Repository navigation
boot: the page-cache probe says what QEMU's own memory is made of - #260
Merged
Merged
Conversation
The probe already had QEMU's resident memory and the guest's RAM mapping at each step, and docs/memory-and-disk.md took "a constant ~50 MB" of QEMU's own from their difference. That is the only number this repository has for what a machine costs beside its guest's RAM, and a comparison with another VMM's overhead (#259) leaned on it without knowing what is in it or what the kernel holds for the machine on top. At the idle step the probe now notes it: resident memory outside the RAM mapping, by what is mapped, QEMU's page tables and KVM's second-level ones. Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
Run 43564203014: 49.1 MB resident outside the guest's RAM, of which 34.7 MB is vmlinux and 7.4 MB the binary, both file pages every machine on a host shares, and 6.9 MB anonymous. Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
The page-cache probe had QEMU's resident memory and the guest's RAM mapping at each step, and
docs/memory-and-disk.mdtook "a constant ~50 MB" of QEMU's own from their difference. It is the only number here for what a machine costs its host beside the guest's RAM, and #259 compared it with another VMM's overhead without knowing what is in it.At the idle step the probe now notes what that memory is: resident memory outside the RAM mapping, by what is mapped, largest first, plus what the kernel holds that no mapping shows - QEMU's page tables (
VmPTE) and KVM's second-level ones (SecPageTables, host-wide; the lab runs one machine at a time).Run with
gh workflow run lab.yml -f experiment=boot:page-cache -f vars='REPS=3; SPIN_PAGE_CACHE_ONLY=^caches$'. The result of the first run on this branch goes in #259 and indocs/memory-and-disk.mdin this PR.Refs #259
🤖 Generated with Claude Code