AaronBallman wrote: > > I think it makes sense to split this off into its own warning group so it > > can be controlled separately from `-Wimplicit-int-conversion`. It is a > > common code pattern, but it can still hide bugs that can be hard to spot > > when the values are on the edge of the range of representable values for > > the type. > > You mean we should keep my previous changes and separate this pattern > > ```c++ > int8_t x = something; > x = -x; > ``` > > to a new warning flag?
Correct. Perhaps something like `-Wimplicit-int-conversion-on-negation` which is grouped under `-Wimplicit-int-conversion`? https://github.com/llvm/llvm-project/pull/139429 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
