Blocked writing with progress bars

See merge request beppp/heka-tooling!4

Changes all writing loops to a generalized function with lambda for serializing the final prv line. This closes #9.

Feature: translate NVTX Start-End ranges too

See merge request beppp/heka-tooling!3

first implementation openacc translate. It includes region type, region with name, function, and bytes transfered.

GPU metrics matching with device now done with lower 8 bits of typeId

Closes #4

See merge request beppp/heka-tooling!1

Adds GPU metrics to trace, as sampling events, in an extra auxiliary thread/stream.  Adds corresponding cfgs to view events.

Major changes to argument parsing. Now path treatment is less prone to errors and you just need to pass the report file.

Adds communication lines for cudaLaunch - kernel relationship. Current status not displaying in all tasks.