[WASM SIMD] Implement primitive shuffles.

[WASM SIMD] Implement primitive shuffles.

- Adds unary Reverse shuffles (swizzles): S32x2Reverse, S16x4Reverse,
  S16x2Reverse, S8x8Reverse, S8x4Reverse, S8x2Reverse. Reversals are
  done within the sub-vectors that prefix the opcode name, e.g. S8x2
  reverses the 8 consecutive pairs in an S8x16 vector.

- Adds binary Zip (interleave) left and right half-shuffles to return a
  single vector: S32x4ZipLeft, S32x4ZipRightS16x8ZipLeft, S16x8ZipRight,
  S8x16ZipLeft, S8x16ZipRight.

- Adds binary Unzip (de-interleave) left and right half shuffles to return
  a single vector: S32x4UnzipLeft, S32x4UnzipRight, S16x8UnzipLeft,
  S16x8UnzipRight, S8x16UnzipLeft, S8x16UnzipRight.

- Adds binary Transpose left and right half shuffles to return
  a single vector: S32x4TransposeLeft, S32x4TransposeRight,
  S16x8TransposeLeft, S16xTransposeRight, S8x16TransposeLeft,
  S8x16TransposeRight.

- Adds binary Concat (concatenate) byte shuffle: S8x16Concat #bytes to
  paste two vectors together.

LOG=N
BUG=v8:6020
Review-Url: https://codereview.chromium.org/2801183002
Cr-Commit-Position: refs/heads/master@{#44734}
Committed: https://chromium.googlesource.com/v8/v8/+/5806d86208c3845b94c8da6953830d428721c112

[bbudge, 2017-04-07 00:04:57]

Description was changed from

==========
SM SIMD] Implement primitive permutations.

- Implements these opcodes for ARM.

LOG=N
BUG=v8:6020
==========

to

==========
[WASM SIMD] Implement primitive shuffles.

- Adds unary Reverse shuffles (swizzles): S64x2Reverse (swap halves),
  S32x2Reverse, S16x4Reverse, S16x2Reverse, S8x8Reverse, S8x4Reverse,
  S8x2Reverse. Reversals are done within the sub-vectors that prefix the
  opcode name, e.g. S8x2 reverses the 8 consecutive pairs in an S8x16
  vector.

- Adds binary Zip (interleave) left and right half-shuffles to return a
  single vector: S32x4ZipLeft, S32x4ZipRightS16x8ZipLeft, S16x8ZipRight,
  S8x16ZipLeft, S8x16ZipRight.

- Adds binary Unzip (de-interleave) left and right half shuffles to return
  a single vector: S32x4UnzipLeft, S32x4UnzipRight, S16x8UnzipLeft,
  S16x8UnzipRight, S8x16UnzipLeft, S8x16UnzipRight.

- Adds binary Transpose left and right half shuffles to return
  a single vector: S32x4TransposeLeft, S32x4TransposeRight,
  S16x8TransposeLeft, S16xTransposeRight, S8x16TransposeLeft,
  S8x16TransposeRight.

- Adds binary Concat (concatenate) byte shuffle: S8x16Concat #bytes to
  paste two vectors together.

LOG=N
BUG=v8:6020
==========

[bbudge, 2017-04-07 12:41:19]

The CQ bit was checked by bbudge@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2017-04-07 12:41:31]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-04-07 12:43:01]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2017-04-07 12:43:02]

Dry run: Try jobs failed on following builders:
  v8_android_arm_compile_rel on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_android_arm_compile_rel/...)
  v8_linux64_asan_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux64_asan_rel_ng/buil...)
  v8_linux64_gyp_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux64_gyp_rel_ng/build...)
  v8_linux64_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux64_rel_ng/builds/24114)
  v8_linux_arm64_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_arm64_rel_ng/build...)
  v8_linux_arm_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_arm_rel_ng/builds/...)
  v8_presubmit on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_presubmit/builds/38717)

[bbudge, 2017-04-07 12:54:51]

Patchset #5 (id:80001) has been deleted

[bbudge, 2017-04-07 12:58:35]

The CQ bit was checked by bbudge@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2017-04-07 12:58:41]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-04-07 13:01:31]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2017-04-07 13:01:32]

Dry run: Try jobs failed on following builders:
  v8_linux64_gyp_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux64_gyp_rel_ng/build...)
  v8_linux64_verify_csa_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux64_verify_csa_rel_n...)
  v8_linux_gcc_compile_rel on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_gcc_compile_rel/bu...)
  v8_win_compile_dbg on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_win_compile_dbg/builds/3...)

[bbudge, 2017-04-07 13:22:19]

The CQ bit was checked by bbudge@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2017-04-07 13:22:28]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-04-07 13:25:58]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2017-04-07 13:25:59]

Dry run: Try jobs failed on following builders:
  v8_linux64_gyp_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux64_gyp_rel_ng/build...)
  v8_linux_mips64el_compile_rel on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_mips64el_compile_r...)
  v8_linux_mipsel_compile_rel on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_mipsel_compile_rel...)
  v8_mac_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_mac_rel_ng/builds/20603)
  v8_win_compile_dbg on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_win_compile_dbg/builds/3...)
  v8_win_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_win_rel_ng/builds/25798)

[bbudge, 2017-04-07 14:23:01]

The CQ bit was checked by bbudge@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2017-04-07 14:23:11]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-04-07 14:29:43]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2017-04-07 14:29:43]

Dry run: Try jobs failed on following builders:
  v8_android_arm_compile_rel on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_android_arm_compile_rel/...)

[bbudge, 2017-04-07 14:39:05]

The CQ bit was checked by bbudge@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2017-04-07 14:39:10]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-04-07 15:06:58]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2017-04-07 15:06:59]

Dry run: This issue passed the CQ dry run.

[bbudge, 2017-04-07 16:30:51]

bbudge@chromium.org changed reviewers:
+ bmeurer@chromium.org, bradnelson@chromium.org, martyn.capewell@arm.com

[bbudge, 2017-04-07 16:30:52]

+Martyn for ARM specific stuff.
+Benedikt for compiler and anything else you want.
+Brad for wasm and SIMD permutation opcodes design.

This CL depends on (and includes) https://codereview.chromium.org/2797923006/

which introduces the D-register versions of vzip, vuzp, and vtrn instructions
for ARM.

[martyn.capewell, 2017-04-10 16:46:56]

My replacement code generation sequences are untested, so I could've got them
wrong, or misinterpreted the intended operation.

https://codereview.chromium.org/2801183002/diff/160001/src/compiler/arm/code-...
File src/compiler/arm/code-generator-arm.cc (right):

https://codereview.chromium.org/2801183002/diff/160001/src/compiler/arm/code-...
src/compiler/arm/code-generator-arm.cc:2031: case kArmS32x4ZipLeft: {
It might be helpful to add a comment to each of these, expressing the
transformation. For example, this is 3210 -> 1100, assuming lowest index lane on
the right. Though maybe your naming implies that lowest index is on the left.

https://codereview.chromium.org/2801183002/diff/160001/src/compiler/arm/code-...
src/compiler/arm/code-generator-arm.cc:2035: __ vmov(dst.high(), src1.low());
If the assembler supports it, you could use:
  vdup.32 dst.high, src.low[1]
  vdup.32 dst.low, src.low[0]

This removes the constraint on dst == src, as do the rest below.

https://codereview.chromium.org/2801183002/diff/160001/src/compiler/arm/code-...
src/compiler/arm/code-generator-arm.cc:2039: case kArmS32x4ZipRight: {
  vdup.32 dst.low, src.high[0]
  vdup.32 dst.high, src.high[1]

https://codereview.chromium.org/2801183002/diff/160001/src/compiler/arm/code-...
src/compiler/arm/code-generator-arm.cc:2047: case kArmS32x4UnzipLeft: {
  vmovn.i64 dst.low, src
  vmov dst.high, dst.low

https://codereview.chromium.org/2801183002/diff/160001/src/compiler/arm/code-...
src/compiler/arm/code-generator-arm.cc:2056: case kArmS32x4UnzipRight: {
  vshrn.i64 dst.low, src, #32
  vmov dst.high, dst.low

https://codereview.chromium.org/2801183002/diff/160001/src/compiler/arm/code-...
src/compiler/arm/code-generator-arm.cc:2065: case kArmS32x4TransposeLeft: {
  vdup.32 dst.low, src.low[0]
  vdup.32 dst.high, src.high[0]

https://codereview.chromium.org/2801183002/diff/160001/src/compiler/arm/code-...
src/compiler/arm/code-generator-arm.cc:2074: case kArmS32x4TransposeRight: {
  vdup dst.low, src.low[1]
  vdup dst.high, src.high[1]

https://codereview.chromium.org/2801183002/diff/160001/src/compiler/arm/code-...
src/compiler/arm/code-generator-arm.cc:2099: case kArmS16x8UnzipLeft: {
  vmovn.i32 dst.low, src
  vmov dst.high, dst.low

https://codereview.chromium.org/2801183002/diff/160001/src/compiler/arm/code-...
src/compiler/arm/code-generator-arm.cc:2108: case kArmS16x8UnzipRight: {
  vshrn.i32 dst.low, src, #16
  vmov dst.high, dst.low

https://codereview.chromium.org/2801183002/diff/160001/src/compiler/arm/code-...
src/compiler/arm/code-generator-arm.cc:2151: case kArmS8x16UnzipLeft: {
  vmovn.i16 dst.low, src
  vmov dst.high, dst.low

https://codereview.chromium.org/2801183002/diff/160001/src/compiler/arm/code-...
src/compiler/arm/code-generator-arm.cc:2160: case kArmS8x16UnzipRight: {
  vshrn.i16, dst.low, src, #8
  vmov dst.high, dst.low

https://codereview.chromium.org/2801183002/diff/160001/src/compiler/arm/code-...
src/compiler/arm/code-generator-arm.cc:2192: case kArmS64x2Reverse: {
vext(dst, src, src, 8);

[bbudge, 2017-04-10 23:26:47]

https://codereview.chromium.org/2801183002/diff/160001/src/compiler/arm/code-...
File src/compiler/arm/code-generator-arm.cc (right):

https://codereview.chromium.org/2801183002/diff/160001/src/compiler/arm/code-...
src/compiler/arm/code-generator-arm.cc:2031: case kArmS32x4ZipLeft: {
On 2017/04/10 16:46:55, martyn.capewell wrote:
> It might be helpful to add a comment to each of these, expressing the
> transformation. For example, this is 3210 -> 1100, assuming lowest index lane
on
> the right. Though maybe your naming implies that lowest index is on the left.

Here's what this is supposed to do:
src0 = 0,1,2,3 (from low to high)
src1 = 4,5,6,7
After the vmov
dst (src0) = 0,1,4,5
After vzip
0,4,1,5

I'll add some comments.

https://codereview.chromium.org/2801183002/diff/160001/src/compiler/arm/code-...
src/compiler/arm/code-generator-arm.cc:2192: case kArmS64x2Reverse: {
On 2017/04/10 16:46:55, martyn.capewell wrote:
> vext(dst, src, src, 8);

Nice. However, since we expose VEXT (as S8x16Concat) this instruction would be
redundant, so I'm removing S64x2Reverse from the prototype.

[bbudge, 2017-04-12 14:10:27]

Addressed some comments about semantics of these ops.

[martyn.capewell, 2017-04-13 17:07:19]

Thanks for the comments - I obviously completely misunderstood what those
shuffles were doing.

Although I've commented on the unzip operations below, perhaps this optimisation
is better to defer to a later patch.

lgtm

https://codereview.chromium.org/2801183002/diff/200001/src/compiler/arm/code-...
File src/compiler/arm/code-generator-arm.cc (right):

https://codereview.chromium.org/2801183002/diff/200001/src/compiler/arm/code-...
src/compiler/arm/code-generator-arm.cc:2049: case kArmS32x4UnzipLeft: {
I think the unzip cases can still be replaced with vmovn for left and vshrn for
right, though I think these instructions aren't yet implemented.

if (dst.is(src0) && dst.is(src1)) {
  __ vmovn(dst.low(), src0);
  __ vmov(dst.low(), dst.high());
} else if (dst.is(src0)) {
  __ vmovn(dst.low(), src0);
  __ vmovn(dst.high(), src1);
} else {
  // dst may be src1.
  __ vmovn(dst.high(), src1);
  __ vmovn(dst.low(), src0);
}

https://codereview.chromium.org/2801183002/diff/200001/src/compiler/arm/code-...
src/compiler/arm/code-generator-arm.cc:2059: case kArmS32x4UnzipRight: {
if (dst.is(src0) && dst.is(src1)) {
  __ vshrn(dst.low(), src0, 32);
  __ vmov(dst.high(), dst.low());
} else if (dst.is(src0)) {
  __ vshrn(dst.low(), src0, 32);
  __ vshrn(dst.high(), src1, 32);
} else etc.

[bbudge, 2017-04-19 18:12:48]

The CQ bit was checked by bbudge@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2017-04-19 18:12:55]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-04-19 18:16:28]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2017-04-19 18:16:28]

Dry run: Try jobs failed on following builders:
  v8_linux64_avx2_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux64_avx2_rel_ng/buil...)
  v8_linux64_verify_csa_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux64_verify_csa_rel_n...)

[bbudge, 2017-04-19 20:01:40]

The CQ bit was checked by bbudge@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2017-04-19 20:01:47]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-04-19 20:27:19]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2017-04-19 20:27:20]

Dry run: This issue passed the CQ dry run.

[bbudge, 2017-04-19 20:30:02]

Description was changed from

==========
[WASM SIMD] Implement primitive shuffles.

- Adds unary Reverse shuffles (swizzles): S64x2Reverse (swap halves),
  S32x2Reverse, S16x4Reverse, S16x2Reverse, S8x8Reverse, S8x4Reverse,
  S8x2Reverse. Reversals are done within the sub-vectors that prefix the
  opcode name, e.g. S8x2 reverses the 8 consecutive pairs in an S8x16
  vector.

- Adds binary Zip (interleave) left and right half-shuffles to return a
  single vector: S32x4ZipLeft, S32x4ZipRightS16x8ZipLeft, S16x8ZipRight,
  S8x16ZipLeft, S8x16ZipRight.

- Adds binary Unzip (de-interleave) left and right half shuffles to return
  a single vector: S32x4UnzipLeft, S32x4UnzipRight, S16x8UnzipLeft,
  S16x8UnzipRight, S8x16UnzipLeft, S8x16UnzipRight.

- Adds binary Transpose left and right half shuffles to return
  a single vector: S32x4TransposeLeft, S32x4TransposeRight,
  S16x8TransposeLeft, S16xTransposeRight, S8x16TransposeLeft,
  S8x16TransposeRight.

- Adds binary Concat (concatenate) byte shuffle: S8x16Concat #bytes to
  paste two vectors together.

LOG=N
BUG=v8:6020
==========

to

==========
[WASM SIMD] Implement primitive shuffles.

- Adds unary Reverse shuffles (swizzles): S32x2Reverse, S16x4Reverse,
  S16x2Reverse, S8x8Reverse, S8x4Reverse, S8x2Reverse. Reversals are
  done within the sub-vectors that prefix the opcode name, e.g. S8x2
  reverses the 8 consecutive pairs in an S8x16 vector.

- Adds binary Zip (interleave) left and right half-shuffles to return a
  single vector: S32x4ZipLeft, S32x4ZipRightS16x8ZipLeft, S16x8ZipRight,
  S8x16ZipLeft, S8x16ZipRight.

- Adds binary Unzip (de-interleave) left and right half shuffles to return
  a single vector: S32x4UnzipLeft, S32x4UnzipRight, S16x8UnzipLeft,
  S16x8UnzipRight, S8x16UnzipLeft, S8x16UnzipRight.

- Adds binary Transpose left and right half shuffles to return
  a single vector: S32x4TransposeLeft, S32x4TransposeRight,
  S16x8TransposeLeft, S16xTransposeRight, S8x16TransposeLeft,
  S8x16TransposeRight.

- Adds binary Concat (concatenate) byte shuffle: S8x16Concat #bytes to
  paste two vectors together.

LOG=N
BUG=v8:6020
==========

[bbudge, 2017-04-19 20:31:49]

On 2017/04/13 17:07:19, martyn.capewell wrote:
> Thanks for the comments - I obviously completely misunderstood what those
> shuffles were doing.
> 
> Although I've commented on the unzip operations below, perhaps this
optimisation
> is better to defer to a later patch.
> 
> lgtm
> 
>
https://codereview.chromium.org/2801183002/diff/200001/src/compiler/arm/code-...
> File src/compiler/arm/code-generator-arm.cc (right):
> 
>
https://codereview.chromium.org/2801183002/diff/200001/src/compiler/arm/code-...
> src/compiler/arm/code-generator-arm.cc:2049: case kArmS32x4UnzipLeft: {
> I think the unzip cases can still be replaced with vmovn for left and vshrn
for
> right, though I think these instructions aren't yet implemented.
> 
> if (dst.is(src0) && dst.is(src1)) {
>   __ vmovn(dst.low(), src0);
>   __ vmov(dst.low(), dst.high());
> } else if (dst.is(src0)) {
>   __ vmovn(dst.low(), src0);
>   __ vmovn(dst.high(), src1);
> } else {
>   // dst may be src1.
>   __ vmovn(dst.high(), src1);
>   __ vmovn(dst.low(), src0);
> }
> 
>
https://codereview.chromium.org/2801183002/diff/200001/src/compiler/arm/code-...
> src/compiler/arm/code-generator-arm.cc:2059: case kArmS32x4UnzipRight: {
> if (dst.is(src0) && dst.is(src1)) {
>   __ vshrn(dst.low(), src0, 32);
>   __ vmov(dst.high(), dst.low());
> } else if (dst.is(src0)) {
>   __ vshrn(dst.low(), src0, 32);
>   __ vshrn(dst.high(), src1, 32);
> } else etc.

These are clever, but since we've added a dedicated scratch q-register, using
the shuffle opcodes are also 2 instructions in all cases.
https://codereview.chromium.org/2827743002/

[bradnelson, 2017-04-19 20:41:22]

lgtm

We probably will need to visit this heavily with all the stakeholders, but seems
fine for now.
I suspect we might need to think about running a pass / some sort of peephole to
merge some of these for intel. On the other hand that can be a later
optimization as long as we get consensus these are the right ops.

[bbudge, 2017-04-19 20:45:51]

The CQ bit was checked by bbudge@chromium.org

[bbudge, 2017-04-19 20:45:52]

The patchset sent to the CQ was uploaded after l-g-t-m from
martyn.capewell@arm.com
Link to the patchset: https://codereview.chromium.org/2801183002/#ps240001
(title: "Fix non-ARM build.")

[commit-bot: I haz the power, 2017-04-19 20:45:55]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-04-19 20:50:25]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2017-04-19 20:50:27]

Try jobs failed on following builders:
  v8_presubmit on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_presubmit/builds/39440)

[bbudge, 2017-04-19 20:54:15]

bbudge@chromium.org changed reviewers:
+ mtrofin@chromium.org

[bbudge, 2017-04-19 20:54:15]

+Mircea for compiler

[Mircea Trofin, 2017-04-19 22:13:42]

On 2017/04/19 20:54:15, bbudge wrote:
> +Mircea for compiler

rubberstamp lgtm

[bbudge, 2017-04-19 22:16:21]

The CQ bit was checked by bbudge@chromium.org

[commit-bot: I haz the power, 2017-04-19 22:16:30]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-04-19 22:18:07]

CQ is committing da patch.

Bot data: {"patchset_id": 240001, "attempt_start_ts": 1492640181237790,
"parent_rev": "c38e88651b68ef8e0be4c51067423f2cd938c157", "commit_rev":
"5806d86208c3845b94c8da6953830d428721c112"}

[commit-bot: I haz the power, 2017-04-19 22:18:14]

Description was changed from

==========
[WASM SIMD] Implement primitive shuffles.

- Adds unary Reverse shuffles (swizzles): S32x2Reverse, S16x4Reverse,
  S16x2Reverse, S8x8Reverse, S8x4Reverse, S8x2Reverse. Reversals are
  done within the sub-vectors that prefix the opcode name, e.g. S8x2
  reverses the 8 consecutive pairs in an S8x16 vector.

- Adds binary Zip (interleave) left and right half-shuffles to return a
  single vector: S32x4ZipLeft, S32x4ZipRightS16x8ZipLeft, S16x8ZipRight,
  S8x16ZipLeft, S8x16ZipRight.

- Adds binary Unzip (de-interleave) left and right half shuffles to return
  a single vector: S32x4UnzipLeft, S32x4UnzipRight, S16x8UnzipLeft,
  S16x8UnzipRight, S8x16UnzipLeft, S8x16UnzipRight.

- Adds binary Transpose left and right half shuffles to return
  a single vector: S32x4TransposeLeft, S32x4TransposeRight,
  S16x8TransposeLeft, S16xTransposeRight, S8x16TransposeLeft,
  S8x16TransposeRight.

- Adds binary Concat (concatenate) byte shuffle: S8x16Concat #bytes to
  paste two vectors together.

LOG=N
BUG=v8:6020
==========

to

==========
[WASM SIMD] Implement primitive shuffles.

- Adds unary Reverse shuffles (swizzles): S32x2Reverse, S16x4Reverse,
  S16x2Reverse, S8x8Reverse, S8x4Reverse, S8x2Reverse. Reversals are
  done within the sub-vectors that prefix the opcode name, e.g. S8x2
  reverses the 8 consecutive pairs in an S8x16 vector.

- Adds binary Zip (interleave) left and right half-shuffles to return a
  single vector: S32x4ZipLeft, S32x4ZipRightS16x8ZipLeft, S16x8ZipRight,
  S8x16ZipLeft, S8x16ZipRight.

- Adds binary Unzip (de-interleave) left and right half shuffles to return
  a single vector: S32x4UnzipLeft, S32x4UnzipRight, S16x8UnzipLeft,
  S16x8UnzipRight, S8x16UnzipLeft, S8x16UnzipRight.

- Adds binary Transpose left and right half shuffles to return
  a single vector: S32x4TransposeLeft, S32x4TransposeRight,
  S16x8TransposeLeft, S16xTransposeRight, S8x16TransposeLeft,
  S8x16TransposeRight.

- Adds binary Concat (concatenate) byte shuffle: S8x16Concat #bytes to
  paste two vectors together.

LOG=N
BUG=v8:6020

Review-Url: https://codereview.chromium.org/2801183002
Cr-Commit-Position: refs/heads/master@{#44734}
Committed:
https://chromium.googlesource.com/v8/v8/+/5806d86208c3845b94c8da6953830d42872...
==========

[commit-bot: I haz the power, 2017-04-19 22:18:15]

Committed patchset #12 (id:240001) as
https://chromium.googlesource.com/v8/v8/+/5806d86208c3845b94c8da6953830d42872...