A concept-centered standard library for C++20, enabling safer and more reliable products and a more modern feel for C++ code.; Also home of Subdoc the code-documentation generator.
I was recently re-reminded that __has_extension isn't the best/right way to test for this (per @zygoloid's Dec 2021 comment in https://reviews.llvm.org/D114732#3165581 ). I suggest either removing this codepath entirely or replacing it with one of the P1144R9 feature-test macros:
Note that Clang trunk's __is_trivially_relocatable(T) builtin is still not usable for your purposes (i.e., https://github.com/llvm/llvm-project/pull/84621 still isn't moving), so this codepath is still relevant only to the P1144 reference implementation on Godbolt.
https://github.com/chromium/subspace/blob/f9c481a241961a7be827d31fadb01badac6ee86a/sus/mem/relocate.h#L60C7-L60C45
I was recently re-reminded that
__has_extension
isn't the best/right way to test for this (per @zygoloid's Dec 2021 comment in https://reviews.llvm.org/D114732#3165581 ). I suggest either removing this codepath entirely or replacing it with one of the P1144R9 feature-test macros:(almost certainly the best option)
or (IMHO a worse option)
Note that Clang trunk's
__is_trivially_relocatable(T)
builtin is still not usable for your purposes (i.e., https://github.com/llvm/llvm-project/pull/84621 still isn't moving), so this codepath is still relevant only to the P1144 reference implementation on Godbolt.attn: @danakj