| FazBrowse GitHub Viewer | Trending | | Home |
| Tools: [Download Repo ZIP] [Original HTTPS Page] |
Sorry, something went wrong.
One of the tests now hits an MSVC error I've seen before, related to using std::source_location via module std import -- it doesn't manifest on any other compiler, or in MSVC using std headers. So the error is spurious, but I'm not able to debug it or find a workaround, and I don't want to record regression test failures just because of that use case where 'module std currently doesn't work.' So this commit also changes all -pure regression tests to use headers, not modules, at least for now until modules work better
|
Thanks! |
Sorry, something went wrong.
|
@hsutter Did you intend to switch from importing std to including std when running the tests? |
Sorry, something went wrong.
That's probably a good idea. @jarzec was maintaining and updating the test runner (thanks again!) -- do those use MSVC modules today which would have required that modules be built somehow (I vaguely recall seeing something like that)? If they do they should probably also switch to use -in to include std headers instead of importing the std module, which is what I did in this commit. |
Sorry, something went wrong.
I have to admit I forgot the details, but clearly there was some effort at some stage to make import usable. @hsutter I will soon try open a quick PR readding imports. You can either merge it or close it. First, however, I will update the regression tests to have a clean start. EDIT |
Sorry, something went wrong.
|
Regression tests are fixed (among other changes) in #1442 |
Sorry, something went wrong.
| Back | FazBrowse Home | New Git URL |
Checks for modulo and modulo-assignment by zero at runtime. Based on a suggestion in #1399.
Resolves #1399.