Joachim

My experience working with ThreadSanitizer for some time matches with @light2yellow: recent gcc + clang compiler versions, at least on x86 architectures, don't need the pi* flags. The flags were...

[compile_commands.json-3.6.0.txt](https://github.com/dash-project/dash/files/2023425/compile_commands.json-3.6.0.txt) [compile_commands.json-2.8.12.2.txt](https://github.com/dash-project/dash/files/2023426/compile_commands.json-2.8.12.2.txt) cmake 3.6.0 adds -std=gnu99 and -std=gnu++14 as compiler flags

I already have a prototype, that fetches the output and redirects the output into MUST's aggregation engine. It's not perfect. I think we can introduce additional filters to specify similarities,...

The function that you register with SetPrintfAndReportCallback(void (*callback)(const char *)); is called multiple times per race report just like you would call printf in your code. The structure of the...

With MUST I refered to our MPI correctness-checking tool: https://www.itc.rwth-aachen.de/MUST It already does aggregation of detected errors across MPI ranks. Currently it only collects MPI specific issues. At some point...

Especially if you look at execution of OpenMP explicit tasks, you will find that tasks, that execute the same lexical code, which might also be created in the same context,...

This benchmark has a high chance to schedule both tasks on the same thread. In that case the tool currently will not detect a race. Only in the rare case...

Can you add a list of the tests that fail on Mac?

I had a quick look, but didn't have an idea why these might fail. I try to have a closer look during the flight back from SC.

I cannot really test this, but I think it should be `||` instead of `&&` in https://github.com/PRUNERS/archer/blob/master/rtl/ompt-tsan.cpp#L173 I still don't understand what is going on with the tests. Are we...