Stream: git-wasmtime

Topic: wasmtime / PR #4360 s390x: use full vector register file ...


view this post on Zulip Wasmtime GitHub notifications bot (Jun 30 2022 at 20:58):

uweigand opened PR #4360 from s390x-vr to main:

This defines the full set of 32 128-bit vector registers on s390x.
(Note that the VRs overlap the existing FPRs.) In addition, this
adds support to use all 32 vector registers to implement floating-
point operations, by using vector floating-point instructions with
the 'W' bit set to operate only on the first element.

This part of the vector instruction set mostly matches the old FP
instruction set, with two exceptions:

For all operations except those two above, we continue to use the
old FP instruction if applicable (i.e. if all operands happen to
have been allocated to the original FP register set), and use the
vector instruction otherwise.

CC @cfallin

<!--

Please ensure that the following steps are all taken care of before submitting
the PR.

Please ensure all communication adheres to the code of conduct.
-->

view this post on Zulip Wasmtime GitHub notifications bot (Jun 30 2022 at 22:24):

cfallin submitted PR review.

view this post on Zulip Wasmtime GitHub notifications bot (Jun 30 2022 at 22:24):

cfallin submitted PR review.

view this post on Zulip Wasmtime GitHub notifications bot (Jun 30 2022 at 22:24):

cfallin created PR review comment:

Likewise here with rule priorities.

view this post on Zulip Wasmtime GitHub notifications bot (Jun 30 2022 at 22:24):

cfallin created PR review comment:

(I do want to have a mode eventually where we randomize rule ordering just to see how many things this breaks; such a test mode would certainly catch this so we don't have to remember the principle manually!)

view this post on Zulip Wasmtime GitHub notifications bot (Jun 30 2022 at 22:24):

cfallin created PR review comment:

We need either rule priorities or an anything-but-these-types extractor here (probably the former is simplest) so that this rule is correct on its own, rather than just as a fallback from the below cases, I think?

view this post on Zulip Wasmtime GitHub notifications bot (Jun 30 2022 at 22:32):

uweigand submitted PR review.

view this post on Zulip Wasmtime GitHub notifications bot (Jun 30 2022 at 22:32):

uweigand created PR review comment:

I don't see the problem here - the first rule only handles the case where source and destination type are equal; all other rules only accept cases where source and destination type are different. Each of the rules should be correct on its own?

view this post on Zulip Wasmtime GitHub notifications bot (Jun 30 2022 at 22:43):

uweigand updated PR #4360 from s390x-vr to main.

view this post on Zulip Wasmtime GitHub notifications bot (Jun 30 2022 at 22:46):

uweigand updated PR #4360 from s390x-vr to main.

view this post on Zulip Wasmtime GitHub notifications bot (Jun 30 2022 at 23:32):

cfallin created PR review comment:

Ah! Yes, I misread it, sorry; ty ty, not ty. All's well as-is here.

view this post on Zulip Wasmtime GitHub notifications bot (Jun 30 2022 at 23:32):

cfallin submitted PR review.

view this post on Zulip Wasmtime GitHub notifications bot (Jun 30 2022 at 23:33):

cfallin submitted PR review.

view this post on Zulip Wasmtime GitHub notifications bot (Jun 30 2022 at 23:33):

cfallin merged PR #4360.


Last updated: Oct 23 2024 at 20:03 UTC