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

[rocprofiler-systems] Update logging to use spdlog library (#2428)

Browse Source 
## Motivation

- Structured logging with proper log levels (TRACE, DEBUG, INFO, WARNING, ERROR, CRITICAL)
- Better performance through compile-time formatting
- Consistent formatting using fmt library
- Runtime log level control via arguments and environment variables
- Easier maintenance and debugging capabilities

## Technical Details

- Added spdlog as a submodule and integrated it into CMake build system
- Created new `rocprofiler-systems-logger` library wrapping spdlog functionality
- Replaced custom logging macros (`ROCPROFSYS_VERBOSE`, `ROCPROFSYS_DEBUG`, `ROCPROFSYS_FATAL`, `ROCPROFSYS_REQUIRE`, `ROCPROFSYS_CI_THROW`, etc.) with spdlog equivalents (`LOG_DEBUG`, `LOG_WARNING`, `LOG_CRITICAL`, etc.)
- Implemented log level control through command-line arguments and environment variables
- Converted assertion macros to proper error handling with exceptions and std::abort()
This commit is contained in:
Milan Radosavljevic
2026-01-14 21:27:51 +01:00
committed by GitHub
parent 499127c0b9
commit 318d13870f
111 changed files with 2657 additions and 2602 deletions
Download Patch File Download Diff File Expand all files Collapse all files
projects/rocprofiler-systems/source/lib/core/gpu.cpp
+15 -16
View File
@@ -35,7 +35,6 @@
# define ROCPROFSYS_USE_ROCM 0
#endif
#include "debug.hpp"
#include "defines.hpp"
#include "gpu.hpp"
@@ -53,6 +52,8 @@
# include <rocprofiler-sdk/fwd.h>
#endif
#include "logger/debug.hpp"
namespace rocprofsys
{
namespace gpu
@@ -70,12 +71,14 @@ check_amdsmi_error(amdsmi_status_t _code, const char* _file, int _line)
const char* _msg = nullptr;
auto _err = amdsmi_status_code_to_string(_code, &_msg);
if(_err != AMDSMI_STATUS_SUCCESS)
ROCPROFSYS_THROW(
{
throw std::runtime_error(fmt::format(
"amdsmi_status_code_to_string failed. No error message available. "
"Error code %i originated at %s:%i\n",
static_cast<int>(_code), _file, _line);
ROCPROFSYS_THROW("[%s:%i] Error code %i :: %s", _file, _line, static_cast<int>(_code),
_msg);
"Error code {} originated at {}:{}",
static_cast<int>(_code), _file, _line));
}
throw std::runtime_error(fmt::format("[{}:{}] Error code {} :: {}", _file, _line,
static_cast<int>(_code), _msg));
}
// Ensures initialization happens only once
@@ -114,8 +117,7 @@ amdsmi_init()
prevent_amdsmi_library_unload();
} catch(std::exception& _e)
{
ROCPROFSYS_BASIC_VERBOSE(1, "Exception thrown initializing amd-smi: %s\n",
_e.what());
LOG_ERROR("Exception thrown initializing amd-smi: {}", _e.what());
_amdsmi_is_initialized() = false; // Mark as not initialized
return false;
}
@@ -165,9 +167,8 @@ query_rocm_agents()
sizeof(rocprofiler_agent_v0_t), nullptr);
} catch(std::exception& _e)
{
ROCPROFSYS_BASIC_VERBOSE(
1, "Exception thrown getting the rocm agents: %s. _dev_cnt=%ld\n", _e.what(),
_dev_cnt);
LOG_ERROR("Exception thrown getting the rocm agents: {}. _dev_cnt={}", _e.what(),
_dev_cnt);
}
_dev_cnt = get_agent_manager_instance().get_gpu_agents_count();
#endif
@@ -230,8 +231,7 @@ add_device_metadata(ArchiveT& ar)
sizeof(rocprofiler_agent_v0_t), &_agents_vec);
} catch(std::exception& _e)
{
ROCPROFSYS_BASIC_VERBOSE(1, "Exception thrown getting the rocm agents: %s.\n",
_e.what());
LOG_ERROR("Exception thrown getting the rocm agents: {}", _e.what());
}
ar(make_nvp("rocm_agents", _agents_vec));
@@ -251,7 +251,7 @@ add_device_metadata()
add_device_metadata(ar);
} catch(std::runtime_error& _e)
{
ROCPROFSYS_VERBOSE(2, "%s\n", _e.what());
LOG_ERROR("Exception thrown adding device metadata: {}", _e.what());
}
});
}
@@ -309,8 +309,7 @@ get_processor_handles()
ret = amdsmi_get_processor_type(processor, &processor_type);
if(processor_type != AMDSMI_PROCESSOR_TYPE_AMD_GPU)
{
ROCPROFSYS_THROW("Not AMD_GPU device type!");
return;
throw std::runtime_error("Not AMD_GPU device type!");
}
processors::processors_list.push_back(processor);