[lld-macho] Parallel input section priorities for -order_file (#241470)

buildInputSectionPriorities() used to look up every symbol in
`priorities`, which does a bunch of work (hash a symbol, do some string
manipulation).

Instead, call getSymbolPriority in parallel over object files, and then
serially merge the results.

(This works because std::min() is associative.)

Takes a Chromium Framework link on my machine from 4.191 s ± 0.020 s to
3.970 s ± 0.045 s.

Chromium uses an order file with just a few entries. I also checked with
a synthetic order file that lists all symbols. There, this phase goes
from 781 ms to 357 ms, so it's still faster (the relative speedup is
just smaller).

Contributes to #222068, depends on #241367.

No behavior change.

GitOrigin-RevId: 82c882f2469f5ca80fa7bd75a42ec9a45d7774f4
1 file changed
tree: 5845f24b247f8d942e0a8e4d5b967665af64e189
  1. cmake/
  2. COFF/
  3. Common/
  4. docs/
  5. ELF/
  6. include/
  7. MachO/
  8. MinGW/
  9. test/
  10. tools/
  11. unittests/
  12. utils/
  13. wasm/
  14. .clang-format
  15. .clang-tidy
  16. .gitignore
  17. CMakeLists.txt
  18. LICENSE.TXT
  19. Maintainers.md
  20. README.md
README.md

LLVM Linker (lld)

This directory and its subdirectories contain source code for the LLVM Linker, a modular cross platform linker which is built as part of the LLVM compiler infrastructure project.

lld is open source software. You may freely distribute it under the terms of the license agreement found in LICENSE.txt.

Benchmarking

In order to make sure various developers can evaluate patches over the same tests, we create a collection of self contained programs.

It is hosted at https://s3-us-west-2.amazonaws.com/linker-tests/lld-speed-test.tar.xz

The current sha256 is 10eec685463d5a8bbf08d77f4ca96282161d396c65bd97dc99dbde644a31610f.