Closed vadz closed 2 years ago
AFAICS the CI build failures have nothing to do with my (trivial) change.
Hi, thanks for the PR. Yes, the failure happens in the CI. I have a tentative fix for it, but now I see a test failure on Windows. I will try to fix this failure first.
While fixing the bug triggering a test failure I got rid of this piece of code, so the PR is not needed anymore.
Thanks, but testing the latest master with MSVC now results in
sigslot\include\sigslot\signal.hpp(203,72): warning C4324: 'sigslot::detail::func_ptr': structure was padded due to alignment specifier
I think this warning is harmless, but it would be nice to disable it. Should I make another PR to do it or will you do it yourself?
Does the warning stop when exchanging the sz and data member variables in func_ptr?
Does the warning stop when exchanging the sz and data member variables in func_ptr?
Yes indeed, exchanging these lines fixes the warning.
Ok thanks for the test, I will fix this tomorrow.
Check that GNUC is defined before testing its value to avoid the equivalent of gcc -Wundef with other compilers, notably MSVC.
No real changes, but just a very minor fix to avoid this MSVC warning (which is disabled by default but that I enable for all my code, as it's quite useful fore detecting typos etc).