Closed Bug 1059710 Opened 10 years ago Closed 10 years ago

OdinMonkey - SIMD: add support for comparisons

Categories

(Core :: JavaScript Engine: JIT, defect)

x86_64
Linux
defect
Not set
normal

Tracking

()

RESOLVED FIXED
mozilla35

People

(Reporter: bbouvier, Assigned: bbouvier)

References

Details

Attachments

(1 file, 2 obsolete files)

This is high priority for the Intel SIMD demo.
Attached patch odin-simd-comparisons.patch (obsolete) — Splinter Review
Attachment #8480494 - Flags: review?(luke)
Comment on attachment 8480494 [details] [diff] [review]
odin-simd-comparisons.patch

Review of attachment 8480494 [details] [diff] [review]:
-----------------------------------------------------------------

So the SIMD spec doesn't provide <= et al for int32x4?  Is that motivated by the hardware?

::: js/src/asmjs/AsmJSValidate.cpp
@@ +4632,5 @@
>      return true;
>  }
>  
>  static bool
> +CheckBinarySimd(FunctionCompiler &f, ParseNode *call, AsmJSSimdType simdType,

I reviewed bug 1059733 before this, but the change I suggested (un-templating CheckBinarySimd and switching on the op at the end) would I think avoid some of this code duplication as well.
Attachment #8480494 - Flags: review?(luke) → review+
Oh yes, that's way nicer! carrying forward r+.
Attachment #8481303 - Flags: review+
Attachment #8480494 - Attachment is obsolete: true
Attached patch bug1059710.patchSplinter Review
Forgot to factor the switch
Attachment #8481303 - Attachment is obsolete: true
Attachment #8481311 - Flags: review+
https://hg.mozilla.org/mozilla-central/rev/c32ca9dc0726
Status: ASSIGNED → RESOLVED
Closed: 10 years ago
Flags: in-testsuite+
Resolution: --- → FIXED
Target Milestone: --- → mozilla35
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: