc8c16c670f88
 2024-01-30 11:12:40 -06:00
Charis Poag 6132074089 Merge rocm-smi/amd-staging into amd-dev 20240119 
Change-Id: Ie706473ff92a91b19e95d2d58f64904cad73a89a
Signed-off-by: Charis Poag <Charis.Poag@amd.com>
 2024-01-19 03:57:00 -05:00
Charis Poag 4575990ae7 GPU Usage/activity update 
CLI:
Every usage field is notated by "activity"
gfx_usage -> gfx_activity
umc_usage -> umc_activity
vcn_activities -> vcn_activity
jpeg_activities[AID#] -> jpeg_activity

Wrapper: fixed metric output, misalignment
with generator

update_wrapper.sh:
DOCKER_BUILDKIT to 0 (if unset)

API:
amdsmi_get_gpu_metrics_info:
1.3: Removed commenting out avg socket power

Signed-off-by: Charis Poag <Charis.Poag@amd.com>
Change-Id: Id3fcc20aef420c7b7a90ba22fa3bc643b2716333
 2024-01-15 23:34:08 -06:00
Charis Poag 5ff5af0b5a Fix GPU metric tests & cleanup test output 
- CLI: Added average_power to display if current_power is empty
    - CLI: fixed PCIe current_speed not displaying GT/s
    - ROCm API: 1.3 & 1.4
                -> commented out setting avg clocks to current clock value
(leave as max uint value, not re-assign; these are not same values)
                    -> commented out setting current_socket_power = average_power
(leave as max uint value, not re-assign; these are not same values)
                    -> For all non-array clocks, placed value in first
                        array[0] to keep outputs consistent
                    (helps xcd calc)
      - ROCm API: rsmi_dev_metrics_curr_gfxclk_get fixed to count
        XCDs using backwards compatible rsmi_dev_gpu_metrics_info_get.
      - ^ Fixes XCD count overall + assigning clock[0] in 1.3 to curr
        freq
      - AMD SMI API: amdsmi_get_gpu_metrics_info() initialized all new
        1.5 metric values for all lower metric tables
      - AMD SMI API: wrapper -> fix is here + returns correct AMD SMI return
      - AMD SMI API: wrapper -> now displays amdsmi return status as
        string in logs
      - gpu_metrics_read.cc -> now has better overview of backwards
        compatible output
      - gpu_metrics_read.cc -> Cleaned up output, added units, and
        display all array output

Signed-off-by: Charis Poag <Charis.Poag@amd.com>
Change-Id: Id5b60ded5b0ed2cdf0f96ca72c79e356f0410960
 2023-12-19 14:18:15 -05:00
Bill(Shuzhou) Liu 59b510de2b Support max_num_cu_shared and num_cache_instance 
Add above fields for cache info. Remove driver_date in CLI and
Remove the disable properties of cache.

Change-Id: I80672490908d9e32a149076cc37459fa56b8b0bf
 2023-12-14 09:59:35 -05:00
Maisam Arif b54086a037 Change xgmi_physical_id to oam_id 
Signed-off-by: Maisam Arif <maisarif@amd.com>
Change-Id: I35fb36ec0e9f72a7135d8bb9070dbdc0e956b93a
 2023-11-22 12:16:38 -06:00
Maisam Arif 02d310e525 Merge rocmsmi/amd-staging into amd-dev 20231121 
Signed-off-by: Maisam Arif <maisarif@amd.com>
Change-Id: I5cc6accced971479583954e0b93cd90c510ca814
Signed-off-by: Maisam Arif <maisarif@amd.com>
 2023-11-22 03:31:35 -06:00
Bill(Shuzhou) Liu ac1ba33371 Add APIs for PM table and register table 
Read the PM table and register table as the name value pair.

Change-Id: Ie44fe67a28af3341bd6beb90d809e90f280351ac
 2023-11-20 12:31:18 -05:00
Bill(Shuzhou) Liu 18d6484c67 Sort GPU index using BDF 
Sort GPU index based on BDF.

Change-Id: I921fc72e534665bd14b9a45a747f72b24f5ba641
 2023-11-08 08:58:15 -05:00
Galantsev, Dmitrii 9cab8c035f Merge rocmsmi/amd-staging into amd-dev 20231103-1 
Change-Id: I82540740892b31232dc9902022b34001ef5b542c
Signed-off-by: Galantsev, Dmitrii <dmitrii.galantsev@amd.com>
 2023-11-03 17:58:30 -05:00
Galantsev, Dmitrii 513dd8a445 Merge rocmsmi/amd-staging into amd-dev 20231103 
Change-Id: Ie70ab54a63b25649b6b9d30620c5546dc66cd766
Signed-off-by: Galantsev, Dmitrii <dmitrii.galantsev@amd.com>
 2023-11-03 02:55:02 -05:00
Bill(Shuzhou) Liu 56b246cc3c Support cache type in cache info 
Add the cache type to the cache info.

Change-Id: Ic13ca9640b65d2b414eeebe7b884530f2036aac8
 2023-11-02 04:53:38 -05:00
Maisam Arif 2b4637ff9f SWDEV-410051 - Updates to board_info struct & CLI 
Signed-off-by: Maisam Arif <maisarif@amd.com>
Change-Id: I8735d8965140ee5da0c35106b388af1dca87ec71
 2023-10-27 16:52:56 -05:00
Galantsev, Dmitrii df4f5e8bf8 Merge rocmsmi/amd-staging into amd-dev 20231016 
Change-Id: I137171162a64af4960d82336cc517c1b34a870f3
Signed-off-by: Galantsev, Dmitrii <dmitrii.galantsev@amd.com>
 2023-10-16 14:31:13 -05:00
Bill(Shuzhou) Liu d92d4e4b38 Add new API for RAS related information 
The API to get the EEPROM version and ECC schema.

Change-Id: Iee6b3c555541a33bf16bf9ac1fd60100dfff5643
 2023-10-13 02:06:14 -04:00
Galantsev, Dmitrii 6d72d65c48 Merge rocmsmi/amd-staging into amd-dev 20231010 
Change-Id: I492562094a004eb78b2cc2b52d14d013d9f97112
Signed-off-by: Galantsev, Dmitrii <dmitrii.galantsev@amd.com>
 2023-10-11 18:58:12 -05:00
Bill(Shuzhou) Liu 6ca95c1a2d Add support to XGMI physical id 
Get XGMI physical id from sysfs.

Change-Id: Ifd9e431bc2fbfd759d888a71b99046a5eb07b6ed
 2023-10-10 09:29:05 -04:00
Bill(Shuzhou) Liu 1a233f93fb APIs for the cache level and size 
Read the cache level and size from topoogy sysfs file.

Change-Id: Id3c558c95bcb79139a19e4adbaa7ff333d06098f
 2023-10-05 11:10:54 -05:00
Bill(Shuzhou) Liu 656f12e0f3 Read PCIe slot type from sysfs 
Read the PCIe slot type from sysfs instead of libdrm.

Change-Id: I9392b9e18a209ac7332f6902bcafb3b6062c86c1
 2023-10-05 09:58:36 -04:00
Galantsev, Dmitrii aa1dd8391e Merge remote-tracking branch 'rocmsmi/amd-staging' into HEAD 
Change-Id: I18c19dc54eedc9a9c851e5d8d0fdfa7b632e23a5
 2023-09-26 14:03:23 -05:00
Galantsev, Dmitrii 31cc2eecfb Merge remote-tracking branch 'rocmsmi/amd-staging' into HEAD 
Change-Id: I0661926c10eef2bc32b83d9a63a3a6eb6991e781
 2023-09-25 04:35:53 -05:00
Galantsev, Dmitrii 5c41319c83 Merge remote-tracking branch 'rocmsmi/amd-staging' into HEAD 
Change-Id: I65ed7f3a0d1b6e58bc8377932d7c39db21d1b422
 2023-09-21 23:43:20 -05:00
Bill(Shuzhou) Liu f86f62b3f7 Return the driver loading status 
When init the library, it could return status whether the driver is
loaded or not.

Change-Id: Id26b8058e32881ebe2514067a639a2a871d1f252
 2023-09-18 08:38:16 -05:00
Bill(Shuzhou) Liu 9021ef96dc Support PCIe vendor name 
Add the support for PCIe vendor name.

Change-Id: Ibc1d289a08731e4c5a14f992f3b0d31b51482396
 2023-08-28 16:46:43 -05:00
Galantsev, Dmitrii 14190c5a94 Merge remote-tracking branch 'rocmsmi/amd-staging' into amd-dev 
Change-Id: I7a35220a2283b92c5b4825ee99d6693401ef8e1e
 2023-08-28 16:01:19 -05:00
Galantsev, Dmitrii 936719eeb6 Merge remote-tracking branch 'rocmsmi/amd-staging' into amd-dev 
Change-Id: I9c38b4facd472b877d1ad133f3176a023c890955
Signed-off-by: Galantsev, Dmitrii <dmitrii.galantsev@amd.com>
 2023-08-23 16:04:15 -05:00
Charis Poag afa174c655 Merge 'rocm-smi/amd-staging' into 'amd-smi/amd-dev' 
Signed-off-by: Charis Poag <Charis.Poag@amd.com>
Change-Id: Id35c9807c45a965c968fb430e3ce4f3c7069c210
 2023-07-19 18:46:28 -05:00
Bill(Shuzhou) Liu 7b107258ff Handle key not exists errors for voltage sensor 
Check the key exists or not before access sensor map.

Change-Id: I0ac5d63ae981ec2f984b1d464e93508560a4d030
 2023-07-05 09:25:27 -04:00
Bill(Shuzhou) Liu 3b1e2f0817 amdsmi_init() fail when miss voltage sysfs 
Fix the error handling when voltage sysfs is missing.

Change-Id: I457289eb4fd45bb05af83a67289b056f7e2ad966
 2023-06-29 17:19:41 -04:00
Bill(Shuzhou) Liu 145868ad72 Return NOT_SUPPORTED if fail to write to sysfs 
Check the error and return NOT_SUPPORTED when write to sysfs.

Change-Id: Idaa91816197413643552e230a483b79500d44e08
 2023-06-28 09:05:51 -05:00
Bill(Shuzhou) Liu 62ce965409 Clean up the APIs 
Remove and rename APIs after review.

Change-Id: I5464f200eb605b366673f8abca95183c3837843b
 2023-05-30 16:08:54 -04:00
Bill(Shuzhou) Liu dc4ba12e00 Return NOT_SUPPORT for set function in VM guest 
Fix the unit tests which are fail in VM guest environment.

Change-Id: Id7c58887692bbdecba54f5d2d8463b292e19b4ad
 2023-05-11 10:42:55 -05:00
Bill(Shuzhou) Liu b39a56fc52 Crash if no hwmon sysfs 
Return NOT_SUPPORTED if no hwmon sysfs.

Change-Id: I8f4bea0d74619150aee7dfc21fe62d24018e2994
 2023-04-19 05:46:02 -04:00
Bill(Shuzhou) Liu ad3b7d9f0d amdsmitst crash when read ECC counter 
If the amdsmi is fail to read ECC counter from sysfs file, it
still assumes the data is read into vector, which causes crash.

Add the check to make sure the data is read into the vector.

Change-Id: Ia24331ee6e75408120248b0ed7bc58581a2f23ad
 2023-03-08 09:13:04 -05:00
Bill(Shuzhou) Liu 552a7403bc SWDEV-373189: build error with g++ v12.1.0 
Fix the g++ error: ‘memset’ was not declared in this scope

Change-Id: I6231f863801f84a5a8c46543c87499058f2ef381
 2022-12-13 08:33:12 -06:00
Bill(Shuzhou) Liu 8d347bb6c4 Return errors when set clock range 
Instead of using assert to abort the application, the fix will
return the error code if the input parameters is incorrect.

Change-Id: I00861ddf1198386fb322ea06232a7178fb5ef4bd
 2022-12-06 12:39:43 -06:00