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

HIP API Tracing (#357)

Telusuri Sumber 
* Update include/rocprofiler-sdk/hip*

- updates for intercept table

* Update lib/common/units.hpp

- clang-tidy fixes

* Add lib/rocprofiler-sdk/hip

- tracing implementation for the HIP intercept table

* Update source/lib/rocprofiler-sdk/CMakeLists.txt

- add_subdirectory(hip)

* Update source/lib/rocprofiler-sdk/hsa

- offset function in hsa_api_info<Idx>
- remove report_activity, set_callback
- Tweak HSA_API_TABLE_LOOKUP_DEFINITION

* Update lib/rocprofiler-sdk/hip

- rocprofiler::hip::copy_table
- stringize_impl print dereferenced pointers when possible

* Update lib/rocprofiler-sdk/hsa/utils.hpp

- stringize_impl print dereferenced pointers when possible

* Update lib/rocprofiler-sdk/tests/intercept_table.cpp

- remove failures for intercepting HIP API tables

* Update include/rocprofiler-sdk/fwd.h

- add ROCPROFILER_HIP_RUNTIME_LIBRARY (== ROCPROFILER_HIP_LIBRARY)
- add ROCPROFILER_HIP_COMPILER_LIBRARY

* Update lib/rocprofiler-sdk/buffer_tracing.cpp

- Support ROCPROFILER_BUFFER_TRACING_HIP_API in rocprofiler_query_buffer_tracing_kind_operation_name
- Support ROCPROFILER_BUFFER_TRACING_HIP_API in rocprofiler_iterate_buffer_tracing_kind_operations

* Update lib/rocprofiler-sdk/callback_tracing.cpp

- Support ROCPROFILER_CALLBACK_TRACING_HIP_API in rocprofiler_query_callback_tracing_kind_operation_name
- Support ROCPROFILER_CALLBACK_TRACING_HIP_API in rocprofiler_iterate_callback_tracing_kind_operations
- Support ROCPROFILER_CALLBACK_TRACING_HIP_API in rocprofiler_iterate_callback_tracing_kind_operation_args

* Update lib/rocprofiler-sdk/intercept_table.cpp

- support HipDispatchTable and HipCompilerDispatchTable

* Update lib/rocprofiler-sdk/internal_threading.cpp

- Support ROCPROFILER_HIP_COMPILER_LIBRARY

* Update lib/rocprofiler-sdk/registration.cpp

- Support "hip" and "hip_compiler" in rocprofiler_set_api_table
- Added some extra logging

* Update samples/api_{buffered,callback}_tracing

- Modifications to demonstrate HIP API tracing

* Update tests/kernel-tracing

- Modifications to handle/test HIP API tracing

* Separate HIP tracing from HIP compiler tracing

* Fix installation of include/rocprofiler-sdk/hip/*

- add compiler and table headers to install

* Fixes to HIP interception

- hip_api_trace.hpp was updated a bit
  - removed hipGetDeviceProperties (generic)
  - added hipGetDevicePropertiesR0600
  - added hipGetDevicePropertiesR0000
  - removed hipRegisterTracerCallback
  - reordered hipCreateChannelDesc, hipExtModuleLaunchKernel, hipHccModuleLaunchKernel
  - added hipDrvGraphAddMemsetNode
- static asserts in hsa_api_info ensuring ordering of pointers

* Update lib/rocprofiler-sdk/hip/hip.*

- use size_t instead of rocprofiler_hip_table_api_id_t as non-type template parameter (smaller binary)
- separated out population of callback_context_data and buffered_context_data into non-template function (significantly smaller binary)

* Update lib/rocprofiler-sdk/hsa/hsa.*

- separated out population of callback_context_data and buffered_context_data into non-template function (significantly smaller binary)

* Update test/kernel-tracing/validate.py

- does not expect any hip_api_traces until libamdhip.so actually starts using rocprofiler-register

* Update tests/tools/json-tool.cpp

- fix context associated with "HIP_API_CALLBACK"

* Update external/CMakeLists.txt

- move misc variables to top of CMakeLists.txt so they apply to all external subprojects
  - BUILD_TESTING (OFF)
  - BUILD_SHARED_LIBS (OFF)
  - BUILD_OBJECT_LIBS (OFF)
  - BUILD_STATIC_LIBS (ON)
  - CMAKE_POSITION_INDEPENDENT_CODE (ON)
  - CMAKE_VISIBILITY_INLINES_HIDDEN (ON)
  - CMAKE_CXX_VISIBILITY_PRESET (hidden)
- disable using libunwind in glog

* Update lib/rocprofiler-{sdk,sdk-tool}/CMakeLists.txt

- remove explicit setting of SKIP_BUILD_RPATH

* Update CMakeLists.txt

- set high-level CMAKE_BUILD_RPATH and CMAKE_INSTALL_RPATH_USE_LINK_PATH

* Update tests/CMakeLists.txt

- include(GNUInstallDirs)

* Update samples/CMakeLists.txt

- include(GNUInstallDirs)

* Update include/rocprofiler-sdk/hip/{compiler_api,api}_args.h

- remove extern "C" due to incompatibility b/t empty struct in C (size 0) vs. empty struct in C++ (size 1)

* Update lib/rocprofiler-sdk/hip/details/ostream.hpp

- clang-tidy fixes

* Update cmake/rocprofiler_linting.cmake

- add a feature for clang tidy exe

* Update lib/rocprofiler-sdk/hip/hip.cpp

- use recursion instead of fold expression due to clang-tidy errors (maximum nesting level exceeded)

* Update lib/rocprofiler-sdk/buffer_tracing.cpp

- fix merge

* Update lib/rocprofiler-sdk/callback_tracing.cpp

- fix merge

* Update bin/rocprofv3

- args for marker, HIP runtime, and HIP compiler tracing

* Update tests/apps/simple-transpose

- use roctx

* Update tests/rocprofv3/tracing

- validate marker API data

* Update lib/rocprofiler-sdk-tool

- support for HIP runtime, HIP compiler, marker API

* Update queue/queue_controller/registration/utility

- call hsa::queue_controller_fini() during finalization
- add a yield function to common/utility.hpp
  - implements a thread yield + sleep
- add a sync function to Queue class
- add a iterate_queues member function to QueueController
  - this is used to sync each queue during queue_controller_fini()

* Fix data races: queue/context/stable_vector

- stable_vector::emplace_back returns reference
- correlation id map uses stable_vector
- queue_info_session has explicit fields for queue id, hsa agent, rocp agent
- use hsa::get_table() in AsyncSignalHandler
- WriteInterceptor does not use TLS for context array

* Update lib/rocprofiler-sdk/hsa/hsa.*

- static object for API subtables
- accessors for API subtables
- google tests for HSA API subtables

* Update lib/rocprofiler-sdk/hsa/{queue,async_copy}.cpp

- use HSA subtable accessors

* Update rocprofiler_memcheck and CI workflow

- use GCC 13 instead of GCC 11 due to suspected false positives in thread sanitizer
  - GCC 13 uses libtsan.so.2

* Update CI workflow

* Update lib/rocprofiler-sdk/counters/{metrics,counters}

- fix possibly dangling reference to a temporary from gcc-13

* Update thread-sanitizer-suppr.txt

- Ignore data races originating in hsa-runtime library

* Update cmake/rocprofiler_memcheck.cmake

- Deduce the sanitizer library to preload by compiling an application and extracting the linked sanitizer library

* Update tests/rocprofv3/tracing/CMakeLists.txt

- add csv files to REQUIRED_FILES and ATTACH_ON_FAIL in validate test

* Update lib/common/container/record_header_buffer.hpp

- fix data race identified by gcc v13 and libtsan.so.2

* Update hip API id, args, and def

- remove hipDrvGraphAddMemsetNode (not part of ROCm 6.0

* Update lib/common/container/record_header_buffer.hpp

- fix deadlock in save/read/reset

* Update source/docs/CMakeLists.txt

- remove COMMAND_ERROR_IS_FATAL ANY to allow for printing of stdout/stderr

* Update lib/rocprofiler-sdk/hip/details/ostream.hpp

- remove overloads for HIP_MEMSET_NODE_PARAMS

* Update docs/CMakeLists.txt

- use find_program for shell instead of hardcoded /bin/bash
This commit is contained in:
Jonathan R. Madsen
2024-01-24 16:32:54 -06:00
committed by GitHub
orang tua f84c2f4386
melakukan c641749fe6
74 mengubah file dengan 10091 tambahan dan 729 penghapusan
Download Patch File Download Diff File Expand all files Collapse all files
source/lib/rocprofiler-sdk/hsa/queue.cpp
+36 -19
Melihat File
@@ -25,6 +25,7 @@
#include "lib/rocprofiler-sdk/buffer.hpp"
#include "lib/rocprofiler-sdk/context/context.hpp"
#include "lib/rocprofiler-sdk/hsa/code_object.hpp"
#include "lib/rocprofiler-sdk/hsa/hsa.hpp"
#include <glog/logging.h>
#include <hsa/hsa.h>
@@ -88,16 +89,16 @@ AsyncSignalHandler(hsa_signal_value_t /*signal_v*/, void* data)
if(!ctxs.empty())
{
// only do the following work if there are contexts that require this info
const auto* _rocp_agent = queue_info_session.queue.get_agent().get_rocp_agent();
auto _hsa_agent = queue_info_session.queue.get_agent().get_hsa_agent();
auto _queue_id = queue_info_session.queue.get_id();
const auto* _rocp_agent = queue_info_session.rocp_agent;
auto _hsa_agent = queue_info_session.hsa_agent;
auto _queue_id = queue_info_session.queue_id;
auto _signal = queue_info_session.interrupt_signal;
auto _kern_id = queue_info_session.kernel_id;
const auto& _extern_corr_ids = queue_info_session.extern_corr_ids;
auto dispatch_time = hsa_amd_profiling_dispatch_time_t{};
auto dispatch_time_status =
queue_info_session.queue.ext_api().hsa_amd_profiling_get_dispatch_time_fn(
hsa::get_amd_ext_table()->hsa_amd_profiling_get_dispatch_time_fn(
_hsa_agent, _signal, &dispatch_time);
// if we encounter this in CI, it will cause test to fail
@@ -185,17 +186,14 @@ AsyncSignalHandler(hsa_signal_value_t /*signal_v*/, void* data)
// Delete signals and packets, signal we have completed.
if(queue_info_session.interrupt_signal.handle != 0u)
{
queue_info_session.queue.core_api().hsa_signal_destroy_fn(
queue_info_session.interrupt_signal);
hsa::get_core_table()->hsa_signal_destroy_fn(queue_info_session.interrupt_signal);
}
if(queue_info_session.kernel_pkt.ext_amd_aql_pm4.completion_signal.handle != 0u)
{
queue_info_session.queue.core_api().hsa_signal_destroy_fn(
hsa::get_core_table()->hsa_signal_destroy_fn(
queue_info_session.kernel_pkt.ext_amd_aql_pm4.completion_signal);
}
queue_info_session.queue.async_complete();
if(_corr_id)
{
LOG_IF(FATAL, _corr_id->ref_count.load() == 0)
@@ -204,7 +202,9 @@ AsyncSignalHandler(hsa_signal_value_t /*signal_v*/, void* data)
_corr_id->ref_count.fetch_sub(1);
}
queue_info_session.queue.async_complete();
delete static_cast<Queue::queue_info_session_t*>(data);
return false;
}
@@ -260,7 +260,7 @@ WriteInterceptor(const void* packets,
LOG_IF(FATAL, data == nullptr) << "WriteInterceptor was not passed a pointer to the queue";
static thread_local auto ctxs = context_array_t{};
auto ctxs = context_array_t{};
context::get_active_contexts(ctxs, context_filter);
auto& queue = *static_cast<Queue*>(data);
@@ -395,6 +395,9 @@ WriteInterceptor(const void* packets,
.interrupt_signal = interrupt_signal,
.tid = thr_id,
.kernel_id = kernel_id,
.queue_id = queue.get_id(),
.hsa_agent = queue.get_agent().get_hsa_agent(),
.rocp_agent = queue.get_agent().get_rocp_agent(),
.correlation_id = corr_id,
.kernel_pkt = kernel_pkt,
.contexts = ctxs,
@@ -405,15 +408,7 @@ WriteInterceptor(const void* packets,
}
} // namespace
Queue::~Queue()
{
// Potentially replace with condition variable at some point
// but performance may not matter here.
while(_active_async_packets.load(std::memory_order_relaxed) > 0)
{
std::this_thread::sleep_for(std::chrono::milliseconds{1});
}
}
Queue::~Queue() { sync(); }
void
Queue::signal_async_handler(const hsa_signal_t& signal, Queue::queue_info_session_t* data) const
@@ -470,6 +465,28 @@ Queue::Queue(const AgentCache& agent,
*queue = _intercept_queue;
}
void
Queue::sync() const
{
// Potentially replace with condition variable at some point
// but performance may not matter here.
constexpr auto max_wait_time = std::chrono::milliseconds{1000};
constexpr auto query_interval = std::chrono::milliseconds{10};
auto _orig_active = _active_async_packets.load(std::memory_order_relaxed);
auto _curr_active = _orig_active;
auto inactive = common::yield(
[this, &_curr_active]() {
return ((_curr_active = _active_async_packets.load(std::memory_order_relaxed)) == 0);
},
max_wait_time,
query_interval);
LOG_IF(WARNING, !inactive)
<< "rocprofiler-sdk Queue (instance=" << this << ") abandoned waiting for " << _orig_active
<< " async completion callbacks after " << max_wait_time.count() << " msecs. There were "
<< _curr_active << " async completion callbacks which were not delivered at that time.";
}
void
Queue::register_callback(ClientID id, queue_cb_t enqueue_cb, completed_cb_t complete_cb)
{