rocm-systems

Ficheiros

T

Xianwei Zhang f95b383dfa pmc: add support of concurrent kernel profiling

The profiling was only enabled in serial mode, i.e., kernels
are serialized in execution, and counters are reset at each
kernel start and read at kernel completion. This patch adds
the concurrent mode, by issuing the process-level start
packet to reset counters, and then reading twice at kernel
start and end time to obtain the counter value difference.
The new concurrent profiling usage needs the integration
with the corresponding augment at aqlprofile side.

Change-Id: I94b4442eadc8c64b8fba51b1e4916fc8b895ad21


[ROCm/rocprofiler commit: 61c9df4631]

2020-07-16 14:39:46 -05:00

bin

Merge "Update to use new bitcode library structure" into amd-master

2020-07-14 15:30:49 -04:00

cmake_modules

cmake update

2020-04-21 00:24:11 -05:00

DEBIAN

pre/post package scripts install dir name fix and CPACK_PACKAGING_INSTALL_PREFIX

2018-11-12 12:39:03 -06:00

doc

latest documentation revision

2018-06-15 10:53:06 -05:00

inc

counters dumping optimization

2020-07-15 09:57:41 -05:00

RPM

pre/post package scripts install dir name fix and CPACK_PACKAGING_INSTALL_PREFIX

2018-11-12 12:39:03 -06:00

src

pmc: add support of concurrent kernel profiling

2020-07-16 14:39:46 -05:00

test

pmc: add support of concurrent kernel profiling

2020-07-16 14:39:46 -05:00

_clang-format

ROC profiler prototype sources importing

2017-11-09 17:26:19 -06:00

build.sh

adding standalone intercept test; queeu_start/stop API fixed as public;

2020-04-08 04:31:52 -05:00

CMakeLists.txt

mem_manager.py packaging fix

2020-06-29 13:27:22 -05:00

LICENSE

minor changes;

2020-01-14 10:25:13 -06:00

Readme.txt

fixing build instructions

2019-09-13 16:41:15 -05:00

sync_hsa_mgr.sh

SQTT local allocation

2018-04-29 03:24:46 -05:00

Readme.txt

ROC Profiler library.
Profiling with metrics and traces based on perfcounters (PMC) and traces (SQTT, PMC).
Implementation is based on AqlProfile HSA extension.
Library supports GFX8/GFX9.

The library source tree:
 - doc  - Documentation
 - inc/rocprofiler.h - Library public API
 - src  - Library sources
   - core - Library API sources
   - util - Library utils sources
   - xml - XML parser
 - test - Library test suite
   - ctrl - Test controll
   - util - Test utils
   - simple_convolution - Simple convolution test kernel

Build environment:

$ export CMAKE_PREFIX_PATH=<path to hsa-runtime includes>:<path to hsa-runtime library>
$ export CMAKE_BUILD_TYPE=<debug|release> # release by default
$ export CMAKE_DEBUG_TRACE=1 # 1 to enable debug tracing

To build with the current installed ROCM:

$ cd .../rocprofiler
$ export CMAKE_PREFIX_PATH=/opt/rocm/include/hsa:/opt/rocm
$ mkdir build
$ cd build
$ cmake ..
$ make

To run the test:

$ cd .../rocprofiler/build
$ export LD_LIBRARY_PATH=.:<other paths> # paths to ROC profiler and oher libraries
$ export HSA_TOOLS_LIB=librocprofiler64.so # ROC profiler library loaded by HSA runtime
$ export ROCP_TOOL_LIB=test/libtool.so # tool library loaded by ROC profiler
$ export ROCP_METRICS=metrics.xml # ROC profiler metrics config file
$ export ROCP_INPUT=input.xml # input file for the tool library
$ export ROCP_OUTPUT_DIR=./ # output directory for the tool library, for metrics results file 'results.txt' and SQTT trace files 'thread_trace.se<n>.out'
$ <your test>

Internal 'simple_convolution' test run script:
$ cd .../rocprofiler/build
$ run.sh

To enabled error messages logging to '/tmp/rocprofiler_log.txt':

$ export ROCPROFILER_LOG=1

To enable verbose tracing:

$ export ROCPROFILER_TRACE=1