vminit: switch root to new tmpfs before serving containers#187
Open
dmcgowan wants to merge 1 commit into
Open
Conversation
There was a problem hiding this comment.
Pull request overview
This PR changes vminitd’s early-boot behavior to replace the initramfs-backed / with a fresh tmpfs “real root” (switch_root via MS_MOVE + chroot) before container services start, enabling a distinct VM root filesystem and allowing container pivot_root behavior to be configurable again.
Changes:
- Add
switchRoot()to build a new tmpfs root, mount pseudo-filesystems into it, copycrun, configure cgroup v2 controllers, and switch/viaMS_MOVE+chroot. - Refactor
systemInitto perform the root switch first, then networking setup, and manage the DHCP renewer goroutine internally. - Change runc create options to set
NoPivotfromp.NoPivotRootinstead of being hardcoded totrue.
Reviewed changes
Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.
| File | Description |
|---|---|
| pkg/vminit/initd/initd.go | Implements tmpfs switch_root flow, updates pseudo-filesystem mounts, and refactors system init/networking startup. |
| internal/vminit/process/init.go | Makes container pivot_root behavior configurable via p.NoPivotRoot. |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
Comment on lines
+234
to
+241
| // avoids the stat walk that os.MkdirAll does per path. /sys/fs and | ||
| // /sys/fs/cgroup are created explicitly because sysfs is not yet mounted. | ||
| for _, dir := range []string{ | ||
| root + "/sbin", | ||
| root + "/proc", | ||
| root + "/sys", | ||
| root + "/sys/fs", | ||
| root + "/sys/fs/cgroup", |
Replace the static initramfs root with a fresh tmpfs using MS_MOVE + chroot (the classic switch_root approach, compatible with any kernel version). - switchRoot mounts a new tmpfs, copies all pseudo-filesystems (proc, sysfs, cgroup2, devtmpfs, /run) directly into it, then replaces / via MS_MOVE + chroot. No intermediate mount-moves are needed since the pseudo-filesystems are mounted into the new root from the start. - /run tmpfs is capped at size=64m,mode=0755; root tmpfs at size=128m,mode=0755. - /tmp is not mounted — nothing in vminitd uses it; containers get their own per-container tmpfs from the OCI spec. - /etc is created inside switchRoot alongside the other directories. - crun is copied into the new root then removed from the initramfs to reclaim its pages before the root switch. - NoPivot is changed from a hardcoded true to p.NoPivotRoot (defaults false), enabling pivot_root inside containers now that the VM root is a distinct tmpfs filesystem separate from the container overlay/erofs rootfs. Signed-off-by: Derek McGowan <derek@mcg.dev>
dmcgowan
force-pushed
the
vminitd-switch-root
branch
from
12678b0 to
4e57ad5
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 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.
Replace the static initramfs root with a fresh tmpfs using MS_MOVE + chroot (the classic switch_root approach, compatible with any kernel version).