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

Merge amd-dev into amd-master 20240808

Browse Source 
Change-Id: I353b1f5219dd67d1d066fcc51d27677447726776
Signed-off-by: Galantsev, Dmitrii <dmitrii.galantsev@amd.com>
This commit is contained in:
Galantsev, Dmitrii
2024-08-08 16:41:50 -05:00
parent ce0c45ebca 40112f5b17
commit d0ff27181b
11 changed files with 97 additions and 57 deletions
Download Patch File Download Diff File Expand all files Collapse all files
CHANGELOG.md
+20
View File
@@ -4,6 +4,26 @@ Full documentation for amd_smi_lib is available at [https://rocm.docs.amd.com/pr
***All information listed below is for reference and subject to change.***
## amd_smi_lib for ROCm 6.2.1
### Additions
- N/A
### Optimizations
- N/A
### Fixes
- **Fixed TypeError in `amd-smi process -G`**.
- **Updated CLI error strings to handle empty and invalid GPU/CPU inputs**.
### Known Issues
- N/A
## amd_smi_lib for ROCm 6.2.0
### Additions
CMakeLists.txt
+4 -1
View File
@@ -28,7 +28,7 @@ find_program(GIT NAMES git)
## Setup the package version based on git tags.
set(PKG_VERSION_GIT_TAG_PREFIX "amdsmi_pkg_ver")
get_package_version_number("24.6.2" ${PKG_VERSION_GIT_TAG_PREFIX} GIT)
get_package_version_number("24.6.3" ${PKG_VERSION_GIT_TAG_PREFIX} GIT)
message("Package version: ${PKG_VERSION_STR}")
set(${AMD_SMI_LIBS_TARGET}_VERSION_MAJOR "${CPACK_PACKAGE_VERSION_MAJOR}")
set(${AMD_SMI_LIBS_TARGET}_VERSION_MINOR "${CPACK_PACKAGE_VERSION_MINOR}")
@@ -286,6 +286,9 @@ set(CPACK_RPM_ASAN_PACKAGE_REQUIRES ${CPACK_RPM_PACKAGE_REQUIRES})
# don't terminate if bytecompile of python files fails
set(CPACK_RPM_SPEC_MORE_DEFINE "%define _python_bytecompile_errors_terminate_build 0")
# Cpack converts !/usr/bin/env python3 to /usr/libexec/platform-python in RHEL8.
# prevent the BRP(buildroot policy) script from checking and modifying interpreter directives
string( APPEND CPACK_RPM_SPEC_MORE_DEFINE "\n%undefine __brp_mangle_shebangs")
# Add rocm-core dependency if -DROCM_DEP_ROCMCORE=ON is passed
if(ROCM_DEP_ROCMCORE)
amdsmi_cli/README.md
+1 -1
View File
@@ -79,7 +79,7 @@ AMD-SMI reports the version and current platform detected when running the comma
~$ amd-smi
usage: amd-smi [-h] ...
AMD System Management Interface | Version: 24.6.2.0 | ROCm version: 6.2.0 | Platform: Linux Baremetal
AMD System Management Interface | Version: 24.6.3.0 | ROCm version: 6.2.1 | Platform: Linux Baremetal
options:
-h, --help show this help message and exit
amdsmi_cli/amdsmi_cli.py
+32 -8
View File
@@ -24,19 +24,43 @@
import logging
import sys
import os
try:
import argcomplete
except ImportError:
logging.debug("argcomplete module not found. Autocomplete will not work.")
from amdsmi_commands import AMDSMICommands
from amdsmi_parser import AMDSMIParser
from amdsmi_logger import AMDSMILogger
import amdsmi_cli_exceptions
from amdsmi import amdsmi_interface
from amdsmi import amdsmi_exception
from typing import TYPE_CHECKING
# only used for type checking
# pyright trips up and cannot find amdsmi scripts without it
if TYPE_CHECKING:
from amdsmi_commands import AMDSMICommands
from amdsmi_parser import AMDSMIParser
from amdsmi_logger import AMDSMILogger
import amdsmi_cli_exceptions
from amdsmi import amdsmi_interface
from amdsmi import amdsmi_exception
try:
from amdsmi_commands import AMDSMICommands
from amdsmi_parser import AMDSMIParser
from amdsmi_logger import AMDSMILogger
import amdsmi_cli_exceptions
from amdsmi import amdsmi_interface
from amdsmi import amdsmi_exception
except ImportError:
current_path = os.path.dirname(os.path.abspath(__file__))
additional_path = f"{current_path}/../libexec/amdsmi_cli"
sys.path.append(additional_path)
try:
from amdsmi_commands import AMDSMICommands
from amdsmi_parser import AMDSMIParser
from amdsmi_logger import AMDSMILogger
import amdsmi_cli_exceptions
from amdsmi import amdsmi_interface
from amdsmi import amdsmi_exception
except ImportError:
print(f"Still couldn't import 'amdsmi related scripts'. Make sure it's installed in {additional_path}")
sys.exit(1)
def _print_error(e, destination):
if destination in ['stdout', 'json', 'csv']:
amdsmi_cli/amdsmi_commands.py
+19 -18
View File
@@ -422,11 +422,11 @@ class AMDSMICommands():
power_limit_error = False
power_cap_info = amdsmi_interface.amdsmi_get_power_cap_info(args.gpu)
max_power_limit = power_cap_info['max_power_cap']
max_power_limit = AMDSMIHelpers.convert_SI_unit(max_power_limit, AMDSMIHelpers.SI_Unit.MICRO)
max_power_limit = self.helpers.convert_SI_unit(max_power_limit, AMDSMIHelpers.SI_Unit.MICRO)
min_power_limit = power_cap_info['min_power_cap']
min_power_limit = AMDSMIHelpers.convert_SI_unit(min_power_limit, AMDSMIHelpers.SI_Unit.MICRO)
min_power_limit = self.helpers.convert_SI_unit(min_power_limit, AMDSMIHelpers.SI_Unit.MICRO)
socket_power_limit = power_cap_info['power_cap']
socket_power_limit = AMDSMIHelpers.convert_SI_unit(socket_power_limit, AMDSMIHelpers.SI_Unit.MICRO)
socket_power_limit = self.helpers.convert_SI_unit(socket_power_limit, AMDSMIHelpers.SI_Unit.MICRO)
except amdsmi_exception.AmdSmiLibraryException as e:
power_limit_error = True
max_power_limit = "N/A"
@@ -2679,14 +2679,15 @@ class AMDSMICommands():
# General and Engine to expose process_info values
if args.general or args.engine:
for process_info in filtered_process_values:
if args.general and args.engine:
del process_info['process_info']['memory_usage']
elif args.general:
del process_info['process_info']['memory_usage']
del process_info['process_info']['usage'] # Used in engine
elif args.engine:
del process_info['process_info']['memory_usage']
del process_info['process_info']['mem_usage'] # Used in general
if not process_info['process_info'] == "N/A":
if args.general and args.engine:
del process_info['process_info']['memory_usage']
elif args.general:
del process_info['process_info']['memory_usage']
del process_info['process_info']['usage'] # Used in engine
elif args.engine:
del process_info['process_info']['memory_usage']
del process_info['process_info']['mem_usage'] # Used in general
# Filter out non specified pids
if args.pid:
@@ -3529,11 +3530,11 @@ class AMDSMICommands():
power_cap_info = amdsmi_interface.amdsmi_get_power_cap_info(args.gpu)
logging.debug(f"Power cap info for gpu {gpu_id} | {power_cap_info}")
min_power_cap = power_cap_info["min_power_cap"]
min_power_cap = AMDSMIHelpers.convert_SI_unit(min_power_cap, AMDSMIHelpers.SI_Unit.MICRO)
min_power_cap = self.helpers.convert_SI_unit(min_power_cap, AMDSMIHelpers.SI_Unit.MICRO)
max_power_cap = power_cap_info["max_power_cap"]
max_power_cap = AMDSMIHelpers.convert_SI_unit(max_power_cap, AMDSMIHelpers.SI_Unit.MICRO)
max_power_cap = self.helpers.convert_SI_unit(max_power_cap, AMDSMIHelpers.SI_Unit.MICRO)
current_power_cap = power_cap_info["power_cap"]
current_power_cap = AMDSMIHelpers.convert_SI_unit(current_power_cap, AMDSMIHelpers.SI_Unit.MICRO)
current_power_cap = self.helpers.convert_SI_unit(current_power_cap, AMDSMIHelpers.SI_Unit.MICRO)
except amdsmi_exception.AmdSmiLibraryException as e:
raise ValueError(f"Unable to get power cap info from {gpu_string}") from e
@@ -3541,7 +3542,7 @@ class AMDSMICommands():
self.logger.store_output(args.gpu, 'powercap', f"Power cap is already set to {args.power_cap}")
elif args.power_cap >= min_power_cap and args.power_cap <= max_power_cap:
try:
new_power_cap = AMDSMIHelpers.convert_SI_unit(args.power_cap, AMDSMIHelpers.SI_Unit.BASE,
new_power_cap = self.helpers.convert_SI_unit(args.power_cap, AMDSMIHelpers.SI_Unit.BASE,
AMDSMIHelpers.SI_Unit.MICRO)
amdsmi_interface.amdsmi_set_power_cap(args.gpu, 0, new_power_cap)
except amdsmi_exception.AmdSmiLibraryException as e:
@@ -3932,9 +3933,9 @@ class AMDSMICommands():
power_cap_info = amdsmi_interface.amdsmi_get_power_cap_info(args.gpu)
logging.debug(f"Power cap info for gpu {gpu_id} | {power_cap_info}")
default_power_cap_in_w = power_cap_info["default_power_cap"]
default_power_cap_in_w = AMDSMIHelpers.convert_SI_unit(default_power_cap_in_w, AMDSMIHelpers.SI_Unit.MICRO)
default_power_cap_in_w = self.helpers.convert_SI_unit(default_power_cap_in_w, AMDSMIHelpers.SI_Unit.MICRO)
current_power_cap_in_w = power_cap_info["power_cap"]
current_power_cap_in_w = AMDSMIHelpers.convert_SI_unit(current_power_cap_in_w, AMDSMIHelpers.SI_Unit.MICRO)
current_power_cap_in_w = self.helpers.convert_SI_unit(current_power_cap_in_w, AMDSMIHelpers.SI_Unit.MICRO)
except amdsmi_exception.AmdSmiLibraryException as e:
raise ValueError(f"Unable to get power cap info from {gpu_id}") from e
@@ -3942,7 +3943,7 @@ class AMDSMICommands():
self.logger.store_output(args.gpu, 'powercap', f"Power cap is already set to {default_power_cap_in_w}")
else:
try:
default_power_cap_in_uw = AMDSMIHelpers.convert_SI_unit(default_power_cap_in_w,
default_power_cap_in_uw = self.helpers.convert_SI_unit(default_power_cap_in_w,
AMDSMIHelpers.SI_Unit.BASE,
AMDSMIHelpers.SI_Unit.MICRO)
amdsmi_interface.amdsmi_set_power_cap(args.gpu, 0, default_power_cap_in_uw)
amdsmi_cli/amdsmi_helpers.py
+13 -26
View File
@@ -28,9 +28,7 @@ import sys
import time
import re
from subprocess import run
from subprocess import PIPE, STDOUT
from typing import List
from typing import List, Union
from enum import Enum
from typing import Set
@@ -333,7 +331,7 @@ class AMDSMIHelpers():
(False, valid_gpu_format, str): Return False, whether the format of the GPU input is valid, and the first input that failed to be converted
"""
if 'all' in gpu_selections:
return (True, amdsmi_interface.amdsmi_get_processor_handles())
return True, True, amdsmi_interface.amdsmi_get_processor_handles()
if isinstance(gpu_selections, str):
gpu_selections = [gpu_selections]
@@ -389,7 +387,7 @@ class AMDSMIHelpers():
(False, str): Return False, and the first input that failed to be converted
"""
if 'all' in cpu_selections:
return (True, amdsmi_interface.amdsmi_get_cpusocket_handles())
return True, True, amdsmi_interface.amdsmi_get_cpusocket_handles()
if isinstance(cpu_selections, str):
cpu_selections = [cpu_selections]
@@ -430,7 +428,7 @@ class AMDSMIHelpers():
(False, str): Return False, and the first input that failed to be converted
"""
if 'all' in core_selections:
return (True, amdsmi_interface.amdsmi_get_cpucore_handles())
return True, True, amdsmi_interface.amdsmi_get_cpucore_handles()
if isinstance(core_selections, str):
core_selections = [core_selections]
@@ -786,35 +784,24 @@ class AMDSMIHelpers():
MICRO = 0.000001 # 10^-6
NANO = 0.000000001 # 10^-9
def convert_SI_unit(val: float, unit_in: SI_Unit, unit_out = SI_Unit.BASE) -> float:
def convert_SI_unit(self, val: Union[int, float], unit_in: SI_Unit, unit_out = SI_Unit.BASE) -> Union[int, float]:
"""This function will convert a value into another
scientific (SI) unit. Defaults unit_out to SI_Unit.BASE
This function returns a float.
params:
val: float unit to convert
val: int or float unit to convert
unit_in: Requires using SI_Unit to set current value's SI unit (eg. SI_Unit.MICRO)
unit_out - Requires using SI_Unit to set current value's SI unit
default value is SI_Unit.BASE (eg. SI_Unit.MICRO)
return:
float : converted SI unit of value requested
int or float : converted SI unit of value requested
"""
return val * unit_in / unit_out
def convert_SI_unit(val: int, unit_in: SI_Unit, unit_out=SI_Unit.BASE) -> int:
"""This function will convert a value into another
scientific (SI) unit. Defaults unit_out to SI_Unit.BASE
This function returns a int.
params:
val: int unit to convert
unit_in: Requires using SI_Unit to set current value's SI unit (eg. SI_Unit.MICRO)
unit_out - Requires using SI_Unit to set current value's SI unit
default value is SI_Unit.BASE (eg. SI_Unit.MICRO)
return:
int : converted SI unit of value requested
"""
return int(float(val) * unit_in / unit_out)
if isinstance(val, float):
return val * unit_in / unit_out
elif isinstance(val, int):
return int(float(val) * unit_in / unit_out)
else:
raise TypeError("val must be an int or float")
def get_pci_device_ids(self) -> Set[str]:
pci_devices_path = "/sys/bus/pci/devices"
amdsmi_cli/amdsmi_parser.py
+4
View File
@@ -171,6 +171,7 @@ class AMDSMIParser(argparse.ArgumentParser):
else:
raise amdsmi_cli_exceptions.AmdSmiInvalidParameterValueException(string_value, outputformat)
def _check_output_file_path(self):
""" Argument action validator:
Returns a path to a file from the output file path provided.
@@ -375,6 +376,9 @@ class AMDSMIParser(argparse.ArgumentParser):
def _validate_cpu_core(self, value):
if value == '':
outputformat = self.helpers.get_output_format()
raise amdsmi_cli_exceptions.AmdSmiMissingParameterValueException(value, outputformat)
if isinstance(value, str):
if value.lower() == "all":
return value
docs/doxygen/Doxyfile
+1 -1
View File
@@ -48,7 +48,7 @@ PROJECT_NAME = AMD SMI
# could be handy for archiving the generated documentation or if some version
# control system is used.
PROJECT_NUMBER = "24.6.2.0"
PROJECT_NUMBER = "24.6.3.0"
# Using the PROJECT_BRIEF tag one can provide an optional one line description
# for a project that appears at the top of each page and should give viewer a
docs/how-to/using-AMD-SMI-CLI-tool.md
+1 -1
View File
@@ -6,7 +6,7 @@ AMD-SMI reports the version and current platform detected when running the comma
~$ amd-smi
usage: amd-smi [-h] ...
AMD System Management Interface | Version: 24.6.2.0 | ROCm version: 6.2.0 | Platform: Linux Baremetal
AMD System Management Interface | Version: 24.6.3.0 | ROCm version: 6.2.1 | Platform: Linux Baremetal
options:
-h, --help show this help message and exit
include/amd_smi/amdsmi.h
+1 -1
View File
@@ -154,7 +154,7 @@ typedef enum {
#define AMDSMI_LIB_VERSION_MAJOR 6
//! Minor version should be updated for each API change, but without changing headers
#define AMDSMI_LIB_VERSION_MINOR 2
#define AMDSMI_LIB_VERSION_MINOR 3
//! Release version should be set to 0 as default and can be updated by the PMs for each CSP point release
#define AMDSMI_LIB_VERSION_RELEASE 0
py-interface/amdsmi_exception.py
+1
View File
@@ -107,6 +107,7 @@ class AmdSmiTimeoutException(AmdSmiLibraryException):
class AmdSmiParameterException(AmdSmiException):
def __init__(self, receivedValue, expectedType, msg=None):
super().__init__(msg)
self.err_code = None
self.actualType = type(receivedValue)
self.expectedType = expectedType
self.set_err_msg()