Skip to content

V
V8
Switch branch/tag
  1. 09 Oct, 2020 15 commits
  3. 08 Oct, 2020 22 commits
  5. 07 Oct, 2020 3 commits
    • Ng Zhi An's avatar
      [wasm-simd][scalar-lowering] Implement the rest of SimdType conversions · b87db517
      Ng Zhi An authored 
      This fills out all possible combinations of SimdType conversions. The
implementation is not the best, it uses existing conversions instead of
creating new specific ones, but is sufficient, since we don't prioritize
performance for lowering.

I added if/else clauses to the existing implementation, so that the diff
is easier to read. I will put up a follow-up to convert it into switch
statements.

Bug: chromium:1134898,v8:10507
Change-Id: I5fcb3597882b0ea3084fdc79eb1dab9b8629044b
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2453462
Commit-Queue: Zhi An Ng <zhin@chromium.org>
Reviewed-by: 's avatarBill Budge <bbudge@chromium.org>
Cr-Commit-Position: refs/heads/master@{#70393}
      b87db517
    • Ng Zhi An's avatar
      [x64] Refactor pinsrb family of instructions · e30c50f3
      Ng Zhi An authored 
      The existing macro assembler define Pinsrb, which expects 3 arguments:

- XMMRegister dst
- Register/Operand src
- uint8_t imm

which overwrites dst with src at lane specified by imm.

That means we cannot use the AVX version, which has 4 arguments, and
does not overwrite dst.

This refactoring defines the 4 argument AVX version instead, and if AVX
is not supported, fall back to the SSE version, and ensure that the
value is copied over into dst first.

For convenience, we define an overload with 3 arguments that duplicates
dst, this replicates the SSE behavior, so that not all callers have to
be updated.

Bug: v8:10975, v8:10933
Change-Id: I6f9b9d37fa08d3f5cff4f040ae7d5e1f0cf36455
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2444096
Commit-Queue: Zhi An Ng <zhin@chromium.org>
Reviewed-by: 's avatarBill Budge <bbudge@chromium.org>
Cr-Commit-Position: refs/heads/master@{#70392}
      e30c50f3
    • Ng Zhi An's avatar
      [wasm-simd] Fix lowering of load transform · f996d50d
      Ng Zhi An authored 
      Load transform nodes can be effect inputs, so we need to update the node
in place. Otherwise, if a load transform node is both an value and an
effect input, only the value input will be updated with replacements,
the effect input will still be the unlowered load transform node.
This is because DefaultLowering only lowers value inputs.

This is consistent with normal loads in the scalar lowering,
and also how int64 does it.

Bug: v8:10984,v8:10507
Change-Id: I04b0f0b1cff95e39589de979ecae6cb4b085a3cf
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2453861Reviewed-by: 's avatarBill Budge <bbudge@chromium.org>
Commit-Queue: Zhi An Ng <zhin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#70391}
      f996d50d