[compiler-rt] [clang] [lldb] [clang-tools-extra] [libc] [flang] [llvm] LLDB Debuginfod usage tests (with fixes) (PR #79181)

[Kevin Frei via cfe-commits]

kevinfrei wrote:

> I'm wondering if shell test are really the best way to test this. For more complex scenarios like are being tested here, we generally prefer [1] API tests because they're more expressive and allow you to build more complicated test binaries with our Makefile system. Things like stripping a binary and generating split DWARF all seems like things that could be done that way, to avoid checking in a (yaml-ized) binary.
> 
> Did you consider/evaluate writing API tests for this?
> 
> [1] https://lldb.llvm.org/resources/test.html

The next addition (async support) *requires* API tests, so I figured getting some shell tests in place to do a more "user experience" kind of testing was reasonable, since there will be API-based tests coming. If you feel like moving this stuff to API tests wholesale is the right thing to do, I'm happy to go figure that out next week.

https://github.com/llvm/llvm-project/pull/79181