Version 0.5.5: Bug fixes and minor improvements

[eyalroz]

Run-time compilation functionality

• Some NVRTC compilation options are not (directly) supported #397 : The NVRTC compilation options class now supports passing extra options to PTXAS, and also supports --dopt

Bug fixes

• scoped_existence_ensurer_t fails if the driver has not been initialized #396 : A gadget for ensuring there is some current context (regardless of which) will no longer assume some other activity has already initialized the NVIDIA driver.
• Starting CUDA profiling: initialization error #395 : Can now start profiling with out nvtx component even if the driver not yet being initialized.

Other changes

• Allow using cuda::wait(my_event) #400 : Added an alias for waiting/synchronizing on an event: You can now execute cuda::wait(my_event), not just cuda::synchronize(my_event).
• Support taking an event pair for time_elapsed_between #399 : The time_elapsed_between() can now accept pairs of events
• Add the bandwidthtest CUDA sample as an example program #398 : Added another example program, the CUDA sample bandwidthtest
• cuda::stream_t::enqueue_t methods are non-const #401 : Made all stream enqueuing methods const (so you can now enqueue on a stream passed by const&)
• overall_dimensions_t should be constructible from a dim3 struct #404 : Can now construct grid::overall_dimensions_t from a dim3 object, so that they're more interoperable with CUDA-related values you obtained elsewhere.

---

This discussion was created from the release Version 0.5.5: Bug fixes and minor improvements.