HashCachedContextLifter: Introduce the lifter and the option #1690
Conversation
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
Using hashcache on contexes makes a big difference in memory usage (churn) and a reasonable improvement on runtime when HashConsing is disabled. This is particularly important for the upcoming parallel solvers
* correct comment on the function of relift * Rename M->Base for consistency * Fix relift
michael-schwarz
approved these changes
Feb 19, 2025
sim642
approved these changes
Feb 19, 2025
Member
sim642
left a comment
sim642
left a comment
There was a problem hiding this comment.
With hashconsing there are separate functors for applying it do D and C. This only applies to C.
Is there a reason that hashcaching C is particularly necessary while on D it doesn't really matter?
Contributor
Author
After analyzing with memtrace, I have found about 90% of memory usage that we can cut down with this comas from |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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Outdated suggestions cannot be applied.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.
Using hashcache on contexes makes a big difference in memory usage (churn) and a reasonable improvement on runtime when HashConsing is disabled. This is particularly important for the upcoming parallel solvers