AI/rocm-systems
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

[SWDEV-463406] Add volation_status current counter/accumulated values

Browse Source 
Changes:
  - amdsmi_violation_status_t now includes current accumulated/counter
   values
  - Tests/wrapper now include added values
  - Removed ASIC references in header for host/bm alignment
  - Fix violation_status->per_hbm_thrm /
    violation_status->active_hbm_thrm
    calculations.

Change-Id: Ic86a7cbad5198a41018f82f6b588b83158d9ba0b
Signed-off-by: Charis Poag <Charis.Poag@amd.com>
This commit is contained in:
Charis Poag
2024-10-02 15:50:41 -05:00
parent e402fe7f36
commit 5eff39915b
6 changed files with 71 additions and 31 deletions
Download Patch File Download Diff File Expand all files Collapse all files
py-interface/amdsmi_interface.py
+6
View File
@@ -2020,6 +2020,12 @@ def amdsmi_get_violation_status(
return {
"reference_timestamp": _validate_if_max_uint(violation_status.reference_timestamp, MaxUIntegerTypes.UINT64_T),
"violation_timestamp": _validate_if_max_uint(violation_status.violation_timestamp, MaxUIntegerTypes.UINT64_T),
"acc_counter": _validate_if_max_uint(violation_status.acc_counter, MaxUIntegerTypes.UINT64_T),
"acc_prochot_thrm": _validate_if_max_uint(violation_status.acc_prochot_thrm, MaxUIntegerTypes.UINT64_T),
"acc_ppt_pwr": _validate_if_max_uint(violation_status.acc_ppt_pwr, MaxUIntegerTypes.UINT64_T), #PVIOL
"acc_socket_thrm": _validate_if_max_uint(violation_status.acc_socket_thrm, MaxUIntegerTypes.UINT64_T), #TVIOL
"acc_vr_thrm": _validate_if_max_uint(violation_status.acc_vr_thrm, MaxUIntegerTypes.UINT64_T),
"acc_hbm_thrm": _validate_if_max_uint(violation_status.acc_hbm_thrm, MaxUIntegerTypes.UINT64_T),
"per_prochot_thrm": _validate_if_max_uint(violation_status.per_prochot_thrm, MaxUIntegerTypes.UINT64_T, isActivity=True),
"per_ppt_pwr": _validate_if_max_uint(violation_status.per_ppt_pwr, MaxUIntegerTypes.UINT64_T, isActivity=True), #PVIOL
"per_socket_thrm": _validate_if_max_uint(violation_status.per_socket_thrm, MaxUIntegerTypes.UINT64_T, isActivity=True), #TVIOL
py-interface/amdsmi_wrapper.py
+20 -14
View File
@@ -734,6 +734,12 @@ struct_amdsmi_violation_status_t._pack_ = 1 # source:False
struct_amdsmi_violation_status_t._fields_ = [
('reference_timestamp', ctypes.c_uint64),
('violation_timestamp', ctypes.c_uint64),
('acc_counter', ctypes.c_uint64),
('acc_prochot_thrm', ctypes.c_uint64),
('acc_ppt_pwr', ctypes.c_uint64),
('acc_socket_thrm', ctypes.c_uint64),
('acc_vr_thrm', ctypes.c_uint64),
('acc_hbm_thrm', ctypes.c_uint64),
('per_prochot_thrm', ctypes.c_uint64),
('per_ppt_pwr', ctypes.c_uint64),
('per_socket_thrm', ctypes.c_uint64),
@@ -745,7 +751,7 @@ struct_amdsmi_violation_status_t._fields_ = [
('active_vr_thrm', ctypes.c_ubyte),
('active_hbm_thrm', ctypes.c_ubyte),
('PADDING_0', ctypes.c_ubyte * 3),
('reserved', ctypes.c_uint64 * 24),
('reserved', ctypes.c_uint64 * 30),
]
amdsmi_violation_status_t = struct_amdsmi_violation_status_t
@@ -798,19 +804,6 @@ amdsmi_card_form_factor_t = ctypes.c_uint32 # enum
class struct_amdsmi_pcie_info_t(Structure):
pass
class struct_pcie_static_(Structure):
pass
struct_pcie_static_._pack_ = 1 # source:False
struct_pcie_static_._fields_ = [
('max_pcie_width', ctypes.c_uint16),
('PADDING_0', ctypes.c_ubyte * 2),
('max_pcie_speed', ctypes.c_uint32),
('pcie_interface_version', ctypes.c_uint32),
('slot_type', amdsmi_card_form_factor_t),
('reserved', ctypes.c_uint64 * 10),
]
class struct_pcie_metric_(Structure):
pass
@@ -831,6 +824,19 @@ struct_pcie_metric_._fields_ = [
('reserved', ctypes.c_uint64 * 12),
]
class struct_pcie_static_(Structure):
pass
struct_pcie_static_._pack_ = 1 # source:False
struct_pcie_static_._fields_ = [
('max_pcie_width', ctypes.c_uint16),
('PADDING_0', ctypes.c_ubyte * 2),
('max_pcie_speed', ctypes.c_uint32),
('pcie_interface_version', ctypes.c_uint32),
('slot_type', amdsmi_card_form_factor_t),
('reserved', ctypes.c_uint64 * 10),
]
struct_amdsmi_pcie_info_t._pack_ = 1 # source:False
struct_amdsmi_pcie_info_t._fields_ = [
('pcie_static', struct_pcie_static_),