You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In #6800, we added some testing which scrapes info-level logs to determine some internal pants state in an integration test -- in that case, we check the exe_filename of the compiler and linker invoked to build some C/C++ sources. This led to some discussion, in which some methods of propagating the compiler's identity to an examinable output build product were brainstormed, with some sound and fury but not many conclusions.
This ticket covers determining significant (i.e. user-visible) differences in compilers (perhaps by identifying some feature supported by one and not the other, and the reverse), and inspecting those instead of polluting normal output with info logs and trying to examine internal state in an integration test.
The text was updated successfully, but these errors were encountered:
In #6800, we added some testing which scrapes info-level logs to determine some internal pants state in an integration test -- in that case, we check the
exe_filename
of the compiler and linker invoked to build some C/C++ sources. This led to some discussion, in which some methods of propagating the compiler's identity to an examinable output build product were brainstormed, with some sound and fury but not many conclusions.This ticket covers determining significant (i.e. user-visible) differences in compilers (perhaps by identifying some feature supported by one and not the other, and the reverse), and inspecting those instead of polluting normal output with info logs and trying to examine internal state in an integration test.
The text was updated successfully, but these errors were encountered: