This should be straightforward given the structure of web-platform tests, and given the volume of web-platform tests it should be a significant improvement.

I'm going to see if I can get anywhere meaningful in the next hour or two...

AFAICT, we don't actually access web-platform test files from the objdir for anything except test packaging. And we already have a mechanism for creating test archives from files directly in the source directory. So, let's stop copying them to the objdir and package them directly from the source directory! The _tests install manifest reports the following change: before: 41,977 files installed after: 24,537 files installed delta: -17,440 files We still copy some WPT files to the objdir. We might be able to eliminate these as well. However, since there are only ~200 files, I'm not too concerned. I manually compared the resulting web-platform zip archives from before and after. No files were removed from the archive. However, the new archive does gain several hundred empty directories with .gitkeep files. This feels weird, but it shouldn't break anything (I would think). I'm inclined to leave them for now. I'll file a follow-up bug to deal with them (preferably by removing them from version control). Review commit: https://reviewboard.mozilla.org/r/31139/diff/#index_header See other reviews: https://reviewboard.mozilla.org/r/31139/

https://reviewboard.mozilla.org/r/31139/#review27947 Significant tests install manifest processing time difference on Windows. This is on a i7-6700K with a fast SSD: hardware is not lacking. Before: ~40s full install manifest After: ~27 Before: ~5.4s no-op install manifest After: ~2.7s

The .gitkeep files are intentional; they're there to mark directories that should contain tests (because they correspond to sections of the specification that have testable assertions), but don't have any yet.

Comment on attachment 8708642 [details] MozReview Request: Bug 1239880 - Don't copy thousands of web-platform test files to objdir; r?chmanchester https://reviewboard.mozilla.org/r/31139/#review27953 Nice.

I believe this change has cut my local test running time by more than 50%. When I filed bug 1234336 I was seeing consistent test runs around 1m8s. Building off of 819fff092e3d I now run the same command and consistently see test runs around 32s Thanks!