Quincey Koziol

Collective CUDA operations have dangerous pointer operation that could corrupt memory

2

Thank you for taking the time to submit an issue! ## Background information While reviewing another PR, I noticed that several routines (mca_coll_cuda_scan, mca_coll_cuda_reduce, mca_coll_cuda_allreduce, mca_coll_cuda_reduce_scatter_block, and mca_coll_cuda_exscan) have similar...

Corresponding changes to make error output for regression tests agnostic to thread setting.

Refactor error handling code to eliminate internal ID calls

3

All calls to the H5I routines are now made in API routines (sometimes in FUNC_ENTER/LEAVE_* macros), except for some calls to H5E_clear_stack() within the library, but I'm planning to remove...

Refactor threading and other concurrency support

95

Complete overhaul of the concurrency-related aspects of the library (threading, atomics, locking, etc.), adding private routines in the H5TS package to allow internal algorithms to use all of these capabilities....

**Describe the bug** The function pointer gets stored, but is never called. **Expected behavior** Call it according to the spec for H5Pset_append_flush() **Platform (please complete the following information)** All **Additional...

Make H5FL package threadsafe

1

Library: - Added mutex that obeys disable-locking-for-this-thread (DLFTT) semantics, to enable threadsafe data structures within the library that "suspend" locking when a thread has exclusive ownership of the library. (in...