Decouple the session loader into reader and writer over the cache #4445
+251
−66
Conversation
This file contains 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
… them to improve performance fix haskell#4381
…p error loading and cradle files
soulomoon
force-pushed
the
batch-load-multi-read
branch
from
ba7379b to
79a43a0
Compare
Renames getOptions to getOptionsWorker for clarity Removes redundant getOptionsLoop function Ensures session loading is called under the same `Action` context
This was referenced Feb 25, 2025
soulomoon
added
the
performance
fendor
pushed a commit
to haskell/hie-bios
that referenced
this pull request
Mar 2, 2025
When we fail to load some files X, we want to know exactly what files Y we actually did try to load. Where Y ⊆ X. In the context of HLS, X is the set of files which we already tried to load with the file that we want to additionally load into our HLS session. Y is a subset of X, chosen based on the Cradle type (bios and stack cradles do not support loading a component with the context information), and GHC and cabal-install version. In particular, cabal cradles can make use of the full set of X for initialising a session if and only if the GHC and cabal-install version is recent enough, *and* the user explicitly requested a multi-repl session. By tracking which files we have actually tried to load, HLS can then try to reduce the number of files we put into sequential loading loop once batch load is failed for haskell/haskell-language-server#4445
|
I don't understand why it fails to build for ghc version before 9.10.1. @fendor Can you help me on this ? I am puzzled. give me |
|
@soulomoon That was a tricky one! The ultimate conflict is that Then, the latest version of Adding to the works for me. |
|
Thank you so much for your help @fendor. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
Follow up of and already include the changes in #4439
Refactored session loader to use architecture that having multiple readers and writer over the cache .
In bigger project, HLS usually loading up slow. One of the main obstacle is that we are loading ghc options for files in strict sequential order.
I have made an attemp to improve the situation by refactoring this loading into reader and writer over cache. This way, the long laoding file won't be blocking already loaded files, making hls more responsive and smooth.
Also it eliminates session loading for the same file from being multiply requested, by lining up in an ordered set in stead of a list.