AI/rocm-systems
2
0
Garpu 0
Kode Masalah Tarik Permintaan Actions Packages Projects Rilis Wiki Kegiatan

Change how cache manager handles child process trace cache for rocpd (#1033)

Telusuri Sumber 
* Change how cache manager handles child process trace cache

* Sampling and backtrace metrics to cache

* Apply cmake formatting

* Fix parsing of metadata json

* Code clean up

* Fix build nlohmann json from source

* Fix storage parsed finished callback

* Revert sampling for child process

* Change cache file name generating

* Fix thread start stop

* Fix process start end timestamp

* Applied suggestions from code review

* Try with late start of flushing task thread

* Change dockerfiles for ci

* Revert changes on github workflows

* Remove json_fwd.hpp include

* fix dump

* Build nlohmann/json by default

Signed-off-by: David Galiffi <David.Galiffi@amd.com>

* Update location of build artifacts for nlohmann/json

Signed-off-by: David Galiffi <David.Galiffi@amd.com>

* Revert use_output_suffix

* Remove unused logs

* Fix cache store inside counter due to structure change

* Remove decode tests from debian ci

* Fix issue where all databases have the same UUID (#1499)

Co-authored-by: Aleksandar Djordjevic <adjordje@amd.com>

* Removing the cpack and install steps to save space

* Revert "Remove decode tests from debian ci"

This reverts commit ddabf6dd142dcf438e6b8997b8abe86f2c868468.

* Revert "Removing the cpack and install steps to save space"

This reverts commit 973da3a1ba99d99d529af5269d30e177092f9bfa.

* Add prepare-runner job as dependency to clean up the space

* Fix formatting

* Free up even more space

* Remove verbose for workflows

* remove hw_counters from ext_data

* move space clean up inside container

* try to remove external folder to free up space

* Check space

* Refactor Cleanup to it's own step

---------

Signed-off-by: David Galiffi <David.Galiffi@amd.com>
Co-authored-by: David Galiffi <David.Galiffi@amd.com>
Co-authored-by: Aleksandar Djordjevic <aleksandar.djordjevic@amd.com>
Co-authored-by: Aleksandar Djordjevic <adjordje@amd.com>
This commit is contained in:
Milan Radosavljevic
2025-10-24 17:47:15 +02:00
committed by GitHub
orang tua 4f075902fc
melakukan 8806be162c
46 mengubah file dengan 1502 tambahan dan 1014 penghapusan
Download Patch File Download Diff File Expand all files Collapse all files
projects/rocprofiler-systems/source/lib/rocprof-sys/library/rocprofiler-sdk.cpp
+12 -30
Melihat File
@@ -29,7 +29,6 @@
#include "core/debug.hpp"
#include "core/gpu.hpp"
#include "core/perfetto.hpp"
#include "core/rocpd/json.hpp"
#include "core/state.hpp"
#include "core/trace_cache/buffer_storage.hpp"
#include "core/trace_cache/cache_manager.hpp"
@@ -65,6 +64,8 @@
#include <timemory/utility/demangle.hpp>
#include <timemory/utility/types.hpp>
#include <nlohmann/json.hpp>
#include <atomic>
#include <cctype>
#include <cstdint>
@@ -368,7 +369,8 @@ consume_args(Tp&&...)
auto
get_backtrace(std::optional<std::vector<tim::unwind::processed_entry>>& _bt_data)
{
auto backtrace = ::rocpd::json::create();
auto backtrace = nlohmann::json();
if(_bt_data && !_bt_data->empty())
{
const std::string _unk = "??";
@@ -384,9 +386,9 @@ get_backtrace(std::optional<std::vector<tim::unwind::processed_entry>>& _bt_data
(_linfo && _linfo.line > 0)
? join("", _linfo.line)
: ((itr.lineno == 0) ? std::string{ "?" } : join("", itr.lineno));
auto _entry = join("", demangle(*_func), " @ ",
join(':', ::basename(_loc->c_str()), _line));
backtrace->set(join("", "frame#", _bt_cnt++), _entry);
auto _entry = join("", demangle(*_func), " @ ",
join(':', ::basename(_loc->c_str()), _line));
backtrace[join("", "frame#", _bt_cnt++)] = _entry;
}
}
return backtrace;
@@ -410,26 +412,6 @@ get_parent_stack_id([[maybe_unused]] const CorrelationIdType& correlation_id)
#endif
}
auto
get_extdata(const rocprofiler_callback_tracing_record_t& record)
{
constexpr auto message_key = "message";
auto args = callback_arg_array_t{};
auto extdata = ::rocpd::json::create();
rocprofiler_iterate_callback_tracing_kind_operation_args(record, save_args, 2, &args);
for(auto [key, val] : args)
{
if(key == message_key)
{
extdata->set(key, val);
}
}
return extdata;
}
struct scope_destructor
{
/// \fn scope_destructor(FuncT&& _fini, InitT&& _init)
@@ -867,7 +849,7 @@ tool_tracing_callback_stop(
cache_category<CategoryT>();
cache_add_thread_info(record.thread_id);
std::string args_str = get_args_string(args);
cache_region(&record, _beg_ts, _end_ts, call_stack->to_string(), args_str,
cache_region(&record, _beg_ts, _end_ts, call_stack.dump(), args_str,
trait::name<CategoryT>::value);
}
}
@@ -952,7 +934,7 @@ ompt_cache_instant_event(
cache_category<category::rocm_ompt_api>();
cache_add_thread_info(record.thread_id);
cache_region(&record, _instant_ts, _instant_ts, call_stack->to_string(),
cache_region(&record, _instant_ts, _instant_ts, call_stack.dump(),
get_args_string(args), trait::name<category::rocm_ompt_api>::value);
}
@@ -966,7 +948,7 @@ ompt_cache_orphan_event(
cache_category<category::rocm_ompt_api>();
cache_add_thread_info(stored_data.record.thread_id);
cache_region(&stored_data.record, stored_data._beg_ts, stored_data._beg_ts,
call_stack->to_string(), get_args_string(stored_data.args),
call_stack.dump(), get_args_string(stored_data.args),
trait::name<category::rocm_ompt_api>::value);
}
@@ -1034,7 +1016,7 @@ ompt_pop_standard_callback(
auto call_stack = get_backtrace(_bt_data);
cache_category<category::rocm_ompt_api>();
cache_add_thread_info(record.thread_id);
cache_region(&record, stored_data._beg_ts, _end_ts, call_stack->to_string(),
cache_region(&record, stored_data._beg_ts, _end_ts, call_stack.dump(),
get_args_string(stored_data.args),
trait::name<category::rocm_ompt_api>::value);
}
@@ -1084,7 +1066,7 @@ ompt_pop_parallel_callback(
cache_category<category::rocm_ompt_api>();
cache_add_thread_info(record.thread_id);
cache_region(&record, stored_data._beg_ts, _end_ts, call_stack->to_string(),
cache_region(&record, stored_data._beg_ts, _end_ts, call_stack.dump(),
get_args_string(stored_data.args),
trait::name<category::rocm_ompt_api>::value);
}