grpc_build_artifacts_multiplatform: passed

While I'm liking that we're finally working on fixing #13856, before we proceed I'd like to get a clarification on whether this will affect performance when C-core library is loaded dynamically? (e.g. some wrapped language do rely on using dlopen / dlsym to load C core).)
https://stackoverflow.com/questions/13106049/what-is-the-performance-penalty-of-c11-thread-local-variables-in-gcc-4-8

CC @jtattermusch

@jtattermusch That's good point. This article is good to understand why dynamic linking has a different performance for TLS in general. This applies for both __thread and thread_local. And we don't worry about the slow case mentioned in Q&A because we don't use extern and class-type. So I don't expect any performance regression because of this case at least for Linux.

grpc_build_artifacts_multiplatform: passed

Was this reverted? Version 1.33.0 works for me but neither 1.33.1 nor 1.33.2 work for me.

@kalbasit It's still here and I don't think there were relevant changes between 1.33.0 and 1.33.1+.