builder: include more info in cache keys, derive inputs#5503
Draft
jakebailey wants to merge 3 commits into
Draft
builder: include more info in cache keys, derive inputs#5503jakebailey wants to merge 3 commits into
jakebailey wants to merge 3 commits into
Conversation
Compute target library cache directories from a content-derived description of the library build instead of relying on manual version bumps. Use that same description to drive compilation, so the cache key covers the inputs consumed by the build. The key includes source hashes, generated headers, header inputs, per-file flags, the compiler argument template, relevant target options, and the clang/LLVM identity. This avoids stale lib.a and crt1.o files after libc, compiler-rt, header, or compiler flag changes.
Stop caching C dependency lists as reusable inputs to object cache lookups. Instead, ask Clang for the current dependency list before each lookup and key the object by the current dependencies, compiler flags, and clang/LLVM identity. This avoids stale object hits when include path resolution changes, such as when a new header is added earlier in an include path. It also avoids using dependency data produced by a different clang binary.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
I've been working on a
GOCACHEPROGimplementation for tinygo, but am pulling stuff out of that branch.One problem I hit in my stack (and, generally) is conditions where I am forced to delete my tinygo cache, usually because I switched branches or rebuilt LLVM. If I want to use a GOCACHEPROG that compresses stuff (my latest obsession), getting the cache key wrong isn't so great since the same cache is also shared by other stuff on my system and cleaning is one thing.
This stack attempts to fix a bunch of caching problems by including more stuff in the cache key, and make it so that when tinygo does some work that's going to be cached, it pulls that info from the cache key instead such that it's less likely to integrate other outside info (something something pure function).