Stream: git-wasmtime

Topic: wasmtime / PR #13851 adds support for components with 64b...


view this post on Zulip Wasmtime GitHub notifications bot (Jul 09 2026 at 10:20):

christoph-dfinity opened PR #13851 from christoph-dfinity:christoph/cm-memory64 to bytecodealliance:main:

With the changes landed in wasm-tools@1.253, we can run a test that roundtrips strings or lists through 32bit and 64bit components. This PR adds said test and the CLI flags/toggles to enable support for 64 bit components.

I added the "incomplete" caveat that exists on some of the other component model toggles as there are still plenty of // FIXME(#4311): needs memory64 support comments in the code. So far I've been unable to actually trigger these faults when playing around with the test case.

Happy to put this into draft and work on these comments first, if we're worried about including this while incomplete.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 09 2026 at 10:21):

christoph-dfinity requested fitzgen for a review on PR #13851.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 09 2026 at 10:21):

christoph-dfinity requested wasmtime-fuzz-reviewers for a review on PR #13851.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 09 2026 at 10:21):

christoph-dfinity requested wasmtime-core-reviewers for a review on PR #13851.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 09 2026 at 10:23):

christoph-dfinity updated PR #13851.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 09 2026 at 13:38):

github-actions[bot] added the label fuzzing on PR #13851.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 09 2026 at 13:38):

github-actions[bot] added the label wasmtime:api on PR #13851.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 09 2026 at 13:38):

github-actions[bot] added the label wasmtime:config on PR #13851.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 09 2026 at 13:39):

github-actions[bot] commented on PR #13851:

Subscribe to Label Action

cc @fitzgen

<details>
This issue or pull request has been labeled: "fuzzing", "wasmtime:api", "wasmtime:config"

Thus the following users have been cc'd because of the following labels:

To subscribe or unsubscribe from this label, edit the <code>.github/subscribe-to-label.json</code> configuration file.

Learn more.
</details>

view this post on Zulip Wasmtime GitHub notifications bot (Jul 09 2026 at 14:37):

github-actions[bot] commented on PR #13851:

Label Messager: wasmtime:config

It looks like you are changing Wasmtime's configuration options. Make sure to
complete this check list:

[fuzzing-config]: https://github.com/bytecodealliance/wasmtime/blob/ca0e8d0a1d8cefc0496dba2f77a670571d8fdcab/crates/fuzzing/src/generators.rs#L182-L194
[fuzzing-docs]: https://docs.wasmtime.dev/contributing-fuzzing.html


<details>

To modify this label's message, edit the <code>.github/label-messager/wasmtime-config.md</code> file.

To add new label messages or remove existing label messages, edit the
<code>.github/label-messager.json</code> configuration file.

Learn more.

</details>

view this post on Zulip Wasmtime GitHub notifications bot (Jul 09 2026 at 16:40):

alexcrichton commented on PR #13851:

Thanks for this! Everything here looks good to me, but it looks like there may be some lingering CI failures (maybe just some test expectation adjustments?)

So far I've been unable to actually trigger these faults when playing around with the test case.

If you've only been using the test case in this PR that's expected because the component-to-component infrastructure doesn't go through the FIXME you're looking at. The FIXME there is for host-to-guest or guest-to-host, so for example if a WASI API was lowered with 64-bit options into a guest you'll probably see issues.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 09 2026 at 16:41):

alexcrichton edited PR #13851:

With the changes landed in wasm-tools@1.253, we can run a test that roundtrips strings or lists through 32bit and 64bit components. This PR adds said test and the CLI flags/toggles to enable support for 64 bit components.

I added the "incomplete" caveat that exists on some of the other component model toggles as there are still plenty of // FIXME(#4311): needs memory64 support comments in the code. So far I've been unable to actually trigger these faults when playing around with the test case.

Happy to put this into draft and work on these comments first, if we're worried about including this while incomplete.

cc #4311

view this post on Zulip Wasmtime GitHub notifications bot (Jul 13 2026 at 07:05):

christoph-dfinity commented on PR #13851:

Thanks for taking a look! The CI failure turns out to be https://github.com/WebAssembly/component-model/pull/676.

The FIXME there is for host-to-guest or guest-to-host, so for example if a WASI API was lowered with 64-bit options into a guest you'll probably see issues.

Ahh, I see. Thanks! I'll see about creating some tests that exercise those APIs then :) Should I do that in this PR, or would you rather land this change and have me work on a follow-up?

view this post on Zulip Wasmtime GitHub notifications bot (Jul 13 2026 at 13:46):

alexcrichton commented on PR #13851:

Thanks for tracking that down, I've merged the spec side over there now, too. If updating the submodule is onerous and/or causes other failure it's ok to leave the test ignored temporarily, too.

For host<->guest interactions I think that'll best be left for a separate PR

view this post on Zulip Wasmtime GitHub notifications bot (Jul 13 2026 at 14:20):

christoph-dfinity updated PR #13851.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 13 2026 at 14:21):

christoph-dfinity commented on PR #13851:

I tried doing the update but am getting failures in other modified/new tests (even after rebasing on main):

component-model/test/wasm-tools/resources.wast
component-model/test/async/big-interleaving-test.wast
component-model/test/values/post-return.wast
component-model/test/values/concat.wast

So I've opted to re-disable the test for now.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 13 2026 at 14:23):

:thumbs_up: alexcrichton submitted PR review.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 13 2026 at 14:23):

alexcrichton has enabled auto merge for PR #13851.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 13 2026 at 14:35):

alexcrichton has disabled auto merge for PR #13851.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 13 2026 at 14:35):

christoph-dfinity updated PR #13851.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 13 2026 at 14:58):

christoph-dfinity commented on PR #13851:

Sorry one change slipped through, I reverted that as well and now CI passes.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 13 2026 at 15:13):

alexcrichton added PR #13851 adds support for components with 64bit memories to the merge queue.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 13 2026 at 15:37):

:check: alexcrichton merged PR #13851.

view this post on Zulip Wasmtime GitHub notifications bot (Jul 13 2026 at 15:37):

alexcrichton removed PR #13851 adds support for components with 64bit memories from the merge queue.


Last updated: Jul 29 2026 at 05:03 UTC