drivers: hv: clear halt when injecting SNP interrupts

[Brian-Perkins]

drivers: hv: better management of the VLT0 HLT/IDLE state. Need to track this properly in order to choose the proper behavior for VTL2 idle behavior (e.g. should we transition to VTL0 or is VTL0 in a halt state).

Signed-off-by: Brian Perkins bperkins@microsoft.com

[Copilot]

Pull request overview

This PR fixes SEV-SNP interrupt wakeup behavior by ensuring a vCPU previously marked as halted is un-halted when pending interrupts are observed in the Secure AVIC (vAPIC) IRR, allowing VTL0 to resume execution instead of remaining in native_safe_halt().

Changes:

• Add an SNP-specific helper to clear HALT_* offload flags (and MSHV_VTL_RUN_FLAG_HALTED when applicable) when IRR indicates a pending interrupt.
• Invoke that helper after pulling proxy_irr into the SNP Secure AVIC page in the interrupt-injection fast path.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated no new comments.

[Copilot]

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated 1 comment.

[Copilot]

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated no new comments.

[namancse]

Hi @Brian-Perkins, could you please elaborate the commit messages of the commits, and reorganize them if required. With ever increasing OOT changes in the kernel, the goal is now to do strict scrutiny for any patches that go in and merge them only when they are absolutely clear on what they are doing, are bug free and are as close to upstream kernel guidelines as possible. It is also going to help track the OOT changes for upstreaming later.

Addressing copilot comments from UI seems to be creating new commits, which we can avoid. I have also noticed that copilot sometimes does not give all the comments at once, so we may have to go through multiple iterations.

[Brian-Perkins]

Hi @Brian-Perkins, could you please elaborate the commit messages of the commits, and reorganize them if required. With ever increasing OOT changes in the kernel, the goal is now to do strict scrutiny for any patches that go in and merge them only when they are absolutely clear on what they are doing, are bug free and are as close to upstream kernel guidelines as possible. It is also going to help track the OOT changes for upstreaming later.

Addressing copilot comments from UI seems to be creating new commits, which we can avoid. I have also noticed that copilot sometimes does not give all the comments at once, so we may have to go through multiple iterations.

Done. Note in other projects we do this by using squash merge so the description is what appears in the PR and the commits are just temporary while the PR is in progress.