[PATCH] D113717: [Symbolizer][Debuginfo] Add debuginfod client to llvm-symbolizer.

[Nico Weber via Phabricator via llvm-commits]

thakis added a comment.

Thinking out loud a bit, maybe there should be a lib/DebuginfoCacheReader that can just read the debuginfod cache but not write on it, and then ~everything could depend on that without problems, and lib/Debuginfod would depend on that but could also write to the cache (and it could depend on libcurl and make http requests etc)? Then in a bot setup, the bot could call the "pull symbols" thing as a dedicated step at the start. And for devs, maybe there could be an `llvm-debuginfod` binary that can pull symbols as part of symbolizing.

(My main concern is that the current design feels that it's very easy to add a dep on an http library to arbitrary llvm tools, and most toolchain tools shouldn't make http requests.)


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D113717/new/

https://reviews.llvm.org/D113717