Stream: git-wasmtime

Topic: wasmtime / PR #10229 aarch64: Add support for the `extr` ...


view this post on Zulip Wasmtime GitHub notifications bot (Feb 13 2025 at 21:06):

alexcrichton requested cfallin for a review on PR #10229.

view this post on Zulip Wasmtime GitHub notifications bot (Feb 13 2025 at 21:06):

alexcrichton requested wasmtime-compiler-reviewers for a review on PR #10229.

view this post on Zulip Wasmtime GitHub notifications bot (Feb 13 2025 at 21:06):

alexcrichton requested dicej for a review on PR #10229.

view this post on Zulip Wasmtime GitHub notifications bot (Feb 13 2025 at 21:06):

alexcrichton requested wasmtime-core-reviewers for a review on PR #10229.

view this post on Zulip Wasmtime GitHub notifications bot (Feb 13 2025 at 21:06):

alexcrichton opened PR #10229 from alexcrichton:a64-extr to bytecodealliance:main:

This is pattern-matched from bor patterns of a specific shape. I found this when doing some benchmarking of Wasmtime on aarch64 and I saw LLVM generating this pattern but Wasmtime didn't. I didn't perform any benchmarking between wasmtime/native though, so I'm just relying on this reducing the number of instructions to probably be a wee bit faster.

<!--
Please make sure you include the following information:

Our development process is documented in the Wasmtime book:
https://docs.wasmtime.dev/contributing-development-process.html

Please ensure all communication follows the code of conduct:
https://github.com/bytecodealliance/wasmtime/blob/main/CODE_OF_CONDUCT.md
-->

view this post on Zulip Wasmtime GitHub notifications bot (Feb 13 2025 at 21:15):

cfallin submitted PR review:

Good find, thanks!

view this post on Zulip Wasmtime GitHub notifications bot (Feb 13 2025 at 21:15):

cfallin created PR review comment:

Is it possible to build this out of the add/eq/... primitives we have? Nothing against custom ctors when needed, just curious if we can reduce our FFI surface somewhat though.

view this post on Zulip Wasmtime GitHub notifications bot (Feb 13 2025 at 21:15):

cfallin created PR review comment:

Can we write the formula in the comment too?

;; (x << xs) | (y >> ys) if (xs + ys == widthof(ty)) => extr(x, y, ys)

view this post on Zulip Wasmtime GitHub notifications bot (Feb 13 2025 at 21:15):

cfallin created PR review comment:

It looks like (at least by my reading) rotr is an alias for / special case of extr, not the other way around, so perhaps we could rename the ALUOp?

view this post on Zulip Wasmtime GitHub notifications bot (Feb 13 2025 at 22:44):

github-actions[bot] commented on PR #10229:

Subscribe to Label Action

cc @cfallin, @fitzgen

<details>
This issue or pull request has been labeled: "cranelift", "cranelift:area:aarch64", "isle"

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 (Feb 14 2025 at 00:48):

alexcrichton submitted PR review.

view this post on Zulip Wasmtime GitHub notifications bot (Feb 14 2025 at 00:48):

alexcrichton created PR review comment:

Good point, I also prefer fewer ISLE FFI functions!

view this post on Zulip Wasmtime GitHub notifications bot (Feb 14 2025 at 00:48):

alexcrichton updated PR #10229.

view this post on Zulip Wasmtime GitHub notifications bot (Feb 14 2025 at 01:05):

alexcrichton updated PR #10229.

view this post on Zulip Wasmtime GitHub notifications bot (Feb 14 2025 at 01:21):

alexcrichton has enabled auto merge for PR #10229.

view this post on Zulip Wasmtime GitHub notifications bot (Feb 14 2025 at 01:44):

alexcrichton merged PR #10229.


Last updated: Feb 28 2025 at 01:30 UTC