OpenMPToolsInterface/LLVM-openmp
This is a stale repository and only there for the commit history. All development moved over to the llvm-project repository! Was: LLVM OpenMP runtime with experimental changes for OMPT (Preview of TR7 features in *_tr branches)
C++NOASSERTION
Issues
- 1
[llvm-openmp@3.9.2b2 [Failed to use a built-in test case] When you run the make ctest/ftest command, the error message "Can' t locate ompts_parserFunctions.pm in @INC " is displayed.
#79 opened by Tom-python0121 - 0
Upstreaming OMPD
#77 opened by jprotze - 0
- 1
Support for the ICVs tool-var and display-affinity-var in ompd_enumerate_icvs() and ompd_get_icv_from_scope()
#74 opened by jinisusan - 1
Support for the ICV nthreads-var in ompd_enumerate_icvs() and ompd_get_icv_from_scope()
#72 opened by jinisusan - 3
dyn-var, stacksize-var, cancel-var, max-task-priority-var, debug-var support in ompd_enumerate_icvs() and ompd_get_icv_from_scope().
#70 opened by jinisusan - 2
[OMPD] The LLVM OMPD ompd_enumerate_icvs() and ompd_enumerate_states() functions do not allocate the strings they returns
#58 opened by jdelsign - 4
- 1
[OMPD] The LLVM OMPD ompd_get_api_version() function returns a non-standard API version
#51 opened by jdelsign - 0
[OMPD] The LLVM OMPD ompd_get_task_function() entry_point address return value is 0 for the outermost implicit task of a parallel region
#69 opened by jdelsign - 2
[OMPD] The LLVM OMPD ompd_get_task_frame() function does not seem to support NVIDIA GPUs
#66 opened by jdelsign - 1
[OMPD] The LLVM OMPD ompd_get_task_function() entry_point address return value for NVIDIA GPUs does not seem to correspond to a valid code address on the GPU
#68 opened by jdelsign - 0
[OMPD] The LLVM OMPD ompd_get_task_in_parallel() function misspells "ompd_nvptx_parallel_info_t"
#64 opened by jdelsign - 0
[OMPD] The LLVM OMPD library does not call the symbol address lookup callback function correctly
#62 opened by jdelsign - 0
[OMPD] The LLVM OMPD ompd_get_api_version_string() function should be named ompd_get_version_string()
#57 opened by jdelsign - 0
- 0
- 0
- 0
[OMPD] The LLVM OMPD library passes the memory read/write callback address argument by value
#53 opened by jdelsign - 0
- 0
- 0
[OMPD] The LLVM OMPD library returns ompd_rc_unsupported or do not provide certain important ICVs for NVIDIA GPU handles
#65 opened by jdelsign - 0
- 0
- 0
- 1
Need to clean up OMPD CMake files for OMPD
#47 opened by mcfadden8 - 4
Chicken/Egg compile problem with libompd
#46 opened by mcfadden8 - 2
- 7
- 4
- 3
Unexpected behavior of idle event callback
#40 opened by madagra - 0
[OMPD] Keeping address spaces separate
#42 opened by jprotze - 12
OMPT on MacOS
#13 opened by simoatze - 4
- 5
- 33
Performance issue
#29 opened by omalyshe - 1
OMPT-placeholders
#33 opened by jprotze - 2
Current source won't build towards_tr4
#37 opened by khuck - 9
codeptr_ra is nil for sync_region_begin/end and sync_region_wait_begin/end callback trigged by non-master threads
#32 opened by yanyh15 - 1
backtraces in ompt-specific.cpp?
#31 opened by jmellorcrummey - 6
Return address for omp_init_lock
#36 opened by sconvent - 1
- 3
barrier_wait_end/barrier_end/implicit_task_end events triggered only after forking the following parallel region (or the thread_end)
#25 opened by yanyh15 - 2
weak attribute on ompt_start_tool
#27 opened by jprotze - 2
ompt_callback_sync_region_wait event name
#24 opened by yanyh15 - 3
Interoperability with pthreads
#16 opened by jprotze - 0
Declare omp_control_tool in fortran header
#21 opened by jprotze - 3
- 15
- 1
Segmentation fault in __kmp_fork_barrier
#14 opened by simoatze