AaronBallman wrote: > OK, so the macro has been copied to some other file, and was not updated at > the same time as the fix to winnt.h. So, how about the next Q: "If so, can > that other file just be updated correspondingly, with the same change?" > > This just seems like a whole lot of workaround cruft, for something that's > trivial to fix in the macro definition -- and, furthermore, already HAS been > fixed, years ago, in most headers on most platforms -- including the > more-commonly-used windows headers.
Another way to think about it is: this blesses an *extremely* common code pattern in the wild that has existed for 40+ years: https://sourcegraph.com/search?q=context:global+lang:C%2B%2B+%5C*%5C%290%5C%29%5C-%5C%3E+count:1000000&patternType=regexp&sm=0 I'm sure some of those 610k+ results are false positives, hits from C code, not used in constant expressions, etc. but this still seems like a very reasonable change IMO. https://github.com/llvm/llvm-project/pull/197005 _______________________________________________ cfe-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
