HighCommander4 wrote: > I think other than the review it would be 100% done? I mean, no big further > modifications would be involved?
In terms of functionality, I'd say the patch is complete. In terms of performance (and specifically memory usage), it's a matter of judgment of what we consider good enough. The current patch increases the memory usage of the index by 2.5% on the example workload of the LLVM codebase. That's not too bad (and a definite improvement over the 8.2% increase for the original version of the patch), but there is a possibility of doing a more involved rework of the index's in-memory representation (discussed under the name "deep lookup optimization" in the Phabricator thread) which would actually result in a net decrease in memory usage even with the feature. One of the key pieces of feedback I'm hoping to get on this patch is whether the 2.5% memory usage increase is acceptable, or if we should consider the deep lookup optimization as a prerequisite. https://github.com/llvm/llvm-project/pull/77556 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
