Repository navigation
fix(bundle): Carry hvi 0.1.0-rc.2 - #37
Merged
Merged
Conversation
hvi 0.1.0-rc.1 advertised RDRAND and RDSEED to x86 guests whatever the host CPU had. On a host without them, as a VM with a generic CPU model (kvm64, qemu64, x86-64-v2-AES) is, the guest kernel triple-faulted at its first rdrand, and every brig run ended in "guest stopped: SystemReset". hvi is the default monitor of a generic boot on Linux, so brig v0.5.0 did not boot there at all. rc.2 passes the host's bits through (brig-sh/hvi-vmm#343). The tag, commit and both tarball shas move together. README, DESIGN.md and docs/variants.md name the new release. Signed-off-by: Anastassios Nanos <ananos@nofire.ai>
ananos
marked this pull request as ready for review
October 31, 2026 23:44
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
hvi 0.1.0-rc.1 advertised RDRAND and RDSEED to x86 guests whatever the host CPU had. On a host without them -- a VM with a generic CPU model (
kvm64,qemu64, Proxmox'sx86-64-v2-AES) -- the guest kernel triple-faulted at its firstrdrand, and everybrig runended inguest stopped: SystemReset. hvi is the default monitor of a generic boot on Linux, so brig v0.5.0 with rc19/rc20 does not boot there at all. A user hit this on a fresh Ubuntu 26.04 VM. hvi 0.1.0-rc.2 passes the host's bits through (brig-sh/hvi-vmm#343, #344).Changes
scripts/build-bundle.sh:HVI_VERSION=0.1.0-rc.2,HVI_COMMIT=2ed8f61d..., both tarball shasTested
scripts/fetch-hvi.shwith the new pins, locally:The arm64 sha given for the amd64 tarball is refused and nothing is written.
tests/fetch-hvi.shpasses (7/7) and shellcheck is clean. The release: the cosign bundle verifies againsttag-release.yml@refs/tags/v0.1.0-rc.2, and is refused for the publish-binaries identity and for the rc.1 tag.The fix was checked end to end before the tag, on a brig v0.5.0 root install in a fresh nested Ubuntu 26.04 VM with only
hviswapped:-cpu host,-rdrand,host,-rdseedandkvm64,+vmxreset or hung with rc.1 and boot with the fix, and the guest sees exactly the host's RDRAND/RDSEED. CI's bundle build on this branch is the gate for the bundle itself.Release sequence
For rc21, together with the rootless installer changes already on main (#36). After the tag: brig's
install.shmovesRUNTIME_VERSIONto rc21, then the brig.sh/installpin.