This repository was archived by the owner on Sep 11, 2024. It is now read-only.
Kernel: [WIP] Use a tree for VMSpace instead of a linked list#57
Draft
byteduck wants to merge 1 commit into
Draft
Kernel: [WIP] Use a tree for VMSpace instead of a linked list#57byteduck wants to merge 1 commit into
byteduck wants to merge 1 commit into
Conversation
This updates VMSpace to use a tree structure to store regions instead of a linked list, which simplifies things and (theoretically) speeds them up as well. In practice, however, this seems to make things *slower* when compiling without optimizations - probably because the VMSpace trees are incredibly unbalanced due to the fact that new regions are pretty much being inserted sequentially. Theoretically, once kstd::map is self-balancing, it should speed things up quite a bit. Then, I'll merge this into the main branch.
1e13d14 to
6f9f442
Compare
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 subscribe to this conversation on GitHub.
Already have an account?
Sign in.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
New PR since I accidentally screwed up the last one by setting its branch to
gh-pagessomehow...This updates VMSpace to use a tree structure to store regions instead of a linked list, which simplifies things and (theoretically) speeds them up as well.
In practice, however, this seems to make things considerably slower. I'm not 100% sure why this is. The first thing that comes to mind is that traversing the map while forking is slower than traversing a linked list. But, usually, the maps are pretty small and forking happens pretty infrequently, so that's most likely not the culprit. I'll have to do some benchmarking to compare allocation and lookup times, probably.