elliottt opened PR #7520 from elliottt:trevor/pcc-xmm-load-fix
to bytecodealliance:main
:
Loads from memory to xmm registers using
vmovss
andvmovsd
don't move a full 128 bits, and instead move either 32 or 64 depending on the instruction. This PR changes the current behavior by special casing these two and adjusting the size of the load, correcting the checks for floating point loads. This doesn't fix the problem generally, as this pr only fixes the Vex case, but it's a step towards stabilization.Co-Authored-By: Nick Fitzgerald <fitzgen@gmail.com>
<!--
Please make sure you include the following information:
If this work has been discussed elsewhere, please include a link to that
conversation. If it was discussed in an issue, just mention "issue #...".Explain why this change is needed. If the details are in an issue already,
this can be brief.Our development process is documented in the Wasmtime book:
https://docs.wasmtime.dev/contributing-development-process.htmlPlease ensure all communication follows the code of conduct:
https://github.com/bytecodealliance/wasmtime/blob/main/CODE_OF_CONDUCT.md
-->
elliottt requested fitzgen for a review on PR #7520.
elliottt requested wasmtime-compiler-reviewers for a review on PR #7520.
fitzgen submitted PR review.
fitzgen merged PR #7520.
Last updated: Dec 23 2024 at 12:05 UTC