dblaikie added a comment. In D101566#2785190 <https://reviews.llvm.org/D101566#2785190>, @aaronpuchert wrote:
> In D101566#2734948 <https://reviews.llvm.org/D101566#2734948>, @dblaikie > wrote: > >> Makes it hard to justify the complexity in the compiler if it's hard to >> justify/support the value of the warning. > > The complexity for `-Wweak-template-vtables` is just 10 lines of code. We're > just using information that's already there. > >> I believe it's compile time/build time, yes - but yeah, it's pretty >> questionable/suspect. LLVM's the only project I know of with it as a coding >> convention/guideline/rule - and even we haven't even remotely tried to >> enforce it. (& when I did do a bit of work to add more key functions people >> reasonably questioned the value of them - and I didn't really have data to >> support it, I could only point to the fact that I was implementing the >> stated policy/style guide) > > Just to make sure I got this right, you're suggesting to remove both warnings? Right - to remove -Wweak-template-vtable in its entirety. The original implementation explicitly didn't warn on implicit instantiations and I think the fact that it warned on explicit instantiations is more bug than feature - and we should treat it that way. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D101566/new/ https://reviews.llvm.org/D101566 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
