AI/rocm-systems
2
0
Forkuj 0
Kod Zgłoszenia Oczekujące zmiany Actions Packages Projekty Wydania Wiki Aktywność

Pull from Github

Przeglądaj źródła 
Squashed commit of the following:

commit f029195705a15700380c6f832ba5d15d46fd6de7
Author: Jonathan R. Madsen <jrmadsen@users.noreply.github.com>
Date:   Thu Jul 13 14:38:56 2023 -0500

    Formatting workflows for source (clang-format) and cmake (cmake-format) (#4)

    * Add .cmake-format.yaml file

    * Add formatting workflow

    * provide base input for creating PR

    * Update scheme for extracting branch name

    - disable running formatting on push to amd-staging branch

    * patch .cmake-format.yaml for find_package signature

    - apparently cmake-format doesn't format the full signature of find_package

    * run formatting (clang-format v11) (#7)

    Co-authored-by: jrmadsen <jrmadsen@users.noreply.github.com>

    * run cmake formatting (cmake-format) (#6)

    Co-authored-by: jrmadsen <jrmadsen@users.noreply.github.com>

    ---------

    Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>

commit bc4d135fdd8a1a9e51235f18a5d575fd2b3735e6
Author: Ammar ELWazir <aelwazir@amd.com>
Date:   Thu Jul 13 12:55:17 2023 -0500

    Removing Build cache for potential issues with auto-generated header files (#5)

    Change-Id: I9e2319f4335e2f88585ffa6fac2bd88a1c952e6e

commit ce86dea6a311d44d880fa684eb78f3329295e2a4
Author: Jonathan R. Madsen <jrmadsen@users.noreply.github.com>
Date:   Thu Jul 13 11:08:58 2023 -0500

    Fix decltype(<hsa-function>) function pointer usage (#3)

    - the following is done in several places:
        decltype(hsa_memory_allocate)* hsa_memory_allocate
    - above can cause compiler errors
    - replace decltype(<hsa-function>) with decltype(::<hsa-function>)
      - this ensures that the type within the decltype is recognized as the global scope HSA function, not the variable
    - in many places, the variable has a "_fn" suffix to prevent this issue but added '::' anyway for consistency

commit ac49fdd92a72e9c99394253a02da413a6c2e3b3a
Merge: a07946a 03a0855
Author: Ammar ELWazir <aelwazir@amd.com>
Date:   Wed Jul 12 11:36:24 2023 -0500

    Merge pull request #2 from ROCm-Developer-Tools/gerrit-amd-staging

    Pull from gerrit

commit 03a085588cffe863e8f466de67be1cfb205b675a
Merge: c26b32b a07946a
Author: Ammar ELWazir <aelwazir@amd.com>
Date:   Wed Jul 12 10:57:30 2023 -0500

    Merge branch 'amd-staging' into gerrit-amd-staging

commit a07946a5cd4c670c83c27ad1a076a9d4567ce6d7
Author: Ammar ELWazir <Ammar.ELWazir@amd.com>
Date:   Wed Jul 12 15:46:04 2023 +0000

    Enabling Cached Builds

commit 525e494a7f13941077a8fd4ad6840904db4d27d4
Author: Ammar ELWazir <Ammar.ELWazir@amd.com>
Date:   Wed Jul 12 04:53:54 2023 +0000

    Updating missed GPU Targets

commit 42c75862f628c9bee7cfb7dc04dff2619430efbc
Author: Ammar ELWazir <Ammar.ELWazir@amd.com>
Date:   Wed Jul 12 04:43:02 2023 +0000

    Adding V1 Testing

commit 9d72fd4aee85e4b0c12e717060d2730fa5b73be1
Author: Ammar ELWazir <Ammar.ELWazir@amd.com>
Date:   Wed Jul 12 03:34:31 2023 +0000

    Fixing Artifacts directory path

commit f4000cc558b3b2e4676f7994f7ce8c8e6f94518e
Author: Ammar ELWazir <Ammar.ELWazir@amd.com>
Date:   Wed Jul 12 03:27:26 2023 +0000

    Fixing CMake for test build job

commit 2ce8115d4c33948c3c8f957f545a95a04e1d6cd2
Author: Ammar ELWazir <Ammar.ELWazir@amd.com>
Date:   Wed Jul 12 03:16:18 2023 +0000

    Fixing Ubuntu CMake for ubuntu test build

commit 6d0ed439191be900748d0c025157f9d689a73ec7
Author: Ammar ELWazir <Ammar.ELWazir@amd.com>
Date:   Wed Jul 12 01:28:41 2023 +0000

    Removing Navi21

commit e349a7642e5ae5eb03ab9fcd0a0f74f09f78cab5
Author: Ammar ELWazir <Ammar.ELWazir@amd.com>
Date:   Wed Jul 12 01:14:14 2023 +0000

    Removing Navi21

commit fefd02fe68d2a4bca7ec2e381960ad004ee9fc5b
Author: Ammar ELWazir <Ammar.ELWazir@amd.com>
Date:   Wed Jul 12 00:42:48 2023 +0000

    Fixing CMake Job

commit 2ea46abf7bf92643efa8c549fa70346ffbd79d65
Author: Ammar ELWazir <Ammar.ELWazir@amd.com>
Date:   Wed Jul 12 00:35:13 2023 +0000

    Fixing CMake Job

commit d99d681ed1999c5fcf291dc678b11a77205fb0f3
Author: Ammar ELWazir <Ammar.ELWazir@amd.com>
Date:   Wed Jul 12 00:32:13 2023 +0000

    Fixing Pull Latest Dockers and CMake Jobs

commit dfc4498072d13b4a1df3a63047d34c682c3d9a29
Author: Ammar ELWazir <Ammar.ELWazir@amd.com>
Date:   Tue Jul 11 23:54:21 2023 +0000

    Fixing CMake job

commit 919efe04de707f7c702031be15c3e2c5f8442cbb
Author: Ammar ELWazir <Ammar.ELWazir@amd.com>
Date:   Tue Jul 11 23:52:13 2023 +0000

    Adding Pull Last dockers job

commit be1b1256e8b0e05308e8f7e7e69bee3acca55281
Author: Ammar ELWazir <aelwazir@amd.com>
Date:   Tue Jul 11 18:25:40 2023 -0500

    Update cmake.yml

commit 212299fa4355ae6ec18f9aaacbb79c51ea6c6f97
Author: Ammar ELWazir <aelwazir@amd.com>
Date:   Tue Jul 11 18:23:35 2023 -0500

    Update cmake.yml

commit 7c2c1327086a61466cc6cac39f70865c051a8bc7
Author: Ammar ELWazir <aelwazir@amd.com>
Date:   Tue Jul 11 18:18:53 2023 -0500

    Update cmake.yml

commit 191b5ce007e612e814c1d7a3afb4ad398f3852e1
Author: Ammar ELWazir <aelwazir@amd.com>
Date:   Tue Jul 11 16:03:22 2023 -0500

    Update cmake.yml

commit 8824113d95f3e13c7ce4d0af8e0d9d8f522a6c4a
Author: Ammar ELWazir <Ammar.ELWazir@amd.com>
Date:   Tue Jul 11 16:28:09 2023 +0000

    Fixing Pull from Gerrit job name

    Change-Id: I9e7ed9a27a13ca49d62c93bdadb30f0057e4d385

commit cc3d5e4b02ffb439e8cc2b3efa53527c376f9982
Author: Ammar ELWazir <Ammar.ELWazir@amd.com>
Date:   Tue Jul 11 16:21:43 2023 +0000

    Adding Staging sync job

    Change-Id: I0551f43878b0678ce4b3e74e27d62357cf95ad95

commit b9be2eee71380a2e6dd34d520e92d0c4209277a0
Author: Ammar ELWazir <Ammar.ELWazir@amd.com>
Date:   Tue Jul 11 15:57:11 2023 +0000

    Fixing build.sh

    Change-Id: Ia987b0244f0875370d5fe69907b3f5e9cea914de

commit 9eee33a95a1abd656a7ac5ca10a9f245e9825431
Author: Ammar ELWazir <aelwazir@amd.com>
Date:   Mon Jul 10 21:39:46 2023 -0500

    Update cmake.yml

commit 7093b85a78497140e8b52632ca2a002bdaeacd62
Author: Ammar ELWazir <aelwazir@amd.com>
Date:   Mon Jul 10 21:33:29 2023 -0500

    Update cmake.yml

commit f54697172c72a67740f9fdfa0c217b6ea6931576
Author: Ammar ELWazir <aelwazir@amd.com>
Date:   Mon Jul 10 21:01:26 2023 -0500

    Update cmake.yml

commit 1b6620e16f8940386b0f4f04e69e2410d21c0e26
Author: Ammar ELWazir <aelwazir@amd.com>
Date:   Mon Jul 10 20:21:02 2023 -0500

    Update cmake.yml

commit a94bec740c6b42c4b79c87bca20fa87b99bf060d
Author: Ammar ELWazir <aelwazir@amd.com>
Date:   Mon Jul 10 19:46:35 2023 -0500

    Update cmake.yml

commit 85d6b29d4375a69d575c18ece8542c50f2ddfcc3
Author: Ammar ELWazir <aelwazir@amd.com>
Date:   Mon Jul 10 19:34:39 2023 -0500

    Update cmake.yml

commit 8c004887cf1435f1a6214c3d2455299a8a27bd4c
Author: Ammar ELWazir <aelwazir@amd.com>
Date:   Mon Jul 10 19:31:17 2023 -0500

    Update cmake.yml

commit a14a9168e17d9348a53c6e9c9a47ba1edb4c4509
Author: Ammar ELWazir <aelwazir@amd.com>
Date:   Mon Jul 10 19:25:46 2023 -0500

    Update cmake.yml

commit 000f2f40b84e6a2f7d4becdbf5aed01436ca4c83
Author: Ammar ELWazir <aelwazir@amd.com>
Date:   Mon Jul 10 19:08:18 2023 -0500

    Update cmake.yml

commit a28a53d56731cad848fa9133d1c4dbaa8fc7afa7
Author: Ammar ELWazir <aelwazir@amd.com>
Date:   Mon Jul 10 19:03:39 2023 -0500

    Update cmake.yml

commit a6a2db01027f0b01fdfbb5997ddb772c7f51b649
Author: Ammar ELWazir <aelwazir@amd.com>
Date:   Mon Jul 10 18:21:53 2023 -0500

    Update cmake.yml

commit 118ef2a88b2d44e3207c31c343da3e5e5ec6f176
Author: Ammar ELWazir <aelwazir@amd.com>
Date:   Mon Jul 10 17:55:57 2023 -0500

    Update cmake.yml

commit 03c4c232396440cd0be6d2dd7baf4ceea1c2589d
Author: Ammar ELWazir <aelwazir@amd.com>
Date:   Mon Jul 10 17:48:49 2023 -0500

    Create cmake.yml

Change-Id: I77992f15694e77cbae49c56f9ff02f4f9079235d


[ROCm/rocprofiler commit: d4a33cf33a]
This commit is contained in:
Ammar ELWazir
2023-07-13 19:48:38 +00:00
zatwierdzone przez Ammar Elwazir
rodzic e599708211
commit 6eb06cf201
144 zmienionych plików z 154836 dodań i 166916 usunięć
Ściągnij plik aktualizacji Ściągnij plik porównania Expand all files Collapse all files
projects/rocprofiler/include/rocprofiler/rocprofiler.h
+165 -165
Wyświetl plik
@@ -164,12 +164,12 @@ typedef struct {
// Profiling feature type
typedef struct {
rocprofiler_feature_kind_t kind; // feature kind
rocprofiler_feature_kind_t kind; // feature kind
union {
const char* name; // feature name
const char* name; // feature name
struct {
const char* block; // counter block name
uint32_t event; // counter event id
const char* block; // counter block name
uint32_t event; // counter event id
} counter;
};
const rocprofiler_parameter_t* parameters; // feature parameters array
@@ -216,23 +216,25 @@ typedef struct {
} rocprofiler_properties_t;
// Create new profiling context
hsa_status_t rocprofiler_open(hsa_agent_t agent, // GPU handle
rocprofiler_feature_t* features, // [in] profiling features array
uint32_t feature_count, // profiling info count
rocprofiler_t** context, // [out] context object
uint32_t mode, // profiling mode mask
hsa_status_t rocprofiler_open(hsa_agent_t agent, // GPU handle
rocprofiler_feature_t* features, // [in] profiling features array
uint32_t feature_count, // profiling info count
rocprofiler_t** context, // [out] context object
uint32_t mode, // profiling mode mask
rocprofiler_properties_t* properties); // profiling properties
// Add feature to a features set
hsa_status_t rocprofiler_add_feature(const rocprofiler_feature_t* feature, // [in]
rocprofiler_feature_set_t* features_set); // [in/out] profiling features set
hsa_status_t rocprofiler_add_feature(
const rocprofiler_feature_t* feature, // [in]
rocprofiler_feature_set_t* features_set); // [in/out] profiling features set
// Create new profiling context
hsa_status_t rocprofiler_features_set_open(hsa_agent_t agent, // GPU handle
rocprofiler_feature_set_t* features_set, // [in] profiling features set
rocprofiler_t** context, // [out] context object
uint32_t mode, // profiling mode mask
rocprofiler_properties_t* properties); // profiling properties
hsa_status_t rocprofiler_features_set_open(
hsa_agent_t agent, // GPU handle
rocprofiler_feature_set_t* features_set, // [in] profiling features set
rocprofiler_t** context, // [out] context object
uint32_t mode, // profiling mode mask
rocprofiler_properties_t* properties); // profiling properties
// Delete profiling info
hsa_status_t rocprofiler_close(rocprofiler_t* context); // [in] profiling context
@@ -242,24 +244,24 @@ hsa_status_t rocprofiler_reset(rocprofiler_t* context, // [in] profiling contex
uint32_t group_index); // group index
// Return context agent
hsa_status_t rocprofiler_get_agent(rocprofiler_t* context, // [in] profiling context
hsa_agent_t* agent); // [out] GPU handle
hsa_status_t rocprofiler_get_agent(rocprofiler_t* context, // [in] profiling context
hsa_agent_t* agent); // [out] GPU handle
// Supported time value ID
typedef enum {
ROCPROFILER_TIME_ID_CLOCK_REALTIME = 0, // Linux realtime clock time
ROCPROFILER_TIME_ID_CLOCK_REALTIME_COARSE = 1, // Linux realtime-coarse clock time
ROCPROFILER_TIME_ID_CLOCK_MONOTONIC = 2, // Linux monotonic clock time
ROCPROFILER_TIME_ID_CLOCK_MONOTONIC_COARSE = 3, // Linux monotonic-coarse clock time
ROCPROFILER_TIME_ID_CLOCK_MONOTONIC_RAW = 4, // Linux monotonic-raw clock time
ROCPROFILER_TIME_ID_CLOCK_REALTIME = 0, // Linux realtime clock time
ROCPROFILER_TIME_ID_CLOCK_REALTIME_COARSE = 1, // Linux realtime-coarse clock time
ROCPROFILER_TIME_ID_CLOCK_MONOTONIC = 2, // Linux monotonic clock time
ROCPROFILER_TIME_ID_CLOCK_MONOTONIC_COARSE = 3, // Linux monotonic-coarse clock time
ROCPROFILER_TIME_ID_CLOCK_MONOTONIC_RAW = 4, // Linux monotonic-raw clock time
} rocprofiler_time_id_t;
// Return time value for a given time ID and profiling timestamp
hsa_status_t rocprofiler_get_time(
rocprofiler_time_id_t time_id, // identifier of the particular time to convert the timesatmp
uint64_t timestamp, // profiling timestamp
uint64_t* value_ns, // [out] returned time 'ns' value, ignored if NULL
uint64_t* error_ns); // [out] returned time error 'ns' value, ignored if NULL
rocprofiler_time_id_t time_id, // identifier of the particular time to convert the timesatmp
uint64_t timestamp, // profiling timestamp
uint64_t* value_ns, // [out] returned time 'ns' value, ignored if NULL
uint64_t* error_ns); // [out] returned time error 'ns' value, ignored if NULL
////////////////////////////////////////////////////////////////////////////////
// Queue callbacks
@@ -269,26 +271,26 @@ hsa_status_t rocprofiler_get_time(
// Dispatch record
typedef struct {
uint64_t dispatch; // dispatch timestamp, ns
uint64_t begin; // kernel begin timestamp, ns
uint64_t end; // kernel end timestamp, ns
uint64_t complete; // completion signal timestamp, ns
uint64_t dispatch; // dispatch timestamp, ns
uint64_t begin; // kernel begin timestamp, ns
uint64_t end; // kernel end timestamp, ns
uint64_t complete; // completion signal timestamp, ns
} rocprofiler_dispatch_record_t;
// Profiling callback data
typedef struct {
hsa_agent_t agent; // GPU agent handle
uint32_t agent_index; // GPU index (GPU Driver Node ID as reported in the sysfs topology)
const hsa_queue_t* queue; // HSA queue
uint64_t queue_index; // Index in the queue
uint32_t queue_id; // Queue id
hsa_signal_t completion_signal; // Completion signal
const hsa_kernel_dispatch_packet_t* packet; // HSA dispatch packet
const char* kernel_name; // Kernel name
uint64_t kernel_object; // Kernel object address
const amd_kernel_code_t* kernel_code; // Kernel code pointer
uint32_t thread_id; // Thread id
const rocprofiler_dispatch_record_t* record; // Dispatch record
hsa_agent_t agent; // GPU agent handle
uint32_t agent_index; // GPU index (GPU Driver Node ID as reported in the sysfs topology)
const hsa_queue_t* queue; // HSA queue
uint64_t queue_index; // Index in the queue
uint32_t queue_id; // Queue id
hsa_signal_t completion_signal; // Completion signal
const hsa_kernel_dispatch_packet_t* packet; // HSA dispatch packet
const char* kernel_name; // Kernel name
uint64_t kernel_object; // Kernel object address
const amd_kernel_code_t* kernel_code; // Kernel code pointer
uint32_t thread_id; // Thread id
const rocprofiler_dispatch_record_t* record; // Dispatch record
} rocprofiler_callback_data_t;
// Profiling callback type
@@ -299,15 +301,14 @@ typedef hsa_status_t (*rocprofiler_callback_t)(
// Queue callbacks
typedef struct {
rocprofiler_callback_t dispatch; // dispatch callback
hsa_status_t (*create)(hsa_queue_t* queue, void* data); // create callback
hsa_status_t (*destroy)(hsa_queue_t* queue, void* data); // destroy callback
rocprofiler_callback_t dispatch; // dispatch callback
hsa_status_t (*create)(hsa_queue_t* queue, void* data); // create callback
hsa_status_t (*destroy)(hsa_queue_t* queue, void* data); // destroy callback
} rocprofiler_queue_callbacks_t;
// Set queue callbacks
hsa_status_t rocprofiler_set_queue_callbacks(
rocprofiler_queue_callbacks_t callbacks, // callbacks
void* data); // [in/out] passed callbacks data
hsa_status_t rocprofiler_set_queue_callbacks(rocprofiler_queue_callbacks_t callbacks, // callbacks
void* data); // [in/out] passed callbacks data
// Remove queue callbacks
hsa_status_t rocprofiler_remove_queue_callbacks();
@@ -323,20 +324,20 @@ hsa_status_t rocprofiler_stop_queue_callbacks();
// contect.invocations' to collect all profiling data
// Start profiling
hsa_status_t rocprofiler_start(rocprofiler_t* context, // [in/out] profiling context
uint32_t group_index); // group index
hsa_status_t rocprofiler_start(rocprofiler_t* context, // [in/out] profiling context
uint32_t group_index); // group index
// Stop profiling
hsa_status_t rocprofiler_stop(rocprofiler_t* context, // [in/out] profiling context
uint32_t group_index); // group index
hsa_status_t rocprofiler_stop(rocprofiler_t* context, // [in/out] profiling context
uint32_t group_index); // group index
// Read profiling
hsa_status_t rocprofiler_read(rocprofiler_t* context, // [in/out] profiling context
uint32_t group_index); // group index
hsa_status_t rocprofiler_read(rocprofiler_t* context, // [in/out] profiling context
uint32_t group_index); // group index
// Read profiling data
hsa_status_t rocprofiler_get_data(rocprofiler_t* context, // [in/out] profiling context
uint32_t group_index); // group index
hsa_status_t rocprofiler_get_data(rocprofiler_t* context, // [in/out] profiling context
uint32_t group_index); // group index
// Get profiling groups count
hsa_status_t rocprofiler_group_count(const rocprofiler_t* context, // [in] profiling context
@@ -379,75 +380,76 @@ hsa_status_t rocprofiler_iterate_trace_data(
// Profiling info kind
typedef enum {
ROCPROFILER_INFO_KIND_METRIC = 0, // metric info
ROCPROFILER_INFO_KIND_METRIC_COUNT = 1, // metric features count, int32
ROCPROFILER_INFO_KIND_TRACE = 2, // trace info
ROCPROFILER_INFO_KIND_TRACE_COUNT = 3, // trace features count, int32
ROCPROFILER_INFO_KIND_TRACE_PARAMETER = 4, // trace parameter info
ROCPROFILER_INFO_KIND_TRACE_PARAMETER_COUNT = 5 // trace parameter count, int32
ROCPROFILER_INFO_KIND_METRIC = 0, // metric info
ROCPROFILER_INFO_KIND_METRIC_COUNT = 1, // metric features count, int32
ROCPROFILER_INFO_KIND_TRACE = 2, // trace info
ROCPROFILER_INFO_KIND_TRACE_COUNT = 3, // trace features count, int32
ROCPROFILER_INFO_KIND_TRACE_PARAMETER = 4, // trace parameter info
ROCPROFILER_INFO_KIND_TRACE_PARAMETER_COUNT = 5 // trace parameter count, int32
} rocprofiler_info_kind_t;
// Profiling info query
typedef union {
rocprofiler_info_kind_t info_kind; // queried profiling info kind
rocprofiler_info_kind_t info_kind; // queried profiling info kind
struct {
const char* trace_name; // queried info trace name
const char* trace_name; // queried info trace name
} trace_parameter;
} rocprofiler_info_query_t;
// Profiling info data
typedef struct {
uint32_t agent_index; // GPU HSA agent index (GPU Driver Node ID as reported in the sysfs topology)
rocprofiler_info_kind_t kind; // info data kind
uint32_t
agent_index; // GPU HSA agent index (GPU Driver Node ID as reported in the sysfs topology)
rocprofiler_info_kind_t kind; // info data kind
union {
struct {
const char* name; // metric name
uint32_t instances; // instances number
const char* expr; // metric expression, NULL for basic counters
const char* description; // metric description
const char* block_name; // block name
uint32_t block_counters; // number of block counters
const char* name; // metric name
uint32_t instances; // instances number
const char* expr; // metric expression, NULL for basic counters
const char* description; // metric description
const char* block_name; // block name
uint32_t block_counters; // number of block counters
} metric;
struct {
const char* name; // trace name
const char* description; // trace description
uint32_t parameter_count; // supported by the trace number parameters
const char* name; // trace name
const char* description; // trace description
uint32_t parameter_count; // supported by the trace number parameters
} trace;
struct {
uint32_t code; // parameter code
const char* trace_name; // trace name
const char* parameter_name; // parameter name
const char* description; // trace parameter description
uint32_t code; // parameter code
const char* trace_name; // trace name
const char* parameter_name; // parameter name
const char* description; // trace parameter description
} trace_parameter;
};
} rocprofiler_info_data_t;
// Return the info for a given info kind
hsa_status_t rocprofiler_get_info(
const hsa_agent_t* agent, // [in] GFXIP handle
rocprofiler_info_kind_t kind, // kind of iterated info
void *data); // [in/out] returned data
hsa_status_t rocprofiler_get_info(const hsa_agent_t* agent, // [in] GFXIP handle
rocprofiler_info_kind_t kind, // kind of iterated info
void* data); // [in/out] returned data
// Iterate over the info for a given info kind, and invoke an application-defined callback on every iteration
hsa_status_t rocprofiler_iterate_info(
const hsa_agent_t* agent, // [in] GFXIP handle
rocprofiler_info_kind_t kind, // kind of iterated info
hsa_status_t (*callback)(const rocprofiler_info_data_t info, void *data), // callback
void *data); // [in/out] data passed to callback
// Iterate over the info for a given info kind, and invoke an application-defined callback on every
// iteration
hsa_status_t rocprofiler_iterate_info(const hsa_agent_t* agent, // [in] GFXIP handle
rocprofiler_info_kind_t kind, // kind of iterated info
hsa_status_t (*callback)(const rocprofiler_info_data_t info,
void* data), // callback
void* data); // [in/out] data passed to callback
// Iterate over the info for a given info query, and invoke an application-defined callback on every iteration
hsa_status_t rocprofiler_query_info(
const hsa_agent_t *agent, // [in] GFXIP handle
rocprofiler_info_query_t query, // iterated info query
hsa_status_t (*callback)(const rocprofiler_info_data_t info, void *data), // callback
void *data); // [in/out] data passed to callback
// Iterate over the info for a given info query, and invoke an application-defined callback on every
// iteration
hsa_status_t rocprofiler_query_info(const hsa_agent_t* agent, // [in] GFXIP handle
rocprofiler_info_query_t query, // iterated info query
hsa_status_t (*callback)(const rocprofiler_info_data_t info,
void* data), // callback
void* data); // [in/out] data passed to callback
// Create a profiled queue. All dispatches on this queue will be profiled
hsa_status_t rocprofiler_queue_create_profiled(
hsa_agent_t agent_handle,uint32_t size, hsa_queue_type32_t type,
void (*callback)(hsa_status_t status, hsa_queue_t* source, void* data),
void* data, uint32_t private_segment_size, uint32_t group_segment_size,
hsa_queue_t** queue);
hsa_agent_t agent_handle, uint32_t size, hsa_queue_type32_t type,
void (*callback)(hsa_status_t status, hsa_queue_t* source, void* data), void* data,
uint32_t private_segment_size, uint32_t group_segment_size, hsa_queue_t** queue);
////////////////////////////////////////////////////////////////////////////////
// Profiling pool
@@ -461,8 +463,8 @@ typedef void rocprofiler_pool_t;
// Profiling pool entry
typedef struct {
rocprofiler_t* context; // context object
void* payload; // payload data object
rocprofiler_t* context; // context object
void* payload; // payload data object
} rocprofiler_pool_entry_t;
// Profiling handler, calling on profiling completion
@@ -478,120 +480,118 @@ typedef struct {
// Open profiling pool
hsa_status_t rocprofiler_pool_open(
hsa_agent_t agent, // GPU handle
rocprofiler_feature_t* features, // [in] profiling features array
uint32_t feature_count, // profiling info count
rocprofiler_pool_t** pool, // [out] context object
uint32_t mode, // profiling mode mask
rocprofiler_pool_properties_t*); // pool properties
hsa_agent_t agent, // GPU handle
rocprofiler_feature_t* features, // [in] profiling features array
uint32_t feature_count, // profiling info count
rocprofiler_pool_t** pool, // [out] context object
uint32_t mode, // profiling mode mask
rocprofiler_pool_properties_t*); // pool properties
// Close profiling pool
hsa_status_t rocprofiler_pool_close(
rocprofiler_pool_t* pool); // profiling pool handle
hsa_status_t rocprofiler_pool_close(rocprofiler_pool_t* pool); // profiling pool handle
// Fetch profiling pool entry
hsa_status_t rocprofiler_pool_fetch(
rocprofiler_pool_t* pool, // profiling pool handle
rocprofiler_pool_entry_t* entry); // [out] empty profiling pool entry
rocprofiler_pool_t* pool, // profiling pool handle
rocprofiler_pool_entry_t* entry); // [out] empty profiling pool entry
// Release profiling pool entry
hsa_status_t rocprofiler_pool_release(
rocprofiler_pool_entry_t* entry); // released profiling pool entry
rocprofiler_pool_entry_t* entry); // released profiling pool entry
// Iterate fetched profiling pool entries
hsa_status_t rocprofiler_pool_iterate(
rocprofiler_pool_t* pool, // profiling pool handle
hsa_status_t (*callback)(rocprofiler_pool_entry_t* entry, void* data), // callback
void *data); // [in/out] data passed to callback
hsa_status_t rocprofiler_pool_iterate(rocprofiler_pool_t* pool, // profiling pool handle
hsa_status_t (*callback)(rocprofiler_pool_entry_t* entry,
void* data), // callback
void* data); // [in/out] data passed to callback
// Flush completed entries in profiling pool
hsa_status_t rocprofiler_pool_flush(
rocprofiler_pool_t* pool); // profiling pool handle
hsa_status_t rocprofiler_pool_flush(rocprofiler_pool_t* pool); // profiling pool handle
////////////////////////////////////////////////////////////////////////////////
// HSA intercepting API
// HSA callbacks ID enumeration
typedef enum {
ROCPROFILER_HSA_CB_ID_ALLOCATE = 0, // Memory allocate callback
ROCPROFILER_HSA_CB_ID_DEVICE = 1, // Device assign callback
ROCPROFILER_HSA_CB_ID_MEMCOPY = 2, // Memcopy callback
ROCPROFILER_HSA_CB_ID_SUBMIT = 3, // Packet submit callback
ROCPROFILER_HSA_CB_ID_KSYMBOL = 4, // Loading/unloading of kernel symbol
ROCPROFILER_HSA_CB_ID_CODEOBJ = 5 // Loading/unloading of kernel symbol
ROCPROFILER_HSA_CB_ID_ALLOCATE = 0, // Memory allocate callback
ROCPROFILER_HSA_CB_ID_DEVICE = 1, // Device assign callback
ROCPROFILER_HSA_CB_ID_MEMCOPY = 2, // Memcopy callback
ROCPROFILER_HSA_CB_ID_SUBMIT = 3, // Packet submit callback
ROCPROFILER_HSA_CB_ID_KSYMBOL = 4, // Loading/unloading of kernel symbol
ROCPROFILER_HSA_CB_ID_CODEOBJ = 5 // Loading/unloading of kernel symbol
} rocprofiler_hsa_cb_id_t;
// HSA callback data type
typedef struct {
union {
struct {
const void* ptr; // allocated area ptr
size_t size; // allocated area size, zero size means 'free' callback
hsa_amd_segment_t segment; // allocated area's memory segment type
const void* ptr; // allocated area ptr
size_t size; // allocated area size, zero size means 'free' callback
hsa_amd_segment_t segment; // allocated area's memory segment type
hsa_amd_memory_pool_global_flag_t global_flag; // allocated area's memory global flag
int is_code; // equal to 1 if code is allocated
} allocate;
struct {
hsa_device_type_t type; // type of assigned device
uint32_t id; // id of assigned device
hsa_agent_t agent; // device HSA agent handle
const void* ptr; // ptr the device is assigned to
hsa_device_type_t type; // type of assigned device
uint32_t id; // id of assigned device
hsa_agent_t agent; // device HSA agent handle
const void* ptr; // ptr the device is assigned to
} device;
struct {
const void* dst; // memcopy dst ptr
const void* src; // memcopy src ptr
size_t size; // memcopy size bytes
const void* dst; // memcopy dst ptr
const void* src; // memcopy src ptr
size_t size; // memcopy size bytes
} memcopy;
struct {
const void* packet; // submitted to GPU packet
const char* kernel_name; // kernel name, not NULL if dispatch
hsa_queue_t* queue; // HSA queue the kernel was submitted to
uint32_t device_type; // type of device the packed is submitted to
uint32_t device_id; // id of device the packed is submitted to
const void* packet; // submitted to GPU packet
const char* kernel_name; // kernel name, not NULL if dispatch
hsa_queue_t* queue; // HSA queue the kernel was submitted to
uint32_t device_type; // type of device the packed is submitted to
uint32_t device_id; // id of device the packed is submitted to
} submit;
struct {
uint64_t object; // kernel symbol object
const char* name; // kernel symbol name
uint32_t name_length; // kernel symbol name length
int unload; // symbol executable destroy
uint64_t object; // kernel symbol object
const char* name; // kernel symbol name
uint32_t name_length; // kernel symbol name length
int unload; // symbol executable destroy
} ksymbol;
struct {
uint32_t storage_type; // code object storage type
int storage_file; // origin file descriptor
uint64_t memory_base; // origin memory base
uint64_t memory_size; // origin memory size
uint64_t load_base; // codeobj load base
uint64_t load_size; // codeobj load size
uint64_t load_delta; // codeobj load size
uint32_t uri_length; // URI string length
char* uri; // URI string
int unload; // unload flag
uint32_t storage_type; // code object storage type
int storage_file; // origin file descriptor
uint64_t memory_base; // origin memory base
uint64_t memory_size; // origin memory size
uint64_t load_base; // codeobj load base
uint64_t load_size; // codeobj load size
uint64_t load_delta; // codeobj load size
uint32_t uri_length; // URI string length
char* uri; // URI string
int unload; // unload flag
} codeobj;
};
} rocprofiler_hsa_callback_data_t;
// HSA callback function type
typedef hsa_status_t (*rocprofiler_hsa_callback_fun_t)(
rocprofiler_hsa_cb_id_t id, // callback id
const rocprofiler_hsa_callback_data_t* data, // [in] callback data
void* arg); // [in/out] user passed data
rocprofiler_hsa_cb_id_t id, // callback id
const rocprofiler_hsa_callback_data_t* data, // [in] callback data
void* arg); // [in/out] user passed data
// HSA callbacks structure
typedef struct {
rocprofiler_hsa_callback_fun_t allocate; // memory allocate callback
rocprofiler_hsa_callback_fun_t device; // agent assign callback
rocprofiler_hsa_callback_fun_t memcopy; // memory copy callback
rocprofiler_hsa_callback_fun_t submit; // packet submit callback
rocprofiler_hsa_callback_fun_t ksymbol; // kernel symbol callback
rocprofiler_hsa_callback_fun_t codeobj; // codeobject load/unload callback
rocprofiler_hsa_callback_fun_t allocate; // memory allocate callback
rocprofiler_hsa_callback_fun_t device; // agent assign callback
rocprofiler_hsa_callback_fun_t memcopy; // memory copy callback
rocprofiler_hsa_callback_fun_t submit; // packet submit callback
rocprofiler_hsa_callback_fun_t ksymbol; // kernel symbol callback
rocprofiler_hsa_callback_fun_t codeobj; // codeobject load/unload callback
} rocprofiler_hsa_callbacks_t;
// Set callbacks. If the callback is NULL then it is disabled.
// If callback returns a value that is not HSA_STATUS_SUCCESS the callback
// will be unregistered.
hsa_status_t rocprofiler_set_hsa_callbacks(
const rocprofiler_hsa_callbacks_t callbacks, // HSA callback function
void* arg); // callback user data
const rocprofiler_hsa_callbacks_t callbacks, // HSA callback function
void* arg); // callback user data
#ifdef __cplusplus
} // extern "C" block
projects/rocprofiler/include/rocprofiler/v2/rocprofiler.h
+1 -1
Wyświetl plik
@@ -1714,7 +1714,7 @@ typedef enum {
ROCPROFILER_ATT_TOKEN_MASK2 = 4,
ROCPROFILER_ATT_SE_MASK = 5,
ROCPROFILER_ATT_SAMPLE_RATE = 6,
ROCPROFILER_ATT_BUFFER_SIZE = 7, //! ATT collection max data size.
ROCPROFILER_ATT_BUFFER_SIZE = 7, //! ATT collection max data size.
ROCPROFILER_ATT_PERF_MASK = 240,
ROCPROFILER_ATT_PERF_CTRL = 241,
ROCPROFILER_ATT_PERFCOUNTER = 242,