feat(aarch64): add WHP backend for ARM64 Windows by cshung · Pull Request #1638 · hyperlight-dev/hyperlight

cshung · 2026-07-12T15:02:55Z

Summary

Implements the WHP (Windows Hypervisor Platform) hypervisor backend for aarch64, enabling hyperlight to run micro-VMs on Windows ARM64 systems.

Resolves #1544

Changes

Structural

Restructure whp.rs into whp/ directory (mod.rs + x86_64.rs) to support per-architecture implementations, matching the existing kvm/ and mshv/ pattern
Fix super::x86_64::hw_interrupts module path after directory restructure

New: `whp/aarch64.rs`

Manual FFI bindings for ARM64 WHP register names, exit reasons, and exit context layout (extracted from Windows SDK WinHvPlatformDefs.h)
Full VirtualMachine trait implementation with:
- MMIO-based I/O handling (ARM64 has no IO ports)
- Register get/set via WHvGet/SetVirtualProcessorRegisters
- Surrogate process support (same pattern as x86_64)

Integration

Wire WhpVm into hyperlight_vm/aarch64.rs for Windows platform
Add WindowsInterruptHandle for aarch64 Windows
Add CpuVendor::current() for aarch64 Windows target

Cross-compilation fix

Move vmm-sys-util to unix-only dependencies (it doesn't compile on Windows)

Verification

Verified compilation on three targets:

x86_64-pc-windows-msvc (native) — just clippy debug/release pass
aarch64-pc-windows-msvc (cross-compile) — cargo check passes
aarch64-unknown-linux-gnu (cross-compile) — cargo check --features kvm passes

Note: No runtime testing was possible — requires actual ARM64 Windows hardware with Hyper-V enabled. CI has no ARM64 Windows runners.

Why manual FFI bindings?

The windows crate (v0.62) does not expose ARM64 WHP types (register names, exit reasons, exit context structs). All definitions were extracted from the Windows SDK header WinHvPlatformDefs.h (SDK 10.0.26100.0) which has full ARM64 support behind #ifdef _ARM64_.

Copilot

Pull request overview

Adds a Windows Hypervisor Platform (WHP) backend for Windows/aarch64 so Hyperlight can run micro-VMs on ARM64 Windows systems, aligning WHP’s structure with existing per-arch hypervisor layouts.

Changes:

Introduces whp/aarch64.rs implementing the VirtualMachine trait for ARM64 WHP, including MMIO-based exit handling and register get/set via WHvGet/SetVirtualProcessorRegisters.
Restructures the WHP backend into whp/ with per-architecture modules and fixes x86_64 interrupt helper module paths.
Wires WHP into the aarch64 Hyperlight VM path on Windows and moves vmm-sys-util into unix-only dependencies to fix Windows builds.

Reviewed changes

Copilot reviewed 6 out of 6 changed files in this pull request and generated 5 comments.

Show a summary per file

File	Description
src/hyperlight_host/src/sandbox/snapshot/file/config.rs	Adds Windows/aarch64 CPU vendor token for snapshot config.
src/hyperlight_host/src/hypervisor/virtual_machine/whp/x86_64.rs	Updates `hw_interrupts` module paths after WHP directory restructure.
src/hyperlight_host/src/hypervisor/virtual_machine/whp/mod.rs	New per-arch WHP module dispatcher (`x86_64` vs `aarch64`).
src/hyperlight_host/src/hypervisor/virtual_machine/whp/aarch64.rs	New ARM64 WHP backend implementation and manual ARM64 WHP FFI bindings.
src/hyperlight_host/src/hypervisor/hyperlight_vm/aarch64.rs	Selects WHP on Windows/aarch64 and adds Windows interrupt handle wiring.
src/hyperlight_host/Cargo.toml	Makes `vmm-sys-util` unix-only to avoid Windows compilation failures.

cshung · 2026-08-05T17:48:19Z

End-to-end cargo-hyperlight validation completed on the ARM64 WHP hardware.

Using this PR's 0.16 crates, a freshly scaffolded sample successfully built aarch64-hyperlight-none guests and ran through WHP in debug and release. Host callbacks, typed calls, persistent state, snapshot, and restore passed. I also ran the release workflow 10 times with surrogate mappings and 10 times with surrogates disabled. All runs produced the expected output.

The backend works. The published scaffold is the remaining user-experience gap: cargo-hyperlight pins Hyperlight 0.15 and hardcodes the generated host's x64 guest path. Filed hyperlight-dev/cargo-hyperlight#70 with the reproduction and results.

syntactically

This looks great! I have a bunch of little nits, and also I (more majorly) think that the addition of reset_vcpu to InteruptHandle seems wrong.

simongdavies · 2026-09-30T10:14:17Z

@cshung the smoke test failed, but is there any reason for no just making this an additional job in the build-test matrix?

Add native Windows ARM64 support through WHP. Integrate debug and release coverage into the standard build-test matrix. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Copilot-Session: 6f20a05d-6bee-4e2e-b320-12f8d9759bbc Signed-off-by: cshung <3410332+cshung@users.noreply.github.com>

cshung · 2026-10-02T16:02:29Z

@simongdavies Done. Windows ARM64 now runs as debug and release entries in the standard build-test matrix using the existing ARM64 guest artifacts. Both entries have passed, and the standalone smoke workflow has been removed.

hyperlight-gh-bot · 2026-10-02T16:32:42Z

Benchmark Results

Measured commit: 827ea49647c8
Baseline commit: 43cf3539bd10

kvm / amd (Linux) (➖ stable)

No benchmark improved or regressed.

Benchmark Results

function_call_codec

	`encode_control`	`decode_vec_bytes_copy`
`byte_chunks`	`767.05 ns` (➖ 1.01x faster)
`vec_bytes`	`585.25 ns` (➖ 1.00x slower)
		`373.26 µs` (➖ 1.00x faster)

payload_allocation

	`slot_pool_segmented`
`262144`	`521.69 ns` (➖ 1.02x faster)
`65536`	`141.51 ns` (➖ 1.02x faster)

sandboxes

	`create_initialized_and_drop`
`medium`	`75.58 ms` (➖ 1.05x faster)

slot_pool

	`alloc_dealloc_1500`	`alloc_dealloc_4096`	`alloc_dealloc_128`
	`7.73 ns` (➖ 1.01x faster)	`7.72 ns` (➖ 1.00x slower)	`7.72 ns` (➖ 1.00x faster)

snapshot_files

	`load_snapshot_unverified`
`small`	`92.53 µs` (➖ 1.00x slower)

virtq_readonly

	`slot_pool_segmented_fragmented`	`slot_pool_segmented`
`262144`	`7.14 µs` (➖ 1.01x faster)	`7.24 µs` (➖ 1.01x faster)
`65536`	`2.07 µs` (➖ 1.01x slower)	`1.92 µs` (➖ 1.00x faster)

virtq_readwrite

	`slot_pool_segmented_fragmented`	`slot_pool_segmented`
`65536`	`6.26 µs` (➖ 1.01x faster)	`6.27 µs` (➖ 1.02x faster)
`8192`	`1.07 µs` (➖ 1.01x slower)	`1.07 µs` (➖ 1.01x slower)
`262144`		`27.00 µs` (➖ 1.01x faster)

kvm / intel (Linux) (➖ stable)

No benchmark improved or regressed.

Benchmark Results

function_call_codec

	`encode_control`	`decode_vec_bytes_copy`
`byte_chunks`	`697.88 ns` (➖ 1.08x faster)
`vec_bytes`	`504.38 ns` (➖ 1.08x faster)
		`658.72 µs` (➖ 1.05x faster)

payload_allocation

	`slot_pool_segmented`
`262144`	`505.77 ns` (➖ 1.13x faster)
`65536`	`136.35 ns` (➖ 1.13x faster)

sandboxes

	`create_initialized_and_drop`
`medium`	`75.16 ms` (➖ 1.06x faster)

slot_pool

	`alloc_dealloc_1500`	`alloc_dealloc_4096`	`alloc_dealloc_128`
	`7.02 ns` (➖ 1.07x faster)	`6.93 ns` (➖ 1.09x faster)	`6.99 ns` (➖ 1.09x faster)

snapshot_files

	`load_snapshot_unverified`
`small`	`45.61 µs` (➖ 1.08x faster)

virtq_readonly

	`slot_pool_segmented_fragmented`	`slot_pool_segmented`
`262144`	`7.53 µs` (➖ 1.13x faster)	`7.59 µs` (➖ 1.13x faster)
`65536`	`2.12 µs` (➖ 1.13x faster)	`2.15 µs` (➖ 1.11x faster)

virtq_readwrite

	`slot_pool_segmented_fragmented`	`slot_pool_segmented`
`65536`	`7.06 µs` (➖ 1.12x faster)	`7.02 µs` (➖ 1.13x faster)
`8192`	`748.89 ns` (➖ 1.12x faster)	`722.89 ns` (➖ 1.09x faster)
`262144`		`29.88 µs` (➖ 1.11x faster)

mshv3 / amd (Linux) (➖ stable)

No benchmark improved or regressed.

Benchmark Results

function_call_codec

	`encode_control`	`decode_vec_bytes_copy`
`byte_chunks`	`989.43 ns` (➖ 1.03x slower)
`vec_bytes`	`718.24 ns` (➖ 1.00x slower)
		`305.32 µs` (➖ 1.14x faster)

payload_allocation

	`slot_pool_segmented`
`262144`	`843.27 ns` (➖ 1.14x slower)
`65536`	`197.67 ns` (➖ 1.00x faster)

sandboxes

	`create_initialized_and_drop`
`medium`	`57.04 ms` (➖ 1.03x slower)

slot_pool

	`alloc_dealloc_1500`	`alloc_dealloc_4096`	`alloc_dealloc_128`
	`9.79 ns` (➖ 1.02x slower)	`9.81 ns` (➖ 1.01x slower)	`9.82 ns` (➖ 1.03x faster)

snapshot_files

	`load_snapshot_unverified`
`small`	`83.90 µs` (➖ 1.00x faster)

virtq_readonly

	`slot_pool_segmented_fragmented`	`slot_pool_segmented`
`262144`	`9.63 µs` (➖ 1.03x faster)	`8.86 µs` (➖ 1.01x faster)
`65536`	`2.25 µs` (➖ 1.00x faster)	`2.45 µs` (➖ 1.01x faster)

virtq_readwrite

	`slot_pool_segmented_fragmented`	`slot_pool_segmented`
`65536`	`8.11 µs` (➖ 1.01x faster)	`8.12 µs` (➖ 1.01x slower)
`8192`	`1.28 µs` (➖ 1.00x slower)	`1.41 µs` (➖ 1.05x slower)
`262144`		`36.73 µs` (➖ 1.01x slower)

mshv3 / intel (Linux) (➖ stable)

No benchmark improved or regressed.

Benchmark Results

function_call_codec

	`encode_control`	`decode_vec_bytes_copy`
`byte_chunks`	`944.75 ns` (➖ 1.01x slower)
`vec_bytes`	`662.76 ns` (➖ 1.00x faster)
		`740.41 µs` (➖ 1.13x slower)

payload_allocation

	`slot_pool_segmented`
`262144`	`630.05 ns` (➖ 1.02x faster)
`65536`	`166.18 ns` (➖ 1.01x slower)

sandboxes

	`create_initialized_and_drop`
`medium`	`77.72 ms` (➖ 1.25x slower)

slot_pool

	`alloc_dealloc_1500`	`alloc_dealloc_4096`	`alloc_dealloc_128`
	`8.80 ns` (➖ 1.00x slower)	`8.78 ns` (➖ 1.00x slower)	`8.81 ns` (➖ 1.00x slower)

snapshot_files

	`load_snapshot_unverified`
`small`	`47.12 µs` (➖ 1.06x slower)

virtq_readonly

	`slot_pool_segmented_fragmented`	`slot_pool_segmented`
`262144`	`7.80 µs` (➖ 1.01x slower)	`8.04 µs` (➖ 1.00x slower)
`65536`	`2.28 µs` (➖ 1.02x slower)	`2.26 µs` (➖ 1.00x faster)

virtq_readwrite

	`slot_pool_segmented_fragmented`	`slot_pool_segmented`
`65536`	`7.62 µs` (➖ 1.01x slower)	`7.59 µs` (➖ 1.00x slower)
`8192`	`880.18 ns` (➖ 1.03x slower)	`885.05 ns` (➖ 1.00x slower)
`262144`		`39.18 µs` (➖ 1.03x slower)

hyperv-ws2025 / amd (Windows) (➖ stable)

No benchmark improved or regressed.

Benchmark Results

function_call_codec

	`encode_control`	`decode_vec_bytes_copy`
`byte_chunks`	`1.13 µs` (➖ 1.01x slower)
`vec_bytes`	`754.20 ns` (➖ 1.26x faster)
		`2.12 ms` (➖ 1.18x faster)

payload_allocation

	`slot_pool_segmented`
`262144`	`783.70 ns` (➖ 1.03x faster)
`65536`	`228.90 ns` (➖ 1.00x slower)

sandboxes

	`create_initialized_and_drop`
`medium`	`90.34 ms` (➖ 1.22x faster)

slot_pool

	`alloc_dealloc_1500`	`alloc_dealloc_4096`	`alloc_dealloc_128`
	`10.62 ns` (➖ 1.30x faster)	`11.03 ns` (➖ 1.02x slower)	`10.55 ns` (➖ 1.00x slower)

snapshot_files

	`load_snapshot_unverified`
`small`	`666.37 µs` (➖ 1.18x faster)

virtq_readonly

	`slot_pool_segmented_fragmented`	`slot_pool_segmented`
`262144`	`8.70 µs` (➖ 1.01x slower)	`9.36 µs` (➖ 1.00x slower)
`65536`	`2.55 µs` (➖ 1.01x slower)	`2.52 µs` (➖ 1.07x faster)

virtq_readwrite

	`slot_pool_segmented_fragmented`	`slot_pool_segmented`
`65536`	`8.62 µs` (➖ 1.01x faster)	`8.99 µs` (➖ 1.02x slower)
`8192`	`1.43 µs` (➖ 1.00x faster)	`1.36 µs` (➖ 1.03x faster)
`262144`		`39.36 µs` (➖ 1.09x slower)

hyperv-ws2025 / intel (Windows) (➖ stable)

No benchmark improved or regressed.

Benchmark Results

function_call_codec

	`encode_control`	`decode_vec_bytes_copy`
`byte_chunks`	`1.30 µs` (➖ 1.08x slower)
`vec_bytes`	`890.56 ns` (➖ 1.15x slower)
		`3.45 ms` (➖ 1.11x slower)

payload_allocation

	`slot_pool_segmented`
`262144`	`741.01 ns` (➖ 1.01x slower)
`65536`	`212.07 ns` (➖ 1.00x faster)

sandboxes

	`create_initialized_and_drop`
`medium`	`112.57 ms` (➖ 1.07x faster)

slot_pool

	`alloc_dealloc_1500`	`alloc_dealloc_4096`	`alloc_dealloc_128`
	`10.09 ns` (➖ 1.01x slower)	`10.16 ns` (➖ 1.01x slower)	`10.58 ns` (➖ 1.00x faster)

snapshot_files

	`load_snapshot_unverified`
`small`	`622.62 µs` (➖ 1.02x slower)

virtq_readonly

	`slot_pool_segmented_fragmented`	`slot_pool_segmented`
`262144`	`7.87 µs` (➖ 1.04x slower)	`7.69 µs` (➖ 1.01x faster)
`65536`	`2.34 µs` (➖ 1.03x slower)	`2.27 µs` (➖ 1.02x faster)

virtq_readwrite

	`slot_pool_segmented_fragmented`	`slot_pool_segmented`
`65536`	`8.52 µs` (➖ 1.02x slower)	`8.36 µs` (➖ 1.01x faster)
`8192`	`1.18 µs` (➖ 1.08x faster)	`1.13 µs` (➖ 1.02x slower)
`262144`		`59.48 µs` (➖ 1.40x slower)

_{Reported by cargo ci bench-report --candidate run:37031177925 --baseline run:36945564806 --config-file bench_report.toml.}

Copilot AI review requested due to automatic review settings July 12, 2026 15:02

cshung added kind/enhancement For PRs adding features, improving functionality, docs, tests, etc. ready-for-review PR is ready for (re-)review labels Jul 12, 2026

cshung requested review from andreiltd, danbugs, dblnz, devigned, jprendes, jsturtevant, ludfjig, simongdavies, squillace and syntactically as code owners July 12, 2026 15:02

Copilot started reviewing on behalf of cshung July 12, 2026 15:06 View session

Copilot AI reviewed Jul 12, 2026

View reviewed changes

cshung marked this pull request as draft July 13, 2026 02:13

github-actions Bot removed the ready-for-review PR is ready for (re-)review label Jul 13, 2026

syntactically reviewed Jul 15, 2026

View reviewed changes

Comment thread src/hyperlight_host/src/sandbox/snapshot/file/config.rs Outdated

cshung force-pushed the cshung/whp-aarch64 branch from 5f992dc to 1d3a299 Compare July 30, 2026 22:14

jsturtevant mentioned this pull request Aug 5, 2026

cargo hyperlight aarch64 support #1546

Closed

2 tasks done

syntactically requested changes Aug 5, 2026

View reviewed changes

cshung force-pushed the cshung/whp-aarch64 branch 5 times, most recently from 7e93813 to 4c2f4e4 Compare August 18, 2026 23:59

cshung force-pushed the cshung/whp-aarch64 branch 3 times, most recently from 404a655 to 189317a Compare August 19, 2026 15:45

This comment has been minimized.

Sign in to view

cshung marked this pull request as ready for review September 29, 2026 17:37

This comment has been minimized.

Sign in to view

cshung force-pushed the cshung/whp-aarch64 branch from 4d08618 to 257930a Compare October 1, 2026 14:54

This comment has been minimized.

Sign in to view

cshung force-pushed the cshung/whp-aarch64 branch from 1d092ab to d1f0697 Compare October 1, 2026 22:53

This comment has been minimized.

Sign in to view

cshung force-pushed the cshung/whp-aarch64 branch from d1f0697 to 827ea49 Compare October 2, 2026 16:02

Conversation

cshung commented Jul 12, 2026

Summary

Changes

Structural

New: whp/aarch64.rs

Integration

Cross-compilation fix

Verification

Why manual FFI bindings?

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

cshung commented Aug 5, 2026

Uh oh!

syntactically left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

simongdavies commented Sep 30, 2026

Uh oh!

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

cshung commented Oct 2, 2026

Uh oh!

hyperlight-gh-bot Bot commented Oct 2, 2026

Benchmark Results

Benchmark Results

function_call_codec

payload_allocation

sandboxes

slot_pool

snapshot_files

virtq_readonly

virtq_readwrite

Benchmark Results

function_call_codec

payload_allocation

sandboxes

slot_pool

snapshot_files

virtq_readonly

New: `whp/aarch64.rs`