Details
-
Improvement
-
Status: Resolved
-
Major
-
Resolution: Fixed
-
0.11.1
Description
Currently, we run Valgrind on a coverage-enabled C++ build on Travis-CI. This means the slowness of Valgrind acts as a multiplier of the overhead of outputting coverage information using the instrumentation added by the compiler.
Instead we should probably emit C++ (and Python) coverage information in a different Travis-CI build without Valgrind enabled.
Attachments
Issue Links
- links to