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

Merge amd-dev into amd-master 20240314

Browse Source 
Signed-off-by: guanyu12 <guanyu12@amd.com>
Change-Id: I120424cd46e7e5cc8ff52df089eec8a6e0ca3399
This commit is contained in:
guanyu12
2024-03-14 11:12:52 +08:00
parent 8e487fa52d b2690fdf1e
commit fc53ab5df5
21 changed files with 747 additions and 175 deletions
Download Patch File Download Diff File Expand all files Collapse all files
amdsmi_cli/README.md
+50 -1
View File
@@ -474,7 +474,7 @@ Command Modifiers:
```bash
usage: amd-smi set [-h] (-g GPU [GPU ...] | -U CPU [CPU ...] | -O CORE [CORE ...]) [-f %]
[-l LEVEL] [-P SETPROFILE] [-d SCLKMAX] [-C PARTITION] [-M PARTITION]
[-o WATTS] [--cpu-pwr-limit PWR_LIMIT]
[-o WATTS] [-p POLICY] [--cpu-pwr-limit PWR_LIMIT]
[--cpu-xgmi-link-width MIN_WIDTH MAX_WIDTH]
[--cpu-lclk-dpm-level NBIOID MIN_DPM MAX_DPM] [--cpu-pwr-eff-mode MODE]
[--cpu-gmi3-link-width MIN_LW MAX_LW] [--cpu-pcie-link-rate LINK_RATE]
@@ -512,6 +512,7 @@ Set Arguments:
-M, --memory-partition PARTITION Set one of the following the memory partition modes:
NPS1, NPS2, NPS4, NPS8
-o, --power-cap WATTS Set power capacity limit
-p, --dpm-policy POLICY_ID Set the GPU DPM policy using policy id
CPU Arguments:
--cpu-pwr-limit PWR_LIMIT Set power limit for the given socket. Input parameter is power limit value.
@@ -674,6 +675,18 @@ GPU: 0
PARTITION:
COMPUTE_PARTITION: SPX
MEMORY_PARTITION: NPS1
POLICY:
NUM_SUPPORTED: 4
CURRENT_ID: 1
POLICIES:
POLICY_ID: 0
POLICY_DESCRIPTION: pstate_default
POLICY_ID: 1
POLICY_DESCRIPTION: soc_pstate_0
POLICY_ID: 2
POLICY_DESCRIPTION: soc_pstate_1
POLICY_ID: 3
POLICY_DESCRIPTION: soc_pstate_2
NUMA:
NODE: 0
AFFINITY: 0
@@ -770,6 +783,18 @@ GPU: 1
PARTITION:
COMPUTE_PARTITION: SPX
MEMORY_PARTITION: NPS1
POLICY:
NUM_SUPPORTED: 4
CURRENT_ID: 1
POLICIES:
POLICY_ID: 0
POLICY_DESCRIPTION: pstate_default
POLICY_ID: 1
POLICY_DESCRIPTION: soc_pstate_0
POLICY_ID: 2
POLICY_DESCRIPTION: soc_pstate_1
POLICY_ID: 3
POLICY_DESCRIPTION: soc_pstate_2
NUMA:
NODE: 1
AFFINITY: 1
@@ -866,6 +891,18 @@ GPU: 2
PARTITION:
COMPUTE_PARTITION: SPX
MEMORY_PARTITION: NPS1
POLICY:
NUM_SUPPORTED: 4
CURRENT_ID: 1
POLICIES:
POLICY_ID: 0
POLICY_DESCRIPTION: pstate_default
POLICY_ID: 1
POLICY_DESCRIPTION: soc_pstate_0
POLICY_ID: 2
POLICY_DESCRIPTION: soc_pstate_1
POLICY_ID: 3
POLICY_DESCRIPTION: soc_pstate_2
NUMA:
NODE: 2
AFFINITY: 2
@@ -962,6 +999,18 @@ GPU: 3
PARTITION:
COMPUTE_PARTITION: SPX
MEMORY_PARTITION: NPS1
POLICY:
NUM_SUPPORTED: 4
CURRENT_ID: 1
POLICIES:
POLICY_ID: 0
POLICY_DESCRIPTION: pstate_default
POLICY_ID: 1
POLICY_DESCRIPTION: soc_pstate_0
POLICY_ID: 2
POLICY_DESCRIPTION: soc_pstate_1
POLICY_ID: 3
POLICY_DESCRIPTION: soc_pstate_2
NUMA:
NODE: 3
AFFINITY: 3
amdsmi_cli/amdsmi_commands.py
+58 -30
View File
@@ -244,7 +244,7 @@ class AMDSMICommands():
def static_gpu(self, args, multiple_devices=False, gpu=None, asic=None, bus=None, vbios=None,
limit=None, driver=None, ras=None, board=None, numa=None, vram=None,
cache=None, partition=None, dfc_ucode=None, fb_info=None, num_vf=None):
cache=None, partition=None, dfc_ucode=None, fb_info=None, num_vf=None, policy=None):
"""Get Static information for target gpu
Args:
@@ -267,7 +267,7 @@ class AMDSMICommands():
dfc_ucode (bool, optional): Value override for args.dfc_ucode. Defaults to None.
fb_info (bool, optional): Value override for args.fb_info. Defaults to None.
num_vf (bool, optional): Value override for args.num_vf. Defaults to None.
policy (bool, optional): Value override for args.policy. Defaults to None.
Returns:
None: Print output via AMDSMILogger to destination
"""
@@ -300,8 +300,10 @@ class AMDSMICommands():
args.partition = partition
if limit:
args.limit = limit
current_platform_args += ["ras", "limit", "partition"]
current_platform_values += [args.ras, args.limit, args.partition]
if policy:
args.policy = policy
current_platform_args += ["ras", "limit", "partition", "policy"]
current_platform_values += [args.ras, args.limit, args.partition, args.policy]
if self.helpers.is_linux() and not self.helpers.is_virtual_os():
if numa:
@@ -344,7 +346,13 @@ class AMDSMICommands():
static_dict['asic'] = "N/A"
logging.debug("Failed to get asic info for gpu %s | %s", gpu_id, e.get_error_info())
if args.bus:
bus_info = {}
bus_info = {
'bdf': "N/A",
'max_pcie_width': "N/A",
'max_pcie_speed': "N/A",
'pcie_interface_version': "N/A",
'slot_type': "N/A"
}
try:
bus_info['bdf'] = amdsmi_interface.amdsmi_get_gpu_device_bdf(args.gpu)
@@ -356,7 +364,6 @@ class AMDSMICommands():
link_caps = amdsmi_interface.amdsmi_get_pcie_info(args.gpu)
bus_info['max_pcie_width'] = link_caps['pcie_static']['max_pcie_width']
bus_info['max_pcie_speed'] = link_caps['pcie_static']['max_pcie_speed']
bus_info['pcie_slot_type'] = link_caps['pcie_static']['slot_type']
bus_info['pcie_interface_version'] = link_caps['pcie_static']['pcie_interface_version']
if bus_info['max_pcie_speed'] % 1000 != 0:
@@ -366,15 +373,13 @@ class AMDSMICommands():
bus_info['max_pcie_speed'] = pcie_speed_GTs_value
slot_type = bus_info.pop('pcie_slot_type')
slot_type = link_caps['pcie_static']['slot_type']
if isinstance(slot_type, int):
slot_types = amdsmi_interface.amdsmi_wrapper.amdsmi_card_form_factor_t__enumvalues
if slot_type in slot_types:
bus_info['slot_type'] = slot_types[slot_type].replace("AMDSMI_CARD_FORM_FACTOR_", "")
else:
bus_info['slot_type'] = "Unknown"
else:
bus_info['slot_type'] = "N/A"
if bus_info['pcie_interface_version'] > 0:
bus_info['pcie_interface_version'] = f"Gen {bus_info['pcie_interface_version']}"
@@ -389,7 +394,6 @@ class AMDSMICommands():
"unit" : pcie_speed_unit}
except amdsmi_exception.AmdSmiLibraryException as e:
bus_info = "N/A"
logging.debug("Failed to get bus info for gpu %s | %s", gpu_id, e.get_error_info())
static_dict['bus'] = bus_info
@@ -484,6 +488,7 @@ class AMDSMICommands():
shutdown_temp_vram_limit = "N/A"
logging.debug("Failed to get vram temperature shutdown metrics for gpu %s | %s", gpu_id, e.get_error_info())
# Assign units
power_unit = 'W'
temp_unit_human_readable = '\N{DEGREE SIGN}C'
@@ -624,6 +629,15 @@ class AMDSMICommands():
static_dict['partition'] = {"compute_partition": compute_partition,
"memory_partition": memory_partition}
if 'policy' in current_platform_args:
if args.policy:
try:
policy_info = amdsmi_interface.amdsmi_get_dpm_policy(args.gpu)
except amdsmi_exception.AmdSmiLibraryException as e:
policy_info = "N/A"
logging.debug("Failed to get policy info for gpu %s | %s", gpu_id, e.get_error_info())
static_dict['dpm_policy'] = policy_info
if 'numa' in current_platform_args:
if args.numa:
try:
@@ -760,7 +774,7 @@ class AMDSMICommands():
bus=None, vbios=None, limit=None, driver=None, ras=None,
board=None, numa=None, vram=None, cache=None, partition=None,
dfc_ucode=None, fb_info=None, num_vf=None, cpu=None,
interface_ver=None):
interface_ver=None, policy=None):
"""Get Static information for target gpu and cpu
Args:
@@ -783,7 +797,7 @@ class AMDSMICommands():
num_vf (bool, optional): Value override for args.num_vf. Defaults to None.
cpu (cpu_handle, optional): cpu_handle for target device. Defaults to None.
interface_ver (bool, optional): Value override for args.interface_ver. Defaults to None
policy (bool, optional): Value override for args.policy. Defaults to None.
Raises:
IndexError: Index error if gpu list is empty
@@ -809,7 +823,7 @@ class AMDSMICommands():
gpu_args_enabled = False
gpu_attributes = ["asic", "bus", "vbios", "limit", "driver", "ras",
"board", "numa", "vram", "cache", "partition",
"dfc_ucode", "fb_info", "num_vf"]
"dfc_ucode", "fb_info", "num_vf", "policy"]
for attr in gpu_attributes:
if hasattr(args, attr):
if getattr(args, attr):
@@ -839,7 +853,7 @@ class AMDSMICommands():
self.static_gpu(args, multiple_devices, gpu, asic,
bus, vbios, limit, driver, ras,
board, numa, vram, cache, partition,
dfc_ucode, fb_info, num_vf)
dfc_ucode, fb_info, num_vf, policy)
elif self.helpers.is_amd_hsmp_initialized(): # Only CPU is initialized
if args.cpu == None:
args.cpu = self.cpu_handles
@@ -853,7 +867,7 @@ class AMDSMICommands():
self.static_gpu(args, multiple_devices, gpu, asic,
bus, vbios, limit, driver, ras,
board, numa, vram, cache, partition,
dfc_ucode, fb_info, num_vf)
dfc_ucode, fb_info, num_vf, policy)
def firmware(self, args, multiple_devices=False, gpu=None, fw_list=True):
@@ -1124,14 +1138,14 @@ class AMDSMICommands():
args.temperature = temperature
if ecc:
args.ecc = ecc
if pcie:
args.pcie = pcie
current_platform_args += ["usage", "power", "clock", "temperature", "ecc", "pcie"]
current_platform_values += [args.usage, args.power, args.clock, args.temperature, args.ecc, args.pcie]
if self.helpers.is_baremetal() and self.helpers.is_linux():
if ecc_blocks:
args.ecc_blocks = ecc_blocks
if pcie:
args.pcie = pcie
current_platform_args += ["usage", "power", "clock", "temperature", "ecc", "ecc_blocks", "pcie"]
current_platform_values += [args.usage, args.power, args.clock, args.temperature, args.ecc, args.ecc_blocks, args.pcie]
if self.helpers.is_baremetal() and self.helpers.is_linux():
if fan:
args.fan = fan
if voltage_curve:
@@ -1144,8 +1158,8 @@ class AMDSMICommands():
args.xgmi_err = xgmi_err
if energy:
args.energy = energy
current_platform_args += ["ecc_blocks", "fan", "voltage_curve", "overdrive", "perf_level", "xgmi_err", "energy"]
current_platform_values += [args.ecc_blocks, args.fan, args.voltage_curve, args.overdrive, args.perf_level, args.xgmi_err, args.energy]
current_platform_args += ["fan", "voltage_curve", "overdrive", "perf_level", "xgmi_err", "energy"]
current_platform_values += [args.fan, args.voltage_curve, args.overdrive, args.perf_level, args.xgmi_err, args.energy]
if self.helpers.is_hypervisor():
if schedule:
@@ -3094,7 +3108,7 @@ class AMDSMICommands():
def set_gpu(self, args, multiple_devices=False, gpu=None, fan=None, perf_level=None,
profile=None, perf_determinism=None, compute_partition=None,
memory_partition=None, power_cap=None):
memory_partition=None, power_cap=None, dpm_policy=None):
"""Issue reset commands to target gpu(s)
Args:
@@ -3108,6 +3122,7 @@ class AMDSMICommands():
compute_partition (amdsmi_interface.AmdSmiComputePartitionType, optional): Value override for args.compute_partition. Defaults to None.
memory_partition (amdsmi_interface.AmdSmiMemoryPartitionType, optional): Value override for args.memory_partition. Defaults to None.
power_cap (int, optional): Value override for args.power_cap. Defaults to None.
dpm_policy (int, optional): Value override for args.dpm_policy. Defaults to None.
Raises:
ValueError: Value error if no gpu value is provided
@@ -3133,7 +3148,8 @@ class AMDSMICommands():
args.memory_partition = memory_partition
if power_cap:
args.power_cap = power_cap
if dpm_policy:
args.dpm_policy = dpm_policy
# Handle No GPU passed
if args.gpu == None:
raise ValueError('No GPU provided, specific GPU target(s) are needed')
@@ -3152,7 +3168,8 @@ class AMDSMICommands():
args.compute_partition,
args.memory_partition,
args.perf_determinism is not None,
args.power_cap]):
args.power_cap,
args.dpm_policy]):
command = " ".join(sys.argv[1:])
raise AmdSmiRequiredCommandException(command, self.logger.format)
@@ -3216,6 +3233,16 @@ class AMDSMICommands():
raise PermissionError('Command requires elevation') from e
raise ValueError(f"Unable to set memory partition to {args.memory_partition} on {gpu_string}") from e
self.logger.store_output(args.gpu, 'memorypartition', f"Successfully set memory partition to {args.memory_partition}")
if args.dpm_policy:
try:
amdsmi_interface.amdsmi_set_dpm_policy(args.gpu, args.dpm_policy)
except amdsmi_exception.AmdSmiLibraryException as e:
if e.get_error_code() == amdsmi_interface.amdsmi_wrapper.AMDSMI_STATUS_NO_PERM:
raise PermissionError('Command requires elevation') from e
raise ValueError(f"Unable to set dpm policy to {args.dpm_policy} on {gpu_string}") from e
self.logger.store_output(args.gpu, 'dpmpolicy', f"Successfully set dpm policy to id {args.dpm_policy}")
if isinstance(args.power_cap, int):
try:
power_cap_info = amdsmi_interface.amdsmi_get_power_cap_info(args.gpu)
@@ -3255,7 +3282,7 @@ class AMDSMICommands():
cpu=None, cpu_pwr_limit=None, cpu_xgmi_link_width=None, cpu_lclk_dpm_level=None,
cpu_pwr_eff_mode=None, cpu_gmi3_link_width=None, cpu_pcie_link_rate=None,
cpu_df_pstate_range=None, cpu_enable_apb=None, cpu_disable_apb=None,
soc_boost_limit=None, core=None, core_boost_limit=None):
soc_boost_limit=None, core=None, core_boost_limit=None, dpm_policy=None):
"""Issue reset commands to target gpu(s)
Args:
@@ -3284,6 +3311,7 @@ class AMDSMICommands():
core (device_handle, optional): device_handle for target core. Defaults to None.
core_boost_limit (int, optional): Value override for args.core_boost_limit. Defaults to None
dpm_policy (int, optional): Value override for args.dpm_policy. Defaults to None.
Raises:
ValueError: Value error if no gpu value is provided
@@ -3304,7 +3332,7 @@ class AMDSMICommands():
# Check if a GPU argument has been set
gpu_args_enabled = False
gpu_attributes = ["fan", "perf_level", "profile", "perf_determinism", "compute_partition",
"memory_partition", "power_cap"]
"memory_partition", "power_cap", "dpm_policy"]
for attr in gpu_attributes:
if hasattr(args, attr):
if getattr(args, attr) is not None:
@@ -3365,7 +3393,7 @@ class AMDSMICommands():
self.logger.clear_multiple_devices_ouput()
self.set_gpu(args, multiple_devices, gpu, fan, perf_level,
profile, perf_determinism, compute_partition,
memory_partition, power_cap)
memory_partition, power_cap, dpm_policy)
elif self.helpers.is_amd_hsmp_initialized(): # Only CPU is initialized
if args.cpu == None and args.core == None:
raise ValueError('No CPU or CORE provided, specific target(s) are needed')
@@ -3384,7 +3412,7 @@ class AMDSMICommands():
self.logger.clear_multiple_devices_ouput()
self.set_gpu(args, multiple_devices, gpu, fan, perf_level,
profile, perf_determinism, compute_partition,
memory_partition, power_cap)
memory_partition, power_cap, dpm_policy)
def reset(self, args, multiple_devices=False, gpu=None, gpureset=None,
amdsmi_cli/amdsmi_helpers.py
+5 -3
View File
@@ -617,8 +617,11 @@ class AMDSMIHelpers():
"""
# Get card vendor id
asic_info = amdsmi_interface.amdsmi_get_gpu_asic_info(device_handle)
return asic_info['vendor_id'] == AMD_VENDOR_ID
try:
vendor_value = int(asic_info['vendor_id'], 16)
return vendor_value == AMD_VENDOR_ID
except:
return False
def get_perf_levels(self):
perf_levels_str = [clock.name for clock in amdsmi_interface.AmdSmiDevPerfLevel]
@@ -632,7 +635,6 @@ class AMDSMIHelpers():
compute_partitions_str.remove('INVALID')
return compute_partitions_str
def get_memory_partition_types(self):
memory_partitions_str = [partition.name for partition in amdsmi_interface.AmdSmiMemoryPartitionType]
if 'UNKNOWN' in memory_partitions_str:
amdsmi_cli/amdsmi_parser.py
+5 -1
View File
@@ -543,6 +543,7 @@ class AMDSMIParser(argparse.ArgumentParser):
vram_help = "All vram information"
cache_help = "All cache information"
board_help = "All board information"
dpm_policy_help = "The available DPM policy"
# Options arguments help text for Hypervisors and Baremetal
ras_help = "Displays RAS features information"
@@ -582,6 +583,7 @@ class AMDSMIParser(argparse.ArgumentParser):
static_parser.add_argument('-r', '--ras', action='store_true', required=False, help=ras_help)
static_parser.add_argument('-p', '--partition', action='store_true', required=False, help=partition_help)
static_parser.add_argument('-l', '--limit', action='store_true', required=False, help=limit_help)
static_parser.add_argument('-P', '--policy', action='store_true', required=False, help=dpm_policy_help)
if self.helpers.is_linux() and not self.helpers.is_virtual_os():
static_parser.add_argument('-u', '--numa', action='store_true', required=False, help=numa_help)
@@ -759,10 +761,10 @@ class AMDSMIParser(argparse.ArgumentParser):
metric_parser.add_argument('-t', '--temperature', action='store_true', required=False, help=temperature_help)
metric_parser.add_argument('-P', '--pcie', action='store_true', required=False, help=pcie_help)
metric_parser.add_argument('-e', '--ecc', action='store_true', required=False, help=ecc_help)
metric_parser.add_argument('-k', '--ecc-blocks', action='store_true', required=False, help=ecc_blocks_help)
# Optional Args for Linux Baremetal Systems
if self.helpers.is_baremetal() and self.helpers.is_linux():
metric_parser.add_argument('-k', '--ecc-blocks', action='store_true', required=False, help=ecc_blocks_help)
metric_parser.add_argument('-f', '--fan', action='store_true', required=False, help=fan_help)
metric_parser.add_argument('-C', '--voltage-curve', action='store_true', required=False, help=vc_help)
metric_parser.add_argument('-o', '--overdrive', action='store_true', required=False, help=overdrive_help)
@@ -963,6 +965,7 @@ class AMDSMIParser(argparse.ArgumentParser):
set_compute_partition_help = f"Set one of the following the compute partition modes:\n\t{compute_partition_choices_str}"
set_memory_partition_help = f"Set one of the following the memory partition modes:\n\t{memory_partition_choices_str}"
set_power_cap_help = "Set power capacity limit"
set_dpm_policy_help = f"Set the GPU DPM policy using policy id\n"
# Help text for CPU set options
set_cpu_pwr_limit_help = "Set power limit for the given socket. Input parameter is power limit value."
@@ -998,6 +1001,7 @@ class AMDSMIParser(argparse.ArgumentParser):
set_value_parser.add_argument('-C', '--compute-partition', action='store', choices=self.helpers.get_compute_partition_types(), type=str.upper, required=False, help=set_compute_partition_help, metavar='PARTITION')
set_value_parser.add_argument('-M', '--memory-partition', action='store', choices=self.helpers.get_memory_partition_types(), type=str.upper, required=False, help=set_memory_partition_help, metavar='PARTITION')
set_value_parser.add_argument('-o', '--power-cap', action='store', type=self._positive_int, required=False, help=set_power_cap_help, metavar='WATTS')
set_value_parser.add_argument('-p', '--dpm-policy', action='store', required=False, type=self._not_negative_int, help=set_dpm_policy_help, metavar='POLICY_ID')
if self.helpers.is_amd_hsmp_initialized():
# Optional CPU Args