Convert OpenMP parallelization to OneAPI::TBB by dbs4261 · Pull Request #6626 · isl-org/Open3D

Commits on Sep 17, 2024

Switched from using OpenMP for parallelism to using oneAPI::TBB (though not using the oneapi scope). Untested but building.
dbs4261
committed
Swichted usage of std::mutex to tbb::mutex for consistency when used in conjunction with tbb parallel constructs.
dbs4261
committed
Fixed bug in CPU reduction
dbs4261
committed
Shift atomic to outside of RW mutex in PointCloudSegmentation.cpp
dbs4261
committed
Switch from using a mutex to a concurrent vector for parallelization of self intersecting triangles.
dbs4261
committed
Get maximum threads from TBB instead of OpenMP
dbs4261
committed
Updates to ProgressBar.h/.cpp Including separating code for writing the progress bar into its own function and a bulk inplace add function operator+=. Also added TBBProgressBar. It does not inherit…
dbs4261
committed
Updated ClusterDBSCAN in PointCloudCluster.cpp to bulk update the progress bar to limit spinning on the mutex.
dbs4261
committed
Applied Open3D style
dbs4261
committed
Marked single argument constructors for reductions as explicit to please codacy
dbs4261
committed
Explicitly load atomics in calls to utility::Log*(...) because newer versions of format wont automatically convert it to its underlying type.
dbs4261
committed
style fix

ssheorey
authored and
dbs4261
committed
Fixed incorrect include in Parallel.h
dbs4261
committed
Updated ProgressBar.h to use size_t in std namespace for consistency
dbs4261
committed
Switched std::lock_guard<std::mutex> for DiscreteGenerator to use tbb::spin_mutex::scoped_lock.
dbs4261
committed
Switch to using tbb::parallel_reduce on all platforms.
dbs4261
committed

Commits on Sep 20, 2024

Commits on Sep 23, 2024

Commits on Oct 2, 2024

Turn TBB into publicly linked library
dbs4261
committed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Convert OpenMP parallelization to OneAPI::TBB#6626

Commits on Sep 17, 2024

Switched from using OpenMP for parallelism to using oneAPI::TBB (though not using the oneapi scope). Untested but building.

Swichted usage of std::mutex to tbb::mutex for consistency when used in conjunction with tbb parallel constructs.

Fixed bug in CPU reduction

Shift atomic to outside of RW mutex in PointCloudSegmentation.cpp

Switch from using a mutex to a concurrent vector for parallelization of self intersecting triangles.

Get maximum threads from TBB instead of OpenMP

Updates to ProgressBar.h/.cpp Including separating code for writing the progress bar into its own function and a bulk inplace add function operator+=. Also added TBBProgressBar. It does not inherit…

Updated ClusterDBSCAN in PointCloudCluster.cpp to bulk update the progress bar to limit spinning on the mutex.

Applied Open3D style

Marked single argument constructors for reductions as explicit to please codacy

Explicitly load atomics in calls to utility::Log*(...) because newer versions of format wont automatically convert it to its underlying type.

style fix

Fixed incorrect include in Parallel.h

Updated ProgressBar.h to use size_t in std namespace for consistency

Switched std::lock_guard<std::mutex> for DiscreteGenerator to use tbb::spin_mutex::scoped_lock.

Switch to using tbb::parallel_reduce on all platforms.

Commits on Sep 20, 2024

Fixed progress bar bug that allow for multiple threads writing to the terminal.

Fixed issue with CPU reduction. Need to exit early if no work is to be done to prevent assignment to the output pointer.

Updated MemoryManagerStatistic to use a tbb::spin_mutex instead of a std::mutex

Updated point cloud segmentation to generate random samples using the global mutex from utilities::random.

Commits on Sep 23, 2024

Updated example usage of global mutex and engine access to reflect the TBB types.

Merge branch 'main' into omp2tbb

Commits on Oct 2, 2024

Turn TBB into publicly linked library