Save sysroot and Coverage caches only on main by ludfjig · Pull Request #1913 · hyperlight-dev/hyperlight

ludfjig · 2026-10-09T00:18:42Z

PR runs save the sysroot and Coverage caches under their PR ref. Only that PR can restore them, but they use the shared 10 GB cache. Right now they take 1.65 GiB.

On main, the code-checks sysroot cache is saved empty. actions/cache saves in a post-job step, which runs after rust-cache's post-job cleanup has deleted the sysroot.

This PR saves both caches only on main. The sysroot is saved by an explicit step at the end of the job, before rust-cache's cleanup.

PR runs saved these caches under their own ref, where no other run can restore them, using shared cache space. Saving the sysroot in a job step also runs before rust-cache's post-job cleanup, which emptied the sysroot before actions/cache saved it on main. Signed-off-by: Ludvig Liljenberg <4257730+ludfjig@users.noreply.github.com>

Copilot

🟢 Approval recommended

The cache conditions and explicit save ordering correctly implement the stated policy.

0 open findings

What changed in this PR

Restricts CI cache writes to main and preserves generated sysroots before cleanup.

Changes:

Splits sysroot cache restore and save operations.
Disables Rust cache saves outside main.

File	Description
`.github/workflows/dep_code_checks.yml`	Saves Linux and Windows sysroots explicitly on `main`.
`.github/workflows/dep_build_guests.yml`	Saves guest sysroots explicitly on `main`.
`.github/workflows/Coverage.yml`	Restricts coverage Rust cache writes to `main`.

🧠 Review effort: Balanced

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

hyperlight-gh-bot · 2026-10-09T01:05:31Z

Benchmark Results

Measured commit: 6518a19cd6d7
Baseline commit: 164728cf08d0

kvm / amd (Linux) (➖ stable)

No benchmark improved or regressed.

Benchmark Results

function_call_codec

	`encode_control`	`decode_vec_bytes_copy`
`byte_chunks`	`796.71 ns` (➖ 1.04x slower)
`vec_bytes`	`586.43 ns` (➖ 1.01x slower)
		`379.72 µs` (➖ 1.02x faster)

payload_allocation

	`slot_pool_segmented`
`262144`	`525.36 ns` (➖ 1.00x slower)
`65536`	`142.36 ns` (➖ 1.00x slower)

sandboxes

	`create_initialized_and_drop`
`medium`	`81.83 ms` (➖ 1.02x faster)

slot_pool

	`alloc_dealloc_1500`	`alloc_dealloc_4096`	`alloc_dealloc_128`
	`7.75 ns` (➖ 1.01x faster)	`7.72 ns` (➖ 1.00x faster)	`7.78 ns` (➖ 1.00x slower)

snapshot_files

	`load_snapshot_unverified`
`small`	`93.23 µs` (➖ 1.01x faster)

virtq_readonly

	`slot_pool_segmented_fragmented`	`slot_pool_segmented`
`262144`	`7.46 µs` (➖ 1.02x slower)	`7.22 µs` (➖ 1.01x slower)
`65536`	`2.14 µs` (➖ 1.10x slower)	`1.92 µs` (➖ 1.09x faster)

virtq_readwrite

	`slot_pool_segmented_fragmented`	`slot_pool_segmented`
`65536`	`6.25 µs` (➖ 1.02x faster)	`6.23 µs` (➖ 1.00x faster)
`8192`	`1.07 µs` (➖ 1.01x slower)	`1.07 µs` (➖ 1.01x faster)
`262144`		`28.42 µs` (➖ 1.02x slower)

kvm / intel (Linux) (➖ stable)

No benchmark improved or regressed.

Benchmark Results

function_call_codec

	`encode_control`	`decode_vec_bytes_copy`
`byte_chunks`	`719.79 ns` (➖ 1.04x slower)
`vec_bytes`	`519.68 ns` (➖ 1.01x faster)
		`686.63 µs` (➖ 1.01x slower)

payload_allocation

	`slot_pool_segmented`
`262144`	`506.39 ns` (➖ 1.00x slower)
`65536`	`136.59 ns` (➖ 1.00x faster)

sandboxes

	`create_initialized_and_drop`
`medium`	`75.62 ms` (➖ 1.00x slower)

slot_pool

	`alloc_dealloc_1500`	`alloc_dealloc_4096`	`alloc_dealloc_128`
	`6.97 ns` (➖ 1.03x faster)	`6.96 ns` (➖ 1.01x faster)	`6.98 ns` (➖ 1.01x faster)

snapshot_files

	`load_snapshot_unverified`
`small`	`46.14 µs` (➖ 1.00x faster)

virtq_readonly

	`slot_pool_segmented_fragmented`	`slot_pool_segmented`
`262144`	`7.66 µs` (➖ 1.02x slower)	`7.69 µs` (➖ 1.02x slower)
`65536`	`2.17 µs` (➖ 1.02x slower)	`2.17 µs` (➖ 1.03x slower)

virtq_readwrite

	`slot_pool_segmented_fragmented`	`slot_pool_segmented`
`65536`	`7.15 µs` (➖ 1.01x slower)	`7.23 µs` (➖ 1.01x slower)
`8192`	`742.54 ns` (➖ 1.02x faster)	`736.51 ns` (➖ 1.02x slower)
`262144`		`29.46 µs` (➖ 1.01x slower)

mshv3 / amd (Linux) (➖ stable)

No benchmark improved or regressed.

Benchmark Results

function_call_codec

	`encode_control`	`decode_vec_bytes_copy`
`byte_chunks`	`1.01 µs` (➖ 1.08x slower)
`vec_bytes`	`702.58 ns` (➖ 1.01x slower)
		`298.30 µs` (➖ 1.02x faster)

payload_allocation

	`slot_pool_segmented`
`262144`	`727.63 ns` (➖ 1.00x slower)
`65536`	`197.09 ns` (➖ 1.01x slower)

sandboxes

	`create_initialized_and_drop`
`medium`	`58.15 ms` (➖ 1.01x faster)

slot_pool

	`alloc_dealloc_1500`	`alloc_dealloc_4096`	`alloc_dealloc_128`
	`10.32 ns` (➖ 1.06x slower)	`9.76 ns` (➖ 1.04x faster)	`9.71 ns` (➖ 1.00x slower)

snapshot_files

	`load_snapshot_unverified`
`small`	`86.78 µs` (➖ 1.02x faster)

virtq_readonly

	`slot_pool_segmented_fragmented`	`slot_pool_segmented`
`262144`	`10.15 µs` (➖ 1.01x faster)	`9.76 µs` (➖ 1.02x faster)
`65536`	`2.45 µs` (➖ 1.01x faster)	`2.28 µs` (➖ 1.01x slower)

virtq_readwrite

	`slot_pool_segmented_fragmented`	`slot_pool_segmented`
`65536`	`8.48 µs` (➖ 1.19x faster)	`8.60 µs` (➖ 1.11x faster)
`8192`	`1.25 µs` (➖ 1.02x faster)	`1.29 µs` (➖ 1.05x slower)
`262144`		`36.32 µs` (➖ 1.04x faster)

mshv3 / intel (Linux) (➖ stable)

No benchmark improved or regressed.

Benchmark Results

function_call_codec

	`encode_control`	`decode_vec_bytes_copy`
`byte_chunks`	`942.23 ns` (➖ 1.03x slower)
`vec_bytes`	`659.91 ns` (➖ 1.00x slower)
		`748.98 µs` (➖ 1.06x slower)

payload_allocation

	`slot_pool_segmented`
`262144`	`627.59 ns` (➖ 1.01x slower)
`65536`	`169.53 ns` (➖ 1.01x slower)

sandboxes

	`create_initialized_and_drop`
`medium`	`60.15 ms` (➖ 1.07x faster)

slot_pool

	`alloc_dealloc_1500`	`alloc_dealloc_4096`	`alloc_dealloc_128`
	`9.60 ns` (➖ 1.01x faster)	`8.75 ns` (➖ 1.01x faster)	`8.80 ns` (➖ 1.00x slower)

snapshot_files

	`load_snapshot_unverified`
`small`	`43.75 µs` (➖ 1.00x slower)

virtq_readonly

	`slot_pool_segmented_fragmented`	`slot_pool_segmented`
`262144`	`7.92 µs` (➖ 1.00x faster)	`8.01 µs` (➖ 1.01x slower)
`65536`	`2.26 µs` (➖ 1.01x faster)	`2.30 µs` (➖ 1.02x slower)

virtq_readwrite

	`slot_pool_segmented_fragmented`	`slot_pool_segmented`
`65536`	`7.79 µs` (➖ 1.04x slower)	`7.45 µs` (➖ 1.01x faster)
`8192`	`909.85 ns` (➖ 1.04x faster)	`928.23 ns` (➖ 1.02x slower)
`262144`		`38.89 µs` (➖ 1.05x faster)

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

No benchmark improved or regressed.

Benchmark Results

function_call_codec

	`encode_control`	`decode_vec_bytes_copy`
`byte_chunks`	`1.33 µs` (➖ 1.12x slower)
`vec_bytes`	`789.57 ns` (➖ 1.00x slower)
		`1.99 ms` (➖ 1.14x faster)

payload_allocation

	`slot_pool_segmented`
`262144`	`800.58 ns` (➖ 1.01x slower)
`65536`	`232.11 ns` (➖ 1.00x faster)

sandboxes

	`create_initialized_and_drop`
`medium`	`72.42 ms` (➖ 1.06x faster)

slot_pool

	`alloc_dealloc_1500`	`alloc_dealloc_4096`	`alloc_dealloc_128`
	`9.94 ns` (➖ 1.02x faster)	`10.11 ns` (➖ 1.01x slower)	`10.18 ns` (➖ 1.01x slower)

snapshot_files

	`load_snapshot_unverified`
`small`	`647.69 µs` (➖ 1.11x slower)

virtq_readonly

	`slot_pool_segmented_fragmented`	`slot_pool_segmented`
`262144`	`9.11 µs` (➖ 1.02x faster)	`9.13 µs` (➖ 1.00x slower)
`65536`	`2.37 µs` (➖ 1.01x faster)	`2.34 µs` (➖ 1.01x faster)

virtq_readwrite

	`slot_pool_segmented_fragmented`	`slot_pool_segmented`
`65536`	`9.04 µs` (➖ 1.06x faster)	`9.35 µs` (➖ 1.07x slower)
`8192`	`1.32 µs` (➖ 1.01x slower)	`1.30 µs` (➖ 1.01x slower)
`262144`		`38.24 µs` (➖ 1.00x faster)

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

No benchmark improved or regressed.

Benchmark Results

function_call_codec

	`encode_control`	`decode_vec_bytes_copy`
`byte_chunks`	`1.22 µs` (➖ 1.01x slower)
`vec_bytes`	`790.07 ns` (➖ 1.00x faster)
		`3.19 ms` (➖ 1.03x slower)

payload_allocation

	`slot_pool_segmented`
`262144`	`740.22 ns` (➖ 1.37x faster)
`65536`	`211.01 ns` (➖ 1.31x faster)

sandboxes

	`create_initialized_and_drop`
`medium`	`104.82 ms` (➖ 1.04x faster)

slot_pool

	`alloc_dealloc_1500`	`alloc_dealloc_4096`	`alloc_dealloc_128`
	`10.27 ns` (➖ 1.40x faster)	`10.03 ns` (➖ 1.42x faster)	`10.55 ns` (➖ 1.36x faster)

snapshot_files

	`load_snapshot_unverified`
`small`	`662.94 µs` (➖ 1.12x slower)

virtq_readonly

	`slot_pool_segmented_fragmented`	`slot_pool_segmented`
`262144`	`7.80 µs` (➖ 1.01x faster)	`7.78 µs` (➖ 1.01x faster)
`65536`	`2.34 µs` (➖ 1.00x slower)	`2.33 µs` (➖ 1.01x faster)

virtq_readwrite

	`slot_pool_segmented_fragmented`	`slot_pool_segmented`
`65536`	`8.51 µs` (➖ 1.03x faster)	`8.49 µs` (➖ 1.02x faster)
`8192`	`1.13 µs` (➖ 1.15x faster)	`1.12 µs` (➖ 1.01x slower)
`262144`		`42.25 µs` (➖ 1.01x faster)

_{Reported by cargo ci bench-report --candidate run:37864121863 --config-file bench_report.toml.}

Copilot AI balanced review requested due to automatic review settings October 9, 2026 00:18

ludfjig added kind/bugfix For PRs that fix bugs ready-for-review PR is ready for (re-)review labels Oct 9, 2026

ludfjig requested review from andreiltd, danbugs, dblnz, devigned, jprendes, jsturtevant, simongdavies, squillace and syntactically as code owners October 9, 2026 00:18

Copilot started reviewing on behalf of ludfjig October 9, 2026 00:23 View session

Copilot AI reviewed Oct 9, 2026

View reviewed changes

jprendes approved these changes Oct 9, 2026

View reviewed changes

ludfjig merged commit dcb53c0 into hyperlight-dev:main Oct 9, 2026
66 checks passed

github-actions Bot removed the ready-for-review PR is ready for (re-)review label Oct 9, 2026

Conversation

ludfjig commented Oct 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

🟢 Approval recommended

Uh oh!

hyperlight-gh-bot Bot commented Oct 9, 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

virtq_readwrite

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

virtq_readwrite

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

virtq_readwrite

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

ludfjig commented Oct 9, 2026 •

edited

Loading