Kevin Huck
Kevin Huck
**Describe the bug** I'm developing and testing a new feature in APEX to report memory leaks on the GPU (cuda, hip, etc.). When I tested with a Kokkos application, it...
## Expected Behavior successful configuration ## Actual Behavior The cmake command results in: ``` -- Configuring done CMake Error in CMakeLists.txt: Evaluation file to be written multiple times with different...
## Expected Behavior The current HPX counter interface expects that all counters are registered during HPX initialization. However, in order for APEX to share counters with HPX, APEX won't register...
Release and RelWithDebInfo configurations fail this test: ``` 177/269 Test #302: tests.unit.plugins.matrixops.clip ................................................***Failed 0.72 sec /var/lib/buildbot/workers/phylanx/x86_64-gcc7-relwithdebinfo/build/tests/unit/plugins/matrixops/clip.cpp(132): test 'phylanx::ir::node_data(expected) == phylanx::execution_tree::extract_numeric_value(f.get())' failed in function 'void test_clip_1d()': '[3, 1, 1, 2, 4,...
The `tests.unit.python.primitives.make_vector` test is failing on buildbot. Using the Debug config. See http://omega.nic.uoregon.edu:8020/#/builders/5/builds/285/steps/16/logs/stdio for details. It is throwing an exception: ``` Traceback (most recent call last): File "/var/lib/buildbot/workers/phylanx/x86_64-gcc7-debug/build/tools/buildbot/build-delphi-x86_64-Linux-gcc/phylanx-Debug/python/build/lib.linux-x86_64-3.6/phylanx/ast/physl.py", line 393,...
The nightly tests are down to just Phylanx-related failures, and one of them is: ` 292 - tests.unit.plugins.matrixops.clip (Failed)` ``` {stack-trace}: 12 frames: 0x7f083c179630 : /lib64/libpthread.so.0(+0xf630) [0x7f083c179630] in /lib64/libpthread.so.0 0x7f0832202cf8...
Buildbot build is failing with the following error: http://omega.nic.uoregon.edu:8020/#/builders/5/builds/258/steps/15/logs/stdio ``` In file included from /var/lib/buildbot/workers/phylanx/x86_64-gcc7-debug/build/phylanx/execution_tree/primitives/base_primitive.hpp:1193:0, from /var/lib/buildbot/workers/phylanx/x86_64-gcc7-debug/build/src/execution_tree/locality_annotation.cpp:9: /var/lib/buildbot/workers/phylanx/x86_64-gcc7-debug/build/phylanx/execution_tree/primitives/primitive_component_base.hpp: In constructor ‘phylanx::execution_tree::match_pattern_type::match_pattern_type(phylanx::execution_tree::match_pattern_type::match_pattern_data&&)’: /var/lib/buildbot/workers/phylanx/x86_64-gcc7-debug/build/phylanx/execution_tree/primitives/primitive_component_base.hpp:188:61: error: no matching function for call to ‘get(phylanx::execution_tree::match_pattern_type::match_pattern_data&)’...
The nightly test of the Release build of tests.unit.python.execution_tree.categorical_crossentropy fails consistently, but when run in isolation it does not crash. The full log is here: http://omega.nic.uoregon.edu:8020/#/builders/2/builds/201/steps/16/logs/stdio . The top of...
Running the lra_csv example results in the following error: ``` [khuck@delphi phylanx]$ /home/users/khuck/nightly/phylanx/phylanx/tools/buildbot/build-delphi-x86_64-Linux-gcc/phylanx-Release/bin/lra_csv -t1 --data_csv=/home/users/khuck/nightly/phylanx/data/10kx10k.csv --hpx:bind=balanced --hpx:numa-sensitive --n=5000 --row_stop=5000 --col_stop=5000 tcmalloc: large alloc 1310851072 bytes == 0x50d18000 @ tcmalloc: large...
The release build has one last failing test: http://omega.nic.uoregon.edu:8020/#/builders/2/builds/167/steps/16/logs/stdio I tracked this down to temporaries `output.tensor()` and `target.tensor()` going out of scope in `cat_cross_operation::cat_cross3d()`: https://github.com/STEllAR-GROUP/phylanx/blob/64540c34dcec64530cc5c9239fb36144b1302bee/src/plugins/keras_support/categorical_crossentropy_operation.cpp#L246-L297 I'll submit a pull request...