mstorsjo wrote: @efriedma-quic What do you think about backporting this one? It's not a regression fix, but it's fixing a longstanding issue that only surfaced recently (from Zig's use). We did backport 5c95f6a859394757b92b0d63ff90b9175056deb6 already, which fixes one aspect of it (until Zig has got more than 4096 bytes of TLS variables in a DLL), while this fixes the rest.
My only concern is that this feels like a slightly bigger and more risky change to backport. But on the other hand, the cases where it does kick in are the rare ones that weren't properly handled before, so any code that worked before shouldn't ever trigger these cases - I think. https://github.com/llvm/llvm-project/pull/202867 _______________________________________________ llvm-branch-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
