diff options
| author | Thomas Lively <7121787+tlively@users.noreply.github.com> | 2021-04-05 12:33:25 -0700 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2021-04-05 12:33:25 -0700 |
| commit | 1bb172c789bb3a61aeaae78f5464d0544627ed3e (patch) | |
| tree | 6bcfd54944e52f9c3f8354a3d3f523f45a5ee670 | |
| parent | c59df4cda843ef11ad261f5c889dddc9a9d59d3b (diff) | |
| download | binaryen-1bb172c789bb3a61aeaae78f5464d0544627ed3e.tar.gz binaryen-1bb172c789bb3a61aeaae78f5464d0544627ed3e.tar.bz2 binaryen-1bb172c789bb3a61aeaae78f5464d0544627ed3e.zip | |
Update SIMD names and opcodes (#3771)
Also removes experimental SIMD instructions that were not included in the final
spec proposal.
45 files changed, 2627 insertions, 3724 deletions
diff --git a/CHANGELOG.md b/CHANGELOG.md index cea49dd55..b177f6efc 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -49,6 +49,8 @@ Current Trunk - `BinaryenGetFunctionTableSegmentData` is replaced with `BinaryenElementSegmentGetData`. - Boolean values in the C API now should use `bool` instead of `int`. +- Experimental SIMD instructions have been removed and the names and opcodes of + the standard instructions have been updated to match the final spec. v100 ---- diff --git a/scripts/gen-s-parser.py b/scripts/gen-s-parser.py index c65e5ddc4..569289427 100755 --- a/scripts/gen-s-parser.py +++ b/scripts/gen-s-parser.py @@ -281,7 +281,7 @@ instructions = [ ("v128.load", "makeLoad(s, Type::v128, /*isAtomic=*/false)"), ("v128.store", "makeStore(s, Type::v128, /*isAtomic=*/false)"), ("v128.const", "makeConst(s, Type::v128)"), - ("v8x16.shuffle", "makeSIMDShuffle(s)"), + ("i8x16.shuffle", "makeSIMDShuffle(s)"), ("i8x16.splat", "makeUnary(s, UnaryOp::SplatVecI8x16)"), ("i8x16.extract_lane_s", "makeSIMDExtract(s, SIMDExtractOp::ExtractLaneSVecI8x16, 16)"), ("i8x16.extract_lane_u", "makeSIMDExtract(s, SIMDExtractOp::ExtractLaneUVecI8x16, 16)"), @@ -333,6 +333,11 @@ instructions = [ ("i32x4.ge_s", "makeBinary(s, BinaryOp::GeSVecI32x4)"), ("i32x4.ge_u", "makeBinary(s, BinaryOp::GeUVecI32x4)"), ("i64x2.eq", "makeBinary(s, BinaryOp::EqVecI64x2)"), + ("i64x2.ne", "makeBinary(s, BinaryOp::NeVecI64x2)"), + ("i64x2.lt_s", "makeBinary(s, BinaryOp::LtSVecI64x2)"), + ("i64x2.gt_s", "makeBinary(s, BinaryOp::GtSVecI64x2)"), + ("i64x2.le_s", "makeBinary(s, BinaryOp::LeSVecI64x2)"), + ("i64x2.ge_s", "makeBinary(s, BinaryOp::GeSVecI64x2)"), ("f32x4.eq", "makeBinary(s, BinaryOp::EqVecF32x4)"), ("f32x4.ne", "makeBinary(s, BinaryOp::NeVecF32x4)"), ("f32x4.lt", "makeBinary(s, BinaryOp::LtVecF32x4)"), @@ -350,11 +355,8 @@ instructions = [ ("v128.or", "makeBinary(s, BinaryOp::OrVec128)"), ("v128.xor", "makeBinary(s, BinaryOp::XorVec128)"), ("v128.andnot", "makeBinary(s, BinaryOp::AndNotVec128)"), + ("v128.any_true", "makeUnary(s, UnaryOp::AnyTrueVec128)"), ("v128.bitselect", "makeSIMDTernary(s, SIMDTernaryOp::Bitselect)"), - ("v8x16.signselect", "makeSIMDTernary(s, SIMDTernaryOp::SignSelectVec8x16)"), - ("v16x8.signselect", "makeSIMDTernary(s, SIMDTernaryOp::SignSelectVec16x8)"), - ("v32x4.signselect", "makeSIMDTernary(s, SIMDTernaryOp::SignSelectVec32x4)"), - ("v64x2.signselect", "makeSIMDTernary(s, SIMDTernaryOp::SignSelectVec64x2)"), ("v128.load8_lane", "makeSIMDLoadStoreLane(s, LoadLaneVec8x16)"), ("v128.load16_lane", "makeSIMDLoadStoreLane(s, LoadLaneVec16x8)"), ("v128.load32_lane", "makeSIMDLoadStoreLane(s, LoadLaneVec32x4)"), @@ -366,19 +368,17 @@ instructions = [ ("i8x16.popcnt", "makeUnary(s, UnaryOp::PopcntVecI8x16)"), ("i8x16.abs", "makeUnary(s, UnaryOp::AbsVecI8x16)"), ("i8x16.neg", "makeUnary(s, UnaryOp::NegVecI8x16)"), - ("i8x16.any_true", "makeUnary(s, UnaryOp::AnyTrueVecI8x16)"), ("i8x16.all_true", "makeUnary(s, UnaryOp::AllTrueVecI8x16)"), ("i8x16.bitmask", "makeUnary(s, UnaryOp::BitmaskVecI8x16)"), ("i8x16.shl", "makeSIMDShift(s, SIMDShiftOp::ShlVecI8x16)"), ("i8x16.shr_s", "makeSIMDShift(s, SIMDShiftOp::ShrSVecI8x16)"), ("i8x16.shr_u", "makeSIMDShift(s, SIMDShiftOp::ShrUVecI8x16)"), ("i8x16.add", "makeBinary(s, BinaryOp::AddVecI8x16)"), - ("i8x16.add_saturate_s", "makeBinary(s, BinaryOp::AddSatSVecI8x16)"), - ("i8x16.add_saturate_u", "makeBinary(s, BinaryOp::AddSatUVecI8x16)"), + ("i8x16.add_sat_s", "makeBinary(s, BinaryOp::AddSatSVecI8x16)"), + ("i8x16.add_sat_u", "makeBinary(s, BinaryOp::AddSatUVecI8x16)"), ("i8x16.sub", "makeBinary(s, BinaryOp::SubVecI8x16)"), - ("i8x16.sub_saturate_s", "makeBinary(s, BinaryOp::SubSatSVecI8x16)"), - ("i8x16.sub_saturate_u", "makeBinary(s, BinaryOp::SubSatUVecI8x16)"), - ("i8x16.mul", "makeBinary(s, BinaryOp::MulVecI8x16)"), + ("i8x16.sub_sat_s", "makeBinary(s, BinaryOp::SubSatSVecI8x16)"), + ("i8x16.sub_sat_u", "makeBinary(s, BinaryOp::SubSatUVecI8x16)"), ("i8x16.min_s", "makeBinary(s, BinaryOp::MinSVecI8x16)"), ("i8x16.min_u", "makeBinary(s, BinaryOp::MinUVecI8x16)"), ("i8x16.max_s", "makeBinary(s, BinaryOp::MaxSVecI8x16)"), @@ -386,18 +386,17 @@ instructions = [ ("i8x16.avgr_u", "makeBinary(s, BinaryOp::AvgrUVecI8x16)"), ("i16x8.abs", "makeUnary(s, UnaryOp::AbsVecI16x8)"), ("i16x8.neg", "makeUnary(s, UnaryOp::NegVecI16x8)"), - ("i16x8.any_true", "makeUnary(s, UnaryOp::AnyTrueVecI16x8)"), ("i16x8.all_true", "makeUnary(s, UnaryOp::AllTrueVecI16x8)"), ("i16x8.bitmask", "makeUnary(s, UnaryOp::BitmaskVecI16x8)"), ("i16x8.shl", "makeSIMDShift(s, SIMDShiftOp::ShlVecI16x8)"), ("i16x8.shr_s", "makeSIMDShift(s, SIMDShiftOp::ShrSVecI16x8)"), ("i16x8.shr_u", "makeSIMDShift(s, SIMDShiftOp::ShrUVecI16x8)"), ("i16x8.add", "makeBinary(s, BinaryOp::AddVecI16x8)"), - ("i16x8.add_saturate_s", "makeBinary(s, BinaryOp::AddSatSVecI16x8)"), - ("i16x8.add_saturate_u", "makeBinary(s, BinaryOp::AddSatUVecI16x8)"), + ("i16x8.add_sat_s", "makeBinary(s, BinaryOp::AddSatSVecI16x8)"), + ("i16x8.add_sat_u", "makeBinary(s, BinaryOp::AddSatUVecI16x8)"), ("i16x8.sub", "makeBinary(s, BinaryOp::SubVecI16x8)"), - ("i16x8.sub_saturate_s", "makeBinary(s, BinaryOp::SubSatSVecI16x8)"), - ("i16x8.sub_saturate_u", "makeBinary(s, BinaryOp::SubSatUVecI16x8)"), + ("i16x8.sub_sat_s", "makeBinary(s, BinaryOp::SubSatSVecI16x8)"), + ("i16x8.sub_sat_u", "makeBinary(s, BinaryOp::SubSatUVecI16x8)"), ("i16x8.mul", "makeBinary(s, BinaryOp::MulVecI16x8)"), ("i16x8.min_s", "makeBinary(s, BinaryOp::MinSVecI16x8)"), ("i16x8.min_u", "makeBinary(s, BinaryOp::MinUVecI16x8)"), @@ -411,7 +410,6 @@ instructions = [ ("i16x8.extmul_high_i8x16_u", "makeBinary(s, BinaryOp::ExtMulHighUVecI16x8)"), ("i32x4.abs", "makeUnary(s, UnaryOp::AbsVecI32x4)"), ("i32x4.neg", "makeUnary(s, UnaryOp::NegVecI32x4)"), - ("i32x4.any_true", "makeUnary(s, UnaryOp::AnyTrueVecI32x4)"), ("i32x4.all_true", "makeUnary(s, UnaryOp::AllTrueVecI32x4)"), ("i32x4.bitmask", "makeUnary(s, UnaryOp::BitmaskVecI32x4)"), ("i32x4.shl", "makeSIMDShift(s, SIMDShiftOp::ShlVecI32x4)"), @@ -429,7 +427,9 @@ instructions = [ ("i32x4.extmul_high_i16x8_s", "makeBinary(s, BinaryOp::ExtMulHighSVecI32x4)"), ("i32x4.extmul_low_i16x8_u", "makeBinary(s, BinaryOp::ExtMulLowUVecI32x4)"), ("i32x4.extmul_high_i16x8_u", "makeBinary(s, BinaryOp::ExtMulHighUVecI32x4)"), + ("i64x2.abs", "makeUnary(s, UnaryOp::AbsVecI64x2)"), ("i64x2.neg", "makeUnary(s, UnaryOp::NegVecI64x2)"), + ("i64x2.all_true", "makeUnary(s, UnaryOp::AllTrueVecI64x2)"), ("i64x2.bitmask", "makeUnary(s, UnaryOp::BitmaskVecI64x2)"), ("i64x2.shl", "makeSIMDShift(s, SIMDShiftOp::ShlVecI64x2)"), ("i64x2.shr_s", "makeSIMDShift(s, SIMDShiftOp::ShrSVecI64x2)"), @@ -444,8 +444,6 @@ instructions = [ ("f32x4.abs", "makeUnary(s, UnaryOp::AbsVecF32x4)"), ("f32x4.neg", "makeUnary(s, UnaryOp::NegVecF32x4)"), ("f32x4.sqrt", "makeUnary(s, UnaryOp::SqrtVecF32x4)"), - ("f32x4.qfma", "makeSIMDTernary(s, SIMDTernaryOp::QFMAF32x4)"), - ("f32x4.qfms", "makeSIMDTernary(s, SIMDTernaryOp::QFMSF32x4)"), ("f32x4.add", "makeBinary(s, BinaryOp::AddVecF32x4)"), ("f32x4.sub", "makeBinary(s, BinaryOp::SubVecF32x4)"), ("f32x4.mul", "makeBinary(s, BinaryOp::MulVecF32x4)"), @@ -461,8 +459,6 @@ instructions = [ ("f64x2.abs", "makeUnary(s, UnaryOp::AbsVecF64x2)"), ("f64x2.neg", "makeUnary(s, UnaryOp::NegVecF64x2)"), ("f64x2.sqrt", "makeUnary(s, UnaryOp::SqrtVecF64x2)"), - ("f64x2.qfma", "makeSIMDTernary(s, SIMDTernaryOp::QFMAF64x2)"), - ("f64x2.qfms", "makeSIMDTernary(s, SIMDTernaryOp::QFMSF64x2)"), ("f64x2.add", "makeBinary(s, BinaryOp::AddVecF64x2)"), ("f64x2.sub", "makeBinary(s, BinaryOp::SubVecF64x2)"), ("f64x2.mul", "makeBinary(s, BinaryOp::MulVecF64x2)"), @@ -477,57 +473,48 @@ instructions = [ ("f64x2.nearest", "makeUnary(s, UnaryOp::NearestVecF64x2)"), ("i32x4.trunc_sat_f32x4_s", "makeUnary(s, UnaryOp::TruncSatSVecF32x4ToVecI32x4)"), ("i32x4.trunc_sat_f32x4_u", "makeUnary(s, UnaryOp::TruncSatUVecF32x4ToVecI32x4)"), - ("i64x2.trunc_sat_f64x2_s", "makeUnary(s, UnaryOp::TruncSatSVecF64x2ToVecI64x2)"), - ("i64x2.trunc_sat_f64x2_u", "makeUnary(s, UnaryOp::TruncSatUVecF64x2ToVecI64x2)"), ("f32x4.convert_i32x4_s", "makeUnary(s, UnaryOp::ConvertSVecI32x4ToVecF32x4)"), ("f32x4.convert_i32x4_u", "makeUnary(s, UnaryOp::ConvertUVecI32x4ToVecF32x4)"), - ("f64x2.convert_i64x2_s", "makeUnary(s, UnaryOp::ConvertSVecI64x2ToVecF64x2)"), - ("f64x2.convert_i64x2_u", "makeUnary(s, UnaryOp::ConvertUVecI64x2ToVecF64x2)"), - ("v8x16.load_splat", "makeSIMDLoad(s, SIMDLoadOp::LoadSplatVec8x16)"), - ("v16x8.load_splat", "makeSIMDLoad(s, SIMDLoadOp::LoadSplatVec16x8)"), - ("v32x4.load_splat", "makeSIMDLoad(s, SIMDLoadOp::LoadSplatVec32x4)"), - ("v64x2.load_splat", "makeSIMDLoad(s, SIMDLoadOp::LoadSplatVec64x2)"), - ("i16x8.load8x8_s", "makeSIMDLoad(s, SIMDLoadOp::LoadExtSVec8x8ToVecI16x8)"), - ("i16x8.load8x8_u", "makeSIMDLoad(s, SIMDLoadOp::LoadExtUVec8x8ToVecI16x8)"), - ("i32x4.load16x4_s", "makeSIMDLoad(s, SIMDLoadOp::LoadExtSVec16x4ToVecI32x4)"), - ("i32x4.load16x4_u", "makeSIMDLoad(s, SIMDLoadOp::LoadExtUVec16x4ToVecI32x4)"), - ("i64x2.load32x2_s", "makeSIMDLoad(s, SIMDLoadOp::LoadExtSVec32x2ToVecI64x2)"), - ("i64x2.load32x2_u", "makeSIMDLoad(s, SIMDLoadOp::LoadExtUVec32x2ToVecI64x2)"), + ("v128.load8_splat", "makeSIMDLoad(s, SIMDLoadOp::LoadSplatVec8x16)"), + ("v128.load16_splat", "makeSIMDLoad(s, SIMDLoadOp::LoadSplatVec16x8)"), + ("v128.load32_splat", "makeSIMDLoad(s, SIMDLoadOp::LoadSplatVec32x4)"), + ("v128.load64_splat", "makeSIMDLoad(s, SIMDLoadOp::LoadSplatVec64x2)"), + ("v128.load8x8_s", "makeSIMDLoad(s, SIMDLoadOp::LoadExtSVec8x8ToVecI16x8)"), + ("v128.load8x8_u", "makeSIMDLoad(s, SIMDLoadOp::LoadExtUVec8x8ToVecI16x8)"), + ("v128.load16x4_s", "makeSIMDLoad(s, SIMDLoadOp::LoadExtSVec16x4ToVecI32x4)"), + ("v128.load16x4_u", "makeSIMDLoad(s, SIMDLoadOp::LoadExtUVec16x4ToVecI32x4)"), + ("v128.load32x2_s", "makeSIMDLoad(s, SIMDLoadOp::LoadExtSVec32x2ToVecI64x2)"), + ("v128.load32x2_u", "makeSIMDLoad(s, SIMDLoadOp::LoadExtUVec32x2ToVecI64x2)"), ("v128.load32_zero", "makeSIMDLoad(s, SIMDLoadOp::Load32Zero)"), ("v128.load64_zero", "makeSIMDLoad(s, SIMDLoadOp::Load64Zero)"), ("i8x16.narrow_i16x8_s", "makeBinary(s, BinaryOp::NarrowSVecI16x8ToVecI8x16)"), ("i8x16.narrow_i16x8_u", "makeBinary(s, BinaryOp::NarrowUVecI16x8ToVecI8x16)"), ("i16x8.narrow_i32x4_s", "makeBinary(s, BinaryOp::NarrowSVecI32x4ToVecI16x8)"), ("i16x8.narrow_i32x4_u", "makeBinary(s, BinaryOp::NarrowUVecI32x4ToVecI16x8)"), - ("i16x8.widen_low_i8x16_s", "makeUnary(s, UnaryOp::WidenLowSVecI8x16ToVecI16x8)"), - ("i16x8.widen_high_i8x16_s", "makeUnary(s, UnaryOp::WidenHighSVecI8x16ToVecI16x8)"), - ("i16x8.widen_low_i8x16_u", "makeUnary(s, UnaryOp::WidenLowUVecI8x16ToVecI16x8)"), - ("i16x8.widen_high_i8x16_u", "makeUnary(s, UnaryOp::WidenHighUVecI8x16ToVecI16x8)"), - ("i32x4.widen_low_i16x8_s", "makeUnary(s, UnaryOp::WidenLowSVecI16x8ToVecI32x4)"), - ("i32x4.widen_high_i16x8_s", "makeUnary(s, UnaryOp::WidenHighSVecI16x8ToVecI32x4)"), - ("i32x4.widen_low_i16x8_u", "makeUnary(s, UnaryOp::WidenLowUVecI16x8ToVecI32x4)"), - ("i32x4.widen_high_i16x8_u", "makeUnary(s, UnaryOp::WidenHighUVecI16x8ToVecI32x4)"), - ("i64x2.widen_low_i32x4_s", "makeUnary(s, UnaryOp::WidenLowSVecI32x4ToVecI64x2)"), - ("i64x2.widen_high_i32x4_s", "makeUnary(s, UnaryOp::WidenHighSVecI32x4ToVecI64x2)"), - ("i64x2.widen_low_i32x4_u", "makeUnary(s, UnaryOp::WidenLowUVecI32x4ToVecI64x2)"), - ("i64x2.widen_high_i32x4_u", "makeUnary(s, UnaryOp::WidenHighUVecI32x4ToVecI64x2)"), - ("v8x16.swizzle", "makeBinary(s, BinaryOp::SwizzleVec8x16)"), + ("i16x8.extend_low_i8x16_s", "makeUnary(s, UnaryOp::ExtendLowSVecI8x16ToVecI16x8)"), + ("i16x8.extend_high_i8x16_s", "makeUnary(s, UnaryOp::ExtendHighSVecI8x16ToVecI16x8)"), + ("i16x8.extend_low_i8x16_u", "makeUnary(s, UnaryOp::ExtendLowUVecI8x16ToVecI16x8)"), + ("i16x8.extend_high_i8x16_u", "makeUnary(s, UnaryOp::ExtendHighUVecI8x16ToVecI16x8)"), + ("i32x4.extend_low_i16x8_s", "makeUnary(s, UnaryOp::ExtendLowSVecI16x8ToVecI32x4)"), + ("i32x4.extend_high_i16x8_s", "makeUnary(s, UnaryOp::ExtendHighSVecI16x8ToVecI32x4)"), + ("i32x4.extend_low_i16x8_u", "makeUnary(s, UnaryOp::ExtendLowUVecI16x8ToVecI32x4)"), + ("i32x4.extend_high_i16x8_u", "makeUnary(s, UnaryOp::ExtendHighUVecI16x8ToVecI32x4)"), + ("i64x2.extend_low_i32x4_s", "makeUnary(s, UnaryOp::ExtendLowSVecI32x4ToVecI64x2)"), + ("i64x2.extend_high_i32x4_s", "makeUnary(s, UnaryOp::ExtendHighSVecI32x4ToVecI64x2)"), + ("i64x2.extend_low_i32x4_u", "makeUnary(s, UnaryOp::ExtendLowUVecI32x4ToVecI64x2)"), + ("i64x2.extend_high_i32x4_u", "makeUnary(s, UnaryOp::ExtendHighUVecI32x4ToVecI64x2)"), + ("i8x16.swizzle", "makeBinary(s, BinaryOp::SwizzleVec8x16)"), ("i16x8.extadd_pairwise_i8x16_s", "makeUnary(s, UnaryOp::ExtAddPairwiseSVecI8x16ToI16x8)"), ("i16x8.extadd_pairwise_i8x16_u", "makeUnary(s, UnaryOp::ExtAddPairwiseUVecI8x16ToI16x8)"), ("i32x4.extadd_pairwise_i16x8_s", "makeUnary(s, UnaryOp::ExtAddPairwiseSVecI16x8ToI32x4)"), ("i32x4.extadd_pairwise_i16x8_u", "makeUnary(s, UnaryOp::ExtAddPairwiseUVecI16x8ToI32x4)"), ("f64x2.convert_low_i32x4_s", "makeUnary(s, UnaryOp::ConvertLowSVecI32x4ToVecF64x2)"), ("f64x2.convert_low_i32x4_u", "makeUnary(s, UnaryOp::ConvertLowUVecI32x4ToVecF64x2)"), - ("i32x4.trunc_sat_f64x2_zero_s", "makeUnary(s, UnaryOp::TruncSatZeroSVecF64x2ToVecI32x4)"), - ("i32x4.trunc_sat_f64x2_zero_u", "makeUnary(s, UnaryOp::TruncSatZeroUVecF64x2ToVecI32x4)"), + ("i32x4.trunc_sat_f64x2_s_zero", "makeUnary(s, UnaryOp::TruncSatZeroSVecF64x2ToVecI32x4)"), + ("i32x4.trunc_sat_f64x2_u_zero", "makeUnary(s, UnaryOp::TruncSatZeroUVecF64x2ToVecI32x4)"), ("f32x4.demote_f64x2_zero", "makeUnary(s, UnaryOp::DemoteZeroVecF64x2ToVecF32x4)"), ("f64x2.promote_low_f32x4", "makeUnary(s, UnaryOp::PromoteLowVecF32x4ToVecF64x2)"), - ("i32x4.widen_i8x16_s", "makeSIMDWiden(s, SIMDWidenOp::WidenSVecI8x16ToVecI32x4)"), - ("i32x4.widen_i8x16_u", "makeSIMDWiden(s, SIMDWidenOp::WidenUVecI8x16ToVecI32x4)"), - # prefetch instructions - ("prefetch.t", "makePrefetch(s, PrefetchOp::PrefetchTemporal)"), - ("prefetch.nt", "makePrefetch(s, PrefetchOp::PrefetchNontemporal)"), # reference types instructions # TODO Add table instructions ("ref.null", "makeRefNull(s)"), diff --git a/src/binaryen-c.cpp b/src/binaryen-c.cpp index aff4ed4fc..e257c976c 100644 --- a/src/binaryen-c.cpp +++ b/src/binaryen-c.cpp @@ -526,9 +526,9 @@ BinaryenOp BinaryenOrVec128(void) { return OrVec128; } BinaryenOp BinaryenXorVec128(void) { return XorVec128; } BinaryenOp BinaryenAndNotVec128(void) { return AndNotVec128; } BinaryenOp BinaryenBitselectVec128(void) { return Bitselect; } +BinaryenOp BinaryenAnyTrueVec128(void) { return AnyTrueVec128; } BinaryenOp BinaryenAbsVecI8x16(void) { return AbsVecI8x16; } BinaryenOp BinaryenNegVecI8x16(void) { return NegVecI8x16; } -BinaryenOp BinaryenAnyTrueVecI8x16(void) { return AnyTrueVecI8x16; } BinaryenOp BinaryenAllTrueVecI8x16(void) { return AllTrueVecI8x16; } BinaryenOp BinaryenBitmaskVecI8x16(void) { return BitmaskVecI8x16; } BinaryenOp BinaryenShlVecI8x16(void) { return ShlVecI8x16; } @@ -540,7 +540,6 @@ BinaryenOp BinaryenAddSatUVecI8x16(void) { return AddSatUVecI8x16; } BinaryenOp BinaryenSubVecI8x16(void) { return SubVecI8x16; } BinaryenOp BinaryenSubSatSVecI8x16(void) { return SubSatSVecI8x16; } BinaryenOp BinaryenSubSatUVecI8x16(void) { return SubSatUVecI8x16; } -BinaryenOp BinaryenMulVecI8x16(void) { return MulVecI8x16; } BinaryenOp BinaryenMinSVecI8x16(void) { return MinSVecI8x16; } BinaryenOp BinaryenMinUVecI8x16(void) { return MinUVecI8x16; } BinaryenOp BinaryenMaxSVecI8x16(void) { return MaxSVecI8x16; } @@ -548,7 +547,6 @@ BinaryenOp BinaryenMaxUVecI8x16(void) { return MaxUVecI8x16; } BinaryenOp BinaryenAvgrUVecI8x16(void) { return AvgrUVecI8x16; } BinaryenOp BinaryenAbsVecI16x8(void) { return AbsVecI16x8; } BinaryenOp BinaryenNegVecI16x8(void) { return NegVecI16x8; } -BinaryenOp BinaryenAnyTrueVecI16x8(void) { return AnyTrueVecI16x8; } BinaryenOp BinaryenAllTrueVecI16x8(void) { return AllTrueVecI16x8; } BinaryenOp BinaryenBitmaskVecI16x8(void) { return BitmaskVecI16x8; } BinaryenOp BinaryenShlVecI16x8(void) { return ShlVecI16x8; } @@ -568,7 +566,6 @@ BinaryenOp BinaryenMaxUVecI16x8(void) { return MaxUVecI16x8; } BinaryenOp BinaryenAvgrUVecI16x8(void) { return AvgrUVecI16x8; } BinaryenOp BinaryenAbsVecI32x4(void) { return AbsVecI32x4; } BinaryenOp BinaryenNegVecI32x4(void) { return NegVecI32x4; } -BinaryenOp BinaryenAnyTrueVecI32x4(void) { return AnyTrueVecI32x4; } BinaryenOp BinaryenAllTrueVecI32x4(void) { return AllTrueVecI32x4; } BinaryenOp BinaryenBitmaskVecI32x4(void) { return BitmaskVecI32x4; } BinaryenOp BinaryenShlVecI32x4(void) { return ShlVecI32x4; } @@ -594,8 +591,6 @@ BinaryenOp BinaryenMulVecI64x2(void) { return MulVecI64x2; } BinaryenOp BinaryenAbsVecF32x4(void) { return AbsVecF32x4; } BinaryenOp BinaryenNegVecF32x4(void) { return NegVecF32x4; } BinaryenOp BinaryenSqrtVecF32x4(void) { return SqrtVecF32x4; } -BinaryenOp BinaryenQFMAVecF32x4(void) { return QFMAF32x4; } -BinaryenOp BinaryenQFMSVecF32x4(void) { return QFMSF32x4; } BinaryenOp BinaryenAddVecF32x4(void) { return AddVecF32x4; } BinaryenOp BinaryenSubVecF32x4(void) { return SubVecF32x4; } BinaryenOp BinaryenMulVecF32x4(void) { return MulVecF32x4; } @@ -611,8 +606,6 @@ BinaryenOp BinaryenPMaxVecF32x4(void) { return PMaxVecF32x4; } BinaryenOp BinaryenAbsVecF64x2(void) { return AbsVecF64x2; } BinaryenOp BinaryenNegVecF64x2(void) { return NegVecF64x2; } BinaryenOp BinaryenSqrtVecF64x2(void) { return SqrtVecF64x2; } -BinaryenOp BinaryenQFMAVecF64x2(void) { return QFMAF64x2; } -BinaryenOp BinaryenQFMSVecF64x2(void) { return QFMSF64x2; } BinaryenOp BinaryenAddVecF64x2(void) { return AddVecF64x2; } BinaryenOp BinaryenSubVecF64x2(void) { return SubVecF64x2; } BinaryenOp BinaryenMulVecF64x2(void) { return MulVecF64x2; } @@ -631,24 +624,12 @@ BinaryenOp BinaryenTruncSatSVecF32x4ToVecI32x4(void) { BinaryenOp BinaryenTruncSatUVecF32x4ToVecI32x4(void) { return TruncSatUVecF32x4ToVecI32x4; } -BinaryenOp BinaryenTruncSatSVecF64x2ToVecI64x2(void) { - return TruncSatSVecF64x2ToVecI64x2; -} -BinaryenOp BinaryenTruncSatUVecF64x2ToVecI64x2(void) { - return TruncSatUVecF64x2ToVecI64x2; -} BinaryenOp BinaryenConvertSVecI32x4ToVecF32x4(void) { return ConvertSVecI32x4ToVecF32x4; } BinaryenOp BinaryenConvertUVecI32x4ToVecF32x4(void) { return ConvertUVecI32x4ToVecF32x4; } -BinaryenOp BinaryenConvertSVecI64x2ToVecF64x2(void) { - return ConvertSVecI64x2ToVecF64x2; -} -BinaryenOp BinaryenConvertUVecI64x2ToVecF64x2(void) { - return ConvertUVecI64x2ToVecF64x2; -} BinaryenOp BinaryenLoadSplatVec8x16(void) { return LoadSplatVec8x16; } BinaryenOp BinaryenLoadSplatVec16x8(void) { return LoadSplatVec16x8; } BinaryenOp BinaryenLoadSplatVec32x4(void) { return LoadSplatVec32x4; } @@ -683,29 +664,29 @@ BinaryenOp BinaryenNarrowSVecI32x4ToVecI16x8(void) { BinaryenOp BinaryenNarrowUVecI32x4ToVecI16x8(void) { return NarrowUVecI32x4ToVecI16x8; } -BinaryenOp BinaryenWidenLowSVecI8x16ToVecI16x8(void) { - return WidenLowSVecI8x16ToVecI16x8; +BinaryenOp BinaryenExtendLowSVecI8x16ToVecI16x8(void) { + return ExtendLowSVecI8x16ToVecI16x8; } -BinaryenOp BinaryenWidenHighSVecI8x16ToVecI16x8(void) { - return WidenHighSVecI8x16ToVecI16x8; +BinaryenOp BinaryenExtendHighSVecI8x16ToVecI16x8(void) { + return ExtendHighSVecI8x16ToVecI16x8; } -BinaryenOp BinaryenWidenLowUVecI8x16ToVecI16x8(void) { - return WidenLowUVecI8x16ToVecI16x8; +BinaryenOp BinaryenExtendLowUVecI8x16ToVecI16x8(void) { + return ExtendLowUVecI8x16ToVecI16x8; } -BinaryenOp BinaryenWidenHighUVecI8x16ToVecI16x8(void) { - return WidenHighUVecI8x16ToVecI16x8; +BinaryenOp BinaryenExtendHighUVecI8x16ToVecI16x8(void) { + return ExtendHighUVecI8x16ToVecI16x8; } -BinaryenOp BinaryenWidenLowSVecI16x8ToVecI32x4(void) { - return WidenLowSVecI16x8ToVecI32x4; +BinaryenOp BinaryenExtendLowSVecI16x8ToVecI32x4(void) { + return ExtendLowSVecI16x8ToVecI32x4; } -BinaryenOp BinaryenWidenHighSVecI16x8ToVecI32x4(void) { - return WidenHighSVecI16x8ToVecI32x4; +BinaryenOp BinaryenExtendHighSVecI16x8ToVecI32x4(void) { + return ExtendHighSVecI16x8ToVecI32x4; } -BinaryenOp BinaryenWidenLowUVecI16x8ToVecI32x4(void) { - return WidenLowUVecI16x8ToVecI32x4; +BinaryenOp BinaryenExtendLowUVecI16x8ToVecI32x4(void) { + return ExtendLowUVecI16x8ToVecI32x4; } -BinaryenOp BinaryenWidenHighUVecI16x8ToVecI32x4(void) { - return WidenHighUVecI16x8ToVecI32x4; +BinaryenOp BinaryenExtendHighUVecI16x8ToVecI32x4(void) { + return ExtendHighUVecI16x8ToVecI32x4; } BinaryenOp BinaryenSwizzleVec8x16(void) { return SwizzleVec8x16; } BinaryenOp BinaryenRefIsNull(void) { return RefIsNull; } diff --git a/src/binaryen-c.h b/src/binaryen-c.h index 0dff91f93..686fb5262 100644 --- a/src/binaryen-c.h +++ b/src/binaryen-c.h @@ -427,10 +427,10 @@ BINARYEN_API BinaryenOp BinaryenOrVec128(void); BINARYEN_API BinaryenOp BinaryenXorVec128(void); BINARYEN_API BinaryenOp BinaryenAndNotVec128(void); BINARYEN_API BinaryenOp BinaryenBitselectVec128(void); +BINARYEN_API BinaryenOp BinaryenAnyTrueVec128(void); // TODO: Add i8x16.popcnt to C and JS APIs once merged to the proposal BINARYEN_API BinaryenOp BinaryenAbsVecI8x16(void); BINARYEN_API BinaryenOp BinaryenNegVecI8x16(void); -BINARYEN_API BinaryenOp BinaryenAnyTrueVecI8x16(void); BINARYEN_API BinaryenOp BinaryenAllTrueVecI8x16(void); BINARYEN_API BinaryenOp BinaryenBitmaskVecI8x16(void); BINARYEN_API BinaryenOp BinaryenShlVecI8x16(void); @@ -442,7 +442,6 @@ BINARYEN_API BinaryenOp BinaryenAddSatUVecI8x16(void); BINARYEN_API BinaryenOp BinaryenSubVecI8x16(void); BINARYEN_API BinaryenOp BinaryenSubSatSVecI8x16(void); BINARYEN_API BinaryenOp BinaryenSubSatUVecI8x16(void); -BINARYEN_API BinaryenOp BinaryenMulVecI8x16(void); BINARYEN_API BinaryenOp BinaryenMinSVecI8x16(void); BINARYEN_API BinaryenOp BinaryenMinUVecI8x16(void); BINARYEN_API BinaryenOp BinaryenMaxSVecI8x16(void); @@ -450,7 +449,6 @@ BINARYEN_API BinaryenOp BinaryenMaxUVecI8x16(void); BINARYEN_API BinaryenOp BinaryenAvgrUVecI8x16(void); BINARYEN_API BinaryenOp BinaryenAbsVecI16x8(void); BINARYEN_API BinaryenOp BinaryenNegVecI16x8(void); -BINARYEN_API BinaryenOp BinaryenAnyTrueVecI16x8(void); BINARYEN_API BinaryenOp BinaryenAllTrueVecI16x8(void); BINARYEN_API BinaryenOp BinaryenBitmaskVecI16x8(void); BINARYEN_API BinaryenOp BinaryenShlVecI16x8(void); @@ -472,7 +470,6 @@ BINARYEN_API BinaryenOp BinaryenAvgrUVecI16x8(void); // TODO: Add extending multiplications to APIs once they are merged as well BINARYEN_API BinaryenOp BinaryenAbsVecI32x4(void); BINARYEN_API BinaryenOp BinaryenNegVecI32x4(void); -BINARYEN_API BinaryenOp BinaryenAnyTrueVecI32x4(void); BINARYEN_API BinaryenOp BinaryenAllTrueVecI32x4(void); BINARYEN_API BinaryenOp BinaryenBitmaskVecI32x4(void); BINARYEN_API BinaryenOp BinaryenShlVecI32x4(void); @@ -496,8 +493,6 @@ BINARYEN_API BinaryenOp BinaryenMulVecI64x2(void); BINARYEN_API BinaryenOp BinaryenAbsVecF32x4(void); BINARYEN_API BinaryenOp BinaryenNegVecF32x4(void); BINARYEN_API BinaryenOp BinaryenSqrtVecF32x4(void); -BINARYEN_API BinaryenOp BinaryenQFMAVecF32x4(void); -BINARYEN_API BinaryenOp BinaryenQFMSVecF32x4(void); BINARYEN_API BinaryenOp BinaryenAddVecF32x4(void); BINARYEN_API BinaryenOp BinaryenSubVecF32x4(void); BINARYEN_API BinaryenOp BinaryenMulVecF32x4(void); @@ -513,8 +508,6 @@ BINARYEN_API BinaryenOp BinaryenNearestVecF32x4(void); BINARYEN_API BinaryenOp BinaryenAbsVecF64x2(void); BINARYEN_API BinaryenOp BinaryenNegVecF64x2(void); BINARYEN_API BinaryenOp BinaryenSqrtVecF64x2(void); -BINARYEN_API BinaryenOp BinaryenQFMAVecF64x2(void); -BINARYEN_API BinaryenOp BinaryenQFMSVecF64x2(void); BINARYEN_API BinaryenOp BinaryenAddVecF64x2(void); BINARYEN_API BinaryenOp BinaryenSubVecF64x2(void); BINARYEN_API BinaryenOp BinaryenMulVecF64x2(void); @@ -530,12 +523,8 @@ BINARYEN_API BinaryenOp BinaryenNearestVecF64x2(void); // TODO: Add extending pairwise adds to C and JS APIs once merged BINARYEN_API BinaryenOp BinaryenTruncSatSVecF32x4ToVecI32x4(void); BINARYEN_API BinaryenOp BinaryenTruncSatUVecF32x4ToVecI32x4(void); -BINARYEN_API BinaryenOp BinaryenTruncSatSVecF64x2ToVecI64x2(void); -BINARYEN_API BinaryenOp BinaryenTruncSatUVecF64x2ToVecI64x2(void); BINARYEN_API BinaryenOp BinaryenConvertSVecI32x4ToVecF32x4(void); BINARYEN_API BinaryenOp BinaryenConvertUVecI32x4ToVecF32x4(void); -BINARYEN_API BinaryenOp BinaryenConvertSVecI64x2ToVecF64x2(void); -BINARYEN_API BinaryenOp BinaryenConvertUVecI64x2ToVecF64x2(void); BINARYEN_API BinaryenOp BinaryenLoadSplatVec8x16(void); BINARYEN_API BinaryenOp BinaryenLoadSplatVec16x8(void); BINARYEN_API BinaryenOp BinaryenLoadSplatVec32x4(void); @@ -551,14 +540,14 @@ BINARYEN_API BinaryenOp BinaryenNarrowSVecI16x8ToVecI8x16(void); BINARYEN_API BinaryenOp BinaryenNarrowUVecI16x8ToVecI8x16(void); BINARYEN_API BinaryenOp BinaryenNarrowSVecI32x4ToVecI16x8(void); BINARYEN_API BinaryenOp BinaryenNarrowUVecI32x4ToVecI16x8(void); -BINARYEN_API BinaryenOp BinaryenWidenLowSVecI8x16ToVecI16x8(void); -BINARYEN_API BinaryenOp BinaryenWidenHighSVecI8x16ToVecI16x8(void); -BINARYEN_API BinaryenOp BinaryenWidenLowUVecI8x16ToVecI16x8(void); -BINARYEN_API BinaryenOp BinaryenWidenHighUVecI8x16ToVecI16x8(void); -BINARYEN_API BinaryenOp BinaryenWidenLowSVecI16x8ToVecI32x4(void); -BINARYEN_API BinaryenOp BinaryenWidenHighSVecI16x8ToVecI32x4(void); -BINARYEN_API BinaryenOp BinaryenWidenLowUVecI16x8ToVecI32x4(void); -BINARYEN_API BinaryenOp BinaryenWidenHighUVecI16x8ToVecI32x4(void); +BINARYEN_API BinaryenOp BinaryenExtendLowSVecI8x16ToVecI16x8(void); +BINARYEN_API BinaryenOp BinaryenExtendHighSVecI8x16ToVecI16x8(void); +BINARYEN_API BinaryenOp BinaryenExtendLowUVecI8x16ToVecI16x8(void); +BINARYEN_API BinaryenOp BinaryenExtendHighUVecI8x16ToVecI16x8(void); +BINARYEN_API BinaryenOp BinaryenExtendLowSVecI16x8ToVecI32x4(void); +BINARYEN_API BinaryenOp BinaryenExtendHighSVecI16x8ToVecI32x4(void); +BINARYEN_API BinaryenOp BinaryenExtendLowUVecI16x8ToVecI32x4(void); +BINARYEN_API BinaryenOp BinaryenExtendHighUVecI16x8ToVecI32x4(void); BINARYEN_API BinaryenOp BinaryenSwizzleVec8x16(void); BINARYEN_API BinaryenOp BinaryenRefIsNull(void); BINARYEN_API BinaryenOp BinaryenRefIsFunc(void); diff --git a/src/gen-s-parser.inc b/src/gen-s-parser.inc index 4e5a61d92..6dfb92f5f 100644 --- a/src/gen-s-parser.inc +++ b/src/gen-s-parser.inc @@ -415,17 +415,6 @@ switch (op[0]) { default: goto parse_error; } } - case 'q': { - switch (op[9]) { - case 'a': - if (strcmp(op, "f32x4.qfma") == 0) { return makeSIMDTernary(s, SIMDTernaryOp::QFMAF32x4); } - goto parse_error; - case 's': - if (strcmp(op, "f32x4.qfms") == 0) { return makeSIMDTernary(s, SIMDTernaryOp::QFMSF32x4); } - goto parse_error; - default: goto parse_error; - } - } case 'r': if (strcmp(op, "f32x4.replace_lane") == 0) { return makeSIMDReplace(s, SIMDReplaceOp::ReplaceLaneVecF32x4, 4); } goto parse_error; @@ -625,29 +614,13 @@ switch (op[0]) { if (strcmp(op, "f64x2.ceil") == 0) { return makeUnary(s, UnaryOp::CeilVecF64x2); } goto parse_error; case 'o': { - switch (op[14]) { - case 'i': { - switch (op[20]) { - case 's': - if (strcmp(op, "f64x2.convert_i64x2_s") == 0) { return makeUnary(s, UnaryOp::ConvertSVecI64x2ToVecF64x2); } - goto parse_error; - case 'u': - if (strcmp(op, "f64x2.convert_i64x2_u") == 0) { return makeUnary(s, UnaryOp::ConvertUVecI64x2ToVecF64x2); } - goto parse_error; - default: goto parse_error; - } - } - case 'l': { - switch (op[24]) { - case 's': - if (strcmp(op, "f64x2.convert_low_i32x4_s") == 0) { return makeUnary(s, UnaryOp::ConvertLowSVecI32x4ToVecF64x2); } - goto parse_error; - case 'u': - if (strcmp(op, "f64x2.convert_low_i32x4_u") == 0) { return makeUnary(s, UnaryOp::ConvertLowUVecI32x4ToVecF64x2); } - goto parse_error; - default: goto parse_error; - } - } + switch (op[24]) { + case 's': + if (strcmp(op, "f64x2.convert_low_i32x4_s") == 0) { return makeUnary(s, UnaryOp::ConvertLowSVecI32x4ToVecF64x2); } + goto parse_error; + case 'u': + if (strcmp(op, "f64x2.convert_low_i32x4_u") == 0) { return makeUnary(s, UnaryOp::ConvertLowUVecI32x4ToVecF64x2); } + goto parse_error; default: goto parse_error; } } @@ -740,17 +713,6 @@ switch (op[0]) { default: goto parse_error; } } - case 'q': { - switch (op[9]) { - case 'a': - if (strcmp(op, "f64x2.qfma") == 0) { return makeSIMDTernary(s, SIMDTernaryOp::QFMAF64x2); } - goto parse_error; - case 's': - if (strcmp(op, "f64x2.qfms") == 0) { return makeSIMDTernary(s, SIMDTernaryOp::QFMSF64x2); } - goto parse_error; - default: goto parse_error; - } - } case 'r': if (strcmp(op, "f64x2.replace_lane") == 0) { return makeSIMDReplace(s, SIMDReplaceOp::ReplaceLaneVecF64x2, 2); } goto parse_error; @@ -806,12 +768,12 @@ switch (op[0]) { if (strcmp(op, "i16x8.add") == 0) { return makeBinary(s, BinaryOp::AddVecI16x8); } goto parse_error; case '_': { - switch (op[19]) { + switch (op[14]) { case 's': - if (strcmp(op, "i16x8.add_saturate_s") == 0) { return makeBinary(s, BinaryOp::AddSatSVecI16x8); } + if (strcmp(op, "i16x8.add_sat_s") == 0) { return makeBinary(s, BinaryOp::AddSatSVecI16x8); } goto parse_error; case 'u': - if (strcmp(op, "i16x8.add_saturate_u") == 0) { return makeBinary(s, BinaryOp::AddSatUVecI16x8); } + if (strcmp(op, "i16x8.add_sat_u") == 0) { return makeBinary(s, BinaryOp::AddSatUVecI16x8); } goto parse_error; default: goto parse_error; } @@ -822,9 +784,6 @@ switch (op[0]) { case 'l': if (strcmp(op, "i16x8.all_true") == 0) { return makeUnary(s, UnaryOp::AllTrueVecI16x8); } goto parse_error; - case 'n': - if (strcmp(op, "i16x8.any_true") == 0) { return makeUnary(s, UnaryOp::AnyTrueVecI16x8); } - goto parse_error; case 'v': if (strcmp(op, "i16x8.avgr_u") == 0) { return makeBinary(s, BinaryOp::AvgrUVecI16x8); } goto parse_error; @@ -852,6 +811,33 @@ switch (op[0]) { default: goto parse_error; } } + case 'e': { + switch (op[13]) { + case 'h': { + switch (op[24]) { + case 's': + if (strcmp(op, "i16x8.extend_high_i8x16_s") == 0) { return makeUnary(s, UnaryOp::ExtendHighSVecI8x16ToVecI16x8); } + goto parse_error; + case 'u': + if (strcmp(op, "i16x8.extend_high_i8x16_u") == 0) { return makeUnary(s, UnaryOp::ExtendHighUVecI8x16ToVecI16x8); } + goto parse_error; + default: goto parse_error; + } + } + case 'l': { + switch (op[23]) { + case 's': + if (strcmp(op, "i16x8.extend_low_i8x16_s") == 0) { return makeUnary(s, UnaryOp::ExtendLowSVecI8x16ToVecI16x8); } + goto parse_error; + case 'u': + if (strcmp(op, "i16x8.extend_low_i8x16_u") == 0) { return makeUnary(s, UnaryOp::ExtendLowUVecI8x16ToVecI16x8); } + goto parse_error; + default: goto parse_error; + } + } + default: goto parse_error; + } + } case 'm': { switch (op[13]) { case 'h': { @@ -936,17 +922,6 @@ switch (op[0]) { default: goto parse_error; } } - case 'o': { - switch (op[14]) { - case 's': - if (strcmp(op, "i16x8.load8x8_s") == 0) { return makeSIMDLoad(s, SIMDLoadOp::LoadExtSVec8x8ToVecI16x8); } - goto parse_error; - case 'u': - if (strcmp(op, "i16x8.load8x8_u") == 0) { return makeSIMDLoad(s, SIMDLoadOp::LoadExtUVec8x8ToVecI16x8); } - goto parse_error; - default: goto parse_error; - } - } case 't': { switch (op[9]) { case 's': @@ -1054,12 +1029,12 @@ switch (op[0]) { if (strcmp(op, "i16x8.sub") == 0) { return makeBinary(s, BinaryOp::SubVecI16x8); } goto parse_error; case '_': { - switch (op[19]) { + switch (op[14]) { case 's': - if (strcmp(op, "i16x8.sub_saturate_s") == 0) { return makeBinary(s, BinaryOp::SubSatSVecI16x8); } + if (strcmp(op, "i16x8.sub_sat_s") == 0) { return makeBinary(s, BinaryOp::SubSatSVecI16x8); } goto parse_error; case 'u': - if (strcmp(op, "i16x8.sub_saturate_u") == 0) { return makeBinary(s, BinaryOp::SubSatUVecI16x8); } + if (strcmp(op, "i16x8.sub_sat_u") == 0) { return makeBinary(s, BinaryOp::SubSatUVecI16x8); } goto parse_error; default: goto parse_error; } @@ -1070,33 +1045,6 @@ switch (op[0]) { default: goto parse_error; } } - case 'w': { - switch (op[12]) { - case 'h': { - switch (op[23]) { - case 's': - if (strcmp(op, "i16x8.widen_high_i8x16_s") == 0) { return makeUnary(s, UnaryOp::WidenHighSVecI8x16ToVecI16x8); } - goto parse_error; - case 'u': - if (strcmp(op, "i16x8.widen_high_i8x16_u") == 0) { return makeUnary(s, UnaryOp::WidenHighUVecI8x16ToVecI16x8); } - goto parse_error; - default: goto parse_error; - } - } - case 'l': { - switch (op[22]) { - case 's': - if (strcmp(op, "i16x8.widen_low_i8x16_s") == 0) { return makeUnary(s, UnaryOp::WidenLowSVecI8x16ToVecI16x8); } - goto parse_error; - case 'u': - if (strcmp(op, "i16x8.widen_low_i8x16_u") == 0) { return makeUnary(s, UnaryOp::WidenLowUVecI8x16ToVecI16x8); } - goto parse_error; - default: goto parse_error; - } - } - default: goto parse_error; - } - } default: goto parse_error; } } @@ -1587,9 +1535,6 @@ switch (op[0]) { case 'l': if (strcmp(op, "i32x4.all_true") == 0) { return makeUnary(s, UnaryOp::AllTrueVecI32x4); } goto parse_error; - case 'n': - if (strcmp(op, "i32x4.any_true") == 0) { return makeUnary(s, UnaryOp::AnyTrueVecI32x4); } - goto parse_error; default: goto parse_error; } } @@ -1617,6 +1562,33 @@ switch (op[0]) { default: goto parse_error; } } + case 'e': { + switch (op[13]) { + case 'h': { + switch (op[24]) { + case 's': + if (strcmp(op, "i32x4.extend_high_i16x8_s") == 0) { return makeUnary(s, UnaryOp::ExtendHighSVecI16x8ToVecI32x4); } + goto parse_error; + case 'u': + if (strcmp(op, "i32x4.extend_high_i16x8_u") == 0) { return makeUnary(s, UnaryOp::ExtendHighUVecI16x8ToVecI32x4); } + goto parse_error; + default: goto parse_error; + } + } + case 'l': { + switch (op[23]) { + case 's': + if (strcmp(op, "i32x4.extend_low_i16x8_s") == 0) { return makeUnary(s, UnaryOp::ExtendLowSVecI16x8ToVecI32x4); } + goto parse_error; + case 'u': + if (strcmp(op, "i32x4.extend_low_i16x8_u") == 0) { return makeUnary(s, UnaryOp::ExtendLowUVecI16x8ToVecI32x4); } + goto parse_error; + default: goto parse_error; + } + } + default: goto parse_error; + } + } case 'm': { switch (op[13]) { case 'h': { @@ -1693,17 +1665,6 @@ switch (op[0]) { default: goto parse_error; } } - case 'o': { - switch (op[15]) { - case 's': - if (strcmp(op, "i32x4.load16x4_s") == 0) { return makeSIMDLoad(s, SIMDLoadOp::LoadExtSVec16x4ToVecI32x4); } - goto parse_error; - case 'u': - if (strcmp(op, "i32x4.load16x4_u") == 0) { return makeSIMDLoad(s, SIMDLoadOp::LoadExtUVec16x4ToVecI32x4); } - goto parse_error; - default: goto parse_error; - } - } case 't': { switch (op[9]) { case 's': @@ -1806,50 +1767,12 @@ switch (op[0]) { } } case '6': { - switch (op[27]) { - case 's': - if (strcmp(op, "i32x4.trunc_sat_f64x2_zero_s") == 0) { return makeUnary(s, UnaryOp::TruncSatZeroSVecF64x2ToVecI32x4); } - goto parse_error; - case 'u': - if (strcmp(op, "i32x4.trunc_sat_f64x2_zero_u") == 0) { return makeUnary(s, UnaryOp::TruncSatZeroUVecF64x2ToVecI32x4); } - goto parse_error; - default: goto parse_error; - } - } - default: goto parse_error; - } - } - case 'w': { - switch (op[12]) { - case 'h': { - switch (op[23]) { - case 's': - if (strcmp(op, "i32x4.widen_high_i16x8_s") == 0) { return makeUnary(s, UnaryOp::WidenHighSVecI16x8ToVecI32x4); } - goto parse_error; - case 'u': - if (strcmp(op, "i32x4.widen_high_i16x8_u") == 0) { return makeUnary(s, UnaryOp::WidenHighUVecI16x8ToVecI32x4); } - goto parse_error; - default: goto parse_error; - } - } - case 'i': { - switch (op[18]) { - case 's': - if (strcmp(op, "i32x4.widen_i8x16_s") == 0) { return makeSIMDWiden(s, SIMDWidenOp::WidenSVecI8x16ToVecI32x4); } - goto parse_error; - case 'u': - if (strcmp(op, "i32x4.widen_i8x16_u") == 0) { return makeSIMDWiden(s, SIMDWidenOp::WidenUVecI8x16ToVecI32x4); } - goto parse_error; - default: goto parse_error; - } - } - case 'l': { switch (op[22]) { case 's': - if (strcmp(op, "i32x4.widen_low_i16x8_s") == 0) { return makeUnary(s, UnaryOp::WidenLowSVecI16x8ToVecI32x4); } + if (strcmp(op, "i32x4.trunc_sat_f64x2_s_zero") == 0) { return makeUnary(s, UnaryOp::TruncSatZeroSVecF64x2ToVecI32x4); } goto parse_error; case 'u': - if (strcmp(op, "i32x4.widen_low_i16x8_u") == 0) { return makeUnary(s, UnaryOp::WidenLowUVecI16x8ToVecI32x4); } + if (strcmp(op, "i32x4.trunc_sat_f64x2_u_zero") == 0) { return makeUnary(s, UnaryOp::TruncSatZeroUVecF64x2ToVecI32x4); } goto parse_error; default: goto parse_error; } @@ -2388,9 +2311,20 @@ switch (op[0]) { } case 'x': { switch (op[6]) { - case 'a': - if (strcmp(op, "i64x2.add") == 0) { return makeBinary(s, BinaryOp::AddVecI64x2); } - goto parse_error; + case 'a': { + switch (op[7]) { + case 'b': + if (strcmp(op, "i64x2.abs") == 0) { return makeUnary(s, UnaryOp::AbsVecI64x2); } + goto parse_error; + case 'd': + if (strcmp(op, "i64x2.add") == 0) { return makeBinary(s, BinaryOp::AddVecI64x2); } + goto parse_error; + case 'l': + if (strcmp(op, "i64x2.all_true") == 0) { return makeUnary(s, UnaryOp::AllTrueVecI64x2); } + goto parse_error; + default: goto parse_error; + } + } case 'b': if (strcmp(op, "i64x2.bitmask") == 0) { return makeUnary(s, UnaryOp::BitmaskVecI64x2); } goto parse_error; @@ -2401,6 +2335,33 @@ switch (op[0]) { goto parse_error; case 'x': { switch (op[9]) { + case 'e': { + switch (op[13]) { + case 'h': { + switch (op[24]) { + case 's': + if (strcmp(op, "i64x2.extend_high_i32x4_s") == 0) { return makeUnary(s, UnaryOp::ExtendHighSVecI32x4ToVecI64x2); } + goto parse_error; + case 'u': + if (strcmp(op, "i64x2.extend_high_i32x4_u") == 0) { return makeUnary(s, UnaryOp::ExtendHighUVecI32x4ToVecI64x2); } + goto parse_error; + default: goto parse_error; + } + } + case 'l': { + switch (op[23]) { + case 's': + if (strcmp(op, "i64x2.extend_low_i32x4_s") == 0) { return makeUnary(s, UnaryOp::ExtendLowSVecI32x4ToVecI64x2); } + goto parse_error; + case 'u': + if (strcmp(op, "i64x2.extend_low_i32x4_u") == 0) { return makeUnary(s, UnaryOp::ExtendLowUVecI32x4ToVecI64x2); } + goto parse_error; + default: goto parse_error; + } + } + default: goto parse_error; + } + } case 'm': { switch (op[13]) { case 'h': { @@ -2437,13 +2398,24 @@ switch (op[0]) { default: goto parse_error; } } + case 'g': { + switch (op[7]) { + case 'e': + if (strcmp(op, "i64x2.ge_s") == 0) { return makeBinary(s, BinaryOp::GeSVecI64x2); } + goto parse_error; + case 't': + if (strcmp(op, "i64x2.gt_s") == 0) { return makeBinary(s, BinaryOp::GtSVecI64x2); } + goto parse_error; + default: goto parse_error; + } + } case 'l': { - switch (op[15]) { - case 's': - if (strcmp(op, "i64x2.load32x2_s") == 0) { return makeSIMDLoad(s, SIMDLoadOp::LoadExtSVec32x2ToVecI64x2); } + switch (op[7]) { + case 'e': + if (strcmp(op, "i64x2.le_s") == 0) { return makeBinary(s, BinaryOp::LeSVecI64x2); } goto parse_error; - case 'u': - if (strcmp(op, "i64x2.load32x2_u") == 0) { return makeSIMDLoad(s, SIMDLoadOp::LoadExtUVec32x2ToVecI64x2); } + case 't': + if (strcmp(op, "i64x2.lt_s") == 0) { return makeBinary(s, BinaryOp::LtSVecI64x2); } goto parse_error; default: goto parse_error; } @@ -2451,9 +2423,17 @@ switch (op[0]) { case 'm': if (strcmp(op, "i64x2.mul") == 0) { return makeBinary(s, BinaryOp::MulVecI64x2); } goto parse_error; - case 'n': - if (strcmp(op, "i64x2.neg") == 0) { return makeUnary(s, UnaryOp::NegVecI64x2); } - goto parse_error; + case 'n': { + switch (op[8]) { + case '\0': + if (strcmp(op, "i64x2.ne") == 0) { return makeBinary(s, BinaryOp::NeVecI64x2); } + goto parse_error; + case 'g': + if (strcmp(op, "i64x2.neg") == 0) { return makeUnary(s, UnaryOp::NegVecI64x2); } + goto parse_error; + default: goto parse_error; + } + } case 'r': if (strcmp(op, "i64x2.replace_lane") == 0) { return makeSIMDReplace(s, SIMDReplaceOp::ReplaceLaneVecI64x2, 2); } goto parse_error; @@ -2487,44 +2467,6 @@ switch (op[0]) { default: goto parse_error; } } - case 't': { - switch (op[22]) { - case 's': - if (strcmp(op, "i64x2.trunc_sat_f64x2_s") == 0) { return makeUnary(s, UnaryOp::TruncSatSVecF64x2ToVecI64x2); } - goto parse_error; - case 'u': - if (strcmp(op, "i64x2.trunc_sat_f64x2_u") == 0) { return makeUnary(s, UnaryOp::TruncSatUVecF64x2ToVecI64x2); } - goto parse_error; - default: goto parse_error; - } - } - case 'w': { - switch (op[12]) { - case 'h': { - switch (op[23]) { - case 's': - if (strcmp(op, "i64x2.widen_high_i32x4_s") == 0) { return makeUnary(s, UnaryOp::WidenHighSVecI32x4ToVecI64x2); } - goto parse_error; - case 'u': - if (strcmp(op, "i64x2.widen_high_i32x4_u") == 0) { return makeUnary(s, UnaryOp::WidenHighUVecI32x4ToVecI64x2); } - goto parse_error; - default: goto parse_error; - } - } - case 'l': { - switch (op[22]) { - case 's': - if (strcmp(op, "i64x2.widen_low_i32x4_s") == 0) { return makeUnary(s, UnaryOp::WidenLowSVecI32x4ToVecI64x2); } - goto parse_error; - case 'u': - if (strcmp(op, "i64x2.widen_low_i32x4_u") == 0) { return makeUnary(s, UnaryOp::WidenLowUVecI32x4ToVecI64x2); } - goto parse_error; - default: goto parse_error; - } - } - default: goto parse_error; - } - } default: goto parse_error; } } @@ -2544,12 +2486,12 @@ switch (op[0]) { if (strcmp(op, "i8x16.add") == 0) { return makeBinary(s, BinaryOp::AddVecI8x16); } goto parse_error; case '_': { - switch (op[19]) { + switch (op[14]) { case 's': - if (strcmp(op, "i8x16.add_saturate_s") == 0) { return makeBinary(s, BinaryOp::AddSatSVecI8x16); } + if (strcmp(op, "i8x16.add_sat_s") == 0) { return makeBinary(s, BinaryOp::AddSatSVecI8x16); } goto parse_error; case 'u': - if (strcmp(op, "i8x16.add_saturate_u") == 0) { return makeBinary(s, BinaryOp::AddSatUVecI8x16); } + if (strcmp(op, "i8x16.add_sat_u") == 0) { return makeBinary(s, BinaryOp::AddSatUVecI8x16); } goto parse_error; default: goto parse_error; } @@ -2560,9 +2502,6 @@ switch (op[0]) { case 'l': if (strcmp(op, "i8x16.all_true") == 0) { return makeUnary(s, UnaryOp::AllTrueVecI8x16); } goto parse_error; - case 'n': - if (strcmp(op, "i8x16.any_true") == 0) { return makeUnary(s, UnaryOp::AnyTrueVecI8x16); } - goto parse_error; case 'v': if (strcmp(op, "i8x16.avgr_u") == 0) { return makeBinary(s, BinaryOp::AvgrUVecI8x16); } goto parse_error; @@ -2669,9 +2608,6 @@ switch (op[0]) { default: goto parse_error; } } - case 'u': - if (strcmp(op, "i8x16.mul") == 0) { return makeBinary(s, BinaryOp::MulVecI8x16); } - goto parse_error; default: goto parse_error; } } @@ -2726,6 +2662,9 @@ switch (op[0]) { default: goto parse_error; } } + case 'u': + if (strcmp(op, "i8x16.shuffle") == 0) { return makeSIMDShuffle(s); } + goto parse_error; default: goto parse_error; } } @@ -2738,12 +2677,12 @@ switch (op[0]) { if (strcmp(op, "i8x16.sub") == 0) { return makeBinary(s, BinaryOp::SubVecI8x16); } goto parse_error; case '_': { - switch (op[19]) { + switch (op[14]) { case 's': - if (strcmp(op, "i8x16.sub_saturate_s") == 0) { return makeBinary(s, BinaryOp::SubSatSVecI8x16); } + if (strcmp(op, "i8x16.sub_sat_s") == 0) { return makeBinary(s, BinaryOp::SubSatSVecI8x16); } goto parse_error; case 'u': - if (strcmp(op, "i8x16.sub_saturate_u") == 0) { return makeBinary(s, BinaryOp::SubSatUVecI8x16); } + if (strcmp(op, "i8x16.sub_sat_u") == 0) { return makeBinary(s, BinaryOp::SubSatUVecI8x16); } goto parse_error; default: goto parse_error; } @@ -2751,6 +2690,9 @@ switch (op[0]) { default: goto parse_error; } } + case 'w': + if (strcmp(op, "i8x16.swizzle") == 0) { return makeBinary(s, BinaryOp::SwizzleVec8x16); } + goto parse_error; default: goto parse_error; } } @@ -2827,25 +2769,9 @@ switch (op[0]) { case 'n': if (strcmp(op, "nop") == 0) { return makeNop(); } goto parse_error; - case 'p': { - switch (op[1]) { - case 'o': - if (strcmp(op, "pop") == 0) { return makePop(s); } - goto parse_error; - case 'r': { - switch (op[9]) { - case 'n': - if (strcmp(op, "prefetch.nt") == 0) { return makePrefetch(s, PrefetchOp::PrefetchNontemporal); } - goto parse_error; - case 't': - if (strcmp(op, "prefetch.t") == 0) { return makePrefetch(s, PrefetchOp::PrefetchTemporal); } - goto parse_error; - default: goto parse_error; - } - } - default: goto parse_error; - } - } + case 'p': + if (strcmp(op, "pop") == 0) { return makePop(s); } + goto parse_error; case 'r': { switch (op[1]) { case 'e': { @@ -3036,154 +2962,167 @@ switch (op[0]) { if (strcmp(op, "unreachable") == 0) { return makeUnreachable(); } goto parse_error; case 'v': { - switch (op[1]) { - case '1': { - switch (op[2]) { - case '2': { - switch (op[5]) { - case 'a': { - switch (op[8]) { - case '\0': - if (strcmp(op, "v128.and") == 0) { return makeBinary(s, BinaryOp::AndVec128); } + switch (op[5]) { + case 'a': { + switch (op[7]) { + case 'd': { + switch (op[8]) { + case '\0': + if (strcmp(op, "v128.and") == 0) { return makeBinary(s, BinaryOp::AndVec128); } + goto parse_error; + case 'n': + if (strcmp(op, "v128.andnot") == 0) { return makeBinary(s, BinaryOp::AndNotVec128); } + goto parse_error; + default: goto parse_error; + } + } + case 'y': + if (strcmp(op, "v128.any_true") == 0) { return makeUnary(s, UnaryOp::AnyTrueVec128); } + goto parse_error; + default: goto parse_error; + } + } + case 'b': + if (strcmp(op, "v128.bitselect") == 0) { return makeSIMDTernary(s, SIMDTernaryOp::Bitselect); } + goto parse_error; + case 'c': + if (strcmp(op, "v128.const") == 0) { return makeConst(s, Type::v128); } + goto parse_error; + case 'l': { + switch (op[9]) { + case '\0': + if (strcmp(op, "v128.load") == 0) { return makeLoad(s, Type::v128, /*isAtomic=*/false); } + goto parse_error; + case '1': { + switch (op[11]) { + case '_': { + switch (op[12]) { + case 'l': + if (strcmp(op, "v128.load16_lane") == 0) { return makeSIMDLoadStoreLane(s, LoadLaneVec16x8); } goto parse_error; - case 'n': - if (strcmp(op, "v128.andnot") == 0) { return makeBinary(s, BinaryOp::AndNotVec128); } + case 's': + if (strcmp(op, "v128.load16_splat") == 0) { return makeSIMDLoad(s, SIMDLoadOp::LoadSplatVec16x8); } goto parse_error; default: goto parse_error; } } - case 'b': - if (strcmp(op, "v128.bitselect") == 0) { return makeSIMDTernary(s, SIMDTernaryOp::Bitselect); } - goto parse_error; - case 'c': - if (strcmp(op, "v128.const") == 0) { return makeConst(s, Type::v128); } - goto parse_error; - case 'l': { - switch (op[9]) { - case '\0': - if (strcmp(op, "v128.load") == 0) { return makeLoad(s, Type::v128, /*isAtomic=*/false); } - goto parse_error; - case '1': - if (strcmp(op, "v128.load16_lane") == 0) { return makeSIMDLoadStoreLane(s, LoadLaneVec16x8); } + case 'x': { + switch (op[14]) { + case 's': + if (strcmp(op, "v128.load16x4_s") == 0) { return makeSIMDLoad(s, SIMDLoadOp::LoadExtSVec16x4ToVecI32x4); } goto parse_error; - case '3': { - switch (op[12]) { - case 'l': - if (strcmp(op, "v128.load32_lane") == 0) { return makeSIMDLoadStoreLane(s, LoadLaneVec32x4); } - goto parse_error; - case 'z': - if (strcmp(op, "v128.load32_zero") == 0) { return makeSIMDLoad(s, SIMDLoadOp::Load32Zero); } - goto parse_error; - default: goto parse_error; - } - } - case '6': { - switch (op[12]) { - case 'l': - if (strcmp(op, "v128.load64_lane") == 0) { return makeSIMDLoadStoreLane(s, LoadLaneVec64x2); } - goto parse_error; - case 'z': - if (strcmp(op, "v128.load64_zero") == 0) { return makeSIMDLoad(s, SIMDLoadOp::Load64Zero); } - goto parse_error; - default: goto parse_error; - } - } - case '8': - if (strcmp(op, "v128.load8_lane") == 0) { return makeSIMDLoadStoreLane(s, LoadLaneVec8x16); } + case 'u': + if (strcmp(op, "v128.load16x4_u") == 0) { return makeSIMDLoad(s, SIMDLoadOp::LoadExtUVec16x4ToVecI32x4); } goto parse_error; default: goto parse_error; } } - case 'n': - if (strcmp(op, "v128.not") == 0) { return makeUnary(s, UnaryOp::NotVec128); } - goto parse_error; - case 'o': - if (strcmp(op, "v128.or") == 0) { return makeBinary(s, BinaryOp::OrVec128); } - goto parse_error; - case 's': { - switch (op[10]) { - case '\0': - if (strcmp(op, "v128.store") == 0) { return makeStore(s, Type::v128, /*isAtomic=*/false); } + default: goto parse_error; + } + } + case '3': { + switch (op[11]) { + case '_': { + switch (op[12]) { + case 'l': + if (strcmp(op, "v128.load32_lane") == 0) { return makeSIMDLoadStoreLane(s, LoadLaneVec32x4); } goto parse_error; - case '1': - if (strcmp(op, "v128.store16_lane") == 0) { return makeSIMDLoadStoreLane(s, StoreLaneVec16x8); } + case 's': + if (strcmp(op, "v128.load32_splat") == 0) { return makeSIMDLoad(s, SIMDLoadOp::LoadSplatVec32x4); } goto parse_error; - case '3': - if (strcmp(op, "v128.store32_lane") == 0) { return makeSIMDLoadStoreLane(s, StoreLaneVec32x4); } + case 'z': + if (strcmp(op, "v128.load32_zero") == 0) { return makeSIMDLoad(s, SIMDLoadOp::Load32Zero); } goto parse_error; - case '6': - if (strcmp(op, "v128.store64_lane") == 0) { return makeSIMDLoadStoreLane(s, StoreLaneVec64x2); } + default: goto parse_error; + } + } + case 'x': { + switch (op[14]) { + case 's': + if (strcmp(op, "v128.load32x2_s") == 0) { return makeSIMDLoad(s, SIMDLoadOp::LoadExtSVec32x2ToVecI64x2); } goto parse_error; - case '8': - if (strcmp(op, "v128.store8_lane") == 0) { return makeSIMDLoadStoreLane(s, StoreLaneVec8x16); } + case 'u': + if (strcmp(op, "v128.load32x2_u") == 0) { return makeSIMDLoad(s, SIMDLoadOp::LoadExtUVec32x2ToVecI64x2); } goto parse_error; default: goto parse_error; } } - case 'x': - if (strcmp(op, "v128.xor") == 0) { return makeBinary(s, BinaryOp::XorVec128); } - goto parse_error; default: goto parse_error; } } case '6': { - switch (op[6]) { + switch (op[12]) { case 'l': - if (strcmp(op, "v16x8.load_splat") == 0) { return makeSIMDLoad(s, SIMDLoadOp::LoadSplatVec16x8); } + if (strcmp(op, "v128.load64_lane") == 0) { return makeSIMDLoadStoreLane(s, LoadLaneVec64x2); } goto parse_error; case 's': - if (strcmp(op, "v16x8.signselect") == 0) { return makeSIMDTernary(s, SIMDTernaryOp::SignSelectVec16x8); } + if (strcmp(op, "v128.load64_splat") == 0) { return makeSIMDLoad(s, SIMDLoadOp::LoadSplatVec64x2); } + goto parse_error; + case 'z': + if (strcmp(op, "v128.load64_zero") == 0) { return makeSIMDLoad(s, SIMDLoadOp::Load64Zero); } goto parse_error; default: goto parse_error; } } + case '8': { + switch (op[10]) { + case '_': { + switch (op[11]) { + case 'l': + if (strcmp(op, "v128.load8_lane") == 0) { return makeSIMDLoadStoreLane(s, LoadLaneVec8x16); } + goto parse_error; + case 's': + if (strcmp(op, "v128.load8_splat") == 0) { return makeSIMDLoad(s, SIMDLoadOp::LoadSplatVec8x16); } + goto parse_error; + default: goto parse_error; + } + } + case 'x': { + switch (op[13]) { + case 's': + if (strcmp(op, "v128.load8x8_s") == 0) { return makeSIMDLoad(s, SIMDLoadOp::LoadExtSVec8x8ToVecI16x8); } + goto parse_error; + case 'u': + if (strcmp(op, "v128.load8x8_u") == 0) { return makeSIMDLoad(s, SIMDLoadOp::LoadExtUVec8x8ToVecI16x8); } + goto parse_error; + default: goto parse_error; + } + } + default: goto parse_error; + } + } default: goto parse_error; } } - case '3': { - switch (op[6]) { - case 'l': - if (strcmp(op, "v32x4.load_splat") == 0) { return makeSIMDLoad(s, SIMDLoadOp::LoadSplatVec32x4); } + case 'n': + if (strcmp(op, "v128.not") == 0) { return makeUnary(s, UnaryOp::NotVec128); } + goto parse_error; + case 'o': + if (strcmp(op, "v128.or") == 0) { return makeBinary(s, BinaryOp::OrVec128); } + goto parse_error; + case 's': { + switch (op[10]) { + case '\0': + if (strcmp(op, "v128.store") == 0) { return makeStore(s, Type::v128, /*isAtomic=*/false); } goto parse_error; - case 's': - if (strcmp(op, "v32x4.signselect") == 0) { return makeSIMDTernary(s, SIMDTernaryOp::SignSelectVec32x4); } + case '1': + if (strcmp(op, "v128.store16_lane") == 0) { return makeSIMDLoadStoreLane(s, StoreLaneVec16x8); } goto parse_error; - default: goto parse_error; - } - } - case '6': { - switch (op[6]) { - case 'l': - if (strcmp(op, "v64x2.load_splat") == 0) { return makeSIMDLoad(s, SIMDLoadOp::LoadSplatVec64x2); } + case '3': + if (strcmp(op, "v128.store32_lane") == 0) { return makeSIMDLoadStoreLane(s, StoreLaneVec32x4); } goto parse_error; - case 's': - if (strcmp(op, "v64x2.signselect") == 0) { return makeSIMDTernary(s, SIMDTernaryOp::SignSelectVec64x2); } + case '6': + if (strcmp(op, "v128.store64_lane") == 0) { return makeSIMDLoadStoreLane(s, StoreLaneVec64x2); } goto parse_error; - default: goto parse_error; - } - } - case '8': { - switch (op[6]) { - case 'l': - if (strcmp(op, "v8x16.load_splat") == 0) { return makeSIMDLoad(s, SIMDLoadOp::LoadSplatVec8x16); } + case '8': + if (strcmp(op, "v128.store8_lane") == 0) { return makeSIMDLoadStoreLane(s, StoreLaneVec8x16); } goto parse_error; - case 's': { - switch (op[7]) { - case 'h': - if (strcmp(op, "v8x16.shuffle") == 0) { return makeSIMDShuffle(s); } - goto parse_error; - case 'i': - if (strcmp(op, "v8x16.signselect") == 0) { return makeSIMDTernary(s, SIMDTernaryOp::SignSelectVec8x16); } - goto parse_error; - case 'w': - if (strcmp(op, "v8x16.swizzle") == 0) { return makeBinary(s, BinaryOp::SwizzleVec8x16); } - goto parse_error; - default: goto parse_error; - } - } default: goto parse_error; } } + case 'x': + if (strcmp(op, "v128.xor") == 0) { return makeBinary(s, BinaryOp::XorVec128); } + goto parse_error; default: goto parse_error; } } diff --git a/src/ir/ReFinalize.cpp b/src/ir/ReFinalize.cpp index 0f7bab653..5582172c3 100644 --- a/src/ir/ReFinalize.cpp +++ b/src/ir/ReFinalize.cpp @@ -115,8 +115,6 @@ void ReFinalize::visitSIMDLoad(SIMDLoad* curr) { curr->finalize(); } void ReFinalize::visitSIMDLoadStoreLane(SIMDLoadStoreLane* curr) { curr->finalize(); } -void ReFinalize::visitSIMDWiden(SIMDWiden* curr) { curr->finalize(); } -void ReFinalize::visitPrefetch(Prefetch* curr) { curr->finalize(); } void ReFinalize::visitMemoryInit(MemoryInit* curr) { curr->finalize(); } void ReFinalize::visitDataDrop(DataDrop* curr) { curr->finalize(); } void ReFinalize::visitMemoryCopy(MemoryCopy* curr) { curr->finalize(); } diff --git a/src/ir/cost.h b/src/ir/cost.h index 07219659c..5d01d2611 100644 --- a/src/ir/cost.h +++ b/src/ir/cost.h @@ -172,23 +172,23 @@ struct CostAnalyzer : public OverriddenVisitor<CostAnalyzer, Index> { case SplatVecF32x4: case SplatVecF64x2: case NotVec128: + case AnyTrueVec128: case AbsVecI8x16: case NegVecI8x16: - case AnyTrueVecI8x16: case AllTrueVecI8x16: case BitmaskVecI8x16: case PopcntVecI8x16: case AbsVecI16x8: case NegVecI16x8: - case AnyTrueVecI16x8: case AllTrueVecI16x8: case BitmaskVecI16x8: case AbsVecI32x4: case NegVecI32x4: - case AnyTrueVecI32x4: case AllTrueVecI32x4: case BitmaskVecI32x4: + case AbsVecI64x2: case NegVecI64x2: + case AllTrueVecI64x2: case BitmaskVecI64x2: case AbsVecF32x4: case NegVecF32x4: @@ -210,24 +210,20 @@ struct CostAnalyzer : public OverriddenVisitor<CostAnalyzer, Index> { case ExtAddPairwiseUVecI16x8ToI32x4: case TruncSatSVecF32x4ToVecI32x4: case TruncSatUVecF32x4ToVecI32x4: - case TruncSatSVecF64x2ToVecI64x2: - case TruncSatUVecF64x2ToVecI64x2: case ConvertSVecI32x4ToVecF32x4: case ConvertUVecI32x4ToVecF32x4: - case ConvertSVecI64x2ToVecF64x2: - case ConvertUVecI64x2ToVecF64x2: - case WidenLowSVecI8x16ToVecI16x8: - case WidenHighSVecI8x16ToVecI16x8: - case WidenLowUVecI8x16ToVecI16x8: - case WidenHighUVecI8x16ToVecI16x8: - case WidenLowSVecI16x8ToVecI32x4: - case WidenHighSVecI16x8ToVecI32x4: - case WidenLowUVecI16x8ToVecI32x4: - case WidenHighUVecI16x8ToVecI32x4: - case WidenLowSVecI32x4ToVecI64x2: - case WidenHighSVecI32x4ToVecI64x2: - case WidenLowUVecI32x4ToVecI64x2: - case WidenHighUVecI32x4ToVecI64x2: + case ExtendLowSVecI8x16ToVecI16x8: + case ExtendHighSVecI8x16ToVecI16x8: + case ExtendLowUVecI8x16ToVecI16x8: + case ExtendHighUVecI8x16ToVecI16x8: + case ExtendLowSVecI16x8ToVecI32x4: + case ExtendHighSVecI16x8ToVecI32x4: + case ExtendLowUVecI16x8ToVecI32x4: + case ExtendHighUVecI16x8ToVecI32x4: + case ExtendLowSVecI32x4ToVecI64x2: + case ExtendHighSVecI32x4ToVecI64x2: + case ExtendLowUVecI32x4ToVecI64x2: + case ExtendHighUVecI32x4ToVecI64x2: case ConvertLowSVecI32x4ToVecF64x2: case ConvertLowUVecI32x4ToVecF64x2: case TruncSatZeroSVecF64x2ToVecI32x4: @@ -377,6 +373,11 @@ struct CostAnalyzer : public OverriddenVisitor<CostAnalyzer, Index> { case GeSVecI32x4: case GeUVecI32x4: case EqVecI64x2: + case NeVecI64x2: + case LtSVecI64x2: + case LeSVecI64x2: + case GtSVecI64x2: + case GeSVecI64x2: case EqVecF32x4: case NeVecF32x4: case LtVecF32x4: @@ -399,11 +400,6 @@ struct CostAnalyzer : public OverriddenVisitor<CostAnalyzer, Index> { case SubVecI8x16: case SubSatSVecI8x16: case SubSatUVecI8x16: - ret = 1; - break; - case MulVecI8x16: - ret = 2; - break; case MinSVecI8x16: case MinUVecI8x16: case MaxSVecI8x16: @@ -522,29 +518,17 @@ struct CostAnalyzer : public OverriddenVisitor<CostAnalyzer, Index> { Index ret = 0; switch (curr->op) { case Bitselect: - case SignSelectVec8x16: - case SignSelectVec16x8: - case SignSelectVec32x4: - case SignSelectVec64x2: ret = 1; break; - case QFMAF32x4: - case QFMSF32x4: - case QFMAF64x2: - case QFMSF64x2: - ret = 2; - break; } return ret + visit(curr->a) + visit(curr->b) + visit(curr->c); } Index visitSIMDShift(SIMDShift* curr) { return 1 + visit(curr->vec) + visit(curr->shift); } - Index visitSIMDWiden(SIMDWiden* curr) { return 1 + visit(curr->vec); } Index visitSIMDShuffle(SIMDShuffle* curr) { return 1 + visit(curr->left) + visit(curr->right); } - Index visitPrefetch(Prefetch* curr) { return 0 + visit(curr->ptr); } Index visitRefNull(RefNull* curr) { return 1; } Index visitRefIs(RefIs* curr) { return 1 + visit(curr->value); } Index visitRefFunc(RefFunc* curr) { return 1; } diff --git a/src/ir/effects.h b/src/ir/effects.h index 444099fe9..de228c3f4 100644 --- a/src/ir/effects.h +++ b/src/ir/effects.h @@ -440,12 +440,6 @@ private: } parent.implicitTrap = true; } - void visitSIMDWiden(SIMDWiden* curr) {} - void visitPrefetch(Prefetch* curr) { - // Do not reorder with respect to other memory ops - parent.writesMemory = true; - parent.readsMemory = true; - } void visitMemoryInit(MemoryInit* curr) { parent.writesMemory = true; parent.implicitTrap = true; diff --git a/src/ir/features.h b/src/ir/features.h index aa87bd5bd..791a4dded 100644 --- a/src/ir/features.h +++ b/src/ir/features.h @@ -48,13 +48,10 @@ inline FeatureSet get(UnaryOp op) { case SplatVecF64x2: case NotVec128: case NegVecI8x16: - case AnyTrueVecI8x16: case AllTrueVecI8x16: case NegVecI16x8: - case AnyTrueVecI16x8: case AllTrueVecI16x8: case NegVecI32x4: - case AnyTrueVecI32x4: case AllTrueVecI32x4: case NegVecI64x2: case AbsVecF32x4: @@ -65,12 +62,8 @@ inline FeatureSet get(UnaryOp op) { case SqrtVecF64x2: case TruncSatSVecF32x4ToVecI32x4: case TruncSatUVecF32x4ToVecI32x4: - case TruncSatSVecF64x2ToVecI64x2: - case TruncSatUVecF64x2ToVecI64x2: case ConvertSVecI32x4ToVecF32x4: - case ConvertUVecI32x4ToVecF32x4: - case ConvertSVecI64x2ToVecF64x2: - case ConvertUVecI64x2ToVecF64x2: { + case ConvertUVecI32x4ToVecF32x4: { ret.setSIMD(); break; } @@ -141,7 +134,6 @@ inline FeatureSet get(BinaryOp op) { case SubVecI8x16: case SubSatSVecI8x16: case SubSatUVecI8x16: - case MulVecI8x16: case AddVecI16x8: case AddSatSVecI16x8: case AddSatUVecI16x8: diff --git a/src/js/binaryen.js-post.js b/src/js/binaryen.js-post.js index bafbaa480..4c8a8c91e 100644 --- a/src/js/binaryen.js-post.js +++ b/src/js/binaryen.js-post.js @@ -359,9 +359,9 @@ function initializeConstants() { 'XorVec128', 'AndNotVec128', 'BitselectVec128', + 'AnyTrueVec128', 'AbsVecI8x16', 'NegVecI8x16', - 'AnyTrueVecI8x16', 'AllTrueVecI8x16', 'BitmaskVecI8x16', 'ShlVecI8x16', @@ -373,7 +373,6 @@ function initializeConstants() { 'SubVecI8x16', 'SubSatSVecI8x16', 'SubSatUVecI8x16', - 'MulVecI8x16', 'MinSVecI8x16', 'MinUVecI8x16', 'MaxSVecI8x16', @@ -381,7 +380,6 @@ function initializeConstants() { 'AvgrUVecI8x16', 'AbsVecI16x8', 'NegVecI16x8', - 'AnyTrueVecI16x8', 'AllTrueVecI16x8', 'BitmaskVecI16x8', 'ShlVecI16x8', @@ -402,7 +400,6 @@ function initializeConstants() { 'DotSVecI16x8ToVecI32x4', 'AbsVecI32x4', 'NegVecI32x4', - 'AnyTrueVecI32x4', 'AllTrueVecI32x4', 'BitmaskVecI32x4', 'ShlVecI32x4', @@ -425,8 +422,6 @@ function initializeConstants() { 'AbsVecF32x4', 'NegVecF32x4', 'SqrtVecF32x4', - 'QFMAVecF32x4', - 'QFMSVecF32x4', 'AddVecF32x4', 'SubVecF32x4', 'MulVecF32x4', @@ -442,8 +437,6 @@ function initializeConstants() { 'AbsVecF64x2', 'NegVecF64x2', 'SqrtVecF64x2', - 'QFMAVecF64x2', - 'QFMSVecF64x2', 'AddVecF64x2', 'SubVecF64x2', 'MulVecF64x2', @@ -458,12 +451,8 @@ function initializeConstants() { 'NearestVecF64x2', 'TruncSatSVecF32x4ToVecI32x4', 'TruncSatUVecF32x4ToVecI32x4', - 'TruncSatSVecF64x2ToVecI64x2', - 'TruncSatUVecF64x2ToVecI64x2', 'ConvertSVecI32x4ToVecF32x4', 'ConvertUVecI32x4ToVecF32x4', - 'ConvertSVecI64x2ToVecF64x2', - 'ConvertUVecI64x2ToVecF64x2', 'LoadSplatVec8x16', 'LoadSplatVec16x8', 'LoadSplatVec32x4', @@ -478,14 +467,14 @@ function initializeConstants() { 'NarrowUVecI16x8ToVecI8x16', 'NarrowSVecI32x4ToVecI16x8', 'NarrowUVecI32x4ToVecI16x8', - 'WidenLowSVecI8x16ToVecI16x8', - 'WidenHighSVecI8x16ToVecI16x8', - 'WidenLowUVecI8x16ToVecI16x8', - 'WidenHighUVecI8x16ToVecI16x8', - 'WidenLowSVecI16x8ToVecI32x4', - 'WidenHighSVecI16x8ToVecI32x4', - 'WidenLowUVecI16x8ToVecI32x4', - 'WidenHighUVecI16x8ToVecI32x4', + 'ExtendLowSVecI8x16ToVecI16x8', + 'ExtendHighSVecI8x16ToVecI16x8', + 'ExtendLowUVecI8x16ToVecI16x8', + 'ExtendHighUVecI8x16ToVecI16x8', + 'ExtendLowSVecI16x8ToVecI32x4', + 'ExtendHighSVecI16x8ToVecI32x4', + 'ExtendLowUVecI16x8ToVecI32x4', + 'ExtendHighUVecI16x8ToVecI32x4', 'SwizzleVec8x16', 'RefIsNull', 'RefIsFunc', @@ -1434,6 +1423,36 @@ function wrapModule(module, self = {}) { 'load'(offset, align, ptr) { return Module['_BinaryenLoad'](module, 16, false, offset, align, Module['v128'], ptr); }, + 'load8_splat'(offset, align, ptr) { + return Module['_BinaryenSIMDLoad'](module, Module['LoadSplatVec8x16'], offset, align, ptr); + }, + 'load16_splat'(offset, align, ptr) { + return Module['_BinaryenSIMDLoad'](module, Module['LoadSplatVec16x8'], offset, align, ptr); + }, + 'load32_splat'(offset, align, ptr) { + return Module['_BinaryenSIMDLoad'](module, Module['LoadSplatVec32x4'], offset, align, ptr); + }, + 'load64_splat'(offset, align, ptr) { + return Module['_BinaryenSIMDLoad'](module, Module['LoadSplatVec64x2'], offset, align, ptr); + }, + 'load8x8_s'(offset, align, ptr) { + return Module['_BinaryenSIMDLoad'](module, Module['LoadExtSVec8x8ToVecI16x8'], offset, align, ptr); + }, + 'load8x8_u'(offset, align, ptr) { + return Module['_BinaryenSIMDLoad'](module, Module['LoadExtUVec8x8ToVecI16x8'], offset, align, ptr); + }, + 'load16x4_s'(offset, align, ptr) { + return Module['_BinaryenSIMDLoad'](module, Module['LoadExtSVec16x4ToVecI32x4'], offset, align, ptr); + }, + 'load16x4_u'(offset, align, ptr) { + return Module['_BinaryenSIMDLoad'](module, Module['LoadExtUVec16x4ToVecI32x4'], offset, align, ptr); + }, + 'load32x2_s'(offset, align, ptr) { + return Module['_BinaryenSIMDLoad'](module, Module['LoadExtSVec32x2ToVecI64x2'], offset, align, ptr); + }, + 'load32x2_u'(offset, align, ptr) { + return Module['_BinaryenSIMDLoad'](module, Module['LoadExtUVec32x2ToVecI64x2'], offset, align, ptr); + }, 'store'(offset, align, ptr, value) { return Module['_BinaryenStore'](module, 16, offset, align, ptr, value, Module['v128']); }, @@ -1447,6 +1466,9 @@ function wrapModule(module, self = {}) { 'not'(value) { return Module['_BinaryenUnary'](module, Module['NotVec128'], value); }, + 'any_true'(value) { + return Module['_BinaryenUnary'](module, Module['AnyTrueVec128'], value); + }, 'and'(left, right) { return Module['_BinaryenBinary'](module, Module['AndVec128'], left, right); }, @@ -1468,6 +1490,12 @@ function wrapModule(module, self = {}) { }; self['i8x16'] = { + 'shuffle'(left, right, mask) { + return preserveStack(() => Module['_BinaryenSIMDShuffle'](module, left, right, i8sToStack(mask))); + }, + 'swizzle'(left, right) { + return Module['_BinaryenBinary'](module, Module['SwizzleVec8x16'], left, right); + }, 'splat'(value) { return Module['_BinaryenUnary'](module, Module['SplatVecI8x16'], value); }, @@ -1516,9 +1544,6 @@ function wrapModule(module, self = {}) { 'neg'(value) { return Module['_BinaryenUnary'](module, Module['NegVecI8x16'], value); }, - 'any_true'(value) { - return Module['_BinaryenUnary'](module, Module['AnyTrueVecI8x16'], value); - }, 'all_true'(value) { return Module['_BinaryenUnary'](module, Module['AllTrueVecI8x16'], value); }, @@ -1552,9 +1577,6 @@ function wrapModule(module, self = {}) { 'sub_saturate_u'(left, right) { return Module['_BinaryenBinary'](module, Module['SubSatUVecI8x16'], left, right); }, - 'mul'(left, right) { - return Module['_BinaryenBinary'](module, Module['MulVecI8x16'], left, right); - }, 'min_s'(left, right) { return Module['_BinaryenBinary'](module, Module['MinSVecI8x16'], left, right); }, @@ -1627,9 +1649,6 @@ function wrapModule(module, self = {}) { 'neg'(value) { return Module['_BinaryenUnary'](module, Module['NegVecI16x8'], value); }, - 'any_true'(value) { - return Module['_BinaryenUnary'](module, Module['AnyTrueVecI16x8'], value); - }, 'all_true'(value) { return Module['_BinaryenUnary'](module, Module['AllTrueVecI16x8'], value); }, @@ -1687,23 +1706,17 @@ function wrapModule(module, self = {}) { 'narrow_i32x4_u'(left, right) { return Module['_BinaryenBinary'](module, Module['NarrowUVecI32x4ToVecI16x8'], left, right); }, - 'widen_low_i8x16_s'(value) { - return Module['_BinaryenUnary'](module, Module['WidenLowSVecI8x16ToVecI16x8'], value); + 'extend_low_i8x16_s'(value) { + return Module['_BinaryenUnary'](module, Module['ExtendLowSVecI8x16ToVecI16x8'], value); }, - 'widen_high_i8x16_s'(value) { - return Module['_BinaryenUnary'](module, Module['WidenHighSVecI8x16ToVecI16x8'], value); + 'extend_high_i8x16_s'(value) { + return Module['_BinaryenUnary'](module, Module['ExtendHighSVecI8x16ToVecI16x8'], value); }, - 'widen_low_i8x16_u'(value) { - return Module['_BinaryenUnary'](module, Module['WidenLowUVecI8x16ToVecI16x8'], value); + 'extend_low_i8x16_u'(value) { + return Module['_BinaryenUnary'](module, Module['ExtendLowUVecI8x16ToVecI16x8'], value); }, - 'widen_high_i8x16_u'(value) { - return Module['_BinaryenUnary'](module, Module['WidenHighUVecI8x16ToVecI16x8'], value); - }, - 'load8x8_s'(offset, align, ptr) { - return Module['_BinaryenSIMDLoad'](module, Module['LoadExtSVec8x8ToVecI16x8'], offset, align, ptr); - }, - 'load8x8_u'(offset, align, ptr) { - return Module['_BinaryenSIMDLoad'](module, Module['LoadExtUVec8x8ToVecI16x8'], offset, align, ptr); + 'extend_high_i8x16_u'(value) { + return Module['_BinaryenUnary'](module, Module['ExtendHighUVecI8x16ToVecI16x8'], value); }, }; @@ -1753,9 +1766,6 @@ function wrapModule(module, self = {}) { 'neg'(value) { return Module['_BinaryenUnary'](module, Module['NegVecI32x4'], value); }, - 'any_true'(value) { - return Module['_BinaryenUnary'](module, Module['AnyTrueVecI32x4'], value); - }, 'all_true'(value) { return Module['_BinaryenUnary'](module, Module['AllTrueVecI32x4'], value); }, @@ -1801,23 +1811,17 @@ function wrapModule(module, self = {}) { 'trunc_sat_f32x4_u'(value) { return Module['_BinaryenUnary'](module, Module['TruncSatUVecF32x4ToVecI32x4'], value); }, - 'widen_low_i16x8_s'(value) { - return Module['_BinaryenUnary'](module, Module['WidenLowSVecI16x8ToVecI32x4'], value); + 'extend_low_i16x8_s'(value) { + return Module['_BinaryenUnary'](module, Module['ExtendLowSVecI16x8ToVecI32x4'], value); }, - 'widen_high_i16x8_s'(value) { - return Module['_BinaryenUnary'](module, Module['WidenHighSVecI16x8ToVecI32x4'], value); + 'extend_high_i16x8_s'(value) { + return Module['_BinaryenUnary'](module, Module['ExtendHighSVecI16x8ToVecI32x4'], value); }, - 'widen_low_i16x8_u'(value) { - return Module['_BinaryenUnary'](module, Module['WidenLowUVecI16x8ToVecI32x4'], value); - }, - 'widen_high_i16x8_u'(value) { - return Module['_BinaryenUnary'](module, Module['WidenHighUVecI16x8ToVecI32x4'], value); - }, - 'load16x4_s'(offset, align, ptr) { - return Module['_BinaryenSIMDLoad'](module, Module['LoadExtSVec16x4ToVecI32x4'], offset, align, ptr); + 'extend_low_i16x8_u'(value) { + return Module['_BinaryenUnary'](module, Module['ExtendLowUVecI16x8ToVecI32x4'], value); }, - 'load16x4_u'(offset, align, ptr) { - return Module['_BinaryenSIMDLoad'](module, Module['LoadExtUVec16x4ToVecI32x4'], offset, align, ptr); + 'extend_high_i16x8_u'(value) { + return Module['_BinaryenUnary'](module, Module['ExtendHighUVecI16x8ToVecI32x4'], value); }, }; @@ -1852,18 +1856,6 @@ function wrapModule(module, self = {}) { 'mul'(left, right) { return Module['_BinaryenBinary'](module, Module['MulVecI64x2'], left, right); }, - 'trunc_sat_f64x2_s'(value) { - return Module['_BinaryenUnary'](module, Module['TruncSatSVecF64x2ToVecI64x2'], value); - }, - 'trunc_sat_f64x2_u'(value) { - return Module['_BinaryenUnary'](module, Module['TruncSatUVecF64x2ToVecI64x2'], value); - }, - 'load32x2_s'(offset, align, ptr) { - return Module['_BinaryenSIMDLoad'](module, Module['LoadExtSVec32x2ToVecI64x2'], offset, align, ptr); - }, - 'load32x2_u'(offset, align, ptr) { - return Module['_BinaryenSIMDLoad'](module, Module['LoadExtUVec32x2ToVecI64x2'], offset, align, ptr); - }, }; self['f32x4'] = { @@ -1903,12 +1895,6 @@ function wrapModule(module, self = {}) { 'sqrt'(value) { return Module['_BinaryenUnary'](module, Module['SqrtVecF32x4'], value); }, - 'qfma'(a, b, c) { - return Module['_BinaryenSIMDTernary'](module, Module['QFMAVecF32x4'], a, b, c); - }, - 'qfms'(a, b, c) { - return Module['_BinaryenSIMDTernary'](module, Module['QFMSVecF32x4'], a, b, c); - }, 'add'(left, right) { return Module['_BinaryenBinary'](module, Module['AddVecF32x4'], left, right); }, @@ -1990,12 +1976,6 @@ function wrapModule(module, self = {}) { 'sqrt'(value) { return Module['_BinaryenUnary'](module, Module['SqrtVecF64x2'], value); }, - 'qfma'(a, b, c) { - return Module['_BinaryenSIMDTernary'](module, Module['QFMAVecF64x2'], a, b, c); - }, - 'qfms'(a, b, c) { - return Module['_BinaryenSIMDTernary'](module, Module['QFMSVecF64x2'], a, b, c); - }, 'add'(left, right) { return Module['_BinaryenBinary'](module, Module['AddVecF64x2'], left, right); }, @@ -2032,42 +2012,6 @@ function wrapModule(module, self = {}) { 'nearest'(value) { return Module['_BinaryenUnary'](module, Module['NearestVecF64x2'], value); }, - 'convert_i64x2_s'(value) { - return Module['_BinaryenUnary'](module, Module['ConvertSVecI64x2ToVecF64x2'], value); - }, - 'convert_i64x2_u'(value) { - return Module['_BinaryenUnary'](module, Module['ConvertUVecI64x2ToVecF64x2'], value); - }, - }; - - self['v8x16'] = { - 'shuffle'(left, right, mask) { - return preserveStack(() => Module['_BinaryenSIMDShuffle'](module, left, right, i8sToStack(mask))); - }, - 'swizzle'(left, right) { - return Module['_BinaryenBinary'](module, Module['SwizzleVec8x16'], left, right); - }, - 'load_splat'(offset, align, ptr) { - return Module['_BinaryenSIMDLoad'](module, Module['LoadSplatVec8x16'], offset, align, ptr); - }, - }; - - self['v16x8'] = { - 'load_splat'(offset, align, ptr) { - return Module['_BinaryenSIMDLoad'](module, Module['LoadSplatVec16x8'], offset, align, ptr); - }, - }; - - self['v32x4'] = { - 'load_splat'(offset, align, ptr) { - return Module['_BinaryenSIMDLoad'](module, Module['LoadSplatVec32x4'], offset, align, ptr); - }, - }; - - self['v64x2'] = { - 'load_splat'(offset, align, ptr) { - return Module['_BinaryenSIMDLoad'](module, Module['LoadSplatVec64x2'], offset, align, ptr); - }, }; self['funcref'] = { diff --git a/src/literal.h b/src/literal.h index 06719cd32..a07e2597c 100644 --- a/src/literal.h +++ b/src/literal.h @@ -482,6 +482,11 @@ public: Literal geSI32x4(const Literal& other) const; Literal geUI32x4(const Literal& other) const; Literal eqI64x2(const Literal& other) const; + Literal neI64x2(const Literal& other) const; + Literal ltSI64x2(const Literal& other) const; + Literal gtSI64x2(const Literal& other) const; + Literal leSI64x2(const Literal& other) const; + Literal geSI64x2(const Literal& other) const; Literal eqF32x4(const Literal& other) const; Literal neF32x4(const Literal& other) const; Literal ltF32x4(const Literal& other) const; @@ -498,10 +503,10 @@ public: Literal andV128(const Literal& other) const; Literal orV128(const Literal& other) const; Literal xorV128(const Literal& other) const; + Literal anyTrueV128() const; Literal bitselectV128(const Literal& left, const Literal& right) const; Literal absI8x16() const; Literal negI8x16() const; - Literal anyTrueI8x16() const; Literal allTrueI8x16() const; Literal bitmaskI8x16() const; Literal shlI8x16(const Literal& other) const; @@ -513,7 +518,6 @@ public: Literal subI8x16(const Literal& other) const; Literal subSaturateSI8x16(const Literal& other) const; Literal subSaturateUI8x16(const Literal& other) const; - Literal mulI8x16(const Literal& other) const; Literal minSI8x16(const Literal& other) const; Literal minUI8x16(const Literal& other) const; Literal maxSI8x16(const Literal& other) const; @@ -522,7 +526,6 @@ public: Literal popcntI8x16() const; Literal absI16x8() const; Literal negI16x8() const; - Literal anyTrueI16x8() const; Literal allTrueI16x8() const; Literal bitmaskI16x8() const; Literal shlI16x8(const Literal& other) const; @@ -547,7 +550,6 @@ public: Literal extMulHighUI16x8(const Literal& other) const; Literal absI32x4() const; Literal negI32x4() const; - Literal anyTrueI32x4() const; Literal allTrueI32x4() const; Literal bitmaskI32x4() const; Literal shlI32x4(const Literal& other) const; @@ -565,7 +567,9 @@ public: Literal extMulHighSI32x4(const Literal& other) const; Literal extMulLowUI32x4(const Literal& other) const; Literal extMulHighUI32x4(const Literal& other) const; + Literal absI64x2() const; Literal negI64x2() const; + Literal allTrueI64x2() const; Literal shlI64x2(const Literal& other) const; Literal shrSI64x2(const Literal& other) const; Literal shrUI64x2(const Literal& other) const; @@ -608,24 +612,20 @@ public: Literal nearestF64x2() const; Literal truncSatToSI32x4() const; Literal truncSatToUI32x4() const; - Literal truncSatToSI64x2() const; - Literal truncSatToUI64x2() const; Literal convertSToF32x4() const; Literal convertUToF32x4() const; - Literal convertSToF64x2() const; - Literal convertUToF64x2() const; Literal narrowSToVecI8x16(const Literal& other) const; Literal narrowUToVecI8x16(const Literal& other) const; Literal narrowSToVecI16x8(const Literal& other) const; Literal narrowUToVecI16x8(const Literal& other) const; - Literal widenLowSToVecI16x8() const; - Literal widenHighSToVecI16x8() const; - Literal widenLowUToVecI16x8() const; - Literal widenHighUToVecI16x8() const; - Literal widenLowSToVecI32x4() const; - Literal widenHighSToVecI32x4() const; - Literal widenLowUToVecI32x4() const; - Literal widenHighUToVecI32x4() const; + Literal extendLowSToVecI16x8() const; + Literal extendHighSToVecI16x8() const; + Literal extendLowUToVecI16x8() const; + Literal extendHighUToVecI16x8() const; + Literal extendLowSToVecI32x4() const; + Literal extendHighSToVecI32x4() const; + Literal extendLowUToVecI32x4() const; + Literal extendHighUToVecI32x4() const; Literal swizzleVec8x16(const Literal& other) const; // Checks if an RTT value is a sub-rtt of another, that is, whether GC data diff --git a/src/passes/Print.cpp b/src/passes/Print.cpp index e3ea04b53..5f8c9f163 100644 --- a/src/passes/Print.cpp +++ b/src/passes/Print.cpp @@ -639,7 +639,7 @@ struct PrintExpressionContents } void visitSIMDShuffle(SIMDShuffle* curr) { prepareColor(o); - o << "v8x16.shuffle"; + o << "i8x16.shuffle"; restoreNormalColor(o); for (uint8_t mask_index : curr->mask) { o << " " << std::to_string(mask_index); @@ -651,30 +651,6 @@ struct PrintExpressionContents case Bitselect: o << "v128.bitselect"; break; - case QFMAF32x4: - o << "f32x4.qfma"; - break; - case QFMSF32x4: - o << "f32x4.qfms"; - break; - case QFMAF64x2: - o << "f64x2.qfma"; - break; - case QFMSF64x2: - o << "f64x2.qfms"; - break; - case SignSelectVec8x16: - o << "v8x16.signselect"; - break; - case SignSelectVec16x8: - o << "v16x8.signselect"; - break; - case SignSelectVec32x4: - o << "v32x4.signselect"; - break; - case SignSelectVec64x2: - o << "v64x2.signselect"; - break; } restoreNormalColor(o); } @@ -724,34 +700,34 @@ struct PrintExpressionContents prepareColor(o); switch (curr->op) { case LoadSplatVec8x16: - o << "v8x16.load_splat"; + o << "v128.load8_splat"; break; case LoadSplatVec16x8: - o << "v16x8.load_splat"; + o << "v128.load16_splat"; break; case LoadSplatVec32x4: - o << "v32x4.load_splat"; + o << "v128.load32_splat"; break; case LoadSplatVec64x2: - o << "v64x2.load_splat"; + o << "v128.load64_splat"; break; case LoadExtSVec8x8ToVecI16x8: - o << "i16x8.load8x8_s"; + o << "v128.load8x8_s"; break; case LoadExtUVec8x8ToVecI16x8: - o << "i16x8.load8x8_u"; + o << "v128.load8x8_u"; break; case LoadExtSVec16x4ToVecI32x4: - o << "i32x4.load16x4_s"; + o << "v128.load16x4_s"; break; case LoadExtUVec16x4ToVecI32x4: - o << "i32x4.load16x4_u"; + o << "v128.load16x4_u"; break; case LoadExtSVec32x2ToVecI64x2: - o << "i64x2.load32x2_s"; + o << "v128.load32x2_s"; break; case LoadExtUVec32x2ToVecI64x2: - o << "i64x2.load32x2_u"; + o << "v128.load32x2_u"; break; case Load32Zero: o << "v128.load32_zero"; @@ -805,37 +781,6 @@ struct PrintExpressionContents } o << " " << int(curr->index); } - void visitSIMDWiden(SIMDWiden* curr) { - prepareColor(o); - switch (curr->op) { - case WidenSVecI8x16ToVecI32x4: - o << "i32x4.widen_i8x16_s "; - break; - case WidenUVecI8x16ToVecI32x4: - o << "i32x4.widen_i8x16_u "; - break; - } - restoreNormalColor(o); - o << int(curr->index); - } - void visitPrefetch(Prefetch* curr) { - prepareColor(o); - switch (curr->op) { - case PrefetchTemporal: - o << "prefetch.t"; - break; - case PrefetchNontemporal: - o << "prefetch.nt"; - break; - } - restoreNormalColor(o); - if (curr->offset) { - o << " offset=" << curr->offset; - } - if (curr->align != 1) { - o << " align=" << curr->align; - } - } void visitMemoryInit(MemoryInit* curr) { prepareColor(o); o << "memory.init"; @@ -1065,15 +1010,15 @@ struct PrintExpressionContents case NotVec128: o << "v128.not"; break; + case AnyTrueVec128: + o << "v128.any_true"; + break; case AbsVecI8x16: o << "i8x16.abs"; break; case NegVecI8x16: o << "i8x16.neg"; break; - case AnyTrueVecI8x16: - o << "i8x16.any_true"; - break; case AllTrueVecI8x16: o << "i8x16.all_true"; break; @@ -1089,9 +1034,6 @@ struct PrintExpressionContents case NegVecI16x8: o << "i16x8.neg"; break; - case AnyTrueVecI16x8: - o << "i16x8.any_true"; - break; case AllTrueVecI16x8: o << "i16x8.all_true"; break; @@ -1104,18 +1046,21 @@ struct PrintExpressionContents case NegVecI32x4: o << "i32x4.neg"; break; - case AnyTrueVecI32x4: - o << "i32x4.any_true"; - break; case AllTrueVecI32x4: o << "i32x4.all_true"; break; case BitmaskVecI32x4: o << "i32x4.bitmask"; break; + case AbsVecI64x2: + o << "i64x2.abs"; + break; case NegVecI64x2: o << "i64x2.neg"; break; + case AllTrueVecI64x2: + o << "i64x2.all_true"; + break; case BitmaskVecI64x2: o << "i64x2.bitmask"; break; @@ -1179,59 +1124,47 @@ struct PrintExpressionContents case TruncSatUVecF32x4ToVecI32x4: o << "i32x4.trunc_sat_f32x4_u"; break; - case TruncSatSVecF64x2ToVecI64x2: - o << "i64x2.trunc_sat_f64x2_s"; - break; - case TruncSatUVecF64x2ToVecI64x2: - o << "i64x2.trunc_sat_f64x2_u"; - break; case ConvertSVecI32x4ToVecF32x4: o << "f32x4.convert_i32x4_s"; break; case ConvertUVecI32x4ToVecF32x4: o << "f32x4.convert_i32x4_u"; break; - case ConvertSVecI64x2ToVecF64x2: - o << "f64x2.convert_i64x2_s"; - break; - case ConvertUVecI64x2ToVecF64x2: - o << "f64x2.convert_i64x2_u"; + case ExtendLowSVecI8x16ToVecI16x8: + o << "i16x8.extend_low_i8x16_s"; break; - case WidenLowSVecI8x16ToVecI16x8: - o << "i16x8.widen_low_i8x16_s"; + case ExtendHighSVecI8x16ToVecI16x8: + o << "i16x8.extend_high_i8x16_s"; break; - case WidenHighSVecI8x16ToVecI16x8: - o << "i16x8.widen_high_i8x16_s"; + case ExtendLowUVecI8x16ToVecI16x8: + o << "i16x8.extend_low_i8x16_u"; break; - case WidenLowUVecI8x16ToVecI16x8: - o << "i16x8.widen_low_i8x16_u"; + case ExtendHighUVecI8x16ToVecI16x8: + o << "i16x8.extend_high_i8x16_u"; break; - case WidenHighUVecI8x16ToVecI16x8: - o << "i16x8.widen_high_i8x16_u"; + case ExtendLowSVecI16x8ToVecI32x4: + o << "i32x4.extend_low_i16x8_s"; break; - case WidenLowSVecI16x8ToVecI32x4: - o << "i32x4.widen_low_i16x8_s"; + case ExtendHighSVecI16x8ToVecI32x4: + o << "i32x4.extend_high_i16x8_s"; break; - case WidenHighSVecI16x8ToVecI32x4: - o << "i32x4.widen_high_i16x8_s"; + case ExtendLowUVecI16x8ToVecI32x4: + o << "i32x4.extend_low_i16x8_u"; break; - case WidenLowUVecI16x8ToVecI32x4: - o << "i32x4.widen_low_i16x8_u"; + case ExtendHighUVecI16x8ToVecI32x4: + o << "i32x4.extend_high_i16x8_u"; break; - case WidenHighUVecI16x8ToVecI32x4: - o << "i32x4.widen_high_i16x8_u"; + case ExtendLowSVecI32x4ToVecI64x2: + o << "i64x2.extend_low_i32x4_s"; break; - case WidenLowSVecI32x4ToVecI64x2: - o << "i64x2.widen_low_i32x4_s"; + case ExtendHighSVecI32x4ToVecI64x2: + o << "i64x2.extend_high_i32x4_s"; break; - case WidenHighSVecI32x4ToVecI64x2: - o << "i64x2.widen_high_i32x4_s"; + case ExtendLowUVecI32x4ToVecI64x2: + o << "i64x2.extend_low_i32x4_u"; break; - case WidenLowUVecI32x4ToVecI64x2: - o << "i64x2.widen_low_i32x4_u"; - break; - case WidenHighUVecI32x4ToVecI64x2: - o << "i64x2.widen_high_i32x4_u"; + case ExtendHighUVecI32x4ToVecI64x2: + o << "i64x2.extend_high_i32x4_u"; break; case ConvertLowSVecI32x4ToVecF64x2: o << "f64x2.convert_low_i32x4_s"; @@ -1240,10 +1173,10 @@ struct PrintExpressionContents o << "f64x2.convert_low_i32x4_u"; break; case TruncSatZeroSVecF64x2ToVecI32x4: - o << "i32x4.trunc_sat_f64x2_zero_s"; + o << "i32x4.trunc_sat_f64x2_s_zero"; break; case TruncSatZeroUVecF64x2ToVecI32x4: - o << "i32x4.trunc_sat_f64x2_zero_u"; + o << "i32x4.trunc_sat_f64x2_u_zero"; break; case DemoteZeroVecF64x2ToVecF32x4: o << "f32x4.demote_f64x2_zero"; @@ -1584,6 +1517,21 @@ struct PrintExpressionContents case EqVecI64x2: o << "i64x2.eq"; break; + case NeVecI64x2: + o << "i64x2.ne"; + break; + case LtSVecI64x2: + o << "i64x2.lt_s"; + break; + case GtSVecI64x2: + o << "i64x2.gt_s"; + break; + case LeSVecI64x2: + o << "i64x2.le_s"; + break; + case GeSVecI64x2: + o << "i64x2.ge_s"; + break; case EqVecF32x4: o << "f32x4.eq"; break; @@ -1638,22 +1586,19 @@ struct PrintExpressionContents o << "i8x16.add"; break; case AddSatSVecI8x16: - o << "i8x16.add_saturate_s"; + o << "i8x16.add_sat_s"; break; case AddSatUVecI8x16: - o << "i8x16.add_saturate_u"; + o << "i8x16.add_sat_u"; break; case SubVecI8x16: o << "i8x16.sub"; break; case SubSatSVecI8x16: - o << "i8x16.sub_saturate_s"; + o << "i8x16.sub_sat_s"; break; case SubSatUVecI8x16: - o << "i8x16.sub_saturate_u"; - break; - case MulVecI8x16: - o << "i8x16.mul"; + o << "i8x16.sub_sat_u"; break; case MinSVecI8x16: o << "i8x16.min_s"; @@ -1674,19 +1619,19 @@ struct PrintExpressionContents o << "i16x8.add"; break; case AddSatSVecI16x8: - o << "i16x8.add_saturate_s"; + o << "i16x8.add_sat_s"; break; case AddSatUVecI16x8: - o << "i16x8.add_saturate_u"; + o << "i16x8.add_sat_u"; break; case SubVecI16x8: o << "i16x8.sub"; break; case SubSatSVecI16x8: - o << "i16x8.sub_saturate_s"; + o << "i16x8.sub_sat_s"; break; case SubSatUVecI16x8: - o << "i16x8.sub_saturate_u"; + o << "i16x8.sub_sat_u"; break; case MulVecI16x8: o << "i16x8.mul"; @@ -1844,7 +1789,7 @@ struct PrintExpressionContents break; case SwizzleVec8x16: - o << "v8x16.swizzle"; + o << "i8x16.swizzle"; break; case InvalidBinary: diff --git a/src/tools/fuzzing.h b/src/tools/fuzzing.h index e6a12543a..81eab1d28 100644 --- a/src/tools/fuzzing.h +++ b/src/tools/fuzzing.h @@ -2210,11 +2210,10 @@ private: } case Type::v128: { assert(wasm.features.hasSIMD()); - return buildUnary({pick(AnyTrueVecI8x16, + // TODO: Add the other SIMD unary ops + return buildUnary({pick(AnyTrueVec128, AllTrueVecI8x16, - AnyTrueVecI16x8, AllTrueVecI16x8, - AnyTrueVecI32x4, AllTrueVecI32x4), make(Type::v128)}); } @@ -2335,7 +2334,7 @@ private: return buildUnary({SplatVecF64x2, make(Type::f64)}); case 4: return buildUnary({pick(NotVec128, - // TODO: i8x16.popcnt once merged + // TODO: add additional SIMD instructions NegVecI8x16, NegVecI16x8, NegVecI32x4, @@ -2348,20 +2347,16 @@ private: SqrtVecF64x2, TruncSatSVecF32x4ToVecI32x4, TruncSatUVecF32x4ToVecI32x4, - TruncSatSVecF64x2ToVecI64x2, - TruncSatUVecF64x2ToVecI64x2, ConvertSVecI32x4ToVecF32x4, ConvertUVecI32x4ToVecF32x4, - ConvertSVecI64x2ToVecF64x2, - ConvertUVecI64x2ToVecF64x2, - WidenLowSVecI8x16ToVecI16x8, - WidenHighSVecI8x16ToVecI16x8, - WidenLowUVecI8x16ToVecI16x8, - WidenHighUVecI8x16ToVecI16x8, - WidenLowSVecI16x8ToVecI32x4, - WidenHighSVecI16x8ToVecI32x4, - WidenLowUVecI16x8ToVecI32x4, - WidenHighUVecI16x8ToVecI32x4), + ExtendLowSVecI8x16ToVecI16x8, + ExtendHighSVecI8x16ToVecI16x8, + ExtendLowUVecI8x16ToVecI16x8, + ExtendHighUVecI8x16ToVecI16x8, + ExtendLowSVecI16x8ToVecI32x4, + ExtendHighSVecI16x8ToVecI32x4, + ExtendLowUVecI16x8ToVecI32x4, + ExtendHighUVecI16x8ToVecI32x4), make(Type::v128)}); } WASM_UNREACHABLE("invalid value"); @@ -2558,7 +2553,6 @@ private: SubVecI8x16, SubSatSVecI8x16, SubSatUVecI8x16, - MulVecI8x16, MinSVecI8x16, MinUVecI8x16, MaxSVecI8x16, diff --git a/src/wasm-binary.h b/src/wasm-binary.h index ec89bb1da..804df5811 100644 --- a/src/wasm-binary.h +++ b/src/wasm-binary.h @@ -803,7 +803,6 @@ enum ASTNodes { I32x4LeU = 0x3e, I32x4GeS = 0x3f, I32x4GeU = 0x40, - I64x2Eq = 0xc0, F32x4Eq = 0x41, F32x4Ne = 0x42, F32x4Lt = 0x43, @@ -823,28 +822,33 @@ enum ASTNodes { V128Or = 0x50, V128Xor = 0x51, V128Bitselect = 0x52, - - V8x16SignSelect = 0x7d, - V16x8SignSelect = 0x7e, - V32x4SignSelect = 0x7f, - V64x2SignSelect = 0x94, - - V128Load8Lane = 0x58, - V128Load16Lane = 0x59, - V128Load32Lane = 0x5a, - V128Load64Lane = 0x5b, - V128Store8Lane = 0x5c, - V128Store16Lane = 0x5d, - V128Store32Lane = 0x5e, - V128Store64Lane = 0x5f, + V128AnyTrue = 0x53, + + V128Load8Lane = 0x54, + V128Load16Lane = 0x55, + V128Load32Lane = 0x56, + V128Load64Lane = 0x57, + V128Store8Lane = 0x58, + V128Store16Lane = 0x59, + V128Store32Lane = 0x5a, + V128Store64Lane = 0x5b, + V128Load32Zero = 0x5c, + V128Load64Zero = 0x5d, + + F32x4DemoteZeroF64x2 = 0x5e, + F64x2PromoteLowF32x4 = 0x5f, I8x16Abs = 0x60, I8x16Neg = 0x61, - I8x16AnyTrue = 0x62, + I8x16Popcnt = 0x62, I8x16AllTrue = 0x63, I8x16Bitmask = 0x64, I8x16NarrowSI16x8 = 0x65, I8x16NarrowUI16x8 = 0x66, + F32x4Ceil = 0x67, + F32x4Floor = 0x68, + F32x4Trunc = 0x69, + F32x4Nearest = 0x6a, I8x16Shl = 0x6b, I8x16ShrS = 0x6c, I8x16ShrU = 0x6d, @@ -854,26 +858,30 @@ enum ASTNodes { I8x16Sub = 0x71, I8x16SubSatS = 0x72, I8x16SubSatU = 0x73, - I8x16Mul = 0x75, + F64x2Ceil = 0x74, + F64x2Floor = 0x75, I8x16MinS = 0x76, I8x16MinU = 0x77, I8x16MaxS = 0x78, I8x16MaxU = 0x79, + F64x2Trunc = 0x7a, I8x16AvgrU = 0x7b, - - I8x16Popcnt = 0x7c, + I16x8ExtAddPairWiseSI8x16 = 0x7c, + I16x8ExtAddPairWiseUI8x16 = 0x7d, + I32x4ExtAddPairWiseSI16x8 = 0x7e, + I32x4ExtAddPairWiseUI16x8 = 0x7f, I16x8Abs = 0x80, I16x8Neg = 0x81, - I16x8AnyTrue = 0x82, + I16x8Q15MulrSatS = 0x82, I16x8AllTrue = 0x83, I16x8Bitmask = 0x84, I16x8NarrowSI32x4 = 0x85, I16x8NarrowUI32x4 = 0x86, - I16x8WidenLowSI8x16 = 0x87, - I16x8WidenHighSI8x16 = 0x88, - I16x8WidenLowUI8x16 = 0x89, - I16x8WidenHighUI8x16 = 0x8a, + I16x8ExtendLowSI8x16 = 0x87, + I16x8ExtendHighSI8x16 = 0x88, + I16x8ExtendLowUI8x16 = 0x89, + I16x8ExtendHighUI8x16 = 0x8a, I16x8Shl = 0x8b, I16x8ShrS = 0x8c, I16x8ShrU = 0x8d, @@ -883,50 +891,88 @@ enum ASTNodes { I16x8Sub = 0x91, I16x8SubSatS = 0x92, I16x8SubSatU = 0x93, + F64x2Nearest = 0x94, I16x8Mul = 0x95, I16x8MinS = 0x96, I16x8MinU = 0x97, I16x8MaxS = 0x98, I16x8MaxU = 0x99, + // 0x9a unused I16x8AvgrU = 0x9b, - I16x8Q15MulrSatS = 0x9c, + I16x8ExtMulLowSI8x16 = 0x9c, + I16x8ExtMulHighSI8x16 = 0x9d, + I16x8ExtMulLowUI8x16 = 0x9e, + I16x8ExtMulHighUI8x16 = 0x9f, I32x4Abs = 0xa0, I32x4Neg = 0xa1, - I32x4AnyTrue = 0xa2, + // 0xa2 unused I32x4AllTrue = 0xa3, I32x4Bitmask = 0xa4, - I32x4WidenLowSI16x8 = 0xa7, - I32x4WidenHighSI16x8 = 0xa8, - I32x4WidenLowUI16x8 = 0xa9, - I32x4WidenHighUI16x8 = 0xaa, + // 0xa5 unused + // 0xa6 unused + I32x4ExtendLowSI16x8 = 0xa7, + I32x4ExtendHighSI16x8 = 0xa8, + I32x4ExtendLowUI16x8 = 0xa9, + I32x4ExtendHighUI16x8 = 0xaa, I32x4Shl = 0xab, I32x4ShrS = 0xac, I32x4ShrU = 0xad, I32x4Add = 0xae, + // 0xaf unused + // 0xb0 unused I32x4Sub = 0xb1, + // 0xb2 unused + // 0xb3 unused + // 0xb4 unused I32x4Mul = 0xb5, I32x4MinS = 0xb6, I32x4MinU = 0xb7, I32x4MaxS = 0xb8, I32x4MaxU = 0xb9, I32x4DotSVecI16x8 = 0xba, + // 0xbb unused + I32x4ExtMulLowSI16x8 = 0xbc, + I32x4ExtMulHighSI16x8 = 0xbd, + I32x4ExtMulLowUI16x8 = 0xbe, + I32x4ExtMulHighUI16x8 = 0xbf, - I64x2Bitmask = 0xc4, - I64x2WidenLowSI32x4 = 0xc7, - I64x2WidenHighSI32x4 = 0xc8, - I64x2WidenLowUI32x4 = 0xc9, - I64x2WidenHighUI32x4 = 0xca, + I64x2Abs = 0xc0, I64x2Neg = 0xc1, + // 0xc2 unused + I64x2AllTrue = 0xc3, + I64x2Bitmask = 0xc4, + // 0xc5 unused + // 0xc6 unused + I64x2ExtendLowSI32x4 = 0xc7, + I64x2ExtendHighSI32x4 = 0xc8, + I64x2ExtendLowUI32x4 = 0xc9, + I64x2ExtendHighUI32x4 = 0xca, I64x2Shl = 0xcb, I64x2ShrS = 0xcc, I64x2ShrU = 0xcd, I64x2Add = 0xce, + // 0xcf unused + // 0xd0 unused I64x2Sub = 0xd1, + // 0xd2 unused + // 0xd3 unused + // 0xd4 unused I64x2Mul = 0xd5, + I64x2Eq = 0xd6, + I64x2Ne = 0xd7, + I64x2LtS = 0xd8, + I64x2GtS = 0xd9, + I64x2LeS = 0xda, + I64x2GeS = 0xdb, + I64x2ExtMulLowSI32x4 = 0xdc, + I64x2ExtMulHighSI32x4 = 0xdd, + I64x2ExtMulLowUI32x4 = 0xde, + I64x2ExtMulHighUI32x4 = 0xdf, F32x4Abs = 0xe0, F32x4Neg = 0xe1, + // 0xe2 unused F32x4Sqrt = 0xe3, F32x4Add = 0xe4, F32x4Sub = 0xe5, @@ -937,17 +983,9 @@ enum ASTNodes { F32x4PMin = 0xea, F32x4PMax = 0xeb, - F32x4Ceil = 0xd8, - F32x4Floor = 0xd9, - F32x4Trunc = 0xda, - F32x4Nearest = 0xdb, - F64x2Ceil = 0xdc, - F64x2Floor = 0xdd, - F64x2Trunc = 0xde, - F64x2Nearest = 0xdf, - F64x2Abs = 0xec, F64x2Neg = 0xed, + // 0xee unused F64x2Sqrt = 0xef, F64x2Add = 0xf0, F64x2Sub = 0xf1, @@ -958,56 +996,14 @@ enum ASTNodes { F64x2PMin = 0xf6, F64x2PMax = 0xf7, - I16x8ExtAddPairWiseSI8x16 = 0xc2, - I16x8ExtAddPairWiseUI8x16 = 0xc3, - I32x4ExtAddPairWiseSI16x8 = 0xa5, - I32x4ExtAddPairWiseUI16x8 = 0xa6, - I32x4TruncSatSF32x4 = 0xf8, I32x4TruncSatUF32x4 = 0xf9, F32x4ConvertSI32x4 = 0xfa, F32x4ConvertUI32x4 = 0xfb, - - V128Load32Zero = 0xfc, - V128Load64Zero = 0xfd, - - F32x4QFMA = 0xb4, - F32x4QFMS = 0xd4, - F64x2QFMA = 0xfe, - F64x2QFMS = 0xff, - - I64x2TruncSatSF64x2 = 0x0100, - I64x2TruncSatUF64x2 = 0x0101, - F64x2ConvertSI64x2 = 0x0102, - F64x2ConvertUI64x2 = 0x0103, - - I16x8ExtMulLowSI8x16 = 0x9a, - I16x8ExtMulHighSI8x16 = 0x9d, - I16x8ExtMulLowUI8x16 = 0x9e, - I16x8ExtMulHighUI8x16 = 0x9f, - I32x4ExtMulLowSI16x8 = 0xbb, - I32x4ExtMulHighSI16x8 = 0xbd, - I32x4ExtMulLowUI16x8 = 0xbe, - I32x4ExtMulHighUI16x8 = 0xbf, - I64x2ExtMulLowSI32x4 = 0xd2, - I64x2ExtMulHighSI32x4 = 0xd3, - I64x2ExtMulLowUI32x4 = 0xd6, - I64x2ExtMulHighUI32x4 = 0xd7, - - F64x2ConvertLowSI32x4 = 0x53, - F64x2ConvertLowUI32x4 = 0x54, - I32x4TruncSatZeroSF64x2 = 0x55, - I32x4TruncSatZeroUF64x2 = 0x56, - F32x4DemoteZeroF64x2 = 0x57, - F64x2PromoteLowF32x4 = 0x69, - - I32x4WidenSI8x16 = 0x67, - I32x4WidenUI8x16 = 0x68, - - // prefetch opcodes - - PrefetchT = 0xc5, - PrefetchNT = 0xc6, + I32x4TruncSatZeroSF64x2 = 0xfc, + I32x4TruncSatZeroUF64x2 = 0xfd, + F64x2ConvertLowSI32x4 = 0xfe, + F64x2ConvertLowUI32x4 = 0xff, // bulk memory opcodes @@ -1595,8 +1591,6 @@ public: bool maybeVisitSIMDShift(Expression*& out, uint32_t code); bool maybeVisitSIMDLoad(Expression*& out, uint32_t code); bool maybeVisitSIMDLoadStoreLane(Expression*& out, uint32_t code); - bool maybeVisitSIMDWiden(Expression*& out, uint32_t code); - bool maybeVisitPrefetch(Expression*& out, uint32_t code); bool maybeVisitMemoryInit(Expression*& out, uint32_t code); bool maybeVisitDataDrop(Expression*& out, uint32_t code); bool maybeVisitMemoryCopy(Expression*& out, uint32_t code); diff --git a/src/wasm-builder.h b/src/wasm-builder.h index 134d42a78..c0d6cbeeb 100644 --- a/src/wasm-builder.h +++ b/src/wasm-builder.h @@ -499,16 +499,6 @@ public: ret->finalize(); return ret; } - Prefetch* - makePrefetch(PrefetchOp op, Address offset, Address align, Expression* ptr) { - auto* ret = wasm.allocator.alloc<Prefetch>(); - ret->op = op; - ret->offset = offset; - ret->align = align; - ret->ptr = ptr; - ret->finalize(); - return ret; - } MemoryInit* makeMemoryInit(uint32_t segment, Expression* dest, Expression* offset, diff --git a/src/wasm-delegations-fields.h b/src/wasm-delegations-fields.h index 63ff235d7..16ef11193 100644 --- a/src/wasm-delegations-fields.h +++ b/src/wasm-delegations-fields.h @@ -387,23 +387,6 @@ switch (DELEGATE_ID) { DELEGATE_END(SIMDLoadStoreLane); break; } - case Expression::Id::SIMDWidenId: { - DELEGATE_START(SIMDWiden); - DELEGATE_FIELD_CHILD(SIMDWiden, vec); - DELEGATE_FIELD_INT(SIMDWiden, op); - DELEGATE_FIELD_INT(SIMDWiden, index); - DELEGATE_END(SIMDWiden); - break; - } - case Expression::Id::PrefetchId: { - DELEGATE_START(Prefetch); - DELEGATE_FIELD_CHILD(Prefetch, ptr); - DELEGATE_FIELD_INT(Prefetch, op); - DELEGATE_FIELD_ADDRESS(Prefetch, offset); - DELEGATE_FIELD_ADDRESS(Prefetch, align); - DELEGATE_END(Prefetch); - break; - } case Expression::Id::MemoryInitId: { DELEGATE_START(MemoryInit); DELEGATE_FIELD_CHILD(MemoryInit, size); diff --git a/src/wasm-delegations.h b/src/wasm-delegations.h index 5a05a731f..b063c81be 100644 --- a/src/wasm-delegations.h +++ b/src/wasm-delegations.h @@ -40,8 +40,6 @@ DELEGATE(SIMDTernary); DELEGATE(SIMDShift); DELEGATE(SIMDLoad); DELEGATE(SIMDLoadStoreLane); -DELEGATE(SIMDWiden); -DELEGATE(Prefetch); DELEGATE(MemoryInit); DELEGATE(DataDrop); DELEGATE(MemoryCopy); diff --git a/src/wasm-interpreter.h b/src/wasm-interpreter.h index 61d29ada1..249ec8fdb 100644 --- a/src/wasm-interpreter.h +++ b/src/wasm-interpreter.h @@ -465,12 +465,12 @@ public: return value.splatF64x2(); case NotVec128: return value.notV128(); + case AnyTrueVec128: + return value.anyTrueV128(); case AbsVecI8x16: return value.absI8x16(); case NegVecI8x16: return value.negI8x16(); - case AnyTrueVecI8x16: - return value.anyTrueI8x16(); case AllTrueVecI8x16: return value.allTrueI8x16(); case BitmaskVecI8x16: @@ -481,8 +481,6 @@ public: return value.absI16x8(); case NegVecI16x8: return value.negI16x8(); - case AnyTrueVecI16x8: - return value.anyTrueI16x8(); case AllTrueVecI16x8: return value.allTrueI16x8(); case BitmaskVecI16x8: @@ -491,14 +489,16 @@ public: return value.absI32x4(); case NegVecI32x4: return value.negI32x4(); - case AnyTrueVecI32x4: - return value.anyTrueI32x4(); case AllTrueVecI32x4: return value.allTrueI32x4(); case BitmaskVecI32x4: return value.bitmaskI32x4(); + case AbsVecI64x2: + return value.absI64x2(); case NegVecI64x2: return value.negI64x2(); + case AllTrueVecI64x2: + return value.allTrueI64x2(); case BitmaskVecI64x2: WASM_UNREACHABLE("unimp"); case AbsVecF32x4: @@ -541,38 +541,30 @@ public: return value.truncSatToSI32x4(); case TruncSatUVecF32x4ToVecI32x4: return value.truncSatToUI32x4(); - case TruncSatSVecF64x2ToVecI64x2: - return value.truncSatToSI64x2(); - case TruncSatUVecF64x2ToVecI64x2: - return value.truncSatToUI64x2(); case ConvertSVecI32x4ToVecF32x4: return value.convertSToF32x4(); case ConvertUVecI32x4ToVecF32x4: return value.convertUToF32x4(); - case ConvertSVecI64x2ToVecF64x2: - return value.convertSToF64x2(); - case ConvertUVecI64x2ToVecF64x2: - return value.convertUToF64x2(); - case WidenLowSVecI8x16ToVecI16x8: - return value.widenLowSToVecI16x8(); - case WidenHighSVecI8x16ToVecI16x8: - return value.widenHighSToVecI16x8(); - case WidenLowUVecI8x16ToVecI16x8: - return value.widenLowUToVecI16x8(); - case WidenHighUVecI8x16ToVecI16x8: - return value.widenHighUToVecI16x8(); - case WidenLowSVecI16x8ToVecI32x4: - return value.widenLowSToVecI32x4(); - case WidenHighSVecI16x8ToVecI32x4: - return value.widenHighSToVecI32x4(); - case WidenLowUVecI16x8ToVecI32x4: - return value.widenLowUToVecI32x4(); - case WidenHighUVecI16x8ToVecI32x4: - return value.widenHighUToVecI32x4(); - case WidenLowSVecI32x4ToVecI64x2: - case WidenHighSVecI32x4ToVecI64x2: - case WidenLowUVecI32x4ToVecI64x2: - case WidenHighUVecI32x4ToVecI64x2: + case ExtendLowSVecI8x16ToVecI16x8: + return value.extendLowSToVecI16x8(); + case ExtendHighSVecI8x16ToVecI16x8: + return value.extendHighSToVecI16x8(); + case ExtendLowUVecI8x16ToVecI16x8: + return value.extendLowUToVecI16x8(); + case ExtendHighUVecI8x16ToVecI16x8: + return value.extendHighUToVecI16x8(); + case ExtendLowSVecI16x8ToVecI32x4: + return value.extendLowSToVecI32x4(); + case ExtendHighSVecI16x8ToVecI32x4: + return value.extendHighSToVecI32x4(); + case ExtendLowUVecI16x8ToVecI32x4: + return value.extendLowUToVecI32x4(); + case ExtendHighUVecI16x8ToVecI32x4: + return value.extendHighUToVecI32x4(); + case ExtendLowSVecI32x4ToVecI64x2: + case ExtendHighSVecI32x4ToVecI64x2: + case ExtendLowUVecI32x4ToVecI64x2: + case ExtendHighUVecI32x4ToVecI64x2: case ConvertLowSVecI32x4ToVecF64x2: case ConvertLowUVecI32x4ToVecF64x2: case TruncSatZeroSVecF64x2ToVecI32x4: @@ -827,6 +819,16 @@ public: return left.geUI32x4(right); case EqVecI64x2: return left.eqI64x2(right); + case NeVecI64x2: + return left.neI64x2(right); + case LtSVecI64x2: + return left.ltSI64x2(right); + case GtSVecI64x2: + return left.gtSI64x2(right); + case LeSVecI64x2: + return left.leSI64x2(right); + case GeSVecI64x2: + return left.geSI64x2(right); case EqVecF32x4: return left.eqF32x4(right); case NeVecF32x4: @@ -873,8 +875,6 @@ public: return left.subSaturateSI8x16(right); case SubSatUVecI8x16: return left.subSaturateUI8x16(right); - case MulVecI8x16: - return left.mulI8x16(right); case MinSVecI8x16: return left.minSI8x16(right); case MinUVecI8x16: @@ -1142,7 +1142,6 @@ public: } WASM_UNREACHABLE("invalid op"); } - Flow visitSIMDWiden(SIMDWiden* curr) { WASM_UNREACHABLE("unimp"); } Flow visitSelect(Select* curr) { NOTE_ENTER("Select"); Flow ifTrue = visit(curr->ifTrue); @@ -1185,14 +1184,6 @@ public: NOTE_ENTER("Nop"); return Flow(); } - Flow visitPrefetch(Prefetch* curr) { - NOTE_ENTER("Prefetch"); - Flow flow = visit(curr->ptr); - if (flow.breaking()) { - return flow; - } - return Flow(); - } Flow visitUnreachable(Unreachable* curr) { NOTE_ENTER("Unreachable"); trap("unreachable"); diff --git a/src/wasm-s-parser.h b/src/wasm-s-parser.h index 03da961bd..c65043d4c 100644 --- a/src/wasm-s-parser.h +++ b/src/wasm-s-parser.h @@ -234,8 +234,6 @@ private: Expression* makeSIMDShift(Element& s, SIMDShiftOp op); Expression* makeSIMDLoad(Element& s, SIMDLoadOp op); Expression* makeSIMDLoadStoreLane(Element& s, SIMDLoadStoreLaneOp op); - Expression* makeSIMDWiden(Element& s, SIMDWidenOp op); - Expression* makePrefetch(Element& s, PrefetchOp op); Expression* makeMemoryInit(Element& s); Expression* makeDataDrop(Element& s); Expression* makeMemoryCopy(Element& s); diff --git a/src/wasm.h b/src/wasm.h index 101f45c81..8e9f93500 100644 --- a/src/wasm.h +++ b/src/wasm.h @@ -157,23 +157,23 @@ enum UnaryOp { // SIMD arithmetic NotVec128, + AnyTrueVec128, AbsVecI8x16, NegVecI8x16, - AnyTrueVecI8x16, AllTrueVecI8x16, BitmaskVecI8x16, PopcntVecI8x16, AbsVecI16x8, NegVecI16x8, - AnyTrueVecI16x8, AllTrueVecI16x8, BitmaskVecI16x8, AbsVecI32x4, NegVecI32x4, - AnyTrueVecI32x4, AllTrueVecI32x4, BitmaskVecI32x4, + AbsVecI64x2, NegVecI64x2, + AllTrueVecI64x2, BitmaskVecI64x2, AbsVecF32x4, NegVecF32x4, @@ -197,24 +197,20 @@ enum UnaryOp { // SIMD conversions TruncSatSVecF32x4ToVecI32x4, TruncSatUVecF32x4ToVecI32x4, - TruncSatSVecF64x2ToVecI64x2, - TruncSatUVecF64x2ToVecI64x2, ConvertSVecI32x4ToVecF32x4, ConvertUVecI32x4ToVecF32x4, - ConvertSVecI64x2ToVecF64x2, - ConvertUVecI64x2ToVecF64x2, - WidenLowSVecI8x16ToVecI16x8, - WidenHighSVecI8x16ToVecI16x8, - WidenLowUVecI8x16ToVecI16x8, - WidenHighUVecI8x16ToVecI16x8, - WidenLowSVecI16x8ToVecI32x4, - WidenHighSVecI16x8ToVecI32x4, - WidenLowUVecI16x8ToVecI32x4, - WidenHighUVecI16x8ToVecI32x4, - WidenLowSVecI32x4ToVecI64x2, - WidenHighSVecI32x4ToVecI64x2, - WidenLowUVecI32x4ToVecI64x2, - WidenHighUVecI32x4ToVecI64x2, + ExtendLowSVecI8x16ToVecI16x8, + ExtendHighSVecI8x16ToVecI16x8, + ExtendLowUVecI8x16ToVecI16x8, + ExtendHighUVecI8x16ToVecI16x8, + ExtendLowSVecI16x8ToVecI32x4, + ExtendHighSVecI16x8ToVecI32x4, + ExtendLowUVecI16x8ToVecI32x4, + ExtendHighUVecI16x8ToVecI32x4, + ExtendLowSVecI32x4ToVecI64x2, + ExtendHighSVecI32x4ToVecI64x2, + ExtendLowUVecI32x4ToVecI64x2, + ExtendHighUVecI32x4ToVecI64x2, ConvertLowSVecI32x4ToVecF64x2, ConvertLowUVecI32x4ToVecF64x2, @@ -367,6 +363,11 @@ enum BinaryOp { GeSVecI32x4, GeUVecI32x4, EqVecI64x2, + NeVecI64x2, + LtSVecI64x2, + GtSVecI64x2, + LeSVecI64x2, + GeSVecI64x2, EqVecF32x4, NeVecF32x4, LtVecF32x4, @@ -391,7 +392,6 @@ enum BinaryOp { SubVecI8x16, SubSatSVecI8x16, SubSatUVecI8x16, - MulVecI8x16, MinSVecI8x16, MinUVecI8x16, MaxSVecI8x16, @@ -527,24 +527,6 @@ enum SIMDLoadStoreLaneOp { enum SIMDTernaryOp { Bitselect, - QFMAF32x4, - QFMSF32x4, - QFMAF64x2, - QFMSF64x2, - SignSelectVec8x16, - SignSelectVec16x8, - SignSelectVec32x4, - SignSelectVec64x2 -}; - -enum SIMDWidenOp { - WidenSVecI8x16ToVecI32x4, - WidenUVecI8x16ToVecI32x4, -}; - -enum PrefetchOp { - PrefetchTemporal, - PrefetchNontemporal, }; enum RefIsOp { @@ -616,7 +598,6 @@ public: MemorySizeId, MemoryGrowId, NopId, - PrefetchId, UnreachableId, AtomicRMWId, AtomicCmpxchgId, @@ -630,7 +611,6 @@ public: SIMDShiftId, SIMDLoadId, SIMDLoadStoreLaneId, - SIMDWidenId, MemoryInitId, DataDropId, MemoryCopyId, @@ -1079,30 +1059,6 @@ public: void finalize(); }; -class SIMDWiden : public SpecificExpression<Expression::SIMDWidenId> { -public: - SIMDWiden() = default; - SIMDWiden(MixedArena& allocator) {} - - SIMDWidenOp op; - uint8_t index; - Expression* vec; - - void finalize(); -}; - -class Prefetch : public SpecificExpression<Expression::PrefetchId> { -public: - Prefetch() = default; - Prefetch(MixedArena& allocator) : Prefetch() {} - - PrefetchOp op; - Address offset; - Address align; - Expression* ptr; - void finalize(); -}; - class MemoryInit : public SpecificExpression<Expression::MemoryInitId> { public: MemoryInit() = default; diff --git a/src/wasm/literal.cpp b/src/wasm/literal.cpp index 54e817189..0c5a48b8e 100644 --- a/src/wasm/literal.cpp +++ b/src/wasm/literal.cpp @@ -1735,6 +1735,9 @@ Literal Literal::absI16x8() const { Literal Literal::absI32x4() const { return unary<4, &Literal::getLanesI32x4, &Literal::abs>(*this); } +Literal Literal::absI64x2() const { + return unary<2, &Literal::getLanesI64x2, &Literal::abs>(*this); +} Literal Literal::negI8x16() const { return unary<16, &Literal::getLanesUI8x16, &Literal::neg>(*this); } @@ -1798,30 +1801,17 @@ Literal Literal::truncSatToSI32x4() const { Literal Literal::truncSatToUI32x4() const { return unary<4, &Literal::getLanesF32x4, &Literal::truncSatToUI32>(*this); } -Literal Literal::truncSatToSI64x2() const { - return unary<2, &Literal::getLanesF64x2, &Literal::truncSatToSI64>(*this); -} -Literal Literal::truncSatToUI64x2() const { - return unary<2, &Literal::getLanesF64x2, &Literal::truncSatToUI64>(*this); -} Literal Literal::convertSToF32x4() const { return unary<4, &Literal::getLanesI32x4, &Literal::convertSIToF32>(*this); } Literal Literal::convertUToF32x4() const { return unary<4, &Literal::getLanesI32x4, &Literal::convertUIToF32>(*this); } -Literal Literal::convertSToF64x2() const { - return unary<2, &Literal::getLanesI64x2, &Literal::convertSIToF64>(*this); -} -Literal Literal::convertUToF64x2() const { - return unary<2, &Literal::getLanesI64x2, &Literal::convertUIToF64>(*this); -} -template<int Lanes, LaneArray<Lanes> (Literal::*IntoLanes)() const> -static Literal any_true(const Literal& val) { - LaneArray<Lanes> lanes = (val.*IntoLanes)(); - for (size_t i = 0; i < Lanes; ++i) { - if (lanes[i] != Literal::makeZero(lanes[i].type)) { +Literal Literal::anyTrueV128() const { + auto lanes = getLanesI32x4(); + for (size_t i = 0; i < 4; ++i) { + if (lanes[i].geti32() != 0) { return Literal(int32_t(1)); } } @@ -1851,33 +1841,27 @@ static Literal bitmask(const Literal& val) { return Literal(result); } -Literal Literal::anyTrueI8x16() const { - return any_true<16, &Literal::getLanesUI8x16>(*this); -} Literal Literal::allTrueI8x16() const { return all_true<16, &Literal::getLanesUI8x16>(*this); } Literal Literal::bitmaskI8x16() const { return bitmask<16, &Literal::getLanesSI8x16>(*this); } -Literal Literal::anyTrueI16x8() const { - return any_true<8, &Literal::getLanesUI16x8>(*this); -} Literal Literal::allTrueI16x8() const { return all_true<8, &Literal::getLanesUI16x8>(*this); } Literal Literal::bitmaskI16x8() const { return bitmask<8, &Literal::getLanesSI16x8>(*this); } -Literal Literal::anyTrueI32x4() const { - return any_true<4, &Literal::getLanesI32x4>(*this); -} Literal Literal::allTrueI32x4() const { return all_true<4, &Literal::getLanesI32x4>(*this); } Literal Literal::bitmaskI32x4() const { return bitmask<4, &Literal::getLanesI32x4>(*this); } +Literal Literal::allTrueI64x2() const { + return all_true<2, &Literal::getLanesI64x2>(*this); +} template<int Lanes, LaneArray<Lanes> (Literal::*IntoLanes)() const, @@ -2039,6 +2023,26 @@ Literal Literal::eqI64x2(const Literal& other) const { return compare<2, &Literal::getLanesI64x2, &Literal::eq, int64_t>(*this, other); } +Literal Literal::neI64x2(const Literal& other) const { + return compare<2, &Literal::getLanesI64x2, &Literal::ne, int64_t>(*this, + other); +} +Literal Literal::ltSI64x2(const Literal& other) const { + return compare<2, &Literal::getLanesI64x2, &Literal::ltS, int64_t>(*this, + other); +} +Literal Literal::gtSI64x2(const Literal& other) const { + return compare<2, &Literal::getLanesI64x2, &Literal::gtS, int64_t>(*this, + other); +} +Literal Literal::leSI64x2(const Literal& other) const { + return compare<2, &Literal::getLanesI64x2, &Literal::leS, int64_t>(*this, + other); +} +Literal Literal::geSI64x2(const Literal& other) const { + return compare<2, &Literal::getLanesI64x2, &Literal::geS, int64_t>(*this, + other); +} Literal Literal::eqF32x4(const Literal& other) const { return compare<4, &Literal::getLanesF32x4, &Literal::eq>(*this, other); } @@ -2125,9 +2129,6 @@ Literal Literal::subSaturateUI8x16(const Literal& other) const { return binary<16, &Literal::getLanesSI8x16, &Literal::subSatUI8>(*this, other); } -Literal Literal::mulI8x16(const Literal& other) const { - return binary<16, &Literal::getLanesUI8x16, &Literal::mul>(*this, other); -} Literal Literal::minSI8x16(const Literal& other) const { return binary<16, &Literal::getLanesSI8x16, &Literal::minInt>(*this, other); } @@ -2329,7 +2330,7 @@ enum class LaneOrder { Low, High }; template<size_t Lanes, LaneArray<Lanes * 2> (Literal::*IntoLanes)() const, LaneOrder Side> -Literal widen(const Literal& vec) { +Literal extend(const Literal& vec) { LaneArray<Lanes* 2> lanes = (vec.*IntoLanes)(); LaneArray<Lanes> result; for (size_t i = 0; i < Lanes; ++i) { @@ -2338,29 +2339,29 @@ Literal widen(const Literal& vec) { return Literal(result); } -Literal Literal::widenLowSToVecI16x8() const { - return widen<8, &Literal::getLanesSI8x16, LaneOrder::Low>(*this); +Literal Literal::extendLowSToVecI16x8() const { + return extend<8, &Literal::getLanesSI8x16, LaneOrder::Low>(*this); } -Literal Literal::widenHighSToVecI16x8() const { - return widen<8, &Literal::getLanesSI8x16, LaneOrder::High>(*this); +Literal Literal::extendHighSToVecI16x8() const { + return extend<8, &Literal::getLanesSI8x16, LaneOrder::High>(*this); } -Literal Literal::widenLowUToVecI16x8() const { - return widen<8, &Literal::getLanesUI8x16, LaneOrder::Low>(*this); +Literal Literal::extendLowUToVecI16x8() const { + return extend<8, &Literal::getLanesUI8x16, LaneOrder::Low>(*this); } -Literal Literal::widenHighUToVecI16x8() const { - return widen<8, &Literal::getLanesUI8x16, LaneOrder::High>(*this); +Literal Literal::extendHighUToVecI16x8() const { + return extend<8, &Literal::getLanesUI8x16, LaneOrder::High>(*this); } -Literal Literal::widenLowSToVecI32x4() const { - return widen<4, &Literal::getLanesSI16x8, LaneOrder::Low>(*this); +Literal Literal::extendLowSToVecI32x4() const { + return extend<4, &Literal::getLanesSI16x8, LaneOrder::Low>(*this); } -Literal Literal::widenHighSToVecI32x4() const { - return widen<4, &Literal::getLanesSI16x8, LaneOrder::High>(*this); +Literal Literal::extendHighSToVecI32x4() const { + return extend<4, &Literal::getLanesSI16x8, LaneOrder::High>(*this); } -Literal Literal::widenLowUToVecI32x4() const { - return widen<4, &Literal::getLanesUI16x8, LaneOrder::Low>(*this); +Literal Literal::extendLowUToVecI32x4() const { + return extend<4, &Literal::getLanesUI16x8, LaneOrder::Low>(*this); } -Literal Literal::widenHighUToVecI32x4() const { - return widen<4, &Literal::getLanesUI16x8, LaneOrder::High>(*this); +Literal Literal::extendHighUToVecI32x4() const { + return extend<4, &Literal::getLanesUI16x8, LaneOrder::High>(*this); } Literal Literal::extMulLowSI16x8(const Literal& other) const { diff --git a/src/wasm/wasm-binary.cpp b/src/wasm/wasm-binary.cpp index 5425c1551..3a9a1eece 100644 --- a/src/wasm/wasm-binary.cpp +++ b/src/wasm/wasm-binary.cpp @@ -3494,12 +3494,6 @@ BinaryConsts::ASTNodes WasmBinaryBuilder::readExpression(Expression*& curr) { if (maybeVisitSIMDLoadStoreLane(curr, opcode)) { break; } - if (maybeVisitSIMDWiden(curr, opcode)) { - break; - } - if (maybeVisitPrefetch(curr, opcode)) { - break; - } throwError("invalid code after SIMD prefix: " + std::to_string(opcode)); break; } @@ -4970,6 +4964,26 @@ bool WasmBinaryBuilder::maybeVisitSIMDBinary(Expression*& out, uint32_t code) { curr = allocator.alloc<Binary>(); curr->op = EqVecI64x2; break; + case BinaryConsts::I64x2Ne: + curr = allocator.alloc<Binary>(); + curr->op = NeVecI64x2; + break; + case BinaryConsts::I64x2LtS: + curr = allocator.alloc<Binary>(); + curr->op = LtSVecI64x2; + break; + case BinaryConsts::I64x2GtS: + curr = allocator.alloc<Binary>(); + curr->op = GtSVecI64x2; + break; + case BinaryConsts::I64x2LeS: + curr = allocator.alloc<Binary>(); + curr->op = LeSVecI64x2; + break; + case BinaryConsts::I64x2GeS: + curr = allocator.alloc<Binary>(); + curr->op = GeSVecI64x2; + break; case BinaryConsts::F32x4Eq: curr = allocator.alloc<Binary>(); curr->op = EqVecF32x4; @@ -5058,10 +5072,6 @@ bool WasmBinaryBuilder::maybeVisitSIMDBinary(Expression*& out, uint32_t code) { curr = allocator.alloc<Binary>(); curr->op = SubSatUVecI8x16; break; - case BinaryConsts::I8x16Mul: - curr = allocator.alloc<Binary>(); - curr->op = MulVecI8x16; - break; case BinaryConsts::I8x16MinS: curr = allocator.alloc<Binary>(); curr->op = MinSVecI8x16; @@ -5351,6 +5361,10 @@ bool WasmBinaryBuilder::maybeVisitSIMDUnary(Expression*& out, uint32_t code) { curr = allocator.alloc<Unary>(); curr->op = NotVec128; break; + case BinaryConsts::V128AnyTrue: + curr = allocator.alloc<Unary>(); + curr->op = AnyTrueVec128; + break; case BinaryConsts::I8x16Popcnt: curr = allocator.alloc<Unary>(); curr->op = PopcntVecI8x16; @@ -5363,10 +5377,6 @@ bool WasmBinaryBuilder::maybeVisitSIMDUnary(Expression*& out, uint32_t code) { curr = allocator.alloc<Unary>(); curr->op = NegVecI8x16; break; - case BinaryConsts::I8x16AnyTrue: - curr = allocator.alloc<Unary>(); - curr->op = AnyTrueVecI8x16; - break; case BinaryConsts::I8x16AllTrue: curr = allocator.alloc<Unary>(); curr->op = AllTrueVecI8x16; @@ -5383,10 +5393,6 @@ bool WasmBinaryBuilder::maybeVisitSIMDUnary(Expression*& out, uint32_t code) { curr = allocator.alloc<Unary>(); curr->op = NegVecI16x8; break; - case BinaryConsts::I16x8AnyTrue: - curr = allocator.alloc<Unary>(); - curr->op = AnyTrueVecI16x8; - break; case BinaryConsts::I16x8AllTrue: curr = allocator.alloc<Unary>(); curr->op = AllTrueVecI16x8; @@ -5403,10 +5409,6 @@ bool WasmBinaryBuilder::maybeVisitSIMDUnary(Expression*& out, uint32_t code) { curr = allocator.alloc<Unary>(); curr->op = NegVecI32x4; break; - case BinaryConsts::I32x4AnyTrue: - curr = allocator.alloc<Unary>(); - curr->op = AnyTrueVecI32x4; - break; case BinaryConsts::I32x4AllTrue: curr = allocator.alloc<Unary>(); curr->op = AllTrueVecI32x4; @@ -5415,10 +5417,18 @@ bool WasmBinaryBuilder::maybeVisitSIMDUnary(Expression*& out, uint32_t code) { curr = allocator.alloc<Unary>(); curr->op = BitmaskVecI32x4; break; + case BinaryConsts::I64x2Abs: + curr = allocator.alloc<Unary>(); + curr->op = AbsVecI64x2; + break; case BinaryConsts::I64x2Neg: curr = allocator.alloc<Unary>(); curr->op = NegVecI64x2; break; + case BinaryConsts::I64x2AllTrue: + curr = allocator.alloc<Unary>(); + curr->op = AllTrueVecI64x2; + break; case BinaryConsts::I64x2Bitmask: curr = allocator.alloc<Unary>(); curr->op = BitmaskVecI64x2; @@ -5503,14 +5513,6 @@ bool WasmBinaryBuilder::maybeVisitSIMDUnary(Expression*& out, uint32_t code) { curr = allocator.alloc<Unary>(); curr->op = TruncSatUVecF32x4ToVecI32x4; break; - case BinaryConsts::I64x2TruncSatSF64x2: - curr = allocator.alloc<Unary>(); - curr->op = TruncSatSVecF64x2ToVecI64x2; - break; - case BinaryConsts::I64x2TruncSatUF64x2: - curr = allocator.alloc<Unary>(); - curr->op = TruncSatUVecF64x2ToVecI64x2; - break; case BinaryConsts::F32x4ConvertSI32x4: curr = allocator.alloc<Unary>(); curr->op = ConvertSVecI32x4ToVecF32x4; @@ -5519,61 +5521,53 @@ bool WasmBinaryBuilder::maybeVisitSIMDUnary(Expression*& out, uint32_t code) { curr = allocator.alloc<Unary>(); curr->op = ConvertUVecI32x4ToVecF32x4; break; - case BinaryConsts::F64x2ConvertSI64x2: + case BinaryConsts::I16x8ExtendLowSI8x16: curr = allocator.alloc<Unary>(); - curr->op = ConvertSVecI64x2ToVecF64x2; + curr->op = ExtendLowSVecI8x16ToVecI16x8; break; - case BinaryConsts::F64x2ConvertUI64x2: + case BinaryConsts::I16x8ExtendHighSI8x16: curr = allocator.alloc<Unary>(); - curr->op = ConvertUVecI64x2ToVecF64x2; + curr->op = ExtendHighSVecI8x16ToVecI16x8; break; - case BinaryConsts::I16x8WidenLowSI8x16: + case BinaryConsts::I16x8ExtendLowUI8x16: curr = allocator.alloc<Unary>(); - curr->op = WidenLowSVecI8x16ToVecI16x8; + curr->op = ExtendLowUVecI8x16ToVecI16x8; break; - case BinaryConsts::I16x8WidenHighSI8x16: + case BinaryConsts::I16x8ExtendHighUI8x16: curr = allocator.alloc<Unary>(); - curr->op = WidenHighSVecI8x16ToVecI16x8; + curr->op = ExtendHighUVecI8x16ToVecI16x8; break; - case BinaryConsts::I16x8WidenLowUI8x16: + case BinaryConsts::I32x4ExtendLowSI16x8: curr = allocator.alloc<Unary>(); - curr->op = WidenLowUVecI8x16ToVecI16x8; + curr->op = ExtendLowSVecI16x8ToVecI32x4; break; - case BinaryConsts::I16x8WidenHighUI8x16: + case BinaryConsts::I32x4ExtendHighSI16x8: curr = allocator.alloc<Unary>(); - curr->op = WidenHighUVecI8x16ToVecI16x8; + curr->op = ExtendHighSVecI16x8ToVecI32x4; break; - case BinaryConsts::I32x4WidenLowSI16x8: + case BinaryConsts::I32x4ExtendLowUI16x8: curr = allocator.alloc<Unary>(); - curr->op = WidenLowSVecI16x8ToVecI32x4; + curr->op = ExtendLowUVecI16x8ToVecI32x4; break; - case BinaryConsts::I32x4WidenHighSI16x8: + case BinaryConsts::I32x4ExtendHighUI16x8: curr = allocator.alloc<Unary>(); - curr->op = WidenHighSVecI16x8ToVecI32x4; + curr->op = ExtendHighUVecI16x8ToVecI32x4; break; - case BinaryConsts::I32x4WidenLowUI16x8: + case BinaryConsts::I64x2ExtendLowSI32x4: curr = allocator.alloc<Unary>(); - curr->op = WidenLowUVecI16x8ToVecI32x4; + curr->op = ExtendLowSVecI32x4ToVecI64x2; break; - case BinaryConsts::I32x4WidenHighUI16x8: + case BinaryConsts::I64x2ExtendHighSI32x4: curr = allocator.alloc<Unary>(); - curr->op = WidenHighUVecI16x8ToVecI32x4; + curr->op = ExtendHighSVecI32x4ToVecI64x2; break; - case BinaryConsts::I64x2WidenLowSI32x4: + case BinaryConsts::I64x2ExtendLowUI32x4: curr = allocator.alloc<Unary>(); - curr->op = WidenLowSVecI32x4ToVecI64x2; + curr->op = ExtendLowUVecI32x4ToVecI64x2; break; - case BinaryConsts::I64x2WidenHighSI32x4: + case BinaryConsts::I64x2ExtendHighUI32x4: curr = allocator.alloc<Unary>(); - curr->op = WidenHighSVecI32x4ToVecI64x2; - break; - case BinaryConsts::I64x2WidenLowUI32x4: - curr = allocator.alloc<Unary>(); - curr->op = WidenLowUVecI32x4ToVecI64x2; - break; - case BinaryConsts::I64x2WidenHighUI32x4: - curr = allocator.alloc<Unary>(); - curr->op = WidenHighUVecI32x4ToVecI64x2; + curr->op = ExtendHighUVecI32x4ToVecI64x2; break; case BinaryConsts::F64x2ConvertLowSI32x4: curr = allocator.alloc<Unary>(); @@ -5752,38 +5746,6 @@ bool WasmBinaryBuilder::maybeVisitSIMDTernary(Expression*& out, uint32_t code) { curr = allocator.alloc<SIMDTernary>(); curr->op = Bitselect; break; - case BinaryConsts::V8x16SignSelect: - curr = allocator.alloc<SIMDTernary>(); - curr->op = SignSelectVec8x16; - break; - case BinaryConsts::V16x8SignSelect: - curr = allocator.alloc<SIMDTernary>(); - curr->op = SignSelectVec16x8; - break; - case BinaryConsts::V32x4SignSelect: - curr = allocator.alloc<SIMDTernary>(); - curr->op = SignSelectVec32x4; - break; - case BinaryConsts::V64x2SignSelect: - curr = allocator.alloc<SIMDTernary>(); - curr->op = SignSelectVec64x2; - break; - case BinaryConsts::F32x4QFMA: - curr = allocator.alloc<SIMDTernary>(); - curr->op = QFMAF32x4; - break; - case BinaryConsts::F32x4QFMS: - curr = allocator.alloc<SIMDTernary>(); - curr->op = QFMSF32x4; - break; - case BinaryConsts::F64x2QFMA: - curr = allocator.alloc<SIMDTernary>(); - curr->op = QFMAF64x2; - break; - case BinaryConsts::F64x2QFMS: - curr = allocator.alloc<SIMDTernary>(); - curr->op = QFMSF64x2; - break; default: return false; } @@ -5979,45 +5941,6 @@ bool WasmBinaryBuilder::maybeVisitSIMDLoadStoreLane(Expression*& out, return true; } -bool WasmBinaryBuilder::maybeVisitSIMDWiden(Expression*& out, uint32_t code) { - SIMDWidenOp op; - switch (code) { - case BinaryConsts::I32x4WidenSI8x16: - op = WidenSVecI8x16ToVecI32x4; - break; - case BinaryConsts::I32x4WidenUI8x16: - op = WidenUVecI8x16ToVecI32x4; - break; - default: - return false; - } - auto* curr = allocator.alloc<SIMDWiden>(); - curr->op = op; - curr->index = getLaneIndex(4); - curr->vec = popNonVoidExpression(); - curr->finalize(); - out = curr; - return true; -} - -bool WasmBinaryBuilder::maybeVisitPrefetch(Expression*& out, uint32_t code) { - PrefetchOp op; - switch (code) { - case BinaryConsts::PrefetchT: - op = PrefetchTemporal; - break; - case BinaryConsts::PrefetchNT: - op = PrefetchNontemporal; - break; - default: - return false; - } - Address align, offset; - readMemoryAccess(align, offset); - out = Builder(wasm).makePrefetch(op, offset, align, popNonVoidExpression()); - return true; -} - void WasmBinaryBuilder::visitSelect(Select* curr, uint8_t code) { BYN_TRACE("zz node: Select, code " << int32_t(code) << std::endl); if (code == BinaryConsts::SelectWithType) { diff --git a/src/wasm/wasm-s-parser.cpp b/src/wasm/wasm-s-parser.cpp index f3ab50f0c..a42b341f6 100644 --- a/src/wasm/wasm-s-parser.cpp +++ b/src/wasm/wasm-s-parser.cpp @@ -2095,21 +2095,6 @@ SExpressionWasmBuilder::makeSIMDLoadStoreLane(Element& s, return ret; } -Expression* SExpressionWasmBuilder::makeSIMDWiden(Element& s, SIMDWidenOp op) { - auto* ret = allocator.alloc<SIMDWiden>(); - ret->op = op; - ret->index = parseLaneIndex(s[1], 4); - ret->vec = parseExpression(s[2]); - ret->finalize(); - return ret; -} - -Expression* SExpressionWasmBuilder::makePrefetch(Element& s, PrefetchOp op) { - Address offset, align; - size_t i = parseMemAttributes(s, offset, align, /*defaultAlign*/ 1); - return Builder(wasm).makePrefetch(op, offset, align, parseExpression(s[i])); -} - Expression* SExpressionWasmBuilder::makeMemoryInit(Element& s) { auto ret = allocator.alloc<MemoryInit>(); ret->segment = atoi(s[1]->str().c_str()); diff --git a/src/wasm/wasm-stack.cpp b/src/wasm/wasm-stack.cpp index b1e9a75cc..87dfe8247 100644 --- a/src/wasm/wasm-stack.cpp +++ b/src/wasm/wasm-stack.cpp @@ -546,30 +546,6 @@ void BinaryInstWriter::visitSIMDTernary(SIMDTernary* curr) { case Bitselect: o << U32LEB(BinaryConsts::V128Bitselect); break; - case QFMAF32x4: - o << U32LEB(BinaryConsts::F32x4QFMA); - break; - case QFMSF32x4: - o << U32LEB(BinaryConsts::F32x4QFMS); - break; - case QFMAF64x2: - o << U32LEB(BinaryConsts::F64x2QFMA); - break; - case QFMSF64x2: - o << U32LEB(BinaryConsts::F64x2QFMS); - break; - case SignSelectVec8x16: - o << U32LEB(BinaryConsts::V8x16SignSelect); - break; - case SignSelectVec16x8: - o << U32LEB(BinaryConsts::V16x8SignSelect); - break; - case SignSelectVec32x4: - o << U32LEB(BinaryConsts::V32x4SignSelect); - break; - case SignSelectVec64x2: - o << U32LEB(BinaryConsts::V64x2SignSelect); - break; } } @@ -692,33 +668,6 @@ void BinaryInstWriter::visitSIMDLoadStoreLane(SIMDLoadStoreLane* curr) { o << curr->index; } -void BinaryInstWriter::visitSIMDWiden(SIMDWiden* curr) { - o << int8_t(BinaryConsts::SIMDPrefix); - switch (curr->op) { - case WidenSVecI8x16ToVecI32x4: - o << U32LEB(BinaryConsts::I32x4WidenSI8x16); - break; - case WidenUVecI8x16ToVecI32x4: - o << U32LEB(BinaryConsts::I32x4WidenUI8x16); - break; - } - o << uint8_t(curr->index); -} - -void BinaryInstWriter::visitPrefetch(Prefetch* curr) { - o << int8_t(BinaryConsts::SIMDPrefix); - switch (curr->op) { - case PrefetchTemporal: - o << U32LEB(BinaryConsts::PrefetchT); - break; - case PrefetchNontemporal: - o << U32LEB(BinaryConsts::PrefetchNT); - break; - } - assert(curr->align); - emitMemoryAccess(curr->align, /*(unused) bytes=*/0, curr->offset); -} - void BinaryInstWriter::visitMemoryInit(MemoryInit* curr) { o << int8_t(BinaryConsts::MiscPrefix); o << U32LEB(BinaryConsts::MemoryInit); @@ -992,16 +941,16 @@ void BinaryInstWriter::visitUnary(Unary* curr) { case NotVec128: o << int8_t(BinaryConsts::SIMDPrefix) << U32LEB(BinaryConsts::V128Not); break; + case AnyTrueVec128: + o << int8_t(BinaryConsts::SIMDPrefix) + << U32LEB(BinaryConsts::V128AnyTrue); + break; case AbsVecI8x16: o << int8_t(BinaryConsts::SIMDPrefix) << U32LEB(BinaryConsts::I8x16Abs); break; case NegVecI8x16: o << int8_t(BinaryConsts::SIMDPrefix) << U32LEB(BinaryConsts::I8x16Neg); break; - case AnyTrueVecI8x16: - o << int8_t(BinaryConsts::SIMDPrefix) - << U32LEB(BinaryConsts::I8x16AnyTrue); - break; case AllTrueVecI8x16: o << int8_t(BinaryConsts::SIMDPrefix) << U32LEB(BinaryConsts::I8x16AllTrue); @@ -1020,10 +969,6 @@ void BinaryInstWriter::visitUnary(Unary* curr) { case NegVecI16x8: o << int8_t(BinaryConsts::SIMDPrefix) << U32LEB(BinaryConsts::I16x8Neg); break; - case AnyTrueVecI16x8: - o << int8_t(BinaryConsts::SIMDPrefix) - << U32LEB(BinaryConsts::I16x8AnyTrue); - break; case AllTrueVecI16x8: o << int8_t(BinaryConsts::SIMDPrefix) << U32LEB(BinaryConsts::I16x8AllTrue); @@ -1038,10 +983,6 @@ void BinaryInstWriter::visitUnary(Unary* curr) { case NegVecI32x4: o << int8_t(BinaryConsts::SIMDPrefix) << U32LEB(BinaryConsts::I32x4Neg); break; - case AnyTrueVecI32x4: - o << int8_t(BinaryConsts::SIMDPrefix) - << U32LEB(BinaryConsts::I32x4AnyTrue); - break; case AllTrueVecI32x4: o << int8_t(BinaryConsts::SIMDPrefix) << U32LEB(BinaryConsts::I32x4AllTrue); @@ -1050,9 +991,16 @@ void BinaryInstWriter::visitUnary(Unary* curr) { o << int8_t(BinaryConsts::SIMDPrefix) << U32LEB(BinaryConsts::I32x4Bitmask); break; + case AbsVecI64x2: + o << int8_t(BinaryConsts::SIMDPrefix) << U32LEB(BinaryConsts::I64x2Abs); + break; case NegVecI64x2: o << int8_t(BinaryConsts::SIMDPrefix) << U32LEB(BinaryConsts::I64x2Neg); break; + case AllTrueVecI64x2: + o << int8_t(BinaryConsts::SIMDPrefix) + << U32LEB(BinaryConsts::I64x2AllTrue); + break; case BitmaskVecI64x2: o << int8_t(BinaryConsts::SIMDPrefix) << U32LEB(BinaryConsts::I64x2Bitmask); @@ -1125,14 +1073,6 @@ void BinaryInstWriter::visitUnary(Unary* curr) { o << int8_t(BinaryConsts::SIMDPrefix) << U32LEB(BinaryConsts::I32x4TruncSatUF32x4); break; - case TruncSatSVecF64x2ToVecI64x2: - o << int8_t(BinaryConsts::SIMDPrefix) - << U32LEB(BinaryConsts::I64x2TruncSatSF64x2); - break; - case TruncSatUVecF64x2ToVecI64x2: - o << int8_t(BinaryConsts::SIMDPrefix) - << U32LEB(BinaryConsts::I64x2TruncSatUF64x2); - break; case ConvertSVecI32x4ToVecF32x4: o << int8_t(BinaryConsts::SIMDPrefix) << U32LEB(BinaryConsts::F32x4ConvertSI32x4); @@ -1141,61 +1081,53 @@ void BinaryInstWriter::visitUnary(Unary* curr) { o << int8_t(BinaryConsts::SIMDPrefix) << U32LEB(BinaryConsts::F32x4ConvertUI32x4); break; - case ConvertSVecI64x2ToVecF64x2: - o << int8_t(BinaryConsts::SIMDPrefix) - << U32LEB(BinaryConsts::F64x2ConvertSI64x2); - break; - case ConvertUVecI64x2ToVecF64x2: - o << int8_t(BinaryConsts::SIMDPrefix) - << U32LEB(BinaryConsts::F64x2ConvertUI64x2); - break; - case WidenLowSVecI8x16ToVecI16x8: + case ExtendLowSVecI8x16ToVecI16x8: o << int8_t(BinaryConsts::SIMDPrefix) - << U32LEB(BinaryConsts::I16x8WidenLowSI8x16); + << U32LEB(BinaryConsts::I16x8ExtendLowSI8x16); break; - case WidenHighSVecI8x16ToVecI16x8: + case ExtendHighSVecI8x16ToVecI16x8: o << int8_t(BinaryConsts::SIMDPrefix) - << U32LEB(BinaryConsts::I16x8WidenHighSI8x16); + << U32LEB(BinaryConsts::I16x8ExtendHighSI8x16); break; - case WidenLowUVecI8x16ToVecI16x8: + case ExtendLowUVecI8x16ToVecI16x8: o << int8_t(BinaryConsts::SIMDPrefix) - << U32LEB(BinaryConsts::I16x8WidenLowUI8x16); + << U32LEB(BinaryConsts::I16x8ExtendLowUI8x16); break; - case WidenHighUVecI8x16ToVecI16x8: + case ExtendHighUVecI8x16ToVecI16x8: o << int8_t(BinaryConsts::SIMDPrefix) - << U32LEB(BinaryConsts::I16x8WidenHighUI8x16); + << U32LEB(BinaryConsts::I16x8ExtendHighUI8x16); break; - case WidenLowSVecI16x8ToVecI32x4: + case ExtendLowSVecI16x8ToVecI32x4: o << int8_t(BinaryConsts::SIMDPrefix) - << U32LEB(BinaryConsts::I32x4WidenLowSI16x8); + << U32LEB(BinaryConsts::I32x4ExtendLowSI16x8); break; - case WidenHighSVecI16x8ToVecI32x4: + case ExtendHighSVecI16x8ToVecI32x4: o << int8_t(BinaryConsts::SIMDPrefix) - << U32LEB(BinaryConsts::I32x4WidenHighSI16x8); + << U32LEB(BinaryConsts::I32x4ExtendHighSI16x8); break; - case WidenLowUVecI16x8ToVecI32x4: + case ExtendLowUVecI16x8ToVecI32x4: o << int8_t(BinaryConsts::SIMDPrefix) - << U32LEB(BinaryConsts::I32x4WidenLowUI16x8); + << U32LEB(BinaryConsts::I32x4ExtendLowUI16x8); break; - case WidenHighUVecI16x8ToVecI32x4: + case ExtendHighUVecI16x8ToVecI32x4: o << int8_t(BinaryConsts::SIMDPrefix) - << U32LEB(BinaryConsts::I32x4WidenHighUI16x8); + << U32LEB(BinaryConsts::I32x4ExtendHighUI16x8); break; - case WidenLowSVecI32x4ToVecI64x2: + case ExtendLowSVecI32x4ToVecI64x2: o << int8_t(BinaryConsts::SIMDPrefix) - << U32LEB(BinaryConsts::I64x2WidenLowSI32x4); + << U32LEB(BinaryConsts::I64x2ExtendLowSI32x4); break; - case WidenHighSVecI32x4ToVecI64x2: + case ExtendHighSVecI32x4ToVecI64x2: o << int8_t(BinaryConsts::SIMDPrefix) - << U32LEB(BinaryConsts::I64x2WidenHighSI32x4); + << U32LEB(BinaryConsts::I64x2ExtendHighSI32x4); break; - case WidenLowUVecI32x4ToVecI64x2: + case ExtendLowUVecI32x4ToVecI64x2: o << int8_t(BinaryConsts::SIMDPrefix) - << U32LEB(BinaryConsts::I64x2WidenLowUI32x4); + << U32LEB(BinaryConsts::I64x2ExtendLowUI32x4); break; - case WidenHighUVecI32x4ToVecI64x2: + case ExtendHighUVecI32x4ToVecI64x2: o << int8_t(BinaryConsts::SIMDPrefix) - << U32LEB(BinaryConsts::I64x2WidenHighUI32x4); + << U32LEB(BinaryConsts::I64x2ExtendHighUI32x4); break; case ConvertLowSVecI32x4ToVecF64x2: o << int8_t(BinaryConsts::SIMDPrefix) @@ -1553,6 +1485,21 @@ void BinaryInstWriter::visitBinary(Binary* curr) { case EqVecI64x2: o << int8_t(BinaryConsts::SIMDPrefix) << U32LEB(BinaryConsts::I64x2Eq); break; + case NeVecI64x2: + o << int8_t(BinaryConsts::SIMDPrefix) << U32LEB(BinaryConsts::I64x2Ne); + break; + case LtSVecI64x2: + o << int8_t(BinaryConsts::SIMDPrefix) << U32LEB(BinaryConsts::I64x2LtS); + break; + case GtSVecI64x2: + o << int8_t(BinaryConsts::SIMDPrefix) << U32LEB(BinaryConsts::I64x2GtS); + break; + case LeSVecI64x2: + o << int8_t(BinaryConsts::SIMDPrefix) << U32LEB(BinaryConsts::I64x2LeS); + break; + case GeSVecI64x2: + o << int8_t(BinaryConsts::SIMDPrefix) << U32LEB(BinaryConsts::I64x2GeS); + break; case EqVecF32x4: o << int8_t(BinaryConsts::SIMDPrefix) << U32LEB(BinaryConsts::F32x4Eq); break; @@ -1623,9 +1570,6 @@ void BinaryInstWriter::visitBinary(Binary* curr) { o << int8_t(BinaryConsts::SIMDPrefix) << U32LEB(BinaryConsts::I8x16SubSatU); break; - case MulVecI8x16: - o << int8_t(BinaryConsts::SIMDPrefix) << U32LEB(BinaryConsts::I8x16Mul); - break; case MinSVecI8x16: o << int8_t(BinaryConsts::SIMDPrefix) << U32LEB(BinaryConsts::I8x16MinS); break; diff --git a/src/wasm/wasm-validator.cpp b/src/wasm/wasm-validator.cpp index 8bc12cadf..a8399bfdf 100644 --- a/src/wasm/wasm-validator.cpp +++ b/src/wasm/wasm-validator.cpp @@ -1137,7 +1137,7 @@ void FunctionValidator::visitSIMDShuffle(SIMDShuffle* curr) { shouldBeTrue( getModule()->features.hasSIMD(), curr, "SIMD operation (SIMD is disabled)"); shouldBeEqualOrFirstIsUnreachable( - curr->type, Type(Type::v128), curr, "v128.shuffle must have type v128"); + curr->type, Type(Type::v128), curr, "i8x16.shuffle must have type v128"); shouldBeEqualOrFirstIsUnreachable( curr->left->type, Type(Type::v128), curr, "expected operand of type v128"); shouldBeEqualOrFirstIsUnreachable( @@ -1524,6 +1524,11 @@ void FunctionValidator::visitBinary(Binary* curr) { case GeSVecI32x4: case GeUVecI32x4: case EqVecI64x2: + case NeVecI64x2: + case LtSVecI64x2: + case LeSVecI64x2: + case GtSVecI64x2: + case GeSVecI64x2: case EqVecF32x4: case NeVecF32x4: case LtVecF32x4: @@ -1546,7 +1551,6 @@ void FunctionValidator::visitBinary(Binary* curr) { case SubVecI8x16: case SubSatSVecI8x16: case SubSatUVecI8x16: - case MulVecI8x16: case MinSVecI8x16: case MinUVecI8x16: case MaxSVecI8x16: @@ -1836,6 +1840,7 @@ void FunctionValidator::visitUnary(Unary* curr) { case AbsVecI8x16: case AbsVecI16x8: case AbsVecI32x4: + case AbsVecI64x2: case NegVecI8x16: case NegVecI16x8: case NegVecI32x4: @@ -1860,24 +1865,20 @@ void FunctionValidator::visitUnary(Unary* curr) { case ExtAddPairwiseUVecI16x8ToI32x4: case TruncSatSVecF32x4ToVecI32x4: case TruncSatUVecF32x4ToVecI32x4: - case TruncSatSVecF64x2ToVecI64x2: - case TruncSatUVecF64x2ToVecI64x2: case ConvertSVecI32x4ToVecF32x4: case ConvertUVecI32x4ToVecF32x4: - case ConvertSVecI64x2ToVecF64x2: - case ConvertUVecI64x2ToVecF64x2: - case WidenLowSVecI8x16ToVecI16x8: - case WidenHighSVecI8x16ToVecI16x8: - case WidenLowUVecI8x16ToVecI16x8: - case WidenHighUVecI8x16ToVecI16x8: - case WidenLowSVecI16x8ToVecI32x4: - case WidenHighSVecI16x8ToVecI32x4: - case WidenLowUVecI16x8ToVecI32x4: - case WidenHighUVecI16x8ToVecI32x4: - case WidenLowSVecI32x4ToVecI64x2: - case WidenHighSVecI32x4ToVecI64x2: - case WidenLowUVecI32x4ToVecI64x2: - case WidenHighUVecI32x4ToVecI64x2: + case ExtendLowSVecI8x16ToVecI16x8: + case ExtendHighSVecI8x16ToVecI16x8: + case ExtendLowUVecI8x16ToVecI16x8: + case ExtendHighUVecI8x16ToVecI16x8: + case ExtendLowSVecI16x8ToVecI32x4: + case ExtendHighSVecI16x8ToVecI32x4: + case ExtendLowUVecI16x8ToVecI32x4: + case ExtendHighUVecI16x8ToVecI32x4: + case ExtendLowSVecI32x4ToVecI64x2: + case ExtendHighSVecI32x4ToVecI64x2: + case ExtendLowUVecI32x4ToVecI64x2: + case ExtendHighUVecI32x4ToVecI64x2: case ConvertLowSVecI32x4ToVecF64x2: case ConvertLowUVecI32x4ToVecF64x2: case TruncSatZeroSVecF64x2ToVecI32x4: @@ -1888,12 +1889,11 @@ void FunctionValidator::visitUnary(Unary* curr) { shouldBeEqual( curr->value->type, Type(Type::v128), curr, "expected v128 operand"); break; - case AnyTrueVecI8x16: - case AnyTrueVecI16x8: - case AnyTrueVecI32x4: + case AnyTrueVec128: case AllTrueVecI8x16: case AllTrueVecI16x8: case AllTrueVecI32x4: + case AllTrueVecI64x2: case BitmaskVecI8x16: case BitmaskVecI16x8: case BitmaskVecI32x4: diff --git a/src/wasm/wasm.cpp b/src/wasm/wasm.cpp index 2ccd9a70a..b227ff99f 100644 --- a/src/wasm/wasm.cpp +++ b/src/wasm/wasm.cpp @@ -499,11 +499,6 @@ void SIMDLoadStoreLane::finalize() { } } -void SIMDWiden::finalize() { - assert(vec); - type = vec->type == Type::unreachable ? Type::unreachable : Type::v128; -} - Index SIMDLoadStoreLane::getMemBytes() { switch (op) { case LoadLaneVec8x16: @@ -538,10 +533,6 @@ bool SIMDLoadStoreLane::isStore() { WASM_UNREACHABLE("unexpected op"); } -void Prefetch::finalize() { - type = ptr->type == Type::unreachable ? Type::unreachable : Type::none; -} - Const* Const::set(Literal value_) { value = value_; type = value.type; @@ -650,6 +641,7 @@ void Unary::finalize() { case AbsVecI8x16: case AbsVecI16x8: case AbsVecI32x4: + case AbsVecI64x2: case PopcntVecI8x16: case NegVecI8x16: case NegVecI16x8: @@ -675,24 +667,20 @@ void Unary::finalize() { case ExtAddPairwiseUVecI16x8ToI32x4: case TruncSatSVecF32x4ToVecI32x4: case TruncSatUVecF32x4ToVecI32x4: - case TruncSatSVecF64x2ToVecI64x2: - case TruncSatUVecF64x2ToVecI64x2: case ConvertSVecI32x4ToVecF32x4: case ConvertUVecI32x4ToVecF32x4: - case ConvertSVecI64x2ToVecF64x2: - case ConvertUVecI64x2ToVecF64x2: - case WidenLowSVecI8x16ToVecI16x8: - case WidenHighSVecI8x16ToVecI16x8: - case WidenLowUVecI8x16ToVecI16x8: - case WidenHighUVecI8x16ToVecI16x8: - case WidenLowSVecI16x8ToVecI32x4: - case WidenHighSVecI16x8ToVecI32x4: - case WidenLowUVecI16x8ToVecI32x4: - case WidenHighUVecI16x8ToVecI32x4: - case WidenLowSVecI32x4ToVecI64x2: - case WidenHighSVecI32x4ToVecI64x2: - case WidenLowUVecI32x4ToVecI64x2: - case WidenHighUVecI32x4ToVecI64x2: + case ExtendLowSVecI8x16ToVecI16x8: + case ExtendHighSVecI8x16ToVecI16x8: + case ExtendLowUVecI8x16ToVecI16x8: + case ExtendHighUVecI8x16ToVecI16x8: + case ExtendLowSVecI16x8ToVecI32x4: + case ExtendHighSVecI16x8ToVecI32x4: + case ExtendLowUVecI16x8ToVecI32x4: + case ExtendHighUVecI16x8ToVecI32x4: + case ExtendLowSVecI32x4ToVecI64x2: + case ExtendHighSVecI32x4ToVecI64x2: + case ExtendLowUVecI32x4ToVecI64x2: + case ExtendHighUVecI32x4ToVecI64x2: case ConvertLowSVecI32x4ToVecF64x2: case ConvertLowUVecI32x4ToVecF64x2: case TruncSatZeroSVecF64x2ToVecI32x4: @@ -701,12 +689,11 @@ void Unary::finalize() { case PromoteLowVecF32x4ToVecF64x2: type = Type::v128; break; - case AnyTrueVecI8x16: - case AnyTrueVecI16x8: - case AnyTrueVecI32x4: + case AnyTrueVec128: case AllTrueVecI8x16: case AllTrueVecI16x8: case AllTrueVecI32x4: + case AllTrueVecI64x2: case BitmaskVecI8x16: case BitmaskVecI16x8: case BitmaskVecI32x4: diff --git a/src/wasm2js.h b/src/wasm2js.h index 6d3f0682c..9a1a93b2b 100644 --- a/src/wasm2js.h +++ b/src/wasm2js.h @@ -1988,8 +1988,6 @@ Ref Wasm2JSBuilder::processFunctionBody(Module* m, } Ref visitNop(Nop* curr) { return ValueBuilder::makeToplevel(); } - Ref visitPrefetch(Prefetch* curr) { return ValueBuilder::makeToplevel(); } - Ref visitUnreachable(Unreachable* curr) { return ValueBuilder::makeCall(ABORT_FUNC); } @@ -2124,10 +2122,6 @@ Ref Wasm2JSBuilder::processFunctionBody(Module* m, unimplemented(curr); WASM_UNREACHABLE("unimp"); } - Ref visitSIMDWiden(SIMDWiden* curr) { - unimplemented(curr); - WASM_UNREACHABLE("unimp"); - } Ref visitMemoryInit(MemoryInit* curr) { ABI::wasm2js::ensureHelpers(module, ABI::wasm2js::MEMORY_INIT); return ValueBuilder::makeCall(ABI::wasm2js::MEMORY_INIT, diff --git a/test/binaryen.js/exception-handling.js.txt b/test/binaryen.js/exception-handling.js.txt index 434546126..386abfb4d 100644 --- a/test/binaryen.js/exception-handling.js.txt +++ b/test/binaryen.js/exception-handling.js.txt @@ -34,7 +34,7 @@ ) ) -getExpressionInfo(throw) = {"id":48,"type":1,"event":"e"} -getExpressionInfo(rethrow) = {"id":49,"type":1,"target":"l0"} -getExpressionInfo(try_catch) = {"id":47,"type":1,"name":"l0","hasCatchAll":0,"delegateTarget":"","isDelegate":0} -getExpressionInfo(try_delegate) = {"id":47,"type":0,"name":"try_outer","hasCatchAll":1,"delegateTarget":"","isDelegate":0} +getExpressionInfo(throw) = {"id":46,"type":1,"event":"e"} +getExpressionInfo(rethrow) = {"id":47,"type":1,"target":"l0"} +getExpressionInfo(try_catch) = {"id":45,"type":1,"name":"l0","hasCatchAll":0,"delegateTarget":"","isDelegate":0} +getExpressionInfo(try_delegate) = {"id":45,"type":0,"name":"try_outer","hasCatchAll":1,"delegateTarget":"","isDelegate":0} diff --git a/test/binaryen.js/expressions.js b/test/binaryen.js/expressions.js index d6256cec1..68f5c8cee 100644 --- a/test/binaryen.js/expressions.js +++ b/test/binaryen.js/expressions.js @@ -1079,7 +1079,7 @@ console.log("# SIMDShuffle"); var left = module.v128.const([1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16]); var right = module.v128.const([2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17]); var mask = [3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18]; - const theSIMDShuffle = binaryen.SIMDShuffle(module.v8x16.shuffle(left, right, mask)); + const theSIMDShuffle = binaryen.SIMDShuffle(module.i8x16.shuffle(left, right, mask)); assert(theSIMDShuffle instanceof binaryen.SIMDShuffle); assert(theSIMDShuffle instanceof binaryen.Expression); assert(theSIMDShuffle.left === left); @@ -1101,7 +1101,7 @@ console.log("# SIMDShuffle"); assert( theSIMDShuffle.toText() == - "(v8x16.shuffle 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3\n (v128.const i32x4 0x01010101 0x01010101 0x01010101 0x01010101)\n (v128.const i32x4 0x02020202 0x02020202 0x02020202 0x02020202)\n)\n" + "(i8x16.shuffle 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3\n (v128.const i32x4 0x01010101 0x01010101 0x01010101 0x01010101)\n (v128.const i32x4 0x02020202 0x02020202 0x02020202 0x02020202)\n)\n" ); module.dispose(); @@ -1124,23 +1124,11 @@ console.log("# SIMDTernary"); assert(theSIMDTernary.c === c); assert(theSIMDTernary.type === binaryen.v128); - theSIMDTernary.op = op = binaryen.Operations.QFMAVecF64x2; - assert(theSIMDTernary.op === op); - theSIMDTernary.a = a = module.v128.const([1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1]); - assert(theSIMDTernary.a === a); - theSIMDTernary.b = b = module.v128.const([2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2]); - assert(theSIMDTernary.b === b); - theSIMDTernary.c = c = module.v128.const([3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3]); - assert(theSIMDTernary.c === c); - theSIMDTernary.type = binaryen.f64; - theSIMDTernary.finalize(); - assert(theSIMDTernary.type === binaryen.v128); - - console.log(theSIMDTernary.toText()); + console.log(theSIMDTernary.toText() + "\n"); assert( theSIMDTernary.toText() == - "(f64x2.qfma\n (v128.const i32x4 0x01010101 0x01010101 0x01010101 0x01010101)\n (v128.const i32x4 0x02020202 0x02020202 0x02020202 0x02020202)\n (v128.const i32x4 0x03030303 0x03030303 0x03030303 0x03030303)\n)\n" + "(v128.bitselect\n (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)\n (v128.const i32x4 0x05040302 0x09080706 0x0d0c0b0a 0x11100f0e)\n (v128.const i32x4 0x06050403 0x0a090807 0x0e0d0c0b 0x1211100f)\n)\n" ); module.dispose(); @@ -1189,7 +1177,7 @@ console.log("# SIMDLoad"); var offset = 16; var align = 2; var ptr = module.i32.const(1); - const theSIMDLoad = binaryen.SIMDLoad(module.i16x8.load8x8_s(offset, align, ptr)); + const theSIMDLoad = binaryen.SIMDLoad(module.v128.load8x8_s(offset, align, ptr)); assert(theSIMDLoad instanceof binaryen.SIMDLoad); assert(theSIMDLoad instanceof binaryen.Expression); assert(theSIMDLoad.offset === offset); @@ -1213,7 +1201,7 @@ console.log("# SIMDLoad"); assert( theSIMDLoad.toText() == - "(v8x16.load_splat offset=32 align=4\n (i32.const 2)\n)\n" + "(v128.load8_splat offset=32 align=4\n (i32.const 2)\n)\n" ); module.dispose(); diff --git a/test/binaryen.js/expressions.js.txt b/test/binaryen.js/expressions.js.txt index 0467e9332..a3e877b0f 100644 --- a/test/binaryen.js/expressions.js.txt +++ b/test/binaryen.js/expressions.js.txt @@ -155,18 +155,19 @@ ) # SIMDShuffle -(v8x16.shuffle 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 +(i8x16.shuffle 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 (v128.const i32x4 0x01010101 0x01010101 0x01010101 0x01010101) (v128.const i32x4 0x02020202 0x02020202 0x02020202 0x02020202) ) # SIMDTernary -(f64x2.qfma - (v128.const i32x4 0x01010101 0x01010101 0x01010101 0x01010101) - (v128.const i32x4 0x02020202 0x02020202 0x02020202 0x02020202) - (v128.const i32x4 0x03030303 0x03030303 0x03030303 0x03030303) +(v128.bitselect + (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) + (v128.const i32x4 0x05040302 0x09080706 0x0d0c0b0a 0x11100f0e) + (v128.const i32x4 0x06050403 0x0a090807 0x0e0d0c0b 0x1211100f) ) + # SIMDShift (i8x16.shr_s (v128.const i32x4 0x01010101 0x01010101 0x01010101 0x01010101) @@ -174,7 +175,7 @@ ) # SIMDLoad -(v8x16.load_splat offset=32 align=4 +(v128.load8_splat offset=32 align=4 (i32.const 2) ) diff --git a/test/binaryen.js/kitchen-sink.js b/test/binaryen.js/kitchen-sink.js index 084a7d47b..cb2cd32ff 100644 --- a/test/binaryen.js/kitchen-sink.js +++ b/test/binaryen.js/kitchen-sink.js @@ -265,19 +265,17 @@ function test_core() { module.f32x4.splat(module.f32.const(42.0)), module.f64x2.splat(module.f64.const(42.0)), module.v128.not(module.v128.const(v128_bytes)), + module.v128.any_true(module.v128.const(v128_bytes)), module.i8x16.abs(module.v128.const(v128_bytes)), module.i8x16.neg(module.v128.const(v128_bytes)), - module.i8x16.any_true(module.v128.const(v128_bytes)), module.i8x16.all_true(module.v128.const(v128_bytes)), module.i8x16.bitmask(module.v128.const(v128_bytes)), module.i16x8.abs(module.v128.const(v128_bytes)), module.i16x8.neg(module.v128.const(v128_bytes)), - module.i16x8.any_true(module.v128.const(v128_bytes)), module.i16x8.all_true(module.v128.const(v128_bytes)), module.i16x8.bitmask(module.v128.const(v128_bytes)), module.i32x4.abs(module.v128.const(v128_bytes)), module.i32x4.neg(module.v128.const(v128_bytes)), - module.i32x4.any_true(module.v128.const(v128_bytes)), module.i32x4.all_true(module.v128.const(v128_bytes)), module.i32x4.bitmask(module.v128.const(v128_bytes)), module.i64x2.neg(module.v128.const(v128_bytes)), @@ -289,20 +287,16 @@ function test_core() { module.f64x2.sqrt(module.v128.const(v128_bytes)), module.i32x4.trunc_sat_f32x4_s(module.v128.const(v128_bytes)), module.i32x4.trunc_sat_f32x4_u(module.v128.const(v128_bytes)), - module.i64x2.trunc_sat_f64x2_s(module.v128.const(v128_bytes)), - module.i64x2.trunc_sat_f64x2_u(module.v128.const(v128_bytes)), module.f32x4.convert_i32x4_s(module.v128.const(v128_bytes)), module.f32x4.convert_i32x4_u(module.v128.const(v128_bytes)), - module.f64x2.convert_i64x2_s(module.v128.const(v128_bytes)), - module.f64x2.convert_i64x2_u(module.v128.const(v128_bytes)), - module.i16x8.widen_low_i8x16_s(module.v128.const(v128_bytes)), - module.i16x8.widen_high_i8x16_s(module.v128.const(v128_bytes)), - module.i16x8.widen_low_i8x16_u(module.v128.const(v128_bytes)), - module.i16x8.widen_high_i8x16_u(module.v128.const(v128_bytes)), - module.i32x4.widen_low_i16x8_s(module.v128.const(v128_bytes)), - module.i32x4.widen_high_i16x8_s(module.v128.const(v128_bytes)), - module.i32x4.widen_low_i16x8_u(module.v128.const(v128_bytes)), - module.i32x4.widen_high_i16x8_u(module.v128.const(v128_bytes)), + module.i16x8.extend_low_i8x16_s(module.v128.const(v128_bytes)), + module.i16x8.extend_high_i8x16_s(module.v128.const(v128_bytes)), + module.i16x8.extend_low_i8x16_u(module.v128.const(v128_bytes)), + module.i16x8.extend_high_i8x16_u(module.v128.const(v128_bytes)), + module.i32x4.extend_low_i16x8_s(module.v128.const(v128_bytes)), + module.i32x4.extend_high_i16x8_s(module.v128.const(v128_bytes)), + module.i32x4.extend_low_i16x8_u(module.v128.const(v128_bytes)), + module.i32x4.extend_high_i16x8_u(module.v128.const(v128_bytes)), // Binary module.i32.add(module.i32.const(-10), module.i32.const(-11)), module.f64.sub(module.f64.const(-9005.841), module.f64.const(-9007.333)), @@ -388,7 +382,6 @@ function test_core() { module.i8x16.sub(module.v128.const(v128_bytes), module.v128.const(v128_bytes)), module.i8x16.sub_saturate_s(module.v128.const(v128_bytes), module.v128.const(v128_bytes)), module.i8x16.sub_saturate_u(module.v128.const(v128_bytes), module.v128.const(v128_bytes)), - module.i8x16.mul(module.v128.const(v128_bytes), module.v128.const(v128_bytes)), module.i8x16.min_s(module.v128.const(v128_bytes), module.v128.const(v128_bytes)), module.i8x16.min_u(module.v128.const(v128_bytes), module.v128.const(v128_bytes)), module.i8x16.max_s(module.v128.const(v128_bytes), module.v128.const(v128_bytes)), @@ -445,7 +438,7 @@ function test_core() { module.i8x16.narrow_i16x8_u(module.v128.const(v128_bytes), module.v128.const(v128_bytes)), module.i16x8.narrow_i32x4_s(module.v128.const(v128_bytes), module.v128.const(v128_bytes)), module.i16x8.narrow_i32x4_u(module.v128.const(v128_bytes), module.v128.const(v128_bytes)), - module.v8x16.swizzle(module.v128.const(v128_bytes), module.v128.const(v128_bytes)), + module.i8x16.swizzle(module.v128.const(v128_bytes), module.v128.const(v128_bytes)), // SIMD lane manipulation module.i8x16.extract_lane_s(module.v128.const(v128_bytes), 1), module.i8x16.extract_lane_u(module.v128.const(v128_bytes), 1), @@ -475,23 +468,19 @@ function test_core() { module.i64x2.shr_s(module.v128.const(v128_bytes), module.i32.const(1)), module.i64x2.shr_u(module.v128.const(v128_bytes), module.i32.const(1)), // SIMD load - module.v8x16.load_splat(0, 1, module.i32.const(128)), - module.v16x8.load_splat(16, 1, module.i32.const(128)), - module.v32x4.load_splat(16, 4, module.i32.const(128)), - module.v64x2.load_splat(0, 4, module.i32.const(128)), - module.i16x8.load8x8_s(0, 8, module.i32.const(128)), - module.i16x8.load8x8_u(0, 8, module.i32.const(128)), - module.i32x4.load16x4_s(0, 8, module.i32.const(128)), - module.i32x4.load16x4_u(0, 8, module.i32.const(128)), - module.i64x2.load32x2_s(0, 8, module.i32.const(128)), - module.i64x2.load32x2_u(0, 8, module.i32.const(128)), + module.v128.load8_splat(0, 1, module.i32.const(128)), + module.v128.load16_splat(16, 1, module.i32.const(128)), + module.v128.load32_splat(16, 4, module.i32.const(128)), + module.v128.load64_splat(0, 4, module.i32.const(128)), + module.v128.load8x8_s(0, 8, module.i32.const(128)), + module.v128.load8x8_u(0, 8, module.i32.const(128)), + module.v128.load16x4_s(0, 8, module.i32.const(128)), + module.v128.load16x4_u(0, 8, module.i32.const(128)), + module.v128.load32x2_s(0, 8, module.i32.const(128)), + module.v128.load32x2_u(0, 8, module.i32.const(128)), // Other SIMD - module.v8x16.shuffle(module.v128.const(v128_bytes), module.v128.const(v128_bytes), v128_bytes), + module.i8x16.shuffle(module.v128.const(v128_bytes), module.v128.const(v128_bytes), v128_bytes), module.v128.bitselect(module.v128.const(v128_bytes), module.v128.const(v128_bytes), module.v128.const(v128_bytes)), - module.f32x4.qfma(module.v128.const(v128_bytes), module.v128.const(v128_bytes), module.v128.const(v128_bytes)), - module.f32x4.qfms(module.v128.const(v128_bytes), module.v128.const(v128_bytes), module.v128.const(v128_bytes)), - module.f64x2.qfma(module.v128.const(v128_bytes), module.v128.const(v128_bytes), module.v128.const(v128_bytes)), - module.f64x2.qfms(module.v128.const(v128_bytes), module.v128.const(v128_bytes), module.v128.const(v128_bytes)), // Bulk memory module.memory.init(0, makeInt32(1024), makeInt32(0), makeInt32(12)), module.data.drop(0), @@ -676,10 +665,10 @@ function test_core() { assert(table.base === ""); assert(table.initial === 0); assert(table.max === 2); - + module.removeTable("t1"); assert(module.getNumTables() === 0); - + module.addTable("t0", 1, 0xffffffff); module.addActiveElementSegment("t0", "e0", [ binaryen.getFunctionInfo(sinker).name ]); assert(module.getNumTables() === 1); diff --git a/test/binaryen.js/kitchen-sink.js.txt b/test/binaryen.js/kitchen-sink.js.txt index a8ea4eccf..9d97b6523 100644 --- a/test/binaryen.js/kitchen-sink.js.txt +++ b/test/binaryen.js/kitchen-sink.js.txt @@ -66,46 +66,46 @@ ReturnId: 19 MemorySizeId: 20 MemoryGrowId: 21 NopId: 22 -UnreachableId: 24 -AtomicCmpxchgId: 26 -AtomicRMWId: 25 -AtomicWaitId: 27 -AtomicNotifyId: 28 -SIMDExtractId: 30 -SIMDReplaceId: 31 -SIMDShuffleId: 32 -SIMDTernaryId: 33 -SIMDShiftId: 34 -SIMDLoadId: 35 -MemoryInitId: 38 -DataDropId: 39 -MemoryCopyId: 40 -MemoryFillId: 41 -PopId: 42 -RefNullId: 43 -RefIsId: 44 -RefFuncId: 45 -RefEqId: 46 -TryId: 47 -ThrowId: 48 -RethrowId: 49 -TupleMakeId: 50 -TupleExtractId: 51 -I31NewId: 52 -I31GetId: 53 -CallRefId: 54 -RefTestId: 55 -RefCastId: 56 -BrOnId: 57 -RttCanonId: 58 -RttSubId: 59 -StructNewId: 60 -StructGetId: 61 -StructSetId: 62 -ArrayNewId: 63 -ArrayGetId: 64 -ArraySetId: 65 -ArrayLenId: 66 +UnreachableId: 23 +AtomicCmpxchgId: 25 +AtomicRMWId: 24 +AtomicWaitId: 26 +AtomicNotifyId: 27 +SIMDExtractId: 29 +SIMDReplaceId: 30 +SIMDShuffleId: 31 +SIMDTernaryId: 32 +SIMDShiftId: 33 +SIMDLoadId: 34 +MemoryInitId: 36 +DataDropId: 37 +MemoryCopyId: 38 +MemoryFillId: 39 +PopId: 40 +RefNullId: 41 +RefIsId: 42 +RefFuncId: 43 +RefEqId: 44 +TryId: 45 +ThrowId: 46 +RethrowId: 47 +TupleMakeId: 48 +TupleExtractId: 49 +I31NewId: 50 +I31GetId: 51 +CallRefId: 52 +RefTestId: 53 +RefCastId: 54 +BrOnId: 55 +RttCanonId: 56 +RttSubId: 57 +StructNewId: 58 +StructGetId: 59 +StructSetId: 60 +ArrayNewId: 61 +ArrayGetId: 62 +ArraySetId: 63 +ArrayLenId: 64 getExpressionInfo={"id":15,"type":4,"op":6} (f32.neg (f32.const -33.61199951171875) @@ -402,17 +402,17 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7} ) ) (drop - (i8x16.abs + (v128.any_true (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i8x16.neg + (i8x16.abs (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i8x16.any_true + (i8x16.neg (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) @@ -437,11 +437,6 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7} ) ) (drop - (i16x8.any_true - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop (i16x8.all_true (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) @@ -462,11 +457,6 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7} ) ) (drop - (i32x4.any_true - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop (i32x4.all_true (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) @@ -522,16 +512,6 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7} ) ) (drop - (i64x2.trunc_sat_f64x2_s - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop - (i64x2.trunc_sat_f64x2_u - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop (f32x4.convert_i32x4_s (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) @@ -542,52 +522,42 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7} ) ) (drop - (f64x2.convert_i64x2_s - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop - (f64x2.convert_i64x2_u + (i16x8.extend_low_i8x16_s (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i16x8.widen_low_i8x16_s + (i16x8.extend_high_i8x16_s (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i16x8.widen_high_i8x16_s + (i16x8.extend_low_i8x16_u (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i16x8.widen_low_i8x16_u + (i16x8.extend_high_i8x16_u (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i16x8.widen_high_i8x16_u + (i32x4.extend_low_i16x8_s (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i32x4.widen_low_i16x8_s + (i32x4.extend_high_i16x8_s (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i32x4.widen_high_i16x8_s + (i32x4.extend_low_i16x8_u (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i32x4.widen_low_i16x8_u - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop - (i32x4.widen_high_i16x8_u + (i32x4.extend_high_i16x8_u (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) @@ -1066,13 +1036,13 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7} ) ) (drop - (i8x16.add_saturate_s + (i8x16.add_sat_s (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i8x16.add_saturate_u + (i8x16.add_sat_u (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) @@ -1084,19 +1054,13 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7} ) ) (drop - (i8x16.sub_saturate_s + (i8x16.sub_sat_s (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i8x16.sub_saturate_u - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop - (i8x16.mul + (i8x16.sub_sat_u (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) @@ -1138,13 +1102,13 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7} ) ) (drop - (i16x8.add_saturate_s + (i16x8.add_sat_s (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i16x8.add_saturate_u + (i16x8.add_sat_u (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) @@ -1156,13 +1120,13 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7} ) ) (drop - (i16x8.sub_saturate_s + (i16x8.sub_sat_s (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i16x8.sub_saturate_u + (i16x8.sub_sat_u (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) @@ -1430,7 +1394,7 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7} ) ) (drop - (v8x16.swizzle + (i8x16.swizzle (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) @@ -1584,57 +1548,57 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7} ) ) (drop - (v8x16.load_splat + (v128.load8_splat (i32.const 128) ) ) (drop - (v16x8.load_splat offset=16 align=1 + (v128.load16_splat offset=16 align=1 (i32.const 128) ) ) (drop - (v32x4.load_splat offset=16 + (v128.load32_splat offset=16 (i32.const 128) ) ) (drop - (v64x2.load_splat align=4 + (v128.load64_splat align=4 (i32.const 128) ) ) (drop - (i16x8.load8x8_s + (v128.load8x8_s (i32.const 128) ) ) (drop - (i16x8.load8x8_u + (v128.load8x8_u (i32.const 128) ) ) (drop - (i32x4.load16x4_s + (v128.load16x4_s (i32.const 128) ) ) (drop - (i32x4.load16x4_u + (v128.load16x4_u (i32.const 128) ) ) (drop - (i64x2.load32x2_s + (v128.load32x2_s (i32.const 128) ) ) (drop - (i64x2.load32x2_u + (v128.load32x2_u (i32.const 128) ) ) (drop - (v8x16.shuffle 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 + (i8x16.shuffle 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) @@ -1646,34 +1610,6 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7} (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) - (drop - (f32x4.qfma - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop - (f32x4.qfms - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop - (f64x2.qfma - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop - (f64x2.qfms - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) (memory.init 0 (i32.const 1024) (i32.const 0) @@ -2264,17 +2200,17 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7} ) ) (drop - (i8x16.abs + (v128.any_true (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i8x16.neg + (i8x16.abs (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i8x16.any_true + (i8x16.neg (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) @@ -2299,11 +2235,6 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7} ) ) (drop - (i16x8.any_true - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop (i16x8.all_true (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) @@ -2324,11 +2255,6 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7} ) ) (drop - (i32x4.any_true - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop (i32x4.all_true (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) @@ -2384,16 +2310,6 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7} ) ) (drop - (i64x2.trunc_sat_f64x2_s - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop - (i64x2.trunc_sat_f64x2_u - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop (f32x4.convert_i32x4_s (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) @@ -2404,52 +2320,42 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7} ) ) (drop - (f64x2.convert_i64x2_s - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop - (f64x2.convert_i64x2_u + (i16x8.extend_low_i8x16_s (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i16x8.widen_low_i8x16_s + (i16x8.extend_high_i8x16_s (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i16x8.widen_high_i8x16_s + (i16x8.extend_low_i8x16_u (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i16x8.widen_low_i8x16_u + (i16x8.extend_high_i8x16_u (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i16x8.widen_high_i8x16_u + (i32x4.extend_low_i16x8_s (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i32x4.widen_low_i16x8_s + (i32x4.extend_high_i16x8_s (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i32x4.widen_high_i16x8_s + (i32x4.extend_low_i16x8_u (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i32x4.widen_low_i16x8_u - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop - (i32x4.widen_high_i16x8_u + (i32x4.extend_high_i16x8_u (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) @@ -2928,13 +2834,13 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7} ) ) (drop - (i8x16.add_saturate_s + (i8x16.add_sat_s (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i8x16.add_saturate_u + (i8x16.add_sat_u (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) @@ -2946,19 +2852,13 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7} ) ) (drop - (i8x16.sub_saturate_s + (i8x16.sub_sat_s (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i8x16.sub_saturate_u - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop - (i8x16.mul + (i8x16.sub_sat_u (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) @@ -3000,13 +2900,13 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7} ) ) (drop - (i16x8.add_saturate_s + (i16x8.add_sat_s (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i16x8.add_saturate_u + (i16x8.add_sat_u (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) @@ -3018,13 +2918,13 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7} ) ) (drop - (i16x8.sub_saturate_s + (i16x8.sub_sat_s (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i16x8.sub_saturate_u + (i16x8.sub_sat_u (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) @@ -3292,7 +3192,7 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7} ) ) (drop - (v8x16.swizzle + (i8x16.swizzle (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) @@ -3446,57 +3346,57 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7} ) ) (drop - (v8x16.load_splat + (v128.load8_splat (i32.const 128) ) ) (drop - (v16x8.load_splat offset=16 align=1 + (v128.load16_splat offset=16 align=1 (i32.const 128) ) ) (drop - (v32x4.load_splat offset=16 + (v128.load32_splat offset=16 (i32.const 128) ) ) (drop - (v64x2.load_splat align=4 + (v128.load64_splat align=4 (i32.const 128) ) ) (drop - (i16x8.load8x8_s + (v128.load8x8_s (i32.const 128) ) ) (drop - (i16x8.load8x8_u + (v128.load8x8_u (i32.const 128) ) ) (drop - (i32x4.load16x4_s + (v128.load16x4_s (i32.const 128) ) ) (drop - (i32x4.load16x4_u + (v128.load16x4_u (i32.const 128) ) ) (drop - (i64x2.load32x2_s + (v128.load32x2_s (i32.const 128) ) ) (drop - (i64x2.load32x2_u + (v128.load32x2_u (i32.const 128) ) ) (drop - (v8x16.shuffle 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 + (i8x16.shuffle 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) @@ -3508,34 +3408,6 @@ getExpressionInfo(tuple[3])={"id":14,"type":5,"value":3.7} (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) - (drop - (f32x4.qfma - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop - (f32x4.qfms - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop - (f64x2.qfma - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop - (f64x2.qfms - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) (memory.init 0 (i32.const 1024) (i32.const 0) diff --git a/test/example/c-api-kitchen-sink.c b/test/example/c-api-kitchen-sink.c index aa82723b4..e4e8c37b0 100644 --- a/test/example/c-api-kitchen-sink.c +++ b/test/example/c-api-kitchen-sink.c @@ -403,19 +403,17 @@ void test_core() { makeUnary(module, BinaryenSplatVecF32x4(), f32), makeUnary(module, BinaryenSplatVecF64x2(), f64), makeUnary(module, BinaryenNotVec128(), v128), + makeUnary(module, BinaryenAnyTrueVec128(), v128), makeUnary(module, BinaryenAbsVecI8x16(), v128), makeUnary(module, BinaryenNegVecI8x16(), v128), - makeUnary(module, BinaryenAnyTrueVecI8x16(), v128), makeUnary(module, BinaryenAllTrueVecI8x16(), v128), makeUnary(module, BinaryenBitmaskVecI8x16(), v128), makeUnary(module, BinaryenAbsVecI16x8(), v128), makeUnary(module, BinaryenNegVecI16x8(), v128), - makeUnary(module, BinaryenAnyTrueVecI16x8(), v128), makeUnary(module, BinaryenAllTrueVecI16x8(), v128), makeUnary(module, BinaryenBitmaskVecI16x8(), v128), makeUnary(module, BinaryenAbsVecI32x4(), v128), makeUnary(module, BinaryenNegVecI32x4(), v128), - makeUnary(module, BinaryenAnyTrueVecI32x4(), v128), makeUnary(module, BinaryenAllTrueVecI32x4(), v128), makeUnary(module, BinaryenBitmaskVecI32x4(), v128), makeUnary(module, BinaryenNegVecI64x2(), v128), @@ -427,20 +425,16 @@ void test_core() { makeUnary(module, BinaryenSqrtVecF64x2(), v128), makeUnary(module, BinaryenTruncSatSVecF32x4ToVecI32x4(), v128), makeUnary(module, BinaryenTruncSatUVecF32x4ToVecI32x4(), v128), - makeUnary(module, BinaryenTruncSatSVecF64x2ToVecI64x2(), v128), - makeUnary(module, BinaryenTruncSatUVecF64x2ToVecI64x2(), v128), makeUnary(module, BinaryenConvertSVecI32x4ToVecF32x4(), v128), makeUnary(module, BinaryenConvertUVecI32x4ToVecF32x4(), v128), - makeUnary(module, BinaryenConvertSVecI64x2ToVecF64x2(), v128), - makeUnary(module, BinaryenConvertUVecI64x2ToVecF64x2(), v128), - makeUnary(module, BinaryenWidenLowSVecI8x16ToVecI16x8(), v128), - makeUnary(module, BinaryenWidenHighSVecI8x16ToVecI16x8(), v128), - makeUnary(module, BinaryenWidenLowUVecI8x16ToVecI16x8(), v128), - makeUnary(module, BinaryenWidenHighUVecI8x16ToVecI16x8(), v128), - makeUnary(module, BinaryenWidenLowSVecI16x8ToVecI32x4(), v128), - makeUnary(module, BinaryenWidenHighSVecI16x8ToVecI32x4(), v128), - makeUnary(module, BinaryenWidenLowUVecI16x8ToVecI32x4(), v128), - makeUnary(module, BinaryenWidenHighUVecI16x8ToVecI32x4(), v128), + makeUnary(module, BinaryenExtendLowSVecI8x16ToVecI16x8(), v128), + makeUnary(module, BinaryenExtendHighSVecI8x16ToVecI16x8(), v128), + makeUnary(module, BinaryenExtendLowUVecI8x16ToVecI16x8(), v128), + makeUnary(module, BinaryenExtendHighUVecI8x16ToVecI16x8(), v128), + makeUnary(module, BinaryenExtendLowSVecI16x8ToVecI32x4(), v128), + makeUnary(module, BinaryenExtendHighSVecI16x8ToVecI32x4(), v128), + makeUnary(module, BinaryenExtendLowUVecI16x8ToVecI32x4(), v128), + makeUnary(module, BinaryenExtendHighUVecI16x8ToVecI32x4(), v128), // Binary makeBinary(module, BinaryenAddInt32(), i32), makeBinary(module, BinaryenSubFloat64(), f64), @@ -526,7 +520,6 @@ void test_core() { makeBinary(module, BinaryenSubVecI8x16(), v128), makeBinary(module, BinaryenSubSatSVecI8x16(), v128), makeBinary(module, BinaryenSubSatUVecI8x16(), v128), - makeBinary(module, BinaryenMulVecI8x16(), v128), makeBinary(module, BinaryenMinSVecI8x16(), v128), makeBinary(module, BinaryenMinUVecI8x16(), v128), makeBinary(module, BinaryenMaxSVecI8x16(), v128), @@ -648,10 +641,6 @@ void test_core() { // Other SIMD makeSIMDShuffle(module), makeSIMDTernary(module, BinaryenBitselectVec128()), - makeSIMDTernary(module, BinaryenQFMAVecF32x4()), - makeSIMDTernary(module, BinaryenQFMSVecF32x4()), - makeSIMDTernary(module, BinaryenQFMAVecF64x2()), - makeSIMDTernary(module, BinaryenQFMSVecF64x2()), // Bulk memory makeMemoryInit(module), makeDataDrop(module), diff --git a/test/example/c-api-kitchen-sink.txt b/test/example/c-api-kitchen-sink.txt index 4cad3e659..e67850376 100644 --- a/test/example/c-api-kitchen-sink.txt +++ b/test/example/c-api-kitchen-sink.txt @@ -312,17 +312,17 @@ BinaryenFeatureAll: 8191 ) ) (drop - (i8x16.abs + (v128.any_true (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i8x16.neg + (i8x16.abs (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i8x16.any_true + (i8x16.neg (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) @@ -347,11 +347,6 @@ BinaryenFeatureAll: 8191 ) ) (drop - (i16x8.any_true - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop (i16x8.all_true (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) @@ -372,11 +367,6 @@ BinaryenFeatureAll: 8191 ) ) (drop - (i32x4.any_true - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop (i32x4.all_true (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) @@ -432,16 +422,6 @@ BinaryenFeatureAll: 8191 ) ) (drop - (i64x2.trunc_sat_f64x2_s - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop - (i64x2.trunc_sat_f64x2_u - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop (f32x4.convert_i32x4_s (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) @@ -452,52 +432,42 @@ BinaryenFeatureAll: 8191 ) ) (drop - (f64x2.convert_i64x2_s - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop - (f64x2.convert_i64x2_u + (i16x8.extend_low_i8x16_s (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i16x8.widen_low_i8x16_s + (i16x8.extend_high_i8x16_s (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i16x8.widen_high_i8x16_s + (i16x8.extend_low_i8x16_u (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i16x8.widen_low_i8x16_u + (i16x8.extend_high_i8x16_u (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i16x8.widen_high_i8x16_u + (i32x4.extend_low_i16x8_s (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i32x4.widen_low_i16x8_s + (i32x4.extend_high_i16x8_s (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i32x4.widen_high_i16x8_s + (i32x4.extend_low_i16x8_u (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i32x4.widen_low_i16x8_u - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop - (i32x4.widen_high_i16x8_u + (i32x4.extend_high_i16x8_u (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) @@ -976,13 +946,13 @@ BinaryenFeatureAll: 8191 ) ) (drop - (i8x16.add_saturate_s + (i8x16.add_sat_s (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i8x16.add_saturate_u + (i8x16.add_sat_u (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) @@ -994,19 +964,13 @@ BinaryenFeatureAll: 8191 ) ) (drop - (i8x16.sub_saturate_s + (i8x16.sub_sat_s (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i8x16.sub_saturate_u - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop - (i8x16.mul + (i8x16.sub_sat_u (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) @@ -1048,13 +1012,13 @@ BinaryenFeatureAll: 8191 ) ) (drop - (i16x8.add_saturate_s + (i16x8.add_sat_s (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i16x8.add_saturate_u + (i16x8.add_sat_u (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) @@ -1066,13 +1030,13 @@ BinaryenFeatureAll: 8191 ) ) (drop - (i16x8.sub_saturate_s + (i16x8.sub_sat_s (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) (drop - (i16x8.sub_saturate_u + (i16x8.sub_sat_u (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) @@ -1340,7 +1304,7 @@ BinaryenFeatureAll: 8191 ) ) (drop - (v8x16.swizzle + (i8x16.swizzle (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) @@ -1494,57 +1458,57 @@ BinaryenFeatureAll: 8191 ) ) (drop - (v8x16.load_splat + (v128.load8_splat (i32.const 128) ) ) (drop - (v16x8.load_splat offset=16 align=1 + (v128.load16_splat offset=16 align=1 (i32.const 128) ) ) (drop - (v32x4.load_splat offset=16 + (v128.load32_splat offset=16 (i32.const 128) ) ) (drop - (v64x2.load_splat align=4 + (v128.load64_splat align=4 (i32.const 128) ) ) (drop - (i16x8.load8x8_s + (v128.load8x8_s (i32.const 128) ) ) (drop - (i16x8.load8x8_u + (v128.load8x8_u (i32.const 128) ) ) (drop - (i32x4.load16x4_s + (v128.load16x4_s (i32.const 128) ) ) (drop - (i32x4.load16x4_u + (v128.load16x4_u (i32.const 128) ) ) (drop - (i64x2.load32x2_s + (v128.load32x2_s (i32.const 128) ) ) (drop - (i64x2.load32x2_u + (v128.load32x2_u (i32.const 128) ) ) (drop - (v8x16.shuffle 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 + (i8x16.shuffle 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) @@ -1556,34 +1520,6 @@ BinaryenFeatureAll: 8191 (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) ) - (drop - (f32x4.qfma - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop - (f32x4.qfms - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop - (f64x2.qfma - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) - (drop - (f64x2.qfms - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) - ) - ) (memory.init 0 (i32.const 1024) (i32.const 0) diff --git a/test/passes/precompute-propagate_all-features.txt b/test/passes/precompute-propagate_all-features.txt index 298b46de9..09f2bc42d 100644 --- a/test/passes/precompute-propagate_all-features.txt +++ b/test/passes/precompute-propagate_all-features.txt @@ -259,7 +259,7 @@ (func $simd-load (result v128) (local $x v128) (local.set $x - (v8x16.load_splat + (v128.load8_splat (i32.const 0) ) ) diff --git a/test/passes/precompute-propagate_all-features.wast b/test/passes/precompute-propagate_all-features.wast index f866c0ea8..bad12eac8 100644 --- a/test/passes/precompute-propagate_all-features.wast +++ b/test/passes/precompute-propagate_all-features.wast @@ -172,7 +172,7 @@ ) (func $simd-load (result v128) (local $x v128) - (local.set $x (v8x16.load_splat (i32.const 0))) + (local.set $x (v128.load8_splat (i32.const 0))) (local.get $x) ) (func $tuple-local (result i32 i64) diff --git a/test/simd.wast b/test/simd.wast index 8d12ea0ac..a89efb384 100644 --- a/test/simd.wast +++ b/test/simd.wast @@ -5,6 +5,56 @@ (local.get $0) ) ) + (func $v128.load8x8_s (param $0 i32) (result v128) + (v128.load8x8_s + (local.get $0) + ) + ) + (func $v128.load8x8_u (param $0 i32) (result v128) + (v128.load8x8_u + (local.get $0) + ) + ) + (func $v128.load16x4_s (param $0 i32) (result v128) + (v128.load16x4_s + (local.get $0) + ) + ) + (func $v128.load16x4_u (param $0 i32) (result v128) + (v128.load16x4_u + (local.get $0) + ) + ) + (func $v128.load32x2_s (param $0 i32) (result v128) + (v128.load32x2_s + (local.get $0) + ) + ) + (func $v128.load32x2_u (param $0 i32) (result v128) + (v128.load32x2_u + (local.get $0) + ) + ) + (func $v128.load8_splat (param $0 i32) (result v128) + (v128.load8_splat + (local.get $0) + ) + ) + (func $v128.load16_splat (param $0 i32) (result v128) + (v128.load16_splat + (local.get $0) + ) + ) + (func $v128.load32_splat (param $0 i32) (result v128) + (v128.load32_splat + (local.get $0) + ) + ) + (func $v128.load64_splat (param $0 i32) (result v128) + (v128.load64_splat + (local.get $0) + ) + ) (func $v128.store (param $0 i32) (param $1 v128) (v128.store offset=0 align=16 (local.get $0) @@ -29,8 +79,14 @@ (func $v128.const.f64x2 (result v128) (v128.const f64x2 1.0 2) ) - (func $v128.shuffle (param $0 v128) (param $1 v128) (result v128) - (v8x16.shuffle 0 17 2 19 4 21 6 23 8 25 10 27 12 29 14 31 + (func $i8x16.shuffle (param $0 v128) (param $1 v128) (result v128) + (i8x16.shuffle 0 17 2 19 4 21 6 23 8 25 10 27 12 29 14 31 + (local.get $0) + (local.get $1) + ) + ) + (func $i8x16.swizzle (param $0 v128) (param $1 v128) (result v128) + (i8x16.swizzle (local.get $0) (local.get $1) ) @@ -40,6 +96,21 @@ (local.get $0) ) ) + (func $i16x8.splat (param $0 i32) (result v128) + (i16x8.splat + (local.get $0) + ) + ) + (func $f32x4.splat (param $0 f32) (result v128) + (f32x4.splat + (local.get $0) + ) + ) + (func $f64x2.splat (param $0 f64) (result v128) + (f64x2.splat + (local.get $0) + ) + ) (func $i8x16.extract_lane_s (param $0 v128) (result i32) (i8x16.extract_lane_s 0 (local.get $0) @@ -56,11 +127,6 @@ (local.get $1) ) ) - (func $i16x8.splat (param $0 i32) (result v128) - (i16x8.splat - (local.get $0) - ) - ) (func $i16x8.extract_lane_s (param $0 v128) (result i32) (i16x8.extract_lane_s 0 (local.get $0) @@ -77,11 +143,6 @@ (local.get $1) ) ) - (func $i32x4.splat (param $0 i32) (result v128) - (i32x4.splat - (local.get $0) - ) - ) (func $i32x4.extract_lane (param $0 v128) (result i32) (i32x4.extract_lane 0 (local.get $0) @@ -104,11 +165,6 @@ (local.get $1) ) ) - (func $f32x4.splat (param $0 f32) (result v128) - (f32x4.splat - (local.get $0) - ) - ) (func $f32x4.extract_lane (param $0 v128) (result f32) (f32x4.extract_lane 0 (local.get $0) @@ -120,11 +176,6 @@ (local.get $1) ) ) - (func $f64x2.splat (param $0 f64) (result v128) - (f64x2.splat - (local.get $0) - ) - ) (func $f64x2.extract_lane (param $0 v128) (result f64) (f64x2.extract_lane 0 (local.get $0) @@ -316,12 +367,6 @@ (local.get $1) ) ) - (func $i64x2.eq (param $0 v128) (param $1 v128) (result v128) - (i64x2.eq - (local.get $0) - (local.get $1) - ) - ) (func $f32x4.eq (param $0 v128) (param $1 v128) (result v128) (f32x4.eq (local.get $0) @@ -405,57 +450,34 @@ (local.get $1) ) ) - (func $v128.or (param $0 v128) (param $1 v128) (result v128) - (v128.or - (local.get $0) - (local.get $1) - ) - ) - (func $v128.xor (param $0 v128) (param $1 v128) (result v128) - (v128.xor - (local.get $0) - (local.get $1) - ) - ) (func $v128.andnot (param $0 v128) (param $1 v128) (result v128) (v128.andnot (local.get $0) (local.get $1) ) ) - (func $v128.bitselect (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (v128.bitselect + (func $v128.or (param $0 v128) (param $1 v128) (result v128) + (v128.or (local.get $0) (local.get $1) - (local.get $2) ) ) - (func $v8x16.signselect (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (v8x16.signselect - (local.get $0) - (local.get $1) - (local.get $2) - ) - ) - (func $v16x8.signselect (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (v16x8.signselect + (func $v128.xor (param $0 v128) (param $1 v128) (result v128) + (v128.xor (local.get $0) (local.get $1) - (local.get $2) ) ) - (func $v32x4.signselect (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (v32x4.signselect + (func $v128.bitselect (param $0 v128) (param $1 v128) (param $2 v128) (result v128) + (v128.bitselect (local.get $0) (local.get $1) (local.get $2) ) ) - (func $v64x2.signselect (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (v64x2.signselect + (func $v128.any_true (param $0 v128) (result i32) + (v128.any_true (local.get $0) - (local.get $1) - (local.get $2) ) ) (func $v128.load8_lane (param $0 i32) (param $1 v128) (result v128) @@ -542,8 +564,23 @@ (local.get $1) ) ) - (func $i8x16.popcnt (param $0 v128) (result v128) - (i8x16.popcnt + (func $v128.load32_zero (param $0 i32) (result v128) + (v128.load32_zero + (local.get $0) + ) + ) + (func $v128.load64_zero (param $0 i32) (result v128) + (v128.load64_zero + (local.get $0) + ) + ) + (func $f32x4.demote_f64x2_zero (param $0 v128) (result v128) + (f32x4.demote_f64x2_zero + (local.get $0) + ) + ) + (func $f64x2.promote_low_f32x4 (param $0 v128) (result v128) + (f64x2.promote_low_f32x4 (local.get $0) ) ) @@ -557,8 +594,8 @@ (local.get $0) ) ) - (func $i8x16.any_true (param $0 v128) (result i32) - (i8x16.any_true + (func $i8x16.popcnt (param $0 v128) (result v128) + (i8x16.popcnt (local.get $0) ) ) @@ -572,6 +609,38 @@ (local.get $0) ) ) + (func $i8x16.narrow_i16x8_s (param $0 v128) (param $1 v128) (result v128) + (i8x16.narrow_i16x8_s + (local.get $0) + (local.get $1) + ) + ) + (func $i8x16.narrow_i16x8_u (param $0 v128) (param $1 v128) (result v128) + (i8x16.narrow_i16x8_u + (local.get $0) + (local.get $1) + ) + ) + (func $f32x4.ceil (param $0 v128) (result v128) + (f32x4.ceil + (local.get $0) + ) + ) + (func $f32x4.floor (param $0 v128) (result v128) + (f32x4.floor + (local.get $0) + ) + ) + (func $f32x4.trunc (param $0 v128) (result v128) + (f32x4.trunc + (local.get $0) + ) + ) + (func $f32x4.nearest (param $0 v128) (result v128) + (f32x4.nearest + (local.get $0) + ) + ) (func $i8x16.shl (param $0 v128) (param $1 i32) (result v128) (i8x16.shl (local.get $0) @@ -596,14 +665,14 @@ (local.get $1) ) ) - (func $i8x16.add_saturate_s (param $0 v128) (param $1 v128) (result v128) - (i8x16.add_saturate_s + (func $i8x16.add_sat_s (param $0 v128) (param $1 v128) (result v128) + (i8x16.add_sat_s (local.get $0) (local.get $1) ) ) - (func $i8x16.add_saturate_u (param $0 v128) (param $1 v128) (result v128) - (i8x16.add_saturate_u + (func $i8x16.add_sat_u (param $0 v128) (param $1 v128) (result v128) + (i8x16.add_sat_u (local.get $0) (local.get $1) ) @@ -614,22 +683,26 @@ (local.get $1) ) ) - (func $i8x16.sub_saturate_s (param $0 v128) (param $1 v128) (result v128) - (i8x16.sub_saturate_s + (func $i8x16.sub_sat_s (param $0 v128) (param $1 v128) (result v128) + (i8x16.sub_sat_s (local.get $0) (local.get $1) ) ) - (func $i8x16.sub_saturate_u (param $0 v128) (param $1 v128) (result v128) - (i8x16.sub_saturate_u + (func $i8x16.sub_sat_u (param $0 v128) (param $1 v128) (result v128) + (i8x16.sub_sat_u (local.get $0) (local.get $1) ) ) - (func $i8x16.mul (param $0 v128) (param $1 v128) (result v128) - (i8x16.mul + (func $f64x2.ceil (param $0 v128) (result v128) + (f64x2.ceil + (local.get $0) + ) + ) + (func $f64x2.floor (param $0 v128) (result v128) + (f64x2.floor (local.get $0) - (local.get $1) ) ) (func $i8x16.min_s (param $0 v128) (param $1 v128) (result v128) @@ -656,12 +729,37 @@ (local.get $1) ) ) + (func $f64x2.trunc (param $0 v128) (result v128) + (f64x2.trunc + (local.get $0) + ) + ) (func $i8x16.avgr_u (param $0 v128) (param $1 v128) (result v128) (i8x16.avgr_u (local.get $0) (local.get $1) ) ) + (func $i16x8.extadd_pairwise_i8x16_s (param $0 v128) (result v128) + (i16x8.extadd_pairwise_i8x16_s + (local.get $0) + ) + ) + (func $i16x8.extadd_pairwise_i8x16_u (param $0 v128) (result v128) + (i16x8.extadd_pairwise_i8x16_u + (local.get $0) + ) + ) + (func $i32x4.extadd_pairwise_i16x8_s (param $0 v128) (result v128) + (i32x4.extadd_pairwise_i16x8_s + (local.get $0) + ) + ) + (func $i32x4.extadd_pairwise_i16x8_u (param $0 v128) (result v128) + (i32x4.extadd_pairwise_i16x8_u + (local.get $0) + ) + ) (func $i16x8.abs (param $0 v128) (result v128) (i16x8.abs (local.get $0) @@ -672,9 +770,10 @@ (local.get $0) ) ) - (func $i16x8.any_true (param $0 v128) (result i32) - (i16x8.any_true + (func $i16x8.q15mulr_sat_s (param $0 v128) (param $1 v128) (result v128) + (i16x8.q15mulr_sat_s (local.get $0) + (local.get $1) ) ) (func $i16x8.all_true (param $0 v128) (result i32) @@ -687,7 +786,39 @@ (local.get $0) ) ) - (func $i16x8.shl (param $0 v128) (param $1 i32) (result v128) + (func $i16x8.narrow_i32x4_s (param $0 v128) (param $1 v128) (result v128) + (i16x8.narrow_i32x4_s + (local.get $0) + (local.get $1) + ) + ) + (func $i16x8.narrow_i32x4_u (param $0 v128) (param $1 v128) (result v128) + (i16x8.narrow_i32x4_u + (local.get $0) + (local.get $1) + ) + ) + (func $i16x8.extend_low_i8x16_s (param $0 v128) (result v128) + (i16x8.extend_low_i8x16_s + (local.get $0) + ) + ) + (func $i16x8.extend_high_i8x16_s (param $0 v128) (result v128) + (i16x8.extend_high_i8x16_s + (local.get $0) + ) + ) + (func $i16x8.extend_low_i8x16_u (param $0 v128) (result v128) + (i16x8.extend_low_i8x16_u + (local.get $0) + ) + ) + (func $i16x8.extend_high_i8x16_u (param $0 v128) (result v128) + (i16x8.extend_high_i8x16_u + (local.get $0) + ) + ) +(func $i16x8.shl (param $0 v128) (param $1 i32) (result v128) (i16x8.shl (local.get $0) (local.get $1) @@ -711,14 +842,14 @@ (local.get $1) ) ) - (func $i16x8.add_saturate_s (param $0 v128) (param $1 v128) (result v128) - (i16x8.add_saturate_s + (func $i16x8.add_sat_s (param $0 v128) (param $1 v128) (result v128) + (i16x8.add_sat_s (local.get $0) (local.get $1) ) ) - (func $i16x8.add_saturate_u (param $0 v128) (param $1 v128) (result v128) - (i16x8.add_saturate_u + (func $i16x8.add_sat_u (param $0 v128) (param $1 v128) (result v128) + (i16x8.add_sat_u (local.get $0) (local.get $1) ) @@ -729,18 +860,23 @@ (local.get $1) ) ) - (func $i16x8.sub_saturate_s (param $0 v128) (param $1 v128) (result v128) - (i16x8.sub_saturate_s + (func $i16x8.sub_sat_s (param $0 v128) (param $1 v128) (result v128) + (i16x8.sub_sat_s (local.get $0) (local.get $1) ) ) - (func $i16x8.sub_saturate_u (param $0 v128) (param $1 v128) (result v128) - (i16x8.sub_saturate_u + (func $i16x8.sub_sat_u (param $0 v128) (param $1 v128) (result v128) + (i16x8.sub_sat_u (local.get $0) (local.get $1) ) ) + (func $f64x2.nearest (param $0 v128) (result v128) + (f64x2.nearest + (local.get $0) + ) + ) (func $i16x8.mul (param $0 v128) (param $1 v128) (result v128) (i16x8.mul (local.get $0) @@ -777,12 +913,6 @@ (local.get $1) ) ) - (func $i16x8.q15mulr_sat_s (param $0 v128) (param $1 v128) (result v128) - (i16x8.q15mulr_sat_s - (local.get $0) - (local.get $1) - ) - ) (func $i16x8.extmul_low_i8x16_s (param $0 v128) (param $1 v128) (result v128) (i16x8.extmul_low_i8x16_s (local.get $0) @@ -817,11 +947,6 @@ (local.get $0) ) ) - (func $i32x4.any_true (param $0 v128) (result i32) - (i32x4.any_true - (local.get $0) - ) - ) (func $i32x4.all_true (param $0 v128) (result i32) (i32x4.all_true (local.get $0) @@ -832,6 +957,26 @@ (local.get $0) ) ) + (func $i32x4.extend_low_i16x8_s (param $0 v128) (result v128) + (i32x4.extend_low_i16x8_s + (local.get $0) + ) + ) + (func $i32x4.extend_high_i16x8_s (param $0 v128) (result v128) + (i32x4.extend_high_i16x8_s + (local.get $0) + ) + ) + (func $i32x4.extend_low_i16x8_u (param $0 v128) (result v128) + (i32x4.extend_low_i16x8_u + (local.get $0) + ) + ) + (func $i32x4.extend_high_i16x8_u (param $0 v128) (result v128) + (i32x4.extend_high_i16x8_u + (local.get $0) + ) + ) (func $i32x4.shl (param $0 v128) (param $1 i32) (result v128) (i32x4.shl (local.get $0) @@ -922,16 +1067,46 @@ (local.get $1) ) ) + (func $i64x2.abs (param $0 v128) (result v128) + (i64x2.abs + (local.get $0) + ) + ) (func $i64x2.neg (param $0 v128) (result v128) (i64x2.neg (local.get $0) ) ) + (func $i64x2.all_true (param $0 v128) (result i32) + (i64x2.all_true + (local.get $0) + ) + ) (func $i64x2.bitmask (param $0 v128) (result i32) (i64x2.bitmask (local.get $0) ) ) + (func $i64x2.extend_low_i32x4_s (param $0 v128) (result v128) + (i64x2.extend_low_i32x4_s + (local.get $0) + ) + ) + (func $i64x2.extend_high_i32x4_s (param $0 v128) (result v128) + (i64x2.extend_high_i32x4_s + (local.get $0) + ) + ) + (func $i64x2.extend_low_i32x4_u (param $0 v128) (result v128) + (i64x2.extend_low_i32x4_u + (local.get $0) + ) + ) + (func $i64x2.extend_high_i32x4_u (param $0 v128) (result v128) + (i64x2.extend_high_i32x4_u + (local.get $0) + ) + ) (func $i64x2.shl (param $0 v128) (param $1 i32) (result v128) (i64x2.shl (local.get $0) @@ -968,6 +1143,42 @@ (local.get $1) ) ) + (func $i64x2.eq (param $0 v128) (param $1 v128) (result v128) + (i64x2.eq + (local.get $0) + (local.get $1) + ) + ) + (func $i64x2.ne (param $0 v128) (param $1 v128) (result v128) + (i64x2.ne + (local.get $0) + (local.get $1) + ) + ) + (func $i64x2.lt_s (param $0 v128) (param $1 v128) (result v128) + (i64x2.lt_s + (local.get $0) + (local.get $1) + ) + ) + (func $i64x2.gt_s (param $0 v128) (param $1 v128) (result v128) + (i64x2.gt_s + (local.get $0) + (local.get $1) + ) + ) + (func $i64x2.le_s (param $0 v128) (param $1 v128) (result v128) + (i64x2.le_s + (local.get $0) + (local.get $1) + ) + ) + (func $i64x2.ge_s (param $0 v128) (param $1 v128) (result v128) + (i64x2.ge_s + (local.get $0) + (local.get $1) + ) + ) (func $i64x2.extmul_low_i32x4_s (param $0 v128) (param $1 v128) (result v128) (i64x2.extmul_low_i32x4_s (local.get $0) @@ -992,6 +1203,21 @@ (local.get $1) ) ) + (func $f32x4.abs (param $0 v128) (result v128) + (f32x4.abs + (local.get $0) + ) + ) + (func $f32x4.neg (param $0 v128) (result v128) + (f32x4.neg + (local.get $0) + ) + ) + (func $f32x4.sqrt (param $0 v128) (result v128) + (f32x4.sqrt + (local.get $0) + ) + ) (func $f32x4.add (param $0 v128) (param $1 v128) (result v128) (f32x4.add (local.get $0) @@ -1040,53 +1266,19 @@ (local.get $1) ) ) - (func $f32x4.ceil (param $0 v128) (result v128) - (f32x4.ceil - (local.get $0) - ) - ) - (func $f32x4.floor (param $0 v128) (result v128) - (f32x4.floor - (local.get $0) - ) - ) - (func $f32x4.trunc (param $0 v128) (result v128) - (f32x4.trunc - (local.get $0) - ) - ) - (func $f32x4.nearest (param $0 v128) (result v128) - (f32x4.nearest - (local.get $0) - ) - ) - (func $f32x4.abs (param $0 v128) (result v128) - (f32x4.abs - (local.get $0) - ) - ) - (func $f32x4.neg (param $0 v128) (result v128) - (f32x4.neg - (local.get $0) - ) - ) - (func $f32x4.sqrt (param $0 v128) (result v128) - (f32x4.sqrt + (func $f64x2.abs (param $0 v128) (result v128) + (f64x2.abs (local.get $0) ) ) - (func $f32x4.qfma (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (f32x4.qfma + (func $f64x2.neg (param $0 v128) (result v128) + (f64x2.neg (local.get $0) - (local.get $1) - (local.get $2) ) ) - (func $f32x4.qfms (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (f32x4.qfms + (func $f64x2.sqrt (param $0 v128) (result v128) + (f64x2.sqrt (local.get $0) - (local.get $1) - (local.get $2) ) ) (func $f64x2.add (param $0 v128) (param $1 v128) (result v128) @@ -1137,75 +1329,6 @@ (local.get $1) ) ) - (func $f64x2.ceil (param $0 v128) (result v128) - (f64x2.ceil - (local.get $0) - ) - ) - (func $f64x2.floor (param $0 v128) (result v128) - (f64x2.floor - (local.get $0) - ) - ) - (func $f64x2.trunc (param $0 v128) (result v128) - (f64x2.trunc - (local.get $0) - ) - ) - (func $f64x2.nearest (param $0 v128) (result v128) - (f64x2.nearest - (local.get $0) - ) - ) - (func $f64x2.abs (param $0 v128) (result v128) - (f64x2.abs - (local.get $0) - ) - ) - (func $f64x2.neg (param $0 v128) (result v128) - (f64x2.neg - (local.get $0) - ) - ) - (func $f64x2.sqrt (param $0 v128) (result v128) - (f64x2.sqrt - (local.get $0) - ) - ) - (func $f64x2.qfma (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (f64x2.qfma - (local.get $0) - (local.get $1) - (local.get $2) - ) - ) - (func $f64x2.qfms (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (f64x2.qfms - (local.get $0) - (local.get $1) - (local.get $2) - ) - ) - (func $i16x8.extadd_pairwise_i8x16_s (param $0 v128) (result v128) - (i16x8.extadd_pairwise_i8x16_s - (local.get $0) - ) - ) - (func $i16x8.extadd_pairwise_i8x16_u (param $0 v128) (result v128) - (i16x8.extadd_pairwise_i8x16_u - (local.get $0) - ) - ) - (func $i32x4.extadd_pairwise_i16x8_s (param $0 v128) (result v128) - (i32x4.extadd_pairwise_i16x8_s - (local.get $0) - ) - ) - (func $i32x4.extadd_pairwise_i16x8_u (param $0 v128) (result v128) - (i32x4.extadd_pairwise_i16x8_u - (local.get $0) - ) - ) (func $i32x4.trunc_sat_f32x4_s (param $0 v128) (result v128) (i32x4.trunc_sat_f32x4_s (local.get $0) @@ -1216,16 +1339,6 @@ (local.get $0) ) ) - (func $i64x2.trunc_sat_f64x2_s (param $0 v128) (result v128) - (i64x2.trunc_sat_f64x2_s - (local.get $0) - ) - ) - (func $i64x2.trunc_sat_f64x2_u (param $0 v128) (result v128) - (i64x2.trunc_sat_f64x2_u - (local.get $0) - ) - ) (func $f32x4.convert_i32x4_s (param $0 v128) (result v128) (f32x4.convert_i32x4_s (local.get $0) @@ -1236,173 +1349,13 @@ (local.get $0) ) ) - (func $f64x2.convert_i64x2_s (param $0 v128) (result v128) - (f64x2.convert_i64x2_s - (local.get $0) - ) - ) - (func $f64x2.convert_i64x2_u (param $0 v128) (result v128) - (f64x2.convert_i64x2_u - (local.get $0) - ) - ) - (func $v8x16.load_splat (param $0 i32) (result v128) - (v8x16.load_splat - (local.get $0) - ) - ) - (func $v16x8.load_splat (param $0 i32) (result v128) - (v16x8.load_splat - (local.get $0) - ) - ) - (func $v32x4.load_splat (param $0 i32) (result v128) - (v32x4.load_splat - (local.get $0) - ) - ) - (func $v64x2.load_splat (param $0 i32) (result v128) - (v64x2.load_splat - (local.get $0) - ) - ) - (func $i8x16.narrow_i16x8_s (param $0 v128) (param $1 v128) (result v128) - (i8x16.narrow_i16x8_s - (local.get $0) - (local.get $1) - ) - ) - (func $i8x16.narrow_i16x8_u (param $0 v128) (param $1 v128) (result v128) - (i8x16.narrow_i16x8_u - (local.get $0) - (local.get $1) - ) - ) - (func $i16x8.narrow_i32x4_s (param $0 v128) (param $1 v128) (result v128) - (i16x8.narrow_i32x4_s - (local.get $0) - (local.get $1) - ) - ) - (func $i16x8.narrow_i32x4_u (param $0 v128) (param $1 v128) (result v128) - (i16x8.narrow_i32x4_u - (local.get $0) - (local.get $1) - ) - ) - (func $i16x8.widen_low_i8x16_s (param $0 v128) (result v128) - (i16x8.widen_low_i8x16_s - (local.get $0) - ) - ) - (func $i16x8.widen_high_i8x16_s (param $0 v128) (result v128) - (i16x8.widen_high_i8x16_s - (local.get $0) - ) - ) - (func $i16x8.widen_low_i8x16_u (param $0 v128) (result v128) - (i16x8.widen_low_i8x16_u + (func $i32x4.trunc_sat_f64x2_s_zero (param $0 v128) (result v128) + (i32x4.trunc_sat_f64x2_s_zero (local.get $0) ) ) - (func $i16x8.widen_high_i8x16_u (param $0 v128) (result v128) - (i16x8.widen_high_i8x16_u - (local.get $0) - ) - ) - (func $i32x4.widen_low_i16x8_s (param $0 v128) (result v128) - (i32x4.widen_low_i16x8_s - (local.get $0) - ) - ) - (func $i32x4.widen_high_i16x8_s (param $0 v128) (result v128) - (i32x4.widen_high_i16x8_s - (local.get $0) - ) - ) - (func $i32x4.widen_low_i16x8_u (param $0 v128) (result v128) - (i32x4.widen_low_i16x8_u - (local.get $0) - ) - ) - (func $i32x4.widen_high_i16x8_u (param $0 v128) (result v128) - (i32x4.widen_high_i16x8_u - (local.get $0) - ) - ) - (func $i64x2.widen_low_i32x4_s (param $0 v128) (result v128) - (i64x2.widen_low_i32x4_s - (local.get $0) - ) - ) - (func $i64x2.widen_high_i32x4_s (param $0 v128) (result v128) - (i64x2.widen_high_i32x4_s - (local.get $0) - ) - ) - (func $i64x2.widen_low_i32x4_u (param $0 v128) (result v128) - (i64x2.widen_low_i32x4_u - (local.get $0) - ) - ) - (func $i64x2.widen_high_i32x4_u (param $0 v128) (result v128) - (i64x2.widen_high_i32x4_u - (local.get $0) - ) - ) - (func $i16x8.load8x8_u (param $0 i32) (result v128) - (i16x8.load8x8_u - (local.get $0) - ) - ) - (func $i16x8.load8x8_s (param $0 i32) (result v128) - (i16x8.load8x8_s - (local.get $0) - ) - ) - (func $i32x4.load16x4_s (param $0 i32) (result v128) - (i32x4.load16x4_s - (local.get $0) - ) - ) - (func $i32x4.load16x4_u (param $0 i32) (result v128) - (i32x4.load16x4_u - (local.get $0) - ) - ) - (func $i64x2.load32x2_s (param $0 i32) (result v128) - (i64x2.load32x2_s - (local.get $0) - ) - ) - (func $i64x2.load32x2_u (param $0 i32) (result v128) - (i64x2.load32x2_u - (local.get $0) - ) - ) - (func $v128.load32_zero (param $0 i32) (result v128) - (v128.load32_zero - (local.get $0) - ) - ) - (func $v128.load64_zero (param $0 i32) (result v128) - (v128.load64_zero - (local.get $0) - ) - ) - (func $v8x16.swizzle (param $0 v128) (param $1 v128) (result v128) - (v8x16.swizzle - (local.get $0) - (local.get $1) - ) - ) - (func $prefetch.t (param $0 i32) - (prefetch.t offset=3 align=2 - (local.get $0) - ) - ) - (func $prefetch.nt (param $0 i32) - (prefetch.nt offset=3 align=2 + (func $i32x4.trunc_sat_f64x2_u_zero (param $0 v128) (result v128) + (i32x4.trunc_sat_f64x2_u_zero (local.get $0) ) ) @@ -1416,34 +1369,4 @@ (local.get $0) ) ) - (func $i32x4.trunc_sat_f64x2_zero_s (param $0 v128) (result v128) - (i32x4.trunc_sat_f64x2_zero_s - (local.get $0) - ) - ) - (func $i32x4.trunc_sat_f64x2_zero_u (param $0 v128) (result v128) - (i32x4.trunc_sat_f64x2_zero_u - (local.get $0) - ) - ) - (func $f32x4.demote_f64x2_zero (param $0 v128) (result v128) - (f32x4.demote_f64x2_zero - (local.get $0) - ) - ) - (func $f64x2.promote_low_f32x4 (param $0 v128) (result v128) - (f64x2.promote_low_f32x4 - (local.get $0) - ) - ) - (func $i32x4.widen_i8x16_s (param $0 v128) (result v128) - (i32x4.widen_i8x16_s 0 - (local.get $0) - ) - ) - (func $i32x4.widen_i8x16_u (param $0 v128) (result v128) - (i32x4.widen_i8x16_u 0 - (local.get $0) - ) - ) ) diff --git a/test/simd.wast.from-wast b/test/simd.wast.from-wast index 6fbacea11..7940a00da 100644 --- a/test/simd.wast.from-wast +++ b/test/simd.wast.from-wast @@ -2,13 +2,11 @@ (type $v128_v128_=>_v128 (func (param v128 v128) (result v128))) (type $v128_=>_v128 (func (param v128) (result v128))) (type $i32_=>_v128 (func (param i32) (result v128))) - (type $v128_=>_i32 (func (param v128) (result i32))) (type $v128_i32_=>_v128 (func (param v128 i32) (result v128))) - (type $v128_v128_v128_=>_v128 (func (param v128 v128 v128) (result v128))) + (type $v128_=>_i32 (func (param v128) (result i32))) (type $i32_v128_=>_none (func (param i32 v128))) (type $i32_v128_=>_v128 (func (param i32 v128) (result v128))) (type $none_=>_v128 (func (result v128))) - (type $i32_=>_none (func (param i32))) (type $v128_=>_i64 (func (param v128) (result i64))) (type $v128_=>_f32 (func (param v128) (result f32))) (type $v128_=>_f64 (func (param v128) (result f64))) @@ -17,12 +15,63 @@ (type $v128_i64_=>_v128 (func (param v128 i64) (result v128))) (type $v128_f32_=>_v128 (func (param v128 f32) (result v128))) (type $v128_f64_=>_v128 (func (param v128 f64) (result v128))) + (type $v128_v128_v128_=>_v128 (func (param v128 v128 v128) (result v128))) (memory $0 1 1) (func $v128.load (param $0 i32) (result v128) (v128.load (local.get $0) ) ) + (func $v128.load8x8_s (param $0 i32) (result v128) + (v128.load8x8_s + (local.get $0) + ) + ) + (func $v128.load8x8_u (param $0 i32) (result v128) + (v128.load8x8_u + (local.get $0) + ) + ) + (func $v128.load16x4_s (param $0 i32) (result v128) + (v128.load16x4_s + (local.get $0) + ) + ) + (func $v128.load16x4_u (param $0 i32) (result v128) + (v128.load16x4_u + (local.get $0) + ) + ) + (func $v128.load32x2_s (param $0 i32) (result v128) + (v128.load32x2_s + (local.get $0) + ) + ) + (func $v128.load32x2_u (param $0 i32) (result v128) + (v128.load32x2_u + (local.get $0) + ) + ) + (func $v128.load8_splat (param $0 i32) (result v128) + (v128.load8_splat + (local.get $0) + ) + ) + (func $v128.load16_splat (param $0 i32) (result v128) + (v128.load16_splat + (local.get $0) + ) + ) + (func $v128.load32_splat (param $0 i32) (result v128) + (v128.load32_splat + (local.get $0) + ) + ) + (func $v128.load64_splat (param $0 i32) (result v128) + (v128.load64_splat + (local.get $0) + ) + ) (func $v128.store (param $0 i32) (param $1 v128) (v128.store (local.get $0) @@ -47,8 +96,14 @@ (func $v128.const.f64x2 (result v128) (v128.const i32x4 0x00000000 0x3ff00000 0x00000000 0x40000000) ) - (func $v128.shuffle (param $0 v128) (param $1 v128) (result v128) - (v8x16.shuffle 0 17 2 19 4 21 6 23 8 25 10 27 12 29 14 31 + (func $i8x16.shuffle (param $0 v128) (param $1 v128) (result v128) + (i8x16.shuffle 0 17 2 19 4 21 6 23 8 25 10 27 12 29 14 31 + (local.get $0) + (local.get $1) + ) + ) + (func $i8x16.swizzle (param $0 v128) (param $1 v128) (result v128) + (i8x16.swizzle (local.get $0) (local.get $1) ) @@ -58,6 +113,21 @@ (local.get $0) ) ) + (func $i16x8.splat (param $0 i32) (result v128) + (i16x8.splat + (local.get $0) + ) + ) + (func $f32x4.splat (param $0 f32) (result v128) + (f32x4.splat + (local.get $0) + ) + ) + (func $f64x2.splat (param $0 f64) (result v128) + (f64x2.splat + (local.get $0) + ) + ) (func $i8x16.extract_lane_s (param $0 v128) (result i32) (i8x16.extract_lane_s 0 (local.get $0) @@ -74,11 +144,6 @@ (local.get $1) ) ) - (func $i16x8.splat (param $0 i32) (result v128) - (i16x8.splat - (local.get $0) - ) - ) (func $i16x8.extract_lane_s (param $0 v128) (result i32) (i16x8.extract_lane_s 0 (local.get $0) @@ -95,11 +160,6 @@ (local.get $1) ) ) - (func $i32x4.splat (param $0 i32) (result v128) - (i32x4.splat - (local.get $0) - ) - ) (func $i32x4.extract_lane (param $0 v128) (result i32) (i32x4.extract_lane 0 (local.get $0) @@ -122,11 +182,6 @@ (local.get $1) ) ) - (func $f32x4.splat (param $0 f32) (result v128) - (f32x4.splat - (local.get $0) - ) - ) (func $f32x4.extract_lane (param $0 v128) (result f32) (f32x4.extract_lane 0 (local.get $0) @@ -138,11 +193,6 @@ (local.get $1) ) ) - (func $f64x2.splat (param $0 f64) (result v128) - (f64x2.splat - (local.get $0) - ) - ) (func $f64x2.extract_lane (param $0 v128) (result f64) (f64x2.extract_lane 0 (local.get $0) @@ -334,12 +384,6 @@ (local.get $1) ) ) - (func $i64x2.eq (param $0 v128) (param $1 v128) (result v128) - (i64x2.eq - (local.get $0) - (local.get $1) - ) - ) (func $f32x4.eq (param $0 v128) (param $1 v128) (result v128) (f32x4.eq (local.get $0) @@ -423,57 +467,34 @@ (local.get $1) ) ) - (func $v128.or (param $0 v128) (param $1 v128) (result v128) - (v128.or - (local.get $0) - (local.get $1) - ) - ) - (func $v128.xor (param $0 v128) (param $1 v128) (result v128) - (v128.xor - (local.get $0) - (local.get $1) - ) - ) (func $v128.andnot (param $0 v128) (param $1 v128) (result v128) (v128.andnot (local.get $0) (local.get $1) ) ) - (func $v128.bitselect (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (v128.bitselect + (func $v128.or (param $0 v128) (param $1 v128) (result v128) + (v128.or (local.get $0) (local.get $1) - (local.get $2) ) ) - (func $v8x16.signselect (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (v8x16.signselect + (func $v128.xor (param $0 v128) (param $1 v128) (result v128) + (v128.xor (local.get $0) (local.get $1) - (local.get $2) ) ) - (func $v16x8.signselect (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (v16x8.signselect + (func $v128.bitselect (param $0 v128) (param $1 v128) (param $2 v128) (result v128) + (v128.bitselect (local.get $0) (local.get $1) (local.get $2) ) ) - (func $v32x4.signselect (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (v32x4.signselect + (func $v128.any_true (param $0 v128) (result i32) + (v128.any_true (local.get $0) - (local.get $1) - (local.get $2) - ) - ) - (func $v64x2.signselect (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (v64x2.signselect - (local.get $0) - (local.get $1) - (local.get $2) ) ) (func $v128.load8_lane (param $0 i32) (param $1 v128) (result v128) @@ -560,8 +581,23 @@ (local.get $1) ) ) - (func $i8x16.popcnt (param $0 v128) (result v128) - (i8x16.popcnt + (func $v128.load32_zero (param $0 i32) (result v128) + (v128.load32_zero + (local.get $0) + ) + ) + (func $v128.load64_zero (param $0 i32) (result v128) + (v128.load64_zero + (local.get $0) + ) + ) + (func $f32x4.demote_f64x2_zero (param $0 v128) (result v128) + (f32x4.demote_f64x2_zero + (local.get $0) + ) + ) + (func $f64x2.promote_low_f32x4 (param $0 v128) (result v128) + (f64x2.promote_low_f32x4 (local.get $0) ) ) @@ -575,8 +611,8 @@ (local.get $0) ) ) - (func $i8x16.any_true (param $0 v128) (result i32) - (i8x16.any_true + (func $i8x16.popcnt (param $0 v128) (result v128) + (i8x16.popcnt (local.get $0) ) ) @@ -590,6 +626,38 @@ (local.get $0) ) ) + (func $i8x16.narrow_i16x8_s (param $0 v128) (param $1 v128) (result v128) + (i8x16.narrow_i16x8_s + (local.get $0) + (local.get $1) + ) + ) + (func $i8x16.narrow_i16x8_u (param $0 v128) (param $1 v128) (result v128) + (i8x16.narrow_i16x8_u + (local.get $0) + (local.get $1) + ) + ) + (func $f32x4.ceil (param $0 v128) (result v128) + (f32x4.ceil + (local.get $0) + ) + ) + (func $f32x4.floor (param $0 v128) (result v128) + (f32x4.floor + (local.get $0) + ) + ) + (func $f32x4.trunc (param $0 v128) (result v128) + (f32x4.trunc + (local.get $0) + ) + ) + (func $f32x4.nearest (param $0 v128) (result v128) + (f32x4.nearest + (local.get $0) + ) + ) (func $i8x16.shl (param $0 v128) (param $1 i32) (result v128) (i8x16.shl (local.get $0) @@ -614,14 +682,14 @@ (local.get $1) ) ) - (func $i8x16.add_saturate_s (param $0 v128) (param $1 v128) (result v128) - (i8x16.add_saturate_s + (func $i8x16.add_sat_s (param $0 v128) (param $1 v128) (result v128) + (i8x16.add_sat_s (local.get $0) (local.get $1) ) ) - (func $i8x16.add_saturate_u (param $0 v128) (param $1 v128) (result v128) - (i8x16.add_saturate_u + (func $i8x16.add_sat_u (param $0 v128) (param $1 v128) (result v128) + (i8x16.add_sat_u (local.get $0) (local.get $1) ) @@ -632,22 +700,26 @@ (local.get $1) ) ) - (func $i8x16.sub_saturate_s (param $0 v128) (param $1 v128) (result v128) - (i8x16.sub_saturate_s + (func $i8x16.sub_sat_s (param $0 v128) (param $1 v128) (result v128) + (i8x16.sub_sat_s (local.get $0) (local.get $1) ) ) - (func $i8x16.sub_saturate_u (param $0 v128) (param $1 v128) (result v128) - (i8x16.sub_saturate_u + (func $i8x16.sub_sat_u (param $0 v128) (param $1 v128) (result v128) + (i8x16.sub_sat_u (local.get $0) (local.get $1) ) ) - (func $i8x16.mul (param $0 v128) (param $1 v128) (result v128) - (i8x16.mul + (func $f64x2.ceil (param $0 v128) (result v128) + (f64x2.ceil + (local.get $0) + ) + ) + (func $f64x2.floor (param $0 v128) (result v128) + (f64x2.floor (local.get $0) - (local.get $1) ) ) (func $i8x16.min_s (param $0 v128) (param $1 v128) (result v128) @@ -674,12 +746,37 @@ (local.get $1) ) ) + (func $f64x2.trunc (param $0 v128) (result v128) + (f64x2.trunc + (local.get $0) + ) + ) (func $i8x16.avgr_u (param $0 v128) (param $1 v128) (result v128) (i8x16.avgr_u (local.get $0) (local.get $1) ) ) + (func $i16x8.extadd_pairwise_i8x16_s (param $0 v128) (result v128) + (i16x8.extadd_pairwise_i8x16_s + (local.get $0) + ) + ) + (func $i16x8.extadd_pairwise_i8x16_u (param $0 v128) (result v128) + (i16x8.extadd_pairwise_i8x16_u + (local.get $0) + ) + ) + (func $i32x4.extadd_pairwise_i16x8_s (param $0 v128) (result v128) + (i32x4.extadd_pairwise_i16x8_s + (local.get $0) + ) + ) + (func $i32x4.extadd_pairwise_i16x8_u (param $0 v128) (result v128) + (i32x4.extadd_pairwise_i16x8_u + (local.get $0) + ) + ) (func $i16x8.abs (param $0 v128) (result v128) (i16x8.abs (local.get $0) @@ -690,9 +787,10 @@ (local.get $0) ) ) - (func $i16x8.any_true (param $0 v128) (result i32) - (i16x8.any_true + (func $i16x8.q15mulr_sat_s (param $0 v128) (param $1 v128) (result v128) + (i16x8.q15mulr_sat_s (local.get $0) + (local.get $1) ) ) (func $i16x8.all_true (param $0 v128) (result i32) @@ -705,6 +803,38 @@ (local.get $0) ) ) + (func $i16x8.narrow_i32x4_s (param $0 v128) (param $1 v128) (result v128) + (i16x8.narrow_i32x4_s + (local.get $0) + (local.get $1) + ) + ) + (func $i16x8.narrow_i32x4_u (param $0 v128) (param $1 v128) (result v128) + (i16x8.narrow_i32x4_u + (local.get $0) + (local.get $1) + ) + ) + (func $i16x8.extend_low_i8x16_s (param $0 v128) (result v128) + (i16x8.extend_low_i8x16_s + (local.get $0) + ) + ) + (func $i16x8.extend_high_i8x16_s (param $0 v128) (result v128) + (i16x8.extend_high_i8x16_s + (local.get $0) + ) + ) + (func $i16x8.extend_low_i8x16_u (param $0 v128) (result v128) + (i16x8.extend_low_i8x16_u + (local.get $0) + ) + ) + (func $i16x8.extend_high_i8x16_u (param $0 v128) (result v128) + (i16x8.extend_high_i8x16_u + (local.get $0) + ) + ) (func $i16x8.shl (param $0 v128) (param $1 i32) (result v128) (i16x8.shl (local.get $0) @@ -729,14 +859,14 @@ (local.get $1) ) ) - (func $i16x8.add_saturate_s (param $0 v128) (param $1 v128) (result v128) - (i16x8.add_saturate_s + (func $i16x8.add_sat_s (param $0 v128) (param $1 v128) (result v128) + (i16x8.add_sat_s (local.get $0) (local.get $1) ) ) - (func $i16x8.add_saturate_u (param $0 v128) (param $1 v128) (result v128) - (i16x8.add_saturate_u + (func $i16x8.add_sat_u (param $0 v128) (param $1 v128) (result v128) + (i16x8.add_sat_u (local.get $0) (local.get $1) ) @@ -747,18 +877,23 @@ (local.get $1) ) ) - (func $i16x8.sub_saturate_s (param $0 v128) (param $1 v128) (result v128) - (i16x8.sub_saturate_s + (func $i16x8.sub_sat_s (param $0 v128) (param $1 v128) (result v128) + (i16x8.sub_sat_s (local.get $0) (local.get $1) ) ) - (func $i16x8.sub_saturate_u (param $0 v128) (param $1 v128) (result v128) - (i16x8.sub_saturate_u + (func $i16x8.sub_sat_u (param $0 v128) (param $1 v128) (result v128) + (i16x8.sub_sat_u (local.get $0) (local.get $1) ) ) + (func $f64x2.nearest (param $0 v128) (result v128) + (f64x2.nearest + (local.get $0) + ) + ) (func $i16x8.mul (param $0 v128) (param $1 v128) (result v128) (i16x8.mul (local.get $0) @@ -795,12 +930,6 @@ (local.get $1) ) ) - (func $i16x8.q15mulr_sat_s (param $0 v128) (param $1 v128) (result v128) - (i16x8.q15mulr_sat_s - (local.get $0) - (local.get $1) - ) - ) (func $i16x8.extmul_low_i8x16_s (param $0 v128) (param $1 v128) (result v128) (i16x8.extmul_low_i8x16_s (local.get $0) @@ -835,11 +964,6 @@ (local.get $0) ) ) - (func $i32x4.any_true (param $0 v128) (result i32) - (i32x4.any_true - (local.get $0) - ) - ) (func $i32x4.all_true (param $0 v128) (result i32) (i32x4.all_true (local.get $0) @@ -850,6 +974,26 @@ (local.get $0) ) ) + (func $i32x4.extend_low_i16x8_s (param $0 v128) (result v128) + (i32x4.extend_low_i16x8_s + (local.get $0) + ) + ) + (func $i32x4.extend_high_i16x8_s (param $0 v128) (result v128) + (i32x4.extend_high_i16x8_s + (local.get $0) + ) + ) + (func $i32x4.extend_low_i16x8_u (param $0 v128) (result v128) + (i32x4.extend_low_i16x8_u + (local.get $0) + ) + ) + (func $i32x4.extend_high_i16x8_u (param $0 v128) (result v128) + (i32x4.extend_high_i16x8_u + (local.get $0) + ) + ) (func $i32x4.shl (param $0 v128) (param $1 i32) (result v128) (i32x4.shl (local.get $0) @@ -940,16 +1084,46 @@ (local.get $1) ) ) + (func $i64x2.abs (param $0 v128) (result v128) + (i64x2.abs + (local.get $0) + ) + ) (func $i64x2.neg (param $0 v128) (result v128) (i64x2.neg (local.get $0) ) ) + (func $i64x2.all_true (param $0 v128) (result i32) + (i64x2.all_true + (local.get $0) + ) + ) (func $i64x2.bitmask (param $0 v128) (result i32) (i64x2.bitmask (local.get $0) ) ) + (func $i64x2.extend_low_i32x4_s (param $0 v128) (result v128) + (i64x2.extend_low_i32x4_s + (local.get $0) + ) + ) + (func $i64x2.extend_high_i32x4_s (param $0 v128) (result v128) + (i64x2.extend_high_i32x4_s + (local.get $0) + ) + ) + (func $i64x2.extend_low_i32x4_u (param $0 v128) (result v128) + (i64x2.extend_low_i32x4_u + (local.get $0) + ) + ) + (func $i64x2.extend_high_i32x4_u (param $0 v128) (result v128) + (i64x2.extend_high_i32x4_u + (local.get $0) + ) + ) (func $i64x2.shl (param $0 v128) (param $1 i32) (result v128) (i64x2.shl (local.get $0) @@ -986,6 +1160,42 @@ (local.get $1) ) ) + (func $i64x2.eq (param $0 v128) (param $1 v128) (result v128) + (i64x2.eq + (local.get $0) + (local.get $1) + ) + ) + (func $i64x2.ne (param $0 v128) (param $1 v128) (result v128) + (i64x2.ne + (local.get $0) + (local.get $1) + ) + ) + (func $i64x2.lt_s (param $0 v128) (param $1 v128) (result v128) + (i64x2.lt_s + (local.get $0) + (local.get $1) + ) + ) + (func $i64x2.gt_s (param $0 v128) (param $1 v128) (result v128) + (i64x2.gt_s + (local.get $0) + (local.get $1) + ) + ) + (func $i64x2.le_s (param $0 v128) (param $1 v128) (result v128) + (i64x2.le_s + (local.get $0) + (local.get $1) + ) + ) + (func $i64x2.ge_s (param $0 v128) (param $1 v128) (result v128) + (i64x2.ge_s + (local.get $0) + (local.get $1) + ) + ) (func $i64x2.extmul_low_i32x4_s (param $0 v128) (param $1 v128) (result v128) (i64x2.extmul_low_i32x4_s (local.get $0) @@ -1010,6 +1220,21 @@ (local.get $1) ) ) + (func $f32x4.abs (param $0 v128) (result v128) + (f32x4.abs + (local.get $0) + ) + ) + (func $f32x4.neg (param $0 v128) (result v128) + (f32x4.neg + (local.get $0) + ) + ) + (func $f32x4.sqrt (param $0 v128) (result v128) + (f32x4.sqrt + (local.get $0) + ) + ) (func $f32x4.add (param $0 v128) (param $1 v128) (result v128) (f32x4.add (local.get $0) @@ -1058,53 +1283,19 @@ (local.get $1) ) ) - (func $f32x4.ceil (param $0 v128) (result v128) - (f32x4.ceil - (local.get $0) - ) - ) - (func $f32x4.floor (param $0 v128) (result v128) - (f32x4.floor - (local.get $0) - ) - ) - (func $f32x4.trunc (param $0 v128) (result v128) - (f32x4.trunc - (local.get $0) - ) - ) - (func $f32x4.nearest (param $0 v128) (result v128) - (f32x4.nearest - (local.get $0) - ) - ) - (func $f32x4.abs (param $0 v128) (result v128) - (f32x4.abs - (local.get $0) - ) - ) - (func $f32x4.neg (param $0 v128) (result v128) - (f32x4.neg + (func $f64x2.abs (param $0 v128) (result v128) + (f64x2.abs (local.get $0) ) ) - (func $f32x4.sqrt (param $0 v128) (result v128) - (f32x4.sqrt + (func $f64x2.neg (param $0 v128) (result v128) + (f64x2.neg (local.get $0) ) ) - (func $f32x4.qfma (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (f32x4.qfma + (func $f64x2.sqrt (param $0 v128) (result v128) + (f64x2.sqrt (local.get $0) - (local.get $1) - (local.get $2) - ) - ) - (func $f32x4.qfms (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (f32x4.qfms - (local.get $0) - (local.get $1) - (local.get $2) ) ) (func $f64x2.add (param $0 v128) (param $1 v128) (result v128) @@ -1155,75 +1346,6 @@ (local.get $1) ) ) - (func $f64x2.ceil (param $0 v128) (result v128) - (f64x2.ceil - (local.get $0) - ) - ) - (func $f64x2.floor (param $0 v128) (result v128) - (f64x2.floor - (local.get $0) - ) - ) - (func $f64x2.trunc (param $0 v128) (result v128) - (f64x2.trunc - (local.get $0) - ) - ) - (func $f64x2.nearest (param $0 v128) (result v128) - (f64x2.nearest - (local.get $0) - ) - ) - (func $f64x2.abs (param $0 v128) (result v128) - (f64x2.abs - (local.get $0) - ) - ) - (func $f64x2.neg (param $0 v128) (result v128) - (f64x2.neg - (local.get $0) - ) - ) - (func $f64x2.sqrt (param $0 v128) (result v128) - (f64x2.sqrt - (local.get $0) - ) - ) - (func $f64x2.qfma (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (f64x2.qfma - (local.get $0) - (local.get $1) - (local.get $2) - ) - ) - (func $f64x2.qfms (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (f64x2.qfms - (local.get $0) - (local.get $1) - (local.get $2) - ) - ) - (func $i16x8.extadd_pairwise_i8x16_s (param $0 v128) (result v128) - (i16x8.extadd_pairwise_i8x16_s - (local.get $0) - ) - ) - (func $i16x8.extadd_pairwise_i8x16_u (param $0 v128) (result v128) - (i16x8.extadd_pairwise_i8x16_u - (local.get $0) - ) - ) - (func $i32x4.extadd_pairwise_i16x8_s (param $0 v128) (result v128) - (i32x4.extadd_pairwise_i16x8_s - (local.get $0) - ) - ) - (func $i32x4.extadd_pairwise_i16x8_u (param $0 v128) (result v128) - (i32x4.extadd_pairwise_i16x8_u - (local.get $0) - ) - ) (func $i32x4.trunc_sat_f32x4_s (param $0 v128) (result v128) (i32x4.trunc_sat_f32x4_s (local.get $0) @@ -1234,16 +1356,6 @@ (local.get $0) ) ) - (func $i64x2.trunc_sat_f64x2_s (param $0 v128) (result v128) - (i64x2.trunc_sat_f64x2_s - (local.get $0) - ) - ) - (func $i64x2.trunc_sat_f64x2_u (param $0 v128) (result v128) - (i64x2.trunc_sat_f64x2_u - (local.get $0) - ) - ) (func $f32x4.convert_i32x4_s (param $0 v128) (result v128) (f32x4.convert_i32x4_s (local.get $0) @@ -1254,173 +1366,13 @@ (local.get $0) ) ) - (func $f64x2.convert_i64x2_s (param $0 v128) (result v128) - (f64x2.convert_i64x2_s - (local.get $0) - ) - ) - (func $f64x2.convert_i64x2_u (param $0 v128) (result v128) - (f64x2.convert_i64x2_u - (local.get $0) - ) - ) - (func $v8x16.load_splat (param $0 i32) (result v128) - (v8x16.load_splat - (local.get $0) - ) - ) - (func $v16x8.load_splat (param $0 i32) (result v128) - (v16x8.load_splat + (func $i32x4.trunc_sat_f64x2_s_zero (param $0 v128) (result v128) + (i32x4.trunc_sat_f64x2_s_zero (local.get $0) ) ) - (func $v32x4.load_splat (param $0 i32) (result v128) - (v32x4.load_splat - (local.get $0) - ) - ) - (func $v64x2.load_splat (param $0 i32) (result v128) - (v64x2.load_splat - (local.get $0) - ) - ) - (func $i8x16.narrow_i16x8_s (param $0 v128) (param $1 v128) (result v128) - (i8x16.narrow_i16x8_s - (local.get $0) - (local.get $1) - ) - ) - (func $i8x16.narrow_i16x8_u (param $0 v128) (param $1 v128) (result v128) - (i8x16.narrow_i16x8_u - (local.get $0) - (local.get $1) - ) - ) - (func $i16x8.narrow_i32x4_s (param $0 v128) (param $1 v128) (result v128) - (i16x8.narrow_i32x4_s - (local.get $0) - (local.get $1) - ) - ) - (func $i16x8.narrow_i32x4_u (param $0 v128) (param $1 v128) (result v128) - (i16x8.narrow_i32x4_u - (local.get $0) - (local.get $1) - ) - ) - (func $i16x8.widen_low_i8x16_s (param $0 v128) (result v128) - (i16x8.widen_low_i8x16_s - (local.get $0) - ) - ) - (func $i16x8.widen_high_i8x16_s (param $0 v128) (result v128) - (i16x8.widen_high_i8x16_s - (local.get $0) - ) - ) - (func $i16x8.widen_low_i8x16_u (param $0 v128) (result v128) - (i16x8.widen_low_i8x16_u - (local.get $0) - ) - ) - (func $i16x8.widen_high_i8x16_u (param $0 v128) (result v128) - (i16x8.widen_high_i8x16_u - (local.get $0) - ) - ) - (func $i32x4.widen_low_i16x8_s (param $0 v128) (result v128) - (i32x4.widen_low_i16x8_s - (local.get $0) - ) - ) - (func $i32x4.widen_high_i16x8_s (param $0 v128) (result v128) - (i32x4.widen_high_i16x8_s - (local.get $0) - ) - ) - (func $i32x4.widen_low_i16x8_u (param $0 v128) (result v128) - (i32x4.widen_low_i16x8_u - (local.get $0) - ) - ) - (func $i32x4.widen_high_i16x8_u (param $0 v128) (result v128) - (i32x4.widen_high_i16x8_u - (local.get $0) - ) - ) - (func $i64x2.widen_low_i32x4_s (param $0 v128) (result v128) - (i64x2.widen_low_i32x4_s - (local.get $0) - ) - ) - (func $i64x2.widen_high_i32x4_s (param $0 v128) (result v128) - (i64x2.widen_high_i32x4_s - (local.get $0) - ) - ) - (func $i64x2.widen_low_i32x4_u (param $0 v128) (result v128) - (i64x2.widen_low_i32x4_u - (local.get $0) - ) - ) - (func $i64x2.widen_high_i32x4_u (param $0 v128) (result v128) - (i64x2.widen_high_i32x4_u - (local.get $0) - ) - ) - (func $i16x8.load8x8_u (param $0 i32) (result v128) - (i16x8.load8x8_u - (local.get $0) - ) - ) - (func $i16x8.load8x8_s (param $0 i32) (result v128) - (i16x8.load8x8_s - (local.get $0) - ) - ) - (func $i32x4.load16x4_s (param $0 i32) (result v128) - (i32x4.load16x4_s - (local.get $0) - ) - ) - (func $i32x4.load16x4_u (param $0 i32) (result v128) - (i32x4.load16x4_u - (local.get $0) - ) - ) - (func $i64x2.load32x2_s (param $0 i32) (result v128) - (i64x2.load32x2_s - (local.get $0) - ) - ) - (func $i64x2.load32x2_u (param $0 i32) (result v128) - (i64x2.load32x2_u - (local.get $0) - ) - ) - (func $v128.load32_zero (param $0 i32) (result v128) - (v128.load32_zero - (local.get $0) - ) - ) - (func $v128.load64_zero (param $0 i32) (result v128) - (v128.load64_zero - (local.get $0) - ) - ) - (func $v8x16.swizzle (param $0 v128) (param $1 v128) (result v128) - (v8x16.swizzle - (local.get $0) - (local.get $1) - ) - ) - (func $prefetch.t (param $0 i32) - (prefetch.t offset=3 align=2 - (local.get $0) - ) - ) - (func $prefetch.nt (param $0 i32) - (prefetch.nt offset=3 align=2 + (func $i32x4.trunc_sat_f64x2_u_zero (param $0 v128) (result v128) + (i32x4.trunc_sat_f64x2_u_zero (local.get $0) ) ) @@ -1434,34 +1386,4 @@ (local.get $0) ) ) - (func $i32x4.trunc_sat_f64x2_zero_s (param $0 v128) (result v128) - (i32x4.trunc_sat_f64x2_zero_s - (local.get $0) - ) - ) - (func $i32x4.trunc_sat_f64x2_zero_u (param $0 v128) (result v128) - (i32x4.trunc_sat_f64x2_zero_u - (local.get $0) - ) - ) - (func $f32x4.demote_f64x2_zero (param $0 v128) (result v128) - (f32x4.demote_f64x2_zero - (local.get $0) - ) - ) - (func $f64x2.promote_low_f32x4 (param $0 v128) (result v128) - (f64x2.promote_low_f32x4 - (local.get $0) - ) - ) - (func $i32x4.widen_i8x16_s (param $0 v128) (result v128) - (i32x4.widen_i8x16_s 0 - (local.get $0) - ) - ) - (func $i32x4.widen_i8x16_u (param $0 v128) (result v128) - (i32x4.widen_i8x16_u 0 - (local.get $0) - ) - ) ) diff --git a/test/simd.wast.fromBinary b/test/simd.wast.fromBinary index 4c84b5b25..56f3fa2cc 100644 --- a/test/simd.wast.fromBinary +++ b/test/simd.wast.fromBinary @@ -2,13 +2,11 @@ (type $v128_v128_=>_v128 (func (param v128 v128) (result v128))) (type $v128_=>_v128 (func (param v128) (result v128))) (type $i32_=>_v128 (func (param i32) (result v128))) - (type $v128_=>_i32 (func (param v128) (result i32))) (type $v128_i32_=>_v128 (func (param v128 i32) (result v128))) - (type $v128_v128_v128_=>_v128 (func (param v128 v128 v128) (result v128))) + (type $v128_=>_i32 (func (param v128) (result i32))) (type $i32_v128_=>_none (func (param i32 v128))) (type $i32_v128_=>_v128 (func (param i32 v128) (result v128))) (type $none_=>_v128 (func (result v128))) - (type $i32_=>_none (func (param i32))) (type $v128_=>_i64 (func (param v128) (result i64))) (type $v128_=>_f32 (func (param v128) (result f32))) (type $v128_=>_f64 (func (param v128) (result f64))) @@ -17,12 +15,63 @@ (type $v128_i64_=>_v128 (func (param v128 i64) (result v128))) (type $v128_f32_=>_v128 (func (param v128 f32) (result v128))) (type $v128_f64_=>_v128 (func (param v128 f64) (result v128))) + (type $v128_v128_v128_=>_v128 (func (param v128 v128 v128) (result v128))) (memory $0 1 1) (func $v128.load (param $0 i32) (result v128) (v128.load (local.get $0) ) ) + (func $v128.load8x8_s (param $0 i32) (result v128) + (v128.load8x8_s + (local.get $0) + ) + ) + (func $v128.load8x8_u (param $0 i32) (result v128) + (v128.load8x8_u + (local.get $0) + ) + ) + (func $v128.load16x4_s (param $0 i32) (result v128) + (v128.load16x4_s + (local.get $0) + ) + ) + (func $v128.load16x4_u (param $0 i32) (result v128) + (v128.load16x4_u + (local.get $0) + ) + ) + (func $v128.load32x2_s (param $0 i32) (result v128) + (v128.load32x2_s + (local.get $0) + ) + ) + (func $v128.load32x2_u (param $0 i32) (result v128) + (v128.load32x2_u + (local.get $0) + ) + ) + (func $v128.load8_splat (param $0 i32) (result v128) + (v128.load8_splat + (local.get $0) + ) + ) + (func $v128.load16_splat (param $0 i32) (result v128) + (v128.load16_splat + (local.get $0) + ) + ) + (func $v128.load32_splat (param $0 i32) (result v128) + (v128.load32_splat + (local.get $0) + ) + ) + (func $v128.load64_splat (param $0 i32) (result v128) + (v128.load64_splat + (local.get $0) + ) + ) (func $v128.store (param $0 i32) (param $1 v128) (v128.store (local.get $0) @@ -47,8 +96,14 @@ (func $v128.const.f64x2 (result v128) (v128.const i32x4 0x00000000 0x3ff00000 0x00000000 0x40000000) ) - (func $v128.shuffle (param $0 v128) (param $1 v128) (result v128) - (v8x16.shuffle 0 17 2 19 4 21 6 23 8 25 10 27 12 29 14 31 + (func $i8x16.shuffle (param $0 v128) (param $1 v128) (result v128) + (i8x16.shuffle 0 17 2 19 4 21 6 23 8 25 10 27 12 29 14 31 + (local.get $0) + (local.get $1) + ) + ) + (func $i8x16.swizzle (param $0 v128) (param $1 v128) (result v128) + (i8x16.swizzle (local.get $0) (local.get $1) ) @@ -58,6 +113,21 @@ (local.get $0) ) ) + (func $i16x8.splat (param $0 i32) (result v128) + (i16x8.splat + (local.get $0) + ) + ) + (func $f32x4.splat (param $0 f32) (result v128) + (f32x4.splat + (local.get $0) + ) + ) + (func $f64x2.splat (param $0 f64) (result v128) + (f64x2.splat + (local.get $0) + ) + ) (func $i8x16.extract_lane_s (param $0 v128) (result i32) (i8x16.extract_lane_s 0 (local.get $0) @@ -74,11 +144,6 @@ (local.get $1) ) ) - (func $i16x8.splat (param $0 i32) (result v128) - (i16x8.splat - (local.get $0) - ) - ) (func $i16x8.extract_lane_s (param $0 v128) (result i32) (i16x8.extract_lane_s 0 (local.get $0) @@ -95,11 +160,6 @@ (local.get $1) ) ) - (func $i32x4.splat (param $0 i32) (result v128) - (i32x4.splat - (local.get $0) - ) - ) (func $i32x4.extract_lane (param $0 v128) (result i32) (i32x4.extract_lane 0 (local.get $0) @@ -122,11 +182,6 @@ (local.get $1) ) ) - (func $f32x4.splat (param $0 f32) (result v128) - (f32x4.splat - (local.get $0) - ) - ) (func $f32x4.extract_lane (param $0 v128) (result f32) (f32x4.extract_lane 0 (local.get $0) @@ -138,11 +193,6 @@ (local.get $1) ) ) - (func $f64x2.splat (param $0 f64) (result v128) - (f64x2.splat - (local.get $0) - ) - ) (func $f64x2.extract_lane (param $0 v128) (result f64) (f64x2.extract_lane 0 (local.get $0) @@ -334,12 +384,6 @@ (local.get $1) ) ) - (func $i64x2.eq (param $0 v128) (param $1 v128) (result v128) - (i64x2.eq - (local.get $0) - (local.get $1) - ) - ) (func $f32x4.eq (param $0 v128) (param $1 v128) (result v128) (f32x4.eq (local.get $0) @@ -423,57 +467,34 @@ (local.get $1) ) ) - (func $v128.or (param $0 v128) (param $1 v128) (result v128) - (v128.or - (local.get $0) - (local.get $1) - ) - ) - (func $v128.xor (param $0 v128) (param $1 v128) (result v128) - (v128.xor - (local.get $0) - (local.get $1) - ) - ) (func $v128.andnot (param $0 v128) (param $1 v128) (result v128) (v128.andnot (local.get $0) (local.get $1) ) ) - (func $v128.bitselect (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (v128.bitselect + (func $v128.or (param $0 v128) (param $1 v128) (result v128) + (v128.or (local.get $0) (local.get $1) - (local.get $2) ) ) - (func $v8x16.signselect (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (v8x16.signselect + (func $v128.xor (param $0 v128) (param $1 v128) (result v128) + (v128.xor (local.get $0) (local.get $1) - (local.get $2) ) ) - (func $v16x8.signselect (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (v16x8.signselect + (func $v128.bitselect (param $0 v128) (param $1 v128) (param $2 v128) (result v128) + (v128.bitselect (local.get $0) (local.get $1) (local.get $2) ) ) - (func $v32x4.signselect (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (v32x4.signselect + (func $v128.any_true (param $0 v128) (result i32) + (v128.any_true (local.get $0) - (local.get $1) - (local.get $2) - ) - ) - (func $v64x2.signselect (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (v64x2.signselect - (local.get $0) - (local.get $1) - (local.get $2) ) ) (func $v128.load8_lane (param $0 i32) (param $1 v128) (result v128) @@ -560,8 +581,23 @@ (local.get $1) ) ) - (func $i8x16.popcnt (param $0 v128) (result v128) - (i8x16.popcnt + (func $v128.load32_zero (param $0 i32) (result v128) + (v128.load32_zero + (local.get $0) + ) + ) + (func $v128.load64_zero (param $0 i32) (result v128) + (v128.load64_zero + (local.get $0) + ) + ) + (func $f32x4.demote_f64x2_zero (param $0 v128) (result v128) + (f32x4.demote_f64x2_zero + (local.get $0) + ) + ) + (func $f64x2.promote_low_f32x4 (param $0 v128) (result v128) + (f64x2.promote_low_f32x4 (local.get $0) ) ) @@ -575,8 +611,8 @@ (local.get $0) ) ) - (func $i8x16.any_true (param $0 v128) (result i32) - (i8x16.any_true + (func $i8x16.popcnt (param $0 v128) (result v128) + (i8x16.popcnt (local.get $0) ) ) @@ -590,6 +626,38 @@ (local.get $0) ) ) + (func $i8x16.narrow_i16x8_s (param $0 v128) (param $1 v128) (result v128) + (i8x16.narrow_i16x8_s + (local.get $0) + (local.get $1) + ) + ) + (func $i8x16.narrow_i16x8_u (param $0 v128) (param $1 v128) (result v128) + (i8x16.narrow_i16x8_u + (local.get $0) + (local.get $1) + ) + ) + (func $f32x4.ceil (param $0 v128) (result v128) + (f32x4.ceil + (local.get $0) + ) + ) + (func $f32x4.floor (param $0 v128) (result v128) + (f32x4.floor + (local.get $0) + ) + ) + (func $f32x4.trunc (param $0 v128) (result v128) + (f32x4.trunc + (local.get $0) + ) + ) + (func $f32x4.nearest (param $0 v128) (result v128) + (f32x4.nearest + (local.get $0) + ) + ) (func $i8x16.shl (param $0 v128) (param $1 i32) (result v128) (i8x16.shl (local.get $0) @@ -614,14 +682,14 @@ (local.get $1) ) ) - (func $i8x16.add_saturate_s (param $0 v128) (param $1 v128) (result v128) - (i8x16.add_saturate_s + (func $i8x16.add_sat_s (param $0 v128) (param $1 v128) (result v128) + (i8x16.add_sat_s (local.get $0) (local.get $1) ) ) - (func $i8x16.add_saturate_u (param $0 v128) (param $1 v128) (result v128) - (i8x16.add_saturate_u + (func $i8x16.add_sat_u (param $0 v128) (param $1 v128) (result v128) + (i8x16.add_sat_u (local.get $0) (local.get $1) ) @@ -632,22 +700,26 @@ (local.get $1) ) ) - (func $i8x16.sub_saturate_s (param $0 v128) (param $1 v128) (result v128) - (i8x16.sub_saturate_s + (func $i8x16.sub_sat_s (param $0 v128) (param $1 v128) (result v128) + (i8x16.sub_sat_s (local.get $0) (local.get $1) ) ) - (func $i8x16.sub_saturate_u (param $0 v128) (param $1 v128) (result v128) - (i8x16.sub_saturate_u + (func $i8x16.sub_sat_u (param $0 v128) (param $1 v128) (result v128) + (i8x16.sub_sat_u (local.get $0) (local.get $1) ) ) - (func $i8x16.mul (param $0 v128) (param $1 v128) (result v128) - (i8x16.mul + (func $f64x2.ceil (param $0 v128) (result v128) + (f64x2.ceil + (local.get $0) + ) + ) + (func $f64x2.floor (param $0 v128) (result v128) + (f64x2.floor (local.get $0) - (local.get $1) ) ) (func $i8x16.min_s (param $0 v128) (param $1 v128) (result v128) @@ -674,12 +746,37 @@ (local.get $1) ) ) + (func $f64x2.trunc (param $0 v128) (result v128) + (f64x2.trunc + (local.get $0) + ) + ) (func $i8x16.avgr_u (param $0 v128) (param $1 v128) (result v128) (i8x16.avgr_u (local.get $0) (local.get $1) ) ) + (func $i16x8.extadd_pairwise_i8x16_s (param $0 v128) (result v128) + (i16x8.extadd_pairwise_i8x16_s + (local.get $0) + ) + ) + (func $i16x8.extadd_pairwise_i8x16_u (param $0 v128) (result v128) + (i16x8.extadd_pairwise_i8x16_u + (local.get $0) + ) + ) + (func $i32x4.extadd_pairwise_i16x8_s (param $0 v128) (result v128) + (i32x4.extadd_pairwise_i16x8_s + (local.get $0) + ) + ) + (func $i32x4.extadd_pairwise_i16x8_u (param $0 v128) (result v128) + (i32x4.extadd_pairwise_i16x8_u + (local.get $0) + ) + ) (func $i16x8.abs (param $0 v128) (result v128) (i16x8.abs (local.get $0) @@ -690,9 +787,10 @@ (local.get $0) ) ) - (func $i16x8.any_true (param $0 v128) (result i32) - (i16x8.any_true + (func $i16x8.q15mulr_sat_s (param $0 v128) (param $1 v128) (result v128) + (i16x8.q15mulr_sat_s (local.get $0) + (local.get $1) ) ) (func $i16x8.all_true (param $0 v128) (result i32) @@ -705,6 +803,38 @@ (local.get $0) ) ) + (func $i16x8.narrow_i32x4_s (param $0 v128) (param $1 v128) (result v128) + (i16x8.narrow_i32x4_s + (local.get $0) + (local.get $1) + ) + ) + (func $i16x8.narrow_i32x4_u (param $0 v128) (param $1 v128) (result v128) + (i16x8.narrow_i32x4_u + (local.get $0) + (local.get $1) + ) + ) + (func $i16x8.extend_low_i8x16_s (param $0 v128) (result v128) + (i16x8.extend_low_i8x16_s + (local.get $0) + ) + ) + (func $i16x8.extend_high_i8x16_s (param $0 v128) (result v128) + (i16x8.extend_high_i8x16_s + (local.get $0) + ) + ) + (func $i16x8.extend_low_i8x16_u (param $0 v128) (result v128) + (i16x8.extend_low_i8x16_u + (local.get $0) + ) + ) + (func $i16x8.extend_high_i8x16_u (param $0 v128) (result v128) + (i16x8.extend_high_i8x16_u + (local.get $0) + ) + ) (func $i16x8.shl (param $0 v128) (param $1 i32) (result v128) (i16x8.shl (local.get $0) @@ -729,14 +859,14 @@ (local.get $1) ) ) - (func $i16x8.add_saturate_s (param $0 v128) (param $1 v128) (result v128) - (i16x8.add_saturate_s + (func $i16x8.add_sat_s (param $0 v128) (param $1 v128) (result v128) + (i16x8.add_sat_s (local.get $0) (local.get $1) ) ) - (func $i16x8.add_saturate_u (param $0 v128) (param $1 v128) (result v128) - (i16x8.add_saturate_u + (func $i16x8.add_sat_u (param $0 v128) (param $1 v128) (result v128) + (i16x8.add_sat_u (local.get $0) (local.get $1) ) @@ -747,18 +877,23 @@ (local.get $1) ) ) - (func $i16x8.sub_saturate_s (param $0 v128) (param $1 v128) (result v128) - (i16x8.sub_saturate_s + (func $i16x8.sub_sat_s (param $0 v128) (param $1 v128) (result v128) + (i16x8.sub_sat_s (local.get $0) (local.get $1) ) ) - (func $i16x8.sub_saturate_u (param $0 v128) (param $1 v128) (result v128) - (i16x8.sub_saturate_u + (func $i16x8.sub_sat_u (param $0 v128) (param $1 v128) (result v128) + (i16x8.sub_sat_u (local.get $0) (local.get $1) ) ) + (func $f64x2.nearest (param $0 v128) (result v128) + (f64x2.nearest + (local.get $0) + ) + ) (func $i16x8.mul (param $0 v128) (param $1 v128) (result v128) (i16x8.mul (local.get $0) @@ -795,12 +930,6 @@ (local.get $1) ) ) - (func $i16x8.q15mulr_sat_s (param $0 v128) (param $1 v128) (result v128) - (i16x8.q15mulr_sat_s - (local.get $0) - (local.get $1) - ) - ) (func $i16x8.extmul_low_i8x16_s (param $0 v128) (param $1 v128) (result v128) (i16x8.extmul_low_i8x16_s (local.get $0) @@ -835,11 +964,6 @@ (local.get $0) ) ) - (func $i32x4.any_true (param $0 v128) (result i32) - (i32x4.any_true - (local.get $0) - ) - ) (func $i32x4.all_true (param $0 v128) (result i32) (i32x4.all_true (local.get $0) @@ -850,6 +974,26 @@ (local.get $0) ) ) + (func $i32x4.extend_low_i16x8_s (param $0 v128) (result v128) + (i32x4.extend_low_i16x8_s + (local.get $0) + ) + ) + (func $i32x4.extend_high_i16x8_s (param $0 v128) (result v128) + (i32x4.extend_high_i16x8_s + (local.get $0) + ) + ) + (func $i32x4.extend_low_i16x8_u (param $0 v128) (result v128) + (i32x4.extend_low_i16x8_u + (local.get $0) + ) + ) + (func $i32x4.extend_high_i16x8_u (param $0 v128) (result v128) + (i32x4.extend_high_i16x8_u + (local.get $0) + ) + ) (func $i32x4.shl (param $0 v128) (param $1 i32) (result v128) (i32x4.shl (local.get $0) @@ -940,16 +1084,46 @@ (local.get $1) ) ) + (func $i64x2.abs (param $0 v128) (result v128) + (i64x2.abs + (local.get $0) + ) + ) (func $i64x2.neg (param $0 v128) (result v128) (i64x2.neg (local.get $0) ) ) + (func $i64x2.all_true (param $0 v128) (result i32) + (i64x2.all_true + (local.get $0) + ) + ) (func $i64x2.bitmask (param $0 v128) (result i32) (i64x2.bitmask (local.get $0) ) ) + (func $i64x2.extend_low_i32x4_s (param $0 v128) (result v128) + (i64x2.extend_low_i32x4_s + (local.get $0) + ) + ) + (func $i64x2.extend_high_i32x4_s (param $0 v128) (result v128) + (i64x2.extend_high_i32x4_s + (local.get $0) + ) + ) + (func $i64x2.extend_low_i32x4_u (param $0 v128) (result v128) + (i64x2.extend_low_i32x4_u + (local.get $0) + ) + ) + (func $i64x2.extend_high_i32x4_u (param $0 v128) (result v128) + (i64x2.extend_high_i32x4_u + (local.get $0) + ) + ) (func $i64x2.shl (param $0 v128) (param $1 i32) (result v128) (i64x2.shl (local.get $0) @@ -986,6 +1160,42 @@ (local.get $1) ) ) + (func $i64x2.eq (param $0 v128) (param $1 v128) (result v128) + (i64x2.eq + (local.get $0) + (local.get $1) + ) + ) + (func $i64x2.ne (param $0 v128) (param $1 v128) (result v128) + (i64x2.ne + (local.get $0) + (local.get $1) + ) + ) + (func $i64x2.lt_s (param $0 v128) (param $1 v128) (result v128) + (i64x2.lt_s + (local.get $0) + (local.get $1) + ) + ) + (func $i64x2.gt_s (param $0 v128) (param $1 v128) (result v128) + (i64x2.gt_s + (local.get $0) + (local.get $1) + ) + ) + (func $i64x2.le_s (param $0 v128) (param $1 v128) (result v128) + (i64x2.le_s + (local.get $0) + (local.get $1) + ) + ) + (func $i64x2.ge_s (param $0 v128) (param $1 v128) (result v128) + (i64x2.ge_s + (local.get $0) + (local.get $1) + ) + ) (func $i64x2.extmul_low_i32x4_s (param $0 v128) (param $1 v128) (result v128) (i64x2.extmul_low_i32x4_s (local.get $0) @@ -1010,6 +1220,21 @@ (local.get $1) ) ) + (func $f32x4.abs (param $0 v128) (result v128) + (f32x4.abs + (local.get $0) + ) + ) + (func $f32x4.neg (param $0 v128) (result v128) + (f32x4.neg + (local.get $0) + ) + ) + (func $f32x4.sqrt (param $0 v128) (result v128) + (f32x4.sqrt + (local.get $0) + ) + ) (func $f32x4.add (param $0 v128) (param $1 v128) (result v128) (f32x4.add (local.get $0) @@ -1058,53 +1283,19 @@ (local.get $1) ) ) - (func $f32x4.ceil (param $0 v128) (result v128) - (f32x4.ceil - (local.get $0) - ) - ) - (func $f32x4.floor (param $0 v128) (result v128) - (f32x4.floor - (local.get $0) - ) - ) - (func $f32x4.trunc (param $0 v128) (result v128) - (f32x4.trunc - (local.get $0) - ) - ) - (func $f32x4.nearest (param $0 v128) (result v128) - (f32x4.nearest - (local.get $0) - ) - ) - (func $f32x4.abs (param $0 v128) (result v128) - (f32x4.abs - (local.get $0) - ) - ) - (func $f32x4.neg (param $0 v128) (result v128) - (f32x4.neg + (func $f64x2.abs (param $0 v128) (result v128) + (f64x2.abs (local.get $0) ) ) - (func $f32x4.sqrt (param $0 v128) (result v128) - (f32x4.sqrt + (func $f64x2.neg (param $0 v128) (result v128) + (f64x2.neg (local.get $0) ) ) - (func $f32x4.qfma (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (f32x4.qfma + (func $f64x2.sqrt (param $0 v128) (result v128) + (f64x2.sqrt (local.get $0) - (local.get $1) - (local.get $2) - ) - ) - (func $f32x4.qfms (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (f32x4.qfms - (local.get $0) - (local.get $1) - (local.get $2) ) ) (func $f64x2.add (param $0 v128) (param $1 v128) (result v128) @@ -1155,75 +1346,6 @@ (local.get $1) ) ) - (func $f64x2.ceil (param $0 v128) (result v128) - (f64x2.ceil - (local.get $0) - ) - ) - (func $f64x2.floor (param $0 v128) (result v128) - (f64x2.floor - (local.get $0) - ) - ) - (func $f64x2.trunc (param $0 v128) (result v128) - (f64x2.trunc - (local.get $0) - ) - ) - (func $f64x2.nearest (param $0 v128) (result v128) - (f64x2.nearest - (local.get $0) - ) - ) - (func $f64x2.abs (param $0 v128) (result v128) - (f64x2.abs - (local.get $0) - ) - ) - (func $f64x2.neg (param $0 v128) (result v128) - (f64x2.neg - (local.get $0) - ) - ) - (func $f64x2.sqrt (param $0 v128) (result v128) - (f64x2.sqrt - (local.get $0) - ) - ) - (func $f64x2.qfma (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (f64x2.qfma - (local.get $0) - (local.get $1) - (local.get $2) - ) - ) - (func $f64x2.qfms (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (f64x2.qfms - (local.get $0) - (local.get $1) - (local.get $2) - ) - ) - (func $i16x8.extadd_pairwise_i8x16_s (param $0 v128) (result v128) - (i16x8.extadd_pairwise_i8x16_s - (local.get $0) - ) - ) - (func $i16x8.extadd_pairwise_i8x16_u (param $0 v128) (result v128) - (i16x8.extadd_pairwise_i8x16_u - (local.get $0) - ) - ) - (func $i32x4.extadd_pairwise_i16x8_s (param $0 v128) (result v128) - (i32x4.extadd_pairwise_i16x8_s - (local.get $0) - ) - ) - (func $i32x4.extadd_pairwise_i16x8_u (param $0 v128) (result v128) - (i32x4.extadd_pairwise_i16x8_u - (local.get $0) - ) - ) (func $i32x4.trunc_sat_f32x4_s (param $0 v128) (result v128) (i32x4.trunc_sat_f32x4_s (local.get $0) @@ -1234,16 +1356,6 @@ (local.get $0) ) ) - (func $i64x2.trunc_sat_f64x2_s (param $0 v128) (result v128) - (i64x2.trunc_sat_f64x2_s - (local.get $0) - ) - ) - (func $i64x2.trunc_sat_f64x2_u (param $0 v128) (result v128) - (i64x2.trunc_sat_f64x2_u - (local.get $0) - ) - ) (func $f32x4.convert_i32x4_s (param $0 v128) (result v128) (f32x4.convert_i32x4_s (local.get $0) @@ -1254,173 +1366,13 @@ (local.get $0) ) ) - (func $f64x2.convert_i64x2_s (param $0 v128) (result v128) - (f64x2.convert_i64x2_s - (local.get $0) - ) - ) - (func $f64x2.convert_i64x2_u (param $0 v128) (result v128) - (f64x2.convert_i64x2_u - (local.get $0) - ) - ) - (func $v8x16.load_splat (param $0 i32) (result v128) - (v8x16.load_splat - (local.get $0) - ) - ) - (func $v16x8.load_splat (param $0 i32) (result v128) - (v16x8.load_splat + (func $i32x4.trunc_sat_f64x2_s_zero (param $0 v128) (result v128) + (i32x4.trunc_sat_f64x2_s_zero (local.get $0) ) ) - (func $v32x4.load_splat (param $0 i32) (result v128) - (v32x4.load_splat - (local.get $0) - ) - ) - (func $v64x2.load_splat (param $0 i32) (result v128) - (v64x2.load_splat - (local.get $0) - ) - ) - (func $i8x16.narrow_i16x8_s (param $0 v128) (param $1 v128) (result v128) - (i8x16.narrow_i16x8_s - (local.get $0) - (local.get $1) - ) - ) - (func $i8x16.narrow_i16x8_u (param $0 v128) (param $1 v128) (result v128) - (i8x16.narrow_i16x8_u - (local.get $0) - (local.get $1) - ) - ) - (func $i16x8.narrow_i32x4_s (param $0 v128) (param $1 v128) (result v128) - (i16x8.narrow_i32x4_s - (local.get $0) - (local.get $1) - ) - ) - (func $i16x8.narrow_i32x4_u (param $0 v128) (param $1 v128) (result v128) - (i16x8.narrow_i32x4_u - (local.get $0) - (local.get $1) - ) - ) - (func $i16x8.widen_low_i8x16_s (param $0 v128) (result v128) - (i16x8.widen_low_i8x16_s - (local.get $0) - ) - ) - (func $i16x8.widen_high_i8x16_s (param $0 v128) (result v128) - (i16x8.widen_high_i8x16_s - (local.get $0) - ) - ) - (func $i16x8.widen_low_i8x16_u (param $0 v128) (result v128) - (i16x8.widen_low_i8x16_u - (local.get $0) - ) - ) - (func $i16x8.widen_high_i8x16_u (param $0 v128) (result v128) - (i16x8.widen_high_i8x16_u - (local.get $0) - ) - ) - (func $i32x4.widen_low_i16x8_s (param $0 v128) (result v128) - (i32x4.widen_low_i16x8_s - (local.get $0) - ) - ) - (func $i32x4.widen_high_i16x8_s (param $0 v128) (result v128) - (i32x4.widen_high_i16x8_s - (local.get $0) - ) - ) - (func $i32x4.widen_low_i16x8_u (param $0 v128) (result v128) - (i32x4.widen_low_i16x8_u - (local.get $0) - ) - ) - (func $i32x4.widen_high_i16x8_u (param $0 v128) (result v128) - (i32x4.widen_high_i16x8_u - (local.get $0) - ) - ) - (func $i64x2.widen_low_i32x4_s (param $0 v128) (result v128) - (i64x2.widen_low_i32x4_s - (local.get $0) - ) - ) - (func $i64x2.widen_high_i32x4_s (param $0 v128) (result v128) - (i64x2.widen_high_i32x4_s - (local.get $0) - ) - ) - (func $i64x2.widen_low_i32x4_u (param $0 v128) (result v128) - (i64x2.widen_low_i32x4_u - (local.get $0) - ) - ) - (func $i64x2.widen_high_i32x4_u (param $0 v128) (result v128) - (i64x2.widen_high_i32x4_u - (local.get $0) - ) - ) - (func $i16x8.load8x8_u (param $0 i32) (result v128) - (i16x8.load8x8_u - (local.get $0) - ) - ) - (func $i16x8.load8x8_s (param $0 i32) (result v128) - (i16x8.load8x8_s - (local.get $0) - ) - ) - (func $i32x4.load16x4_s (param $0 i32) (result v128) - (i32x4.load16x4_s - (local.get $0) - ) - ) - (func $i32x4.load16x4_u (param $0 i32) (result v128) - (i32x4.load16x4_u - (local.get $0) - ) - ) - (func $i64x2.load32x2_s (param $0 i32) (result v128) - (i64x2.load32x2_s - (local.get $0) - ) - ) - (func $i64x2.load32x2_u (param $0 i32) (result v128) - (i64x2.load32x2_u - (local.get $0) - ) - ) - (func $v128.load32_zero (param $0 i32) (result v128) - (v128.load32_zero - (local.get $0) - ) - ) - (func $v128.load64_zero (param $0 i32) (result v128) - (v128.load64_zero - (local.get $0) - ) - ) - (func $v8x16.swizzle (param $0 v128) (param $1 v128) (result v128) - (v8x16.swizzle - (local.get $0) - (local.get $1) - ) - ) - (func $prefetch.t (param $0 i32) - (prefetch.t offset=3 align=2 - (local.get $0) - ) - ) - (func $prefetch.nt (param $0 i32) - (prefetch.nt offset=3 align=2 + (func $i32x4.trunc_sat_f64x2_u_zero (param $0 v128) (result v128) + (i32x4.trunc_sat_f64x2_u_zero (local.get $0) ) ) @@ -1434,35 +1386,5 @@ (local.get $0) ) ) - (func $i32x4.trunc_sat_f64x2_zero_s (param $0 v128) (result v128) - (i32x4.trunc_sat_f64x2_zero_s - (local.get $0) - ) - ) - (func $i32x4.trunc_sat_f64x2_zero_u (param $0 v128) (result v128) - (i32x4.trunc_sat_f64x2_zero_u - (local.get $0) - ) - ) - (func $f32x4.demote_f64x2_zero (param $0 v128) (result v128) - (f32x4.demote_f64x2_zero - (local.get $0) - ) - ) - (func $f64x2.promote_low_f32x4 (param $0 v128) (result v128) - (f64x2.promote_low_f32x4 - (local.get $0) - ) - ) - (func $i32x4.widen_i8x16_s (param $0 v128) (result v128) - (i32x4.widen_i8x16_s 0 - (local.get $0) - ) - ) - (func $i32x4.widen_i8x16_u (param $0 v128) (result v128) - (i32x4.widen_i8x16_u 0 - (local.get $0) - ) - ) ) diff --git a/test/simd.wast.fromBinary.noDebugInfo b/test/simd.wast.fromBinary.noDebugInfo index bb5cdd5e0..78a6576d5 100644 --- a/test/simd.wast.fromBinary.noDebugInfo +++ b/test/simd.wast.fromBinary.noDebugInfo @@ -2,13 +2,11 @@ (type $v128_v128_=>_v128 (func (param v128 v128) (result v128))) (type $v128_=>_v128 (func (param v128) (result v128))) (type $i32_=>_v128 (func (param i32) (result v128))) - (type $v128_=>_i32 (func (param v128) (result i32))) (type $v128_i32_=>_v128 (func (param v128 i32) (result v128))) - (type $v128_v128_v128_=>_v128 (func (param v128 v128 v128) (result v128))) + (type $v128_=>_i32 (func (param v128) (result i32))) (type $i32_v128_=>_none (func (param i32 v128))) (type $i32_v128_=>_v128 (func (param i32 v128) (result v128))) (type $none_=>_v128 (func (result v128))) - (type $i32_=>_none (func (param i32))) (type $v128_=>_i64 (func (param v128) (result i64))) (type $v128_=>_f32 (func (param v128) (result f32))) (type $v128_=>_f64 (func (param v128) (result f64))) @@ -17,1452 +15,1376 @@ (type $v128_i64_=>_v128 (func (param v128 i64) (result v128))) (type $v128_f32_=>_v128 (func (param v128 f32) (result v128))) (type $v128_f64_=>_v128 (func (param v128 f64) (result v128))) + (type $v128_v128_v128_=>_v128 (func (param v128 v128 v128) (result v128))) (memory $0 1 1) (func $0 (param $0 i32) (result v128) (v128.load (local.get $0) ) ) - (func $1 (param $0 i32) (param $1 v128) + (func $1 (param $0 i32) (result v128) + (v128.load8x8_s + (local.get $0) + ) + ) + (func $2 (param $0 i32) (result v128) + (v128.load8x8_u + (local.get $0) + ) + ) + (func $3 (param $0 i32) (result v128) + (v128.load16x4_s + (local.get $0) + ) + ) + (func $4 (param $0 i32) (result v128) + (v128.load16x4_u + (local.get $0) + ) + ) + (func $5 (param $0 i32) (result v128) + (v128.load32x2_s + (local.get $0) + ) + ) + (func $6 (param $0 i32) (result v128) + (v128.load32x2_u + (local.get $0) + ) + ) + (func $7 (param $0 i32) (result v128) + (v128.load8_splat + (local.get $0) + ) + ) + (func $8 (param $0 i32) (result v128) + (v128.load16_splat + (local.get $0) + ) + ) + (func $9 (param $0 i32) (result v128) + (v128.load32_splat + (local.get $0) + ) + ) + (func $10 (param $0 i32) (result v128) + (v128.load64_splat + (local.get $0) + ) + ) + (func $11 (param $0 i32) (param $1 v128) (v128.store (local.get $0) (local.get $1) ) ) - (func $2 (result v128) + (func $12 (result v128) (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d) ) - (func $3 (result v128) + (func $13 (result v128) (v128.const i32x4 0x00020001 0x00040003 0x00060005 0x00080007) ) - (func $4 (result v128) + (func $14 (result v128) (v128.const i32x4 0x00000001 0x00000002 0x00000003 0x00000004) ) - (func $5 (result v128) + (func $15 (result v128) (v128.const i32x4 0x00000001 0x00000000 0x00000002 0x00000000) ) - (func $6 (result v128) + (func $16 (result v128) (v128.const i32x4 0x3f800000 0x40000000 0x40400000 0x40800000) ) - (func $7 (result v128) + (func $17 (result v128) (v128.const i32x4 0x00000000 0x3ff00000 0x00000000 0x40000000) ) - (func $8 (param $0 v128) (param $1 v128) (result v128) - (v8x16.shuffle 0 17 2 19 4 21 6 23 8 25 10 27 12 29 14 31 + (func $18 (param $0 v128) (param $1 v128) (result v128) + (i8x16.shuffle 0 17 2 19 4 21 6 23 8 25 10 27 12 29 14 31 (local.get $0) (local.get $1) ) ) - (func $9 (param $0 i32) (result v128) + (func $19 (param $0 v128) (param $1 v128) (result v128) + (i8x16.swizzle + (local.get $0) + (local.get $1) + ) + ) + (func $20 (param $0 i32) (result v128) (i8x16.splat (local.get $0) ) ) - (func $10 (param $0 v128) (result i32) + (func $21 (param $0 i32) (result v128) + (i16x8.splat + (local.get $0) + ) + ) + (func $22 (param $0 f32) (result v128) + (f32x4.splat + (local.get $0) + ) + ) + (func $23 (param $0 f64) (result v128) + (f64x2.splat + (local.get $0) + ) + ) + (func $24 (param $0 v128) (result i32) (i8x16.extract_lane_s 0 (local.get $0) ) ) - (func $11 (param $0 v128) (result i32) + (func $25 (param $0 v128) (result i32) (i8x16.extract_lane_u 0 (local.get $0) ) ) - (func $12 (param $0 v128) (param $1 i32) (result v128) + (func $26 (param $0 v128) (param $1 i32) (result v128) (i8x16.replace_lane 0 (local.get $0) (local.get $1) ) ) - (func $13 (param $0 i32) (result v128) - (i16x8.splat - (local.get $0) - ) - ) - (func $14 (param $0 v128) (result i32) + (func $27 (param $0 v128) (result i32) (i16x8.extract_lane_s 0 (local.get $0) ) ) - (func $15 (param $0 v128) (result i32) + (func $28 (param $0 v128) (result i32) (i16x8.extract_lane_u 0 (local.get $0) ) ) - (func $16 (param $0 v128) (param $1 i32) (result v128) + (func $29 (param $0 v128) (param $1 i32) (result v128) (i16x8.replace_lane 0 (local.get $0) (local.get $1) ) ) - (func $17 (param $0 i32) (result v128) - (i32x4.splat - (local.get $0) - ) - ) - (func $18 (param $0 v128) (result i32) + (func $30 (param $0 v128) (result i32) (i32x4.extract_lane 0 (local.get $0) ) ) - (func $19 (param $0 v128) (param $1 i32) (result v128) + (func $31 (param $0 v128) (param $1 i32) (result v128) (i32x4.replace_lane 0 (local.get $0) (local.get $1) ) ) - (func $20 (param $0 v128) (result i64) + (func $32 (param $0 v128) (result i64) (i64x2.extract_lane 0 (local.get $0) ) ) - (func $21 (param $0 v128) (param $1 i64) (result v128) + (func $33 (param $0 v128) (param $1 i64) (result v128) (i64x2.replace_lane 0 (local.get $0) (local.get $1) ) ) - (func $22 (param $0 f32) (result v128) - (f32x4.splat - (local.get $0) - ) - ) - (func $23 (param $0 v128) (result f32) + (func $34 (param $0 v128) (result f32) (f32x4.extract_lane 0 (local.get $0) ) ) - (func $24 (param $0 v128) (param $1 f32) (result v128) + (func $35 (param $0 v128) (param $1 f32) (result v128) (f32x4.replace_lane 0 (local.get $0) (local.get $1) ) ) - (func $25 (param $0 f64) (result v128) - (f64x2.splat - (local.get $0) - ) - ) - (func $26 (param $0 v128) (result f64) + (func $36 (param $0 v128) (result f64) (f64x2.extract_lane 0 (local.get $0) ) ) - (func $27 (param $0 v128) (param $1 f64) (result v128) + (func $37 (param $0 v128) (param $1 f64) (result v128) (f64x2.replace_lane 0 (local.get $0) (local.get $1) ) ) - (func $28 (param $0 v128) (param $1 v128) (result v128) + (func $38 (param $0 v128) (param $1 v128) (result v128) (i8x16.eq (local.get $0) (local.get $1) ) ) - (func $29 (param $0 v128) (param $1 v128) (result v128) + (func $39 (param $0 v128) (param $1 v128) (result v128) (i8x16.ne (local.get $0) (local.get $1) ) ) - (func $30 (param $0 v128) (param $1 v128) (result v128) + (func $40 (param $0 v128) (param $1 v128) (result v128) (i8x16.lt_s (local.get $0) (local.get $1) ) ) - (func $31 (param $0 v128) (param $1 v128) (result v128) + (func $41 (param $0 v128) (param $1 v128) (result v128) (i8x16.lt_u (local.get $0) (local.get $1) ) ) - (func $32 (param $0 v128) (param $1 v128) (result v128) + (func $42 (param $0 v128) (param $1 v128) (result v128) (i8x16.gt_s (local.get $0) (local.get $1) ) ) - (func $33 (param $0 v128) (param $1 v128) (result v128) + (func $43 (param $0 v128) (param $1 v128) (result v128) (i8x16.gt_u (local.get $0) (local.get $1) ) ) - (func $34 (param $0 v128) (param $1 v128) (result v128) + (func $44 (param $0 v128) (param $1 v128) (result v128) (i8x16.le_s (local.get $0) (local.get $1) ) ) - (func $35 (param $0 v128) (param $1 v128) (result v128) + (func $45 (param $0 v128) (param $1 v128) (result v128) (i8x16.le_u (local.get $0) (local.get $1) ) ) - (func $36 (param $0 v128) (param $1 v128) (result v128) + (func $46 (param $0 v128) (param $1 v128) (result v128) (i8x16.ge_s (local.get $0) (local.get $1) ) ) - (func $37 (param $0 v128) (param $1 v128) (result v128) + (func $47 (param $0 v128) (param $1 v128) (result v128) (i8x16.ge_u (local.get $0) (local.get $1) ) ) - (func $38 (param $0 v128) (param $1 v128) (result v128) + (func $48 (param $0 v128) (param $1 v128) (result v128) (i16x8.eq (local.get $0) (local.get $1) ) ) - (func $39 (param $0 v128) (param $1 v128) (result v128) + (func $49 (param $0 v128) (param $1 v128) (result v128) (i16x8.ne (local.get $0) (local.get $1) ) ) - (func $40 (param $0 v128) (param $1 v128) (result v128) + (func $50 (param $0 v128) (param $1 v128) (result v128) (i16x8.lt_s (local.get $0) (local.get $1) ) ) - (func $41 (param $0 v128) (param $1 v128) (result v128) + (func $51 (param $0 v128) (param $1 v128) (result v128) (i16x8.lt_u (local.get $0) (local.get $1) ) ) - (func $42 (param $0 v128) (param $1 v128) (result v128) + (func $52 (param $0 v128) (param $1 v128) (result v128) (i16x8.gt_s (local.get $0) (local.get $1) ) ) - (func $43 (param $0 v128) (param $1 v128) (result v128) + (func $53 (param $0 v128) (param $1 v128) (result v128) (i16x8.gt_u (local.get $0) (local.get $1) ) ) - (func $44 (param $0 v128) (param $1 v128) (result v128) + (func $54 (param $0 v128) (param $1 v128) (result v128) (i16x8.le_s (local.get $0) (local.get $1) ) ) - (func $45 (param $0 v128) (param $1 v128) (result v128) + (func $55 (param $0 v128) (param $1 v128) (result v128) (i16x8.le_u (local.get $0) (local.get $1) ) ) - (func $46 (param $0 v128) (param $1 v128) (result v128) + (func $56 (param $0 v128) (param $1 v128) (result v128) (i16x8.ge_s (local.get $0) (local.get $1) ) ) - (func $47 (param $0 v128) (param $1 v128) (result v128) + (func $57 (param $0 v128) (param $1 v128) (result v128) (i16x8.ge_u (local.get $0) (local.get $1) ) ) - (func $48 (param $0 v128) (param $1 v128) (result v128) + (func $58 (param $0 v128) (param $1 v128) (result v128) (i32x4.eq (local.get $0) (local.get $1) ) ) - (func $49 (param $0 v128) (param $1 v128) (result v128) + (func $59 (param $0 v128) (param $1 v128) (result v128) (i32x4.ne (local.get $0) (local.get $1) ) ) - (func $50 (param $0 v128) (param $1 v128) (result v128) + (func $60 (param $0 v128) (param $1 v128) (result v128) (i32x4.lt_s (local.get $0) (local.get $1) ) ) - (func $51 (param $0 v128) (param $1 v128) (result v128) + (func $61 (param $0 v128) (param $1 v128) (result v128) (i32x4.lt_u (local.get $0) (local.get $1) ) ) - (func $52 (param $0 v128) (param $1 v128) (result v128) + (func $62 (param $0 v128) (param $1 v128) (result v128) (i32x4.gt_s (local.get $0) (local.get $1) ) ) - (func $53 (param $0 v128) (param $1 v128) (result v128) + (func $63 (param $0 v128) (param $1 v128) (result v128) (i32x4.gt_u (local.get $0) (local.get $1) ) ) - (func $54 (param $0 v128) (param $1 v128) (result v128) + (func $64 (param $0 v128) (param $1 v128) (result v128) (i32x4.le_s (local.get $0) (local.get $1) ) ) - (func $55 (param $0 v128) (param $1 v128) (result v128) + (func $65 (param $0 v128) (param $1 v128) (result v128) (i32x4.le_u (local.get $0) (local.get $1) ) ) - (func $56 (param $0 v128) (param $1 v128) (result v128) + (func $66 (param $0 v128) (param $1 v128) (result v128) (i32x4.ge_s (local.get $0) (local.get $1) ) ) - (func $57 (param $0 v128) (param $1 v128) (result v128) + (func $67 (param $0 v128) (param $1 v128) (result v128) (i32x4.ge_u (local.get $0) (local.get $1) ) ) - (func $58 (param $0 v128) (param $1 v128) (result v128) - (i64x2.eq - (local.get $0) - (local.get $1) - ) - ) - (func $59 (param $0 v128) (param $1 v128) (result v128) + (func $68 (param $0 v128) (param $1 v128) (result v128) (f32x4.eq (local.get $0) (local.get $1) ) ) - (func $60 (param $0 v128) (param $1 v128) (result v128) + (func $69 (param $0 v128) (param $1 v128) (result v128) (f32x4.ne (local.get $0) (local.get $1) ) ) - (func $61 (param $0 v128) (param $1 v128) (result v128) + (func $70 (param $0 v128) (param $1 v128) (result v128) (f32x4.lt (local.get $0) (local.get $1) ) ) - (func $62 (param $0 v128) (param $1 v128) (result v128) + (func $71 (param $0 v128) (param $1 v128) (result v128) (f32x4.gt (local.get $0) (local.get $1) ) ) - (func $63 (param $0 v128) (param $1 v128) (result v128) + (func $72 (param $0 v128) (param $1 v128) (result v128) (f32x4.le (local.get $0) (local.get $1) ) ) - (func $64 (param $0 v128) (param $1 v128) (result v128) + (func $73 (param $0 v128) (param $1 v128) (result v128) (f32x4.ge (local.get $0) (local.get $1) ) ) - (func $65 (param $0 v128) (param $1 v128) (result v128) + (func $74 (param $0 v128) (param $1 v128) (result v128) (f64x2.eq (local.get $0) (local.get $1) ) ) - (func $66 (param $0 v128) (param $1 v128) (result v128) + (func $75 (param $0 v128) (param $1 v128) (result v128) (f64x2.ne (local.get $0) (local.get $1) ) ) - (func $67 (param $0 v128) (param $1 v128) (result v128) + (func $76 (param $0 v128) (param $1 v128) (result v128) (f64x2.lt (local.get $0) (local.get $1) ) ) - (func $68 (param $0 v128) (param $1 v128) (result v128) + (func $77 (param $0 v128) (param $1 v128) (result v128) (f64x2.gt (local.get $0) (local.get $1) ) ) - (func $69 (param $0 v128) (param $1 v128) (result v128) + (func $78 (param $0 v128) (param $1 v128) (result v128) (f64x2.le (local.get $0) (local.get $1) ) ) - (func $70 (param $0 v128) (param $1 v128) (result v128) + (func $79 (param $0 v128) (param $1 v128) (result v128) (f64x2.ge (local.get $0) (local.get $1) ) ) - (func $71 (param $0 v128) (result v128) + (func $80 (param $0 v128) (result v128) (v128.not (local.get $0) ) ) - (func $72 (param $0 v128) (param $1 v128) (result v128) + (func $81 (param $0 v128) (param $1 v128) (result v128) (v128.and (local.get $0) (local.get $1) ) ) - (func $73 (param $0 v128) (param $1 v128) (result v128) - (v128.or - (local.get $0) - (local.get $1) - ) - ) - (func $74 (param $0 v128) (param $1 v128) (result v128) - (v128.xor - (local.get $0) - (local.get $1) - ) - ) - (func $75 (param $0 v128) (param $1 v128) (result v128) + (func $82 (param $0 v128) (param $1 v128) (result v128) (v128.andnot (local.get $0) (local.get $1) ) ) - (func $76 (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (v128.bitselect - (local.get $0) - (local.get $1) - (local.get $2) - ) - ) - (func $77 (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (v8x16.signselect + (func $83 (param $0 v128) (param $1 v128) (result v128) + (v128.or (local.get $0) (local.get $1) - (local.get $2) ) ) - (func $78 (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (v16x8.signselect + (func $84 (param $0 v128) (param $1 v128) (result v128) + (v128.xor (local.get $0) (local.get $1) - (local.get $2) ) ) - (func $79 (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (v32x4.signselect + (func $85 (param $0 v128) (param $1 v128) (param $2 v128) (result v128) + (v128.bitselect (local.get $0) (local.get $1) (local.get $2) ) ) - (func $80 (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (v64x2.signselect + (func $86 (param $0 v128) (result i32) + (v128.any_true (local.get $0) - (local.get $1) - (local.get $2) ) ) - (func $81 (param $0 i32) (param $1 v128) (result v128) + (func $87 (param $0 i32) (param $1 v128) (result v128) (v128.load8_lane 0 (local.get $0) (local.get $1) ) ) - (func $82 (param $0 i32) (param $1 v128) (result v128) + (func $88 (param $0 i32) (param $1 v128) (result v128) (v128.load16_lane 0 (local.get $0) (local.get $1) ) ) - (func $83 (param $0 i32) (param $1 v128) (result v128) + (func $89 (param $0 i32) (param $1 v128) (result v128) (v128.load32_lane 0 (local.get $0) (local.get $1) ) ) - (func $84 (param $0 i32) (param $1 v128) (result v128) + (func $90 (param $0 i32) (param $1 v128) (result v128) (v128.load64_lane 0 (local.get $0) (local.get $1) ) ) - (func $85 (param $0 i32) (param $1 v128) (result v128) + (func $91 (param $0 i32) (param $1 v128) (result v128) (v128.load64_lane align=1 0 (local.get $0) (local.get $1) ) ) - (func $86 (param $0 i32) (param $1 v128) (result v128) + (func $92 (param $0 i32) (param $1 v128) (result v128) (v128.load64_lane offset=32 0 (local.get $0) (local.get $1) ) ) - (func $87 (param $0 i32) (param $1 v128) (result v128) + (func $93 (param $0 i32) (param $1 v128) (result v128) (v128.load64_lane offset=32 align=1 0 (local.get $0) (local.get $1) ) ) - (func $88 (param $0 i32) (param $1 v128) + (func $94 (param $0 i32) (param $1 v128) (v128.store8_lane 0 (local.get $0) (local.get $1) ) ) - (func $89 (param $0 i32) (param $1 v128) + (func $95 (param $0 i32) (param $1 v128) (v128.store16_lane 0 (local.get $0) (local.get $1) ) ) - (func $90 (param $0 i32) (param $1 v128) + (func $96 (param $0 i32) (param $1 v128) (v128.store32_lane 0 (local.get $0) (local.get $1) ) ) - (func $91 (param $0 i32) (param $1 v128) + (func $97 (param $0 i32) (param $1 v128) (v128.store64_lane 0 (local.get $0) (local.get $1) ) ) - (func $92 (param $0 i32) (param $1 v128) + (func $98 (param $0 i32) (param $1 v128) (v128.store64_lane align=1 0 (local.get $0) (local.get $1) ) ) - (func $93 (param $0 i32) (param $1 v128) + (func $99 (param $0 i32) (param $1 v128) (v128.store64_lane offset=32 0 (local.get $0) (local.get $1) ) ) - (func $94 (param $0 i32) (param $1 v128) + (func $100 (param $0 i32) (param $1 v128) (v128.store64_lane offset=32 align=1 0 (local.get $0) (local.get $1) ) ) - (func $95 (param $0 v128) (result v128) - (i8x16.popcnt - (local.get $0) - ) - ) - (func $96 (param $0 v128) (result v128) - (i8x16.abs + (func $101 (param $0 i32) (result v128) + (v128.load32_zero (local.get $0) ) ) - (func $97 (param $0 v128) (result v128) - (i8x16.neg + (func $102 (param $0 i32) (result v128) + (v128.load64_zero (local.get $0) ) ) - (func $98 (param $0 v128) (result i32) - (i8x16.any_true + (func $103 (param $0 v128) (result v128) + (f32x4.demote_f64x2_zero (local.get $0) ) ) - (func $99 (param $0 v128) (result i32) - (i8x16.all_true + (func $104 (param $0 v128) (result v128) + (f64x2.promote_low_f32x4 (local.get $0) ) ) - (func $100 (param $0 v128) (result i32) - (i8x16.bitmask + (func $105 (param $0 v128) (result v128) + (i8x16.abs (local.get $0) ) ) - (func $101 (param $0 v128) (param $1 i32) (result v128) - (i8x16.shl + (func $106 (param $0 v128) (result v128) + (i8x16.neg (local.get $0) - (local.get $1) ) ) - (func $102 (param $0 v128) (param $1 i32) (result v128) - (i8x16.shr_s + (func $107 (param $0 v128) (result v128) + (i8x16.popcnt (local.get $0) - (local.get $1) ) ) - (func $103 (param $0 v128) (param $1 i32) (result v128) - (i8x16.shr_u + (func $108 (param $0 v128) (result i32) + (i8x16.all_true (local.get $0) - (local.get $1) ) ) - (func $104 (param $0 v128) (param $1 v128) (result v128) - (i8x16.add + (func $109 (param $0 v128) (result i32) + (i8x16.bitmask (local.get $0) - (local.get $1) ) ) - (func $105 (param $0 v128) (param $1 v128) (result v128) - (i8x16.add_saturate_s + (func $110 (param $0 v128) (param $1 v128) (result v128) + (i8x16.narrow_i16x8_s (local.get $0) (local.get $1) ) ) - (func $106 (param $0 v128) (param $1 v128) (result v128) - (i8x16.add_saturate_u + (func $111 (param $0 v128) (param $1 v128) (result v128) + (i8x16.narrow_i16x8_u (local.get $0) (local.get $1) ) ) - (func $107 (param $0 v128) (param $1 v128) (result v128) - (i8x16.sub + (func $112 (param $0 v128) (result v128) + (f32x4.ceil (local.get $0) - (local.get $1) ) ) - (func $108 (param $0 v128) (param $1 v128) (result v128) - (i8x16.sub_saturate_s + (func $113 (param $0 v128) (result v128) + (f32x4.floor (local.get $0) - (local.get $1) ) ) - (func $109 (param $0 v128) (param $1 v128) (result v128) - (i8x16.sub_saturate_u + (func $114 (param $0 v128) (result v128) + (f32x4.trunc (local.get $0) - (local.get $1) ) ) - (func $110 (param $0 v128) (param $1 v128) (result v128) - (i8x16.mul + (func $115 (param $0 v128) (result v128) + (f32x4.nearest (local.get $0) - (local.get $1) ) ) - (func $111 (param $0 v128) (param $1 v128) (result v128) - (i8x16.min_s + (func $116 (param $0 v128) (param $1 i32) (result v128) + (i8x16.shl (local.get $0) (local.get $1) ) ) - (func $112 (param $0 v128) (param $1 v128) (result v128) - (i8x16.min_u + (func $117 (param $0 v128) (param $1 i32) (result v128) + (i8x16.shr_s (local.get $0) (local.get $1) ) ) - (func $113 (param $0 v128) (param $1 v128) (result v128) - (i8x16.max_s + (func $118 (param $0 v128) (param $1 i32) (result v128) + (i8x16.shr_u (local.get $0) (local.get $1) ) ) - (func $114 (param $0 v128) (param $1 v128) (result v128) - (i8x16.max_u + (func $119 (param $0 v128) (param $1 v128) (result v128) + (i8x16.add (local.get $0) (local.get $1) ) ) - (func $115 (param $0 v128) (param $1 v128) (result v128) - (i8x16.avgr_u + (func $120 (param $0 v128) (param $1 v128) (result v128) + (i8x16.add_sat_s (local.get $0) (local.get $1) ) ) - (func $116 (param $0 v128) (result v128) - (i16x8.abs - (local.get $0) - ) - ) - (func $117 (param $0 v128) (result v128) - (i16x8.neg - (local.get $0) - ) - ) - (func $118 (param $0 v128) (result i32) - (i16x8.any_true - (local.get $0) - ) - ) - (func $119 (param $0 v128) (result i32) - (i16x8.all_true - (local.get $0) - ) - ) - (func $120 (param $0 v128) (result i32) - (i16x8.bitmask - (local.get $0) - ) - ) - (func $121 (param $0 v128) (param $1 i32) (result v128) - (i16x8.shl + (func $121 (param $0 v128) (param $1 v128) (result v128) + (i8x16.add_sat_u (local.get $0) (local.get $1) ) ) - (func $122 (param $0 v128) (param $1 i32) (result v128) - (i16x8.shr_s + (func $122 (param $0 v128) (param $1 v128) (result v128) + (i8x16.sub (local.get $0) (local.get $1) ) ) - (func $123 (param $0 v128) (param $1 i32) (result v128) - (i16x8.shr_u + (func $123 (param $0 v128) (param $1 v128) (result v128) + (i8x16.sub_sat_s (local.get $0) (local.get $1) ) ) (func $124 (param $0 v128) (param $1 v128) (result v128) - (i16x8.add + (i8x16.sub_sat_u (local.get $0) (local.get $1) ) ) - (func $125 (param $0 v128) (param $1 v128) (result v128) - (i16x8.add_saturate_s + (func $125 (param $0 v128) (result v128) + (f64x2.ceil (local.get $0) - (local.get $1) ) ) - (func $126 (param $0 v128) (param $1 v128) (result v128) - (i16x8.add_saturate_u + (func $126 (param $0 v128) (result v128) + (f64x2.floor (local.get $0) - (local.get $1) ) ) (func $127 (param $0 v128) (param $1 v128) (result v128) - (i16x8.sub + (i8x16.min_s (local.get $0) (local.get $1) ) ) (func $128 (param $0 v128) (param $1 v128) (result v128) - (i16x8.sub_saturate_s + (i8x16.min_u (local.get $0) (local.get $1) ) ) (func $129 (param $0 v128) (param $1 v128) (result v128) - (i16x8.sub_saturate_u + (i8x16.max_s (local.get $0) (local.get $1) ) ) (func $130 (param $0 v128) (param $1 v128) (result v128) - (i16x8.mul + (i8x16.max_u (local.get $0) (local.get $1) ) ) - (func $131 (param $0 v128) (param $1 v128) (result v128) - (i16x8.min_s + (func $131 (param $0 v128) (result v128) + (f64x2.trunc (local.get $0) - (local.get $1) ) ) (func $132 (param $0 v128) (param $1 v128) (result v128) - (i16x8.min_u + (i8x16.avgr_u (local.get $0) (local.get $1) ) ) - (func $133 (param $0 v128) (param $1 v128) (result v128) - (i16x8.max_s + (func $133 (param $0 v128) (result v128) + (i16x8.extadd_pairwise_i8x16_s (local.get $0) - (local.get $1) ) ) - (func $134 (param $0 v128) (param $1 v128) (result v128) - (i16x8.max_u + (func $134 (param $0 v128) (result v128) + (i16x8.extadd_pairwise_i8x16_u (local.get $0) - (local.get $1) ) ) - (func $135 (param $0 v128) (param $1 v128) (result v128) - (i16x8.avgr_u + (func $135 (param $0 v128) (result v128) + (i32x4.extadd_pairwise_i16x8_s (local.get $0) - (local.get $1) ) ) - (func $136 (param $0 v128) (param $1 v128) (result v128) - (i16x8.q15mulr_sat_s + (func $136 (param $0 v128) (result v128) + (i32x4.extadd_pairwise_i16x8_u (local.get $0) - (local.get $1) ) ) - (func $137 (param $0 v128) (param $1 v128) (result v128) - (i16x8.extmul_low_i8x16_s + (func $137 (param $0 v128) (result v128) + (i16x8.abs (local.get $0) - (local.get $1) ) ) - (func $138 (param $0 v128) (param $1 v128) (result v128) - (i16x8.extmul_high_i8x16_s + (func $138 (param $0 v128) (result v128) + (i16x8.neg (local.get $0) - (local.get $1) ) ) (func $139 (param $0 v128) (param $1 v128) (result v128) - (i16x8.extmul_low_i8x16_u + (i16x8.q15mulr_sat_s (local.get $0) (local.get $1) ) ) - (func $140 (param $0 v128) (param $1 v128) (result v128) - (i16x8.extmul_high_i8x16_u + (func $140 (param $0 v128) (result i32) + (i16x8.all_true (local.get $0) - (local.get $1) ) ) - (func $141 (param $0 v128) (result v128) - (i32x4.abs + (func $141 (param $0 v128) (result i32) + (i16x8.bitmask (local.get $0) ) ) - (func $142 (param $0 v128) (result v128) - (i32x4.neg + (func $142 (param $0 v128) (param $1 v128) (result v128) + (i16x8.narrow_i32x4_s (local.get $0) + (local.get $1) ) ) - (func $143 (param $0 v128) (result i32) - (i32x4.any_true + (func $143 (param $0 v128) (param $1 v128) (result v128) + (i16x8.narrow_i32x4_u (local.get $0) + (local.get $1) ) ) - (func $144 (param $0 v128) (result i32) - (i32x4.all_true + (func $144 (param $0 v128) (result v128) + (i16x8.extend_low_i8x16_s (local.get $0) ) ) - (func $145 (param $0 v128) (result i32) - (i32x4.bitmask + (func $145 (param $0 v128) (result v128) + (i16x8.extend_high_i8x16_s (local.get $0) ) ) - (func $146 (param $0 v128) (param $1 i32) (result v128) - (i32x4.shl + (func $146 (param $0 v128) (result v128) + (i16x8.extend_low_i8x16_u (local.get $0) - (local.get $1) ) ) - (func $147 (param $0 v128) (param $1 i32) (result v128) - (i32x4.shr_s + (func $147 (param $0 v128) (result v128) + (i16x8.extend_high_i8x16_u (local.get $0) - (local.get $1) ) ) (func $148 (param $0 v128) (param $1 i32) (result v128) - (i32x4.shr_u + (i16x8.shl (local.get $0) (local.get $1) ) ) - (func $149 (param $0 v128) (param $1 v128) (result v128) - (i32x4.add + (func $149 (param $0 v128) (param $1 i32) (result v128) + (i16x8.shr_s (local.get $0) (local.get $1) ) ) - (func $150 (param $0 v128) (param $1 v128) (result v128) - (i32x4.sub + (func $150 (param $0 v128) (param $1 i32) (result v128) + (i16x8.shr_u (local.get $0) (local.get $1) ) ) (func $151 (param $0 v128) (param $1 v128) (result v128) - (i32x4.mul + (i16x8.add (local.get $0) (local.get $1) ) ) (func $152 (param $0 v128) (param $1 v128) (result v128) - (i32x4.min_s + (i16x8.add_sat_s (local.get $0) (local.get $1) ) ) (func $153 (param $0 v128) (param $1 v128) (result v128) - (i32x4.min_u + (i16x8.add_sat_u (local.get $0) (local.get $1) ) ) (func $154 (param $0 v128) (param $1 v128) (result v128) - (i32x4.max_s + (i16x8.sub (local.get $0) (local.get $1) ) ) (func $155 (param $0 v128) (param $1 v128) (result v128) - (i32x4.max_u + (i16x8.sub_sat_s (local.get $0) (local.get $1) ) ) (func $156 (param $0 v128) (param $1 v128) (result v128) - (i32x4.dot_i16x8_s + (i16x8.sub_sat_u (local.get $0) (local.get $1) ) ) - (func $157 (param $0 v128) (param $1 v128) (result v128) - (i32x4.extmul_low_i16x8_s + (func $157 (param $0 v128) (result v128) + (f64x2.nearest (local.get $0) - (local.get $1) ) ) (func $158 (param $0 v128) (param $1 v128) (result v128) - (i32x4.extmul_high_i16x8_s + (i16x8.mul (local.get $0) (local.get $1) ) ) (func $159 (param $0 v128) (param $1 v128) (result v128) - (i32x4.extmul_low_i16x8_u + (i16x8.min_s (local.get $0) (local.get $1) ) ) (func $160 (param $0 v128) (param $1 v128) (result v128) - (i32x4.extmul_high_i16x8_u + (i16x8.min_u (local.get $0) (local.get $1) ) ) - (func $161 (param $0 v128) (result v128) - (i64x2.neg + (func $161 (param $0 v128) (param $1 v128) (result v128) + (i16x8.max_s (local.get $0) + (local.get $1) ) ) - (func $162 (param $0 v128) (result i32) - (i64x2.bitmask + (func $162 (param $0 v128) (param $1 v128) (result v128) + (i16x8.max_u (local.get $0) + (local.get $1) ) ) - (func $163 (param $0 v128) (param $1 i32) (result v128) - (i64x2.shl + (func $163 (param $0 v128) (param $1 v128) (result v128) + (i16x8.avgr_u (local.get $0) (local.get $1) ) ) - (func $164 (param $0 v128) (param $1 i32) (result v128) - (i64x2.shr_s + (func $164 (param $0 v128) (param $1 v128) (result v128) + (i16x8.extmul_low_i8x16_s (local.get $0) (local.get $1) ) ) - (func $165 (param $0 v128) (param $1 i32) (result v128) - (i64x2.shr_u + (func $165 (param $0 v128) (param $1 v128) (result v128) + (i16x8.extmul_high_i8x16_s (local.get $0) (local.get $1) ) ) (func $166 (param $0 v128) (param $1 v128) (result v128) - (i64x2.add + (i16x8.extmul_low_i8x16_u (local.get $0) (local.get $1) ) ) (func $167 (param $0 v128) (param $1 v128) (result v128) - (i64x2.sub + (i16x8.extmul_high_i8x16_u (local.get $0) (local.get $1) ) ) - (func $168 (param $0 v128) (param $1 v128) (result v128) - (i64x2.mul + (func $168 (param $0 v128) (result v128) + (i32x4.abs (local.get $0) - (local.get $1) ) ) - (func $169 (param $0 v128) (param $1 v128) (result v128) - (i64x2.extmul_low_i32x4_s + (func $169 (param $0 v128) (result v128) + (i32x4.neg (local.get $0) - (local.get $1) ) ) - (func $170 (param $0 v128) (param $1 v128) (result v128) - (i64x2.extmul_high_i32x4_s + (func $170 (param $0 v128) (result i32) + (i32x4.all_true (local.get $0) - (local.get $1) ) ) - (func $171 (param $0 v128) (param $1 v128) (result v128) - (i64x2.extmul_low_i32x4_u + (func $171 (param $0 v128) (result i32) + (i32x4.bitmask (local.get $0) - (local.get $1) ) ) - (func $172 (param $0 v128) (param $1 v128) (result v128) - (i64x2.extmul_high_i32x4_u + (func $172 (param $0 v128) (result v128) + (i32x4.extend_low_i16x8_s (local.get $0) - (local.get $1) ) ) - (func $173 (param $0 v128) (param $1 v128) (result v128) - (f32x4.add + (func $173 (param $0 v128) (result v128) + (i32x4.extend_high_i16x8_s (local.get $0) - (local.get $1) ) ) - (func $174 (param $0 v128) (param $1 v128) (result v128) - (f32x4.sub + (func $174 (param $0 v128) (result v128) + (i32x4.extend_low_i16x8_u (local.get $0) - (local.get $1) ) ) - (func $175 (param $0 v128) (param $1 v128) (result v128) - (f32x4.mul + (func $175 (param $0 v128) (result v128) + (i32x4.extend_high_i16x8_u (local.get $0) - (local.get $1) ) ) - (func $176 (param $0 v128) (param $1 v128) (result v128) - (f32x4.div + (func $176 (param $0 v128) (param $1 i32) (result v128) + (i32x4.shl (local.get $0) (local.get $1) ) ) - (func $177 (param $0 v128) (param $1 v128) (result v128) - (f32x4.min + (func $177 (param $0 v128) (param $1 i32) (result v128) + (i32x4.shr_s (local.get $0) (local.get $1) ) ) - (func $178 (param $0 v128) (param $1 v128) (result v128) - (f32x4.max + (func $178 (param $0 v128) (param $1 i32) (result v128) + (i32x4.shr_u (local.get $0) (local.get $1) ) ) (func $179 (param $0 v128) (param $1 v128) (result v128) - (f32x4.pmin + (i32x4.add (local.get $0) (local.get $1) ) ) (func $180 (param $0 v128) (param $1 v128) (result v128) - (f32x4.pmax + (i32x4.sub (local.get $0) (local.get $1) ) ) - (func $181 (param $0 v128) (result v128) - (f32x4.ceil + (func $181 (param $0 v128) (param $1 v128) (result v128) + (i32x4.mul (local.get $0) + (local.get $1) ) ) - (func $182 (param $0 v128) (result v128) - (f32x4.floor + (func $182 (param $0 v128) (param $1 v128) (result v128) + (i32x4.min_s (local.get $0) + (local.get $1) ) ) - (func $183 (param $0 v128) (result v128) - (f32x4.trunc + (func $183 (param $0 v128) (param $1 v128) (result v128) + (i32x4.min_u (local.get $0) + (local.get $1) ) ) - (func $184 (param $0 v128) (result v128) - (f32x4.nearest + (func $184 (param $0 v128) (param $1 v128) (result v128) + (i32x4.max_s (local.get $0) + (local.get $1) ) ) - (func $185 (param $0 v128) (result v128) - (f32x4.abs + (func $185 (param $0 v128) (param $1 v128) (result v128) + (i32x4.max_u (local.get $0) + (local.get $1) ) ) - (func $186 (param $0 v128) (result v128) - (f32x4.neg + (func $186 (param $0 v128) (param $1 v128) (result v128) + (i32x4.dot_i16x8_s (local.get $0) + (local.get $1) ) ) - (func $187 (param $0 v128) (result v128) - (f32x4.sqrt + (func $187 (param $0 v128) (param $1 v128) (result v128) + (i32x4.extmul_low_i16x8_s (local.get $0) + (local.get $1) ) ) - (func $188 (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (f32x4.qfma + (func $188 (param $0 v128) (param $1 v128) (result v128) + (i32x4.extmul_high_i16x8_s (local.get $0) (local.get $1) - (local.get $2) ) ) - (func $189 (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (f32x4.qfms + (func $189 (param $0 v128) (param $1 v128) (result v128) + (i32x4.extmul_low_i16x8_u (local.get $0) (local.get $1) - (local.get $2) ) ) (func $190 (param $0 v128) (param $1 v128) (result v128) - (f64x2.add + (i32x4.extmul_high_i16x8_u (local.get $0) (local.get $1) ) ) - (func $191 (param $0 v128) (param $1 v128) (result v128) - (f64x2.sub + (func $191 (param $0 v128) (result v128) + (i64x2.abs (local.get $0) - (local.get $1) ) ) - (func $192 (param $0 v128) (param $1 v128) (result v128) - (f64x2.mul + (func $192 (param $0 v128) (result v128) + (i64x2.neg (local.get $0) - (local.get $1) ) ) - (func $193 (param $0 v128) (param $1 v128) (result v128) - (f64x2.div + (func $193 (param $0 v128) (result i32) + (i64x2.all_true (local.get $0) - (local.get $1) ) ) - (func $194 (param $0 v128) (param $1 v128) (result v128) - (f64x2.min + (func $194 (param $0 v128) (result i32) + (i64x2.bitmask (local.get $0) - (local.get $1) ) ) - (func $195 (param $0 v128) (param $1 v128) (result v128) - (f64x2.max + (func $195 (param $0 v128) (result v128) + (i64x2.extend_low_i32x4_s (local.get $0) - (local.get $1) ) ) - (func $196 (param $0 v128) (param $1 v128) (result v128) - (f64x2.pmin + (func $196 (param $0 v128) (result v128) + (i64x2.extend_high_i32x4_s (local.get $0) - (local.get $1) ) ) - (func $197 (param $0 v128) (param $1 v128) (result v128) - (f64x2.pmax + (func $197 (param $0 v128) (result v128) + (i64x2.extend_low_i32x4_u (local.get $0) - (local.get $1) ) ) (func $198 (param $0 v128) (result v128) - (f64x2.ceil + (i64x2.extend_high_i32x4_u (local.get $0) ) ) - (func $199 (param $0 v128) (result v128) - (f64x2.floor + (func $199 (param $0 v128) (param $1 i32) (result v128) + (i64x2.shl (local.get $0) + (local.get $1) ) ) - (func $200 (param $0 v128) (result v128) - (f64x2.trunc + (func $200 (param $0 v128) (param $1 i32) (result v128) + (i64x2.shr_s (local.get $0) + (local.get $1) ) ) - (func $201 (param $0 v128) (result v128) - (f64x2.nearest + (func $201 (param $0 v128) (param $1 i32) (result v128) + (i64x2.shr_u (local.get $0) + (local.get $1) ) ) - (func $202 (param $0 v128) (result v128) - (f64x2.abs + (func $202 (param $0 v128) (param $1 v128) (result v128) + (i64x2.add (local.get $0) + (local.get $1) ) ) - (func $203 (param $0 v128) (result v128) - (f64x2.neg + (func $203 (param $0 v128) (param $1 v128) (result v128) + (i64x2.sub (local.get $0) + (local.get $1) ) ) - (func $204 (param $0 v128) (result v128) - (f64x2.sqrt + (func $204 (param $0 v128) (param $1 v128) (result v128) + (i64x2.mul (local.get $0) + (local.get $1) ) ) - (func $205 (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (f64x2.qfma + (func $205 (param $0 v128) (param $1 v128) (result v128) + (i64x2.eq (local.get $0) (local.get $1) - (local.get $2) ) ) - (func $206 (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (f64x2.qfms + (func $206 (param $0 v128) (param $1 v128) (result v128) + (i64x2.ne (local.get $0) (local.get $1) - (local.get $2) ) ) - (func $207 (param $0 v128) (result v128) - (i16x8.extadd_pairwise_i8x16_s + (func $207 (param $0 v128) (param $1 v128) (result v128) + (i64x2.lt_s (local.get $0) + (local.get $1) ) ) - (func $208 (param $0 v128) (result v128) - (i16x8.extadd_pairwise_i8x16_u + (func $208 (param $0 v128) (param $1 v128) (result v128) + (i64x2.gt_s (local.get $0) + (local.get $1) ) ) - (func $209 (param $0 v128) (result v128) - (i32x4.extadd_pairwise_i16x8_s + (func $209 (param $0 v128) (param $1 v128) (result v128) + (i64x2.le_s (local.get $0) + (local.get $1) ) ) - (func $210 (param $0 v128) (result v128) - (i32x4.extadd_pairwise_i16x8_u + (func $210 (param $0 v128) (param $1 v128) (result v128) + (i64x2.ge_s (local.get $0) + (local.get $1) ) ) - (func $211 (param $0 v128) (result v128) - (i32x4.trunc_sat_f32x4_s + (func $211 (param $0 v128) (param $1 v128) (result v128) + (i64x2.extmul_low_i32x4_s (local.get $0) + (local.get $1) ) ) - (func $212 (param $0 v128) (result v128) - (i32x4.trunc_sat_f32x4_u + (func $212 (param $0 v128) (param $1 v128) (result v128) + (i64x2.extmul_high_i32x4_s (local.get $0) + (local.get $1) ) ) - (func $213 (param $0 v128) (result v128) - (i64x2.trunc_sat_f64x2_s + (func $213 (param $0 v128) (param $1 v128) (result v128) + (i64x2.extmul_low_i32x4_u (local.get $0) + (local.get $1) ) ) - (func $214 (param $0 v128) (result v128) - (i64x2.trunc_sat_f64x2_u + (func $214 (param $0 v128) (param $1 v128) (result v128) + (i64x2.extmul_high_i32x4_u (local.get $0) + (local.get $1) ) ) (func $215 (param $0 v128) (result v128) - (f32x4.convert_i32x4_s + (f32x4.abs (local.get $0) ) ) (func $216 (param $0 v128) (result v128) - (f32x4.convert_i32x4_u + (f32x4.neg (local.get $0) ) ) (func $217 (param $0 v128) (result v128) - (f64x2.convert_i64x2_s + (f32x4.sqrt (local.get $0) ) ) - (func $218 (param $0 v128) (result v128) - (f64x2.convert_i64x2_u + (func $218 (param $0 v128) (param $1 v128) (result v128) + (f32x4.add (local.get $0) + (local.get $1) ) ) - (func $219 (param $0 i32) (result v128) - (v8x16.load_splat + (func $219 (param $0 v128) (param $1 v128) (result v128) + (f32x4.sub (local.get $0) + (local.get $1) ) ) - (func $220 (param $0 i32) (result v128) - (v16x8.load_splat + (func $220 (param $0 v128) (param $1 v128) (result v128) + (f32x4.mul (local.get $0) + (local.get $1) ) ) - (func $221 (param $0 i32) (result v128) - (v32x4.load_splat + (func $221 (param $0 v128) (param $1 v128) (result v128) + (f32x4.div (local.get $0) + (local.get $1) ) ) - (func $222 (param $0 i32) (result v128) - (v64x2.load_splat + (func $222 (param $0 v128) (param $1 v128) (result v128) + (f32x4.min (local.get $0) + (local.get $1) ) ) (func $223 (param $0 v128) (param $1 v128) (result v128) - (i8x16.narrow_i16x8_s + (f32x4.max (local.get $0) (local.get $1) ) ) (func $224 (param $0 v128) (param $1 v128) (result v128) - (i8x16.narrow_i16x8_u + (f32x4.pmin (local.get $0) (local.get $1) ) ) (func $225 (param $0 v128) (param $1 v128) (result v128) - (i16x8.narrow_i32x4_s + (f32x4.pmax (local.get $0) (local.get $1) ) ) - (func $226 (param $0 v128) (param $1 v128) (result v128) - (i16x8.narrow_i32x4_u + (func $226 (param $0 v128) (result v128) + (f64x2.abs (local.get $0) - (local.get $1) ) ) (func $227 (param $0 v128) (result v128) - (i16x8.widen_low_i8x16_s + (f64x2.neg (local.get $0) ) ) (func $228 (param $0 v128) (result v128) - (i16x8.widen_high_i8x16_s + (f64x2.sqrt (local.get $0) ) ) - (func $229 (param $0 v128) (result v128) - (i16x8.widen_low_i8x16_u + (func $229 (param $0 v128) (param $1 v128) (result v128) + (f64x2.add (local.get $0) + (local.get $1) ) ) - (func $230 (param $0 v128) (result v128) - (i16x8.widen_high_i8x16_u + (func $230 (param $0 v128) (param $1 v128) (result v128) + (f64x2.sub (local.get $0) + (local.get $1) ) ) - (func $231 (param $0 v128) (result v128) - (i32x4.widen_low_i16x8_s + (func $231 (param $0 v128) (param $1 v128) (result v128) + (f64x2.mul (local.get $0) + (local.get $1) ) ) - (func $232 (param $0 v128) (result v128) - (i32x4.widen_high_i16x8_s + (func $232 (param $0 v128) (param $1 v128) (result v128) + (f64x2.div (local.get $0) + (local.get $1) ) ) - (func $233 (param $0 v128) (result v128) - (i32x4.widen_low_i16x8_u + (func $233 (param $0 v128) (param $1 v128) (result v128) + (f64x2.min (local.get $0) + (local.get $1) ) ) - (func $234 (param $0 v128) (result v128) - (i32x4.widen_high_i16x8_u + (func $234 (param $0 v128) (param $1 v128) (result v128) + (f64x2.max (local.get $0) + (local.get $1) ) ) - (func $235 (param $0 v128) (result v128) - (i64x2.widen_low_i32x4_s + (func $235 (param $0 v128) (param $1 v128) (result v128) + (f64x2.pmin (local.get $0) + (local.get $1) ) ) - (func $236 (param $0 v128) (result v128) - (i64x2.widen_high_i32x4_s + (func $236 (param $0 v128) (param $1 v128) (result v128) + (f64x2.pmax (local.get $0) + (local.get $1) ) ) (func $237 (param $0 v128) (result v128) - (i64x2.widen_low_i32x4_u + (i32x4.trunc_sat_f32x4_s (local.get $0) ) ) (func $238 (param $0 v128) (result v128) - (i64x2.widen_high_i32x4_u - (local.get $0) - ) - ) - (func $239 (param $0 i32) (result v128) - (i16x8.load8x8_u - (local.get $0) - ) - ) - (func $240 (param $0 i32) (result v128) - (i16x8.load8x8_s - (local.get $0) - ) - ) - (func $241 (param $0 i32) (result v128) - (i32x4.load16x4_s - (local.get $0) - ) - ) - (func $242 (param $0 i32) (result v128) - (i32x4.load16x4_u - (local.get $0) - ) - ) - (func $243 (param $0 i32) (result v128) - (i64x2.load32x2_s - (local.get $0) - ) - ) - (func $244 (param $0 i32) (result v128) - (i64x2.load32x2_u - (local.get $0) - ) - ) - (func $245 (param $0 i32) (result v128) - (v128.load32_zero + (i32x4.trunc_sat_f32x4_u (local.get $0) ) ) - (func $246 (param $0 i32) (result v128) - (v128.load64_zero + (func $239 (param $0 v128) (result v128) + (f32x4.convert_i32x4_s (local.get $0) ) ) - (func $247 (param $0 v128) (param $1 v128) (result v128) - (v8x16.swizzle + (func $240 (param $0 v128) (result v128) + (f32x4.convert_i32x4_u (local.get $0) - (local.get $1) ) ) - (func $248 (param $0 i32) - (prefetch.t offset=3 align=2 + (func $241 (param $0 v128) (result v128) + (i32x4.trunc_sat_f64x2_s_zero (local.get $0) ) ) - (func $249 (param $0 i32) - (prefetch.nt offset=3 align=2 + (func $242 (param $0 v128) (result v128) + (i32x4.trunc_sat_f64x2_u_zero (local.get $0) ) ) - (func $250 (param $0 v128) (result v128) + (func $243 (param $0 v128) (result v128) (f64x2.convert_low_i32x4_s (local.get $0) ) ) - (func $251 (param $0 v128) (result v128) + (func $244 (param $0 v128) (result v128) (f64x2.convert_low_i32x4_u (local.get $0) ) ) - (func $252 (param $0 v128) (result v128) - (i32x4.trunc_sat_f64x2_zero_s - (local.get $0) - ) - ) - (func $253 (param $0 v128) (result v128) - (i32x4.trunc_sat_f64x2_zero_u - (local.get $0) - ) - ) - (func $254 (param $0 v128) (result v128) - (f32x4.demote_f64x2_zero - (local.get $0) - ) - ) - (func $255 (param $0 v128) (result v128) - (f64x2.promote_low_f32x4 - (local.get $0) - ) - ) - (func $256 (param $0 v128) (result v128) - (i32x4.widen_i8x16_s 0 - (local.get $0) - ) - ) - (func $257 (param $0 v128) (result v128) - (i32x4.widen_i8x16_u 0 - (local.get $0) - ) - ) ) diff --git a/test/simd64.wast b/test/simd64.wast index 65b1ef8b7..f5cf85055 100644 --- a/test/simd64.wast +++ b/test/simd64.wast @@ -11,53 +11,53 @@ (local.get $1) ) ) - (func $v8x16.load_splat (param $0 i64) (result v128) - (v8x16.load_splat + (func $v128.load8_splat (param $0 i64) (result v128) + (v128.load8_splat (local.get $0) ) ) - (func $v16x8.load_splat (param $0 i64) (result v128) - (v16x8.load_splat + (func $v128.load16_splat (param $0 i64) (result v128) + (v128.load16_splat (local.get $0) ) ) - (func $v32x4.load_splat (param $0 i64) (result v128) - (v32x4.load_splat + (func $v128.load32_splat (param $0 i64) (result v128) + (v128.load32_splat (local.get $0) ) ) - (func $v64x2.load_splat (param $0 i64) (result v128) - (v64x2.load_splat + (func $v128.load64_splat (param $0 i64) (result v128) + (v128.load64_splat (local.get $0) ) ) - (func $i16x8.load8x8_u (param $0 i64) (result v128) - (i16x8.load8x8_u + (func $v128.load8x8_u (param $0 i64) (result v128) + (v128.load8x8_u (local.get $0) ) ) - (func $i16x8.load8x8_s (param $0 i64) (result v128) - (i16x8.load8x8_s + (func $v128.load8x8_s (param $0 i64) (result v128) + (v128.load8x8_s (local.get $0) ) ) - (func $i32x4.load16x4_s (param $0 i64) (result v128) - (i32x4.load16x4_s + (func $v128.load16x4_s (param $0 i64) (result v128) + (v128.load16x4_s (local.get $0) ) ) - (func $i32x4.load16x4_u (param $0 i64) (result v128) - (i32x4.load16x4_u + (func $v128.load16x4_u (param $0 i64) (result v128) + (v128.load16x4_u (local.get $0) ) ) - (func $i64x2.load32x2_s (param $0 i64) (result v128) - (i64x2.load32x2_s + (func $v128.load32x2_s (param $0 i64) (result v128) + (v128.load32x2_s (local.get $0) ) ) - (func $i64x2.load32x2_u (param $0 i64) (result v128) - (i64x2.load32x2_u + (func $v128.load32x2_u (param $0 i64) (result v128) + (v128.load32x2_u (local.get $0) ) ) diff --git a/test/simd64.wast.from-wast b/test/simd64.wast.from-wast index cd08c79ed..a12e5d08f 100644 --- a/test/simd64.wast.from-wast +++ b/test/simd64.wast.from-wast @@ -13,53 +13,53 @@ (local.get $1) ) ) - (func $v8x16.load_splat (param $0 i64) (result v128) - (v8x16.load_splat + (func $v128.load8_splat (param $0 i64) (result v128) + (v128.load8_splat (local.get $0) ) ) - (func $v16x8.load_splat (param $0 i64) (result v128) - (v16x8.load_splat + (func $v128.load16_splat (param $0 i64) (result v128) + (v128.load16_splat (local.get $0) ) ) - (func $v32x4.load_splat (param $0 i64) (result v128) - (v32x4.load_splat + (func $v128.load32_splat (param $0 i64) (result v128) + (v128.load32_splat (local.get $0) ) ) - (func $v64x2.load_splat (param $0 i64) (result v128) - (v64x2.load_splat + (func $v128.load64_splat (param $0 i64) (result v128) + (v128.load64_splat (local.get $0) ) ) - (func $i16x8.load8x8_u (param $0 i64) (result v128) - (i16x8.load8x8_u + (func $v128.load8x8_u (param $0 i64) (result v128) + (v128.load8x8_u (local.get $0) ) ) - (func $i16x8.load8x8_s (param $0 i64) (result v128) - (i16x8.load8x8_s + (func $v128.load8x8_s (param $0 i64) (result v128) + (v128.load8x8_s (local.get $0) ) ) - (func $i32x4.load16x4_s (param $0 i64) (result v128) - (i32x4.load16x4_s + (func $v128.load16x4_s (param $0 i64) (result v128) + (v128.load16x4_s (local.get $0) ) ) - (func $i32x4.load16x4_u (param $0 i64) (result v128) - (i32x4.load16x4_u + (func $v128.load16x4_u (param $0 i64) (result v128) + (v128.load16x4_u (local.get $0) ) ) - (func $i64x2.load32x2_s (param $0 i64) (result v128) - (i64x2.load32x2_s + (func $v128.load32x2_s (param $0 i64) (result v128) + (v128.load32x2_s (local.get $0) ) ) - (func $i64x2.load32x2_u (param $0 i64) (result v128) - (i64x2.load32x2_u + (func $v128.load32x2_u (param $0 i64) (result v128) + (v128.load32x2_u (local.get $0) ) ) diff --git a/test/simd64.wast.fromBinary b/test/simd64.wast.fromBinary index 6202b7c06..988f678b8 100644 --- a/test/simd64.wast.fromBinary +++ b/test/simd64.wast.fromBinary @@ -13,53 +13,53 @@ (local.get $1) ) ) - (func $v8x16.load_splat (param $0 i64) (result v128) - (v8x16.load_splat + (func $v128.load8_splat (param $0 i64) (result v128) + (v128.load8_splat (local.get $0) ) ) - (func $v16x8.load_splat (param $0 i64) (result v128) - (v16x8.load_splat + (func $v128.load16_splat (param $0 i64) (result v128) + (v128.load16_splat (local.get $0) ) ) - (func $v32x4.load_splat (param $0 i64) (result v128) - (v32x4.load_splat + (func $v128.load32_splat (param $0 i64) (result v128) + (v128.load32_splat (local.get $0) ) ) - (func $v64x2.load_splat (param $0 i64) (result v128) - (v64x2.load_splat + (func $v128.load64_splat (param $0 i64) (result v128) + (v128.load64_splat (local.get $0) ) ) - (func $i16x8.load8x8_u (param $0 i64) (result v128) - (i16x8.load8x8_u + (func $v128.load8x8_u (param $0 i64) (result v128) + (v128.load8x8_u (local.get $0) ) ) - (func $i16x8.load8x8_s (param $0 i64) (result v128) - (i16x8.load8x8_s + (func $v128.load8x8_s (param $0 i64) (result v128) + (v128.load8x8_s (local.get $0) ) ) - (func $i32x4.load16x4_s (param $0 i64) (result v128) - (i32x4.load16x4_s + (func $v128.load16x4_s (param $0 i64) (result v128) + (v128.load16x4_s (local.get $0) ) ) - (func $i32x4.load16x4_u (param $0 i64) (result v128) - (i32x4.load16x4_u + (func $v128.load16x4_u (param $0 i64) (result v128) + (v128.load16x4_u (local.get $0) ) ) - (func $i64x2.load32x2_s (param $0 i64) (result v128) - (i64x2.load32x2_s + (func $v128.load32x2_s (param $0 i64) (result v128) + (v128.load32x2_s (local.get $0) ) ) - (func $i64x2.load32x2_u (param $0 i64) (result v128) - (i64x2.load32x2_u + (func $v128.load32x2_u (param $0 i64) (result v128) + (v128.load32x2_u (local.get $0) ) ) diff --git a/test/simd64.wast.fromBinary.noDebugInfo b/test/simd64.wast.fromBinary.noDebugInfo index 67ce14b29..2e96dfe64 100644 --- a/test/simd64.wast.fromBinary.noDebugInfo +++ b/test/simd64.wast.fromBinary.noDebugInfo @@ -14,52 +14,52 @@ ) ) (func $2 (param $0 i64) (result v128) - (v8x16.load_splat + (v128.load8_splat (local.get $0) ) ) (func $3 (param $0 i64) (result v128) - (v16x8.load_splat + (v128.load16_splat (local.get $0) ) ) (func $4 (param $0 i64) (result v128) - (v32x4.load_splat + (v128.load32_splat (local.get $0) ) ) (func $5 (param $0 i64) (result v128) - (v64x2.load_splat + (v128.load64_splat (local.get $0) ) ) (func $6 (param $0 i64) (result v128) - (i16x8.load8x8_u + (v128.load8x8_u (local.get $0) ) ) (func $7 (param $0 i64) (result v128) - (i16x8.load8x8_s + (v128.load8x8_s (local.get $0) ) ) (func $8 (param $0 i64) (result v128) - (i32x4.load16x4_s + (v128.load16x4_s (local.get $0) ) ) (func $9 (param $0 i64) (result v128) - (i32x4.load16x4_u + (v128.load16x4_u (local.get $0) ) ) (func $10 (param $0 i64) (result v128) - (i64x2.load32x2_s + (v128.load32x2_s (local.get $0) ) ) (func $11 (param $0 i64) (result v128) - (i64x2.load32x2_u + (v128.load32x2_u (local.get $0) ) ) diff --git a/test/spec/simd.wast b/test/spec/simd.wast index 87bbb96cb..4876a032d 100644 --- a/test/spec/simd.wast +++ b/test/spec/simd.wast @@ -14,11 +14,11 @@ (func (export "v128.const.i64x2") (result v128) (v128.const i64x2 1 2)) (func (export "v128.const.f32x4") (result v128) (v128.const f32x4 1.0 2 3 4)) (func (export "v128.const.f64x2") (result v128) (v128.const f64x2 1.0 2)) - (func (export "v128.shuffle_interleave_bytes") (param $0 v128) (param $1 v128) (result v128) - (v8x16.shuffle 0 17 2 19 4 21 6 23 8 25 10 27 12 29 14 31 (local.get $0) (local.get $1)) + (func (export "i8x16.shuffle_interleave_bytes") (param $0 v128) (param $1 v128) (result v128) + (i8x16.shuffle 0 17 2 19 4 21 6 23 8 25 10 27 12 29 14 31 (local.get $0) (local.get $1)) ) - (func (export "v128.shuffle_reverse_i32s") (param $0 v128) (result v128) - (v8x16.shuffle 12 13 14 15 8 9 10 11 4 5 6 7 0 1 2 3 (local.get $0) (local.get $0)) + (func (export "i8x16.shuffle_reverse_i32s") (param $0 v128) (result v128) + (i8x16.shuffle 12 13 14 15 8 9 10 11 4 5 6 7 0 1 2 3 (local.get $0) (local.get $0)) ) (func (export "i8x16.splat") (param $0 i32) (result v128) (i8x16.splat (local.get $0))) (func (export "i8x16.extract_lane_s_first") (param $0 v128) (result i32) (i8x16.extract_lane_s 0 (local.get $0))) @@ -105,18 +105,6 @@ (func (export "v128.bitselect") (param $0 v128) (param $1 v128) (param $2 v128) (result v128) (v128.bitselect (local.get $0) (local.get $1) (local.get $2)) ) - (func (export "v8x16.signselect") (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (v8x16.signselect (local.get $0) (local.get $1) (local.get $2)) - ) - (func (export "v16x8.signselect") (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (v16x8.signselect (local.get $0) (local.get $1) (local.get $2)) - ) - (func (export "v32x4.signselect") (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (v32x4.signselect (local.get $0) (local.get $1) (local.get $2)) - ) - (func (export "v64x2.signselect") (param $0 v128) (param $1 v128) (param $2 v128) (result v128) - (v64x2.signselect (local.get $0) (local.get $1) (local.get $2)) - ) (func (export "v128.load8_lane") (param $0 i32) (param $1 v128) (result v128) (v128.load8_lane 0 (local.get $0) (local.get $1))) (func (export "v128.load16_lane") (param $0 i32) (param $1 v128) (result v128) (v128.load16_lane 0 (local.get $0) (local.get $1))) (func (export "v128.load32_lane") (param $0 i32) (param $1 v128) (result v128) (v128.load32_lane 0 (local.get $0) (local.get $1))) @@ -128,19 +116,17 @@ (func (export "i8x16.popcnt") (param $0 v128) (result v128) (i8x16.popcnt (local.get $0))) (func (export "i8x16.abs") (param $0 v128) (result v128) (i8x16.abs (local.get $0))) (func (export "i8x16.neg") (param $0 v128) (result v128) (i8x16.neg (local.get $0))) - (func (export "i8x16.any_true") (param $0 v128) (result i32) (i8x16.any_true (local.get $0))) (func (export "i8x16.all_true") (param $0 v128) (result i32) (i8x16.all_true (local.get $0))) (func (export "i8x16.bitmask") (param $0 v128) (result i32) (i8x16.bitmask (local.get $0))) (func (export "i8x16.shl") (param $0 v128) (param $1 i32) (result v128) (i8x16.shl (local.get $0) (local.get $1))) (func (export "i8x16.shr_s") (param $0 v128) (param $1 i32) (result v128) (i8x16.shr_s (local.get $0) (local.get $1))) (func (export "i8x16.shr_u") (param $0 v128) (param $1 i32) (result v128) (i8x16.shr_u (local.get $0) (local.get $1))) (func (export "i8x16.add") (param $0 v128) (param $1 v128) (result v128) (i8x16.add (local.get $0) (local.get $1))) - (func (export "i8x16.add_saturate_s") (param $0 v128) (param $1 v128) (result v128) (i8x16.add_saturate_s (local.get $0) (local.get $1))) - (func (export "i8x16.add_saturate_u") (param $0 v128) (param $1 v128) (result v128) (i8x16.add_saturate_u (local.get $0) (local.get $1))) + (func (export "i8x16.add_sat_s") (param $0 v128) (param $1 v128) (result v128) (i8x16.add_sat_s (local.get $0) (local.get $1))) + (func (export "i8x16.add_sat_u") (param $0 v128) (param $1 v128) (result v128) (i8x16.add_sat_u (local.get $0) (local.get $1))) (func (export "i8x16.sub") (param $0 v128) (param $1 v128) (result v128) (i8x16.sub (local.get $0) (local.get $1))) - (func (export "i8x16.sub_saturate_s") (param $0 v128) (param $1 v128) (result v128) (i8x16.sub_saturate_s (local.get $0) (local.get $1))) - (func (export "i8x16.sub_saturate_u") (param $0 v128) (param $1 v128) (result v128) (i8x16.sub_saturate_u (local.get $0) (local.get $1))) - (func (export "i8x16.mul") (param $0 v128) (param $1 v128) (result v128) (i8x16.mul (local.get $0) (local.get $1))) + (func (export "i8x16.sub_sat_s") (param $0 v128) (param $1 v128) (result v128) (i8x16.sub_sat_s (local.get $0) (local.get $1))) + (func (export "i8x16.sub_sat_u") (param $0 v128) (param $1 v128) (result v128) (i8x16.sub_sat_u (local.get $0) (local.get $1))) (func (export "i8x16.min_s") (param $0 v128) (param $1 v128) (result v128) (i8x16.min_s (local.get $0) (local.get $1))) (func (export "i8x16.min_u") (param $0 v128) (param $1 v128) (result v128) (i8x16.min_u (local.get $0) (local.get $1))) (func (export "i8x16.max_s") (param $0 v128) (param $1 v128) (result v128) (i8x16.max_s (local.get $0) (local.get $1))) @@ -148,18 +134,17 @@ (func (export "i8x16.avgr_u") (param $0 v128) (param $1 v128) (result v128) (i8x16.avgr_u (local.get $0) (local.get $1))) (func (export "i16x8.abs") (param $0 v128) (result v128) (i16x8.abs (local.get $0))) (func (export "i16x8.neg") (param $0 v128) (result v128) (i16x8.neg (local.get $0))) - (func (export "i16x8.any_true") (param $0 v128) (result i32) (i16x8.any_true (local.get $0))) (func (export "i16x8.all_true") (param $0 v128) (result i32) (i16x8.all_true (local.get $0))) (func (export "i16x8.bitmask") (param $0 v128) (result i32) (i16x8.bitmask (local.get $0))) (func (export "i16x8.shl") (param $0 v128) (param $1 i32) (result v128) (i16x8.shl (local.get $0) (local.get $1))) (func (export "i16x8.shr_s") (param $0 v128) (param $1 i32) (result v128) (i16x8.shr_s (local.get $0) (local.get $1))) (func (export "i16x8.shr_u") (param $0 v128) (param $1 i32) (result v128) (i16x8.shr_u (local.get $0) (local.get $1))) (func (export "i16x8.add") (param $0 v128) (param $1 v128) (result v128) (i16x8.add (local.get $0) (local.get $1))) - (func (export "i16x8.add_saturate_s") (param $0 v128) (param $1 v128) (result v128) (i16x8.add_saturate_s (local.get $0) (local.get $1))) - (func (export "i16x8.add_saturate_u") (param $0 v128) (param $1 v128) (result v128) (i16x8.add_saturate_u (local.get $0) (local.get $1))) + (func (export "i16x8.add_sat_s") (param $0 v128) (param $1 v128) (result v128) (i16x8.add_sat_s (local.get $0) (local.get $1))) + (func (export "i16x8.add_sat_u") (param $0 v128) (param $1 v128) (result v128) (i16x8.add_sat_u (local.get $0) (local.get $1))) (func (export "i16x8.sub") (param $0 v128) (param $1 v128) (result v128) (i16x8.sub (local.get $0) (local.get $1))) - (func (export "i16x8.sub_saturate_s") (param $0 v128) (param $1 v128) (result v128) (i16x8.sub_saturate_s (local.get $0) (local.get $1))) - (func (export "i16x8.sub_saturate_u") (param $0 v128) (param $1 v128) (result v128) (i16x8.sub_saturate_u (local.get $0) (local.get $1))) + (func (export "i16x8.sub_sat_s") (param $0 v128) (param $1 v128) (result v128) (i16x8.sub_sat_s (local.get $0) (local.get $1))) + (func (export "i16x8.sub_sat_u") (param $0 v128) (param $1 v128) (result v128) (i16x8.sub_sat_u (local.get $0) (local.get $1))) (func (export "i16x8.mul") (param $0 v128) (param $1 v128) (result v128) (i16x8.mul (local.get $0) (local.get $1))) (func (export "i16x8.min_s") (param $0 v128) (param $1 v128) (result v128) (i16x8.min_s (local.get $0) (local.get $1))) (func (export "i16x8.min_u") (param $0 v128) (param $1 v128) (result v128) (i16x8.min_u (local.get $0) (local.get $1))) @@ -170,7 +155,6 @@ ;; TODO: extending multiplications (func (export "i32x4.abs") (param $0 v128) (result v128) (i32x4.abs (local.get $0))) (func (export "i32x4.neg") (param $0 v128) (result v128) (i32x4.neg (local.get $0))) - (func (export "i32x4.any_true") (param $0 v128) (result i32) (i32x4.any_true (local.get $0))) (func (export "i32x4.all_true") (param $0 v128) (result i32) (i32x4.all_true (local.get $0))) (func (export "i32x4.bitmask") (param $0 v128) (result i32) (i32x4.bitmask (local.get $0))) (func (export "i32x4.shl") (param $0 v128) (param $1 i32) (result v128) (i32x4.shl (local.get $0) (local.get $1))) @@ -195,8 +179,6 @@ (func (export "f32x4.abs") (param $0 v128) (result v128) (f32x4.abs (local.get $0))) (func (export "f32x4.neg") (param $0 v128) (result v128) (f32x4.neg (local.get $0))) (func (export "f32x4.sqrt") (param $0 v128) (result v128) (f32x4.sqrt (local.get $0))) - (func (export "f32x4.qfma") (param $0 v128) (param $1 v128) (param $2 v128) (result v128) (f32x4.qfma (local.get $0) (local.get $1) (local.get $2))) - (func (export "f32x4.qfms") (param $0 v128) (param $1 v128) (param $2 v128) (result v128) (f32x4.qfms (local.get $0) (local.get $1) (local.get $2))) (func (export "f32x4.add") (param $0 v128) (param $1 v128) (result v128) (f32x4.add (local.get $0) (local.get $1))) (func (export "f32x4.sub") (param $0 v128) (param $1 v128) (result v128) (f32x4.sub (local.get $0) (local.get $1))) (func (export "f32x4.mul") (param $0 v128) (param $1 v128) (result v128) (f32x4.mul (local.get $0) (local.get $1))) @@ -212,8 +194,6 @@ (func (export "f64x2.abs") (param $0 v128) (result v128) (f64x2.abs (local.get $0))) (func (export "f64x2.neg") (param $0 v128) (result v128) (f64x2.neg (local.get $0))) (func (export "f64x2.sqrt") (param $0 v128) (result v128) (f64x2.sqrt (local.get $0))) - (func (export "f64x2.qfma") (param $0 v128) (param $1 v128) (param $2 v128) (result v128) (f64x2.qfma (local.get $0) (local.get $1) (local.get $2))) - (func (export "f64x2.qfms") (param $0 v128) (param $1 v128) (param $2 v128) (result v128) (f64x2.qfms (local.get $0) (local.get $1) (local.get $2))) (func (export "f64x2.add") (param $0 v128) (param $1 v128) (result v128) (f64x2.add (local.get $0) (local.get $1))) (func (export "f64x2.sub") (param $0 v128) (param $1 v128) (result v128) (f64x2.sub (local.get $0) (local.get $1))) (func (export "f64x2.mul") (param $0 v128) (param $1 v128) (result v128) (f64x2.mul (local.get $0) (local.get $1))) @@ -229,51 +209,47 @@ ;; TODO: Extending pairwise adds once they have interpreter support (func (export "i32x4.trunc_sat_f32x4_s") (param $0 v128) (result v128) (i32x4.trunc_sat_f32x4_s (local.get $0))) (func (export "i32x4.trunc_sat_f32x4_u") (param $0 v128) (result v128) (i32x4.trunc_sat_f32x4_u (local.get $0))) - (func (export "i64x2.trunc_sat_f64x2_s") (param $0 v128) (result v128) (i64x2.trunc_sat_f64x2_s (local.get $0))) - (func (export "i64x2.trunc_sat_f64x2_u") (param $0 v128) (result v128) (i64x2.trunc_sat_f64x2_u (local.get $0))) (func (export "f32x4.convert_i32x4_s") (param $0 v128) (result v128) (f32x4.convert_i32x4_s (local.get $0))) (func (export "f32x4.convert_i32x4_u") (param $0 v128) (result v128) (f32x4.convert_i32x4_u (local.get $0))) - (func (export "f64x2.convert_i64x2_s") (param $0 v128) (result v128) (f64x2.convert_i64x2_s (local.get $0))) - (func (export "f64x2.convert_i64x2_u") (param $0 v128) (result v128) (f64x2.convert_i64x2_u (local.get $0))) - (func (export "v8x16.load_splat") (param $0 i32) (result v128) (v8x16.load_splat (local.get $0))) - (func (export "v16x8.load_splat") (param $0 i32) (result v128) (v16x8.load_splat (local.get $0))) - (func (export "v32x4.load_splat") (param $0 i32) (result v128) (v32x4.load_splat (local.get $0))) - (func (export "v64x2.load_splat") (param $0 i32) (result v128) (v64x2.load_splat (local.get $0))) + (func (export "v128.load8_splat") (param $0 i32) (result v128) (v128.load8_splat (local.get $0))) + (func (export "v128.load16_splat") (param $0 i32) (result v128) (v128.load16_splat (local.get $0))) + (func (export "v128.load32_splat") (param $0 i32) (result v128) (v128.load32_splat (local.get $0))) + (func (export "v128.load64_splat") (param $0 i32) (result v128) (v128.load64_splat (local.get $0))) (func (export "i8x16.narrow_i16x8_s") (param $0 v128) (param $1 v128) (result v128) (i8x16.narrow_i16x8_s (local.get $0) (local.get $1))) (func (export "i8x16.narrow_i16x8_u") (param $0 v128) (param $1 v128) (result v128) (i8x16.narrow_i16x8_u (local.get $0) (local.get $1))) (func (export "i16x8.narrow_i32x4_s") (param $0 v128) (param $1 v128) (result v128) (i16x8.narrow_i32x4_s (local.get $0) (local.get $1))) (func (export "i16x8.narrow_i32x4_u") (param $0 v128) (param $1 v128) (result v128) (i16x8.narrow_i32x4_u (local.get $0) (local.get $1))) - (func (export "i16x8.widen_low_i8x16_s") (param $0 v128) (result v128) (i16x8.widen_low_i8x16_s (local.get $0))) - (func (export "i16x8.widen_high_i8x16_s") (param $0 v128) (result v128) (i16x8.widen_high_i8x16_s (local.get $0))) - (func (export "i16x8.widen_low_i8x16_u") (param $0 v128) (result v128) (i16x8.widen_low_i8x16_u (local.get $0))) - (func (export "i16x8.widen_high_i8x16_u") (param $0 v128) (result v128) (i16x8.widen_high_i8x16_u (local.get $0))) - (func (export "i32x4.widen_low_i16x8_s") (param $0 v128) (result v128) (i32x4.widen_low_i16x8_s (local.get $0))) - (func (export "i32x4.widen_high_i16x8_s") (param $0 v128) (result v128) (i32x4.widen_high_i16x8_s (local.get $0))) - (func (export "i32x4.widen_low_i16x8_u") (param $0 v128) (result v128) (i32x4.widen_low_i16x8_u (local.get $0))) - (func (export "i32x4.widen_high_i16x8_u") (param $0 v128) (result v128) (i32x4.widen_high_i16x8_u (local.get $0))) - (func (export "i64x2.widen_low_i32x4_s") (param $0 v128) (result v128) (i64x2.widen_low_i32x4_s (local.get $0))) - (func (export "i64x2.widen_high_i32x4_s") (param $0 v128) (result v128) (i64x2.widen_high_i32x4_s (local.get $0))) - (func (export "i64x2.widen_low_i32x4_u") (param $0 v128) (result v128) (i64x2.widen_low_i32x4_u (local.get $0))) - (func (export "i64x2.widen_high_i32x4_u") (param $0 v128) (result v128) (i64x2.widen_high_i32x4_u (local.get $0))) - (func (export "i16x8.load8x8_u") (param $0 i32) (result v128) (i16x8.load8x8_u (local.get $0))) - (func (export "i16x8.load8x8_s") (param $0 i32) (result v128) (i16x8.load8x8_s (local.get $0))) - (func (export "i32x4.load16x4_u") (param $0 i32) (result v128) (i32x4.load16x4_u (local.get $0))) - (func (export "i32x4.load16x4_s") (param $0 i32) (result v128) (i32x4.load16x4_s (local.get $0))) - (func (export "i64x2.load32x2_u") (param $0 i32) (result v128) (i64x2.load32x2_u (local.get $0))) - (func (export "i64x2.load32x2_s") (param $0 i32) (result v128) (i64x2.load32x2_s (local.get $0))) + (func (export "i16x8.extend_low_i8x16_s") (param $0 v128) (result v128) (i16x8.extend_low_i8x16_s (local.get $0))) + (func (export "i16x8.extend_high_i8x16_s") (param $0 v128) (result v128) (i16x8.extend_high_i8x16_s (local.get $0))) + (func (export "i16x8.extend_low_i8x16_u") (param $0 v128) (result v128) (i16x8.extend_low_i8x16_u (local.get $0))) + (func (export "i16x8.extend_high_i8x16_u") (param $0 v128) (result v128) (i16x8.extend_high_i8x16_u (local.get $0))) + (func (export "i32x4.extend_low_i16x8_s") (param $0 v128) (result v128) (i32x4.extend_low_i16x8_s (local.get $0))) + (func (export "i32x4.extend_high_i16x8_s") (param $0 v128) (result v128) (i32x4.extend_high_i16x8_s (local.get $0))) + (func (export "i32x4.extend_low_i16x8_u") (param $0 v128) (result v128) (i32x4.extend_low_i16x8_u (local.get $0))) + (func (export "i32x4.extend_high_i16x8_u") (param $0 v128) (result v128) (i32x4.extend_high_i16x8_u (local.get $0))) + (func (export "i64x2.extend_low_i32x4_s") (param $0 v128) (result v128) (i64x2.extend_low_i32x4_s (local.get $0))) + (func (export "i64x2.extend_high_i32x4_s") (param $0 v128) (result v128) (i64x2.extend_high_i32x4_s (local.get $0))) + (func (export "i64x2.extend_low_i32x4_u") (param $0 v128) (result v128) (i64x2.extend_low_i32x4_u (local.get $0))) + (func (export "i64x2.extend_high_i32x4_u") (param $0 v128) (result v128) (i64x2.extend_high_i32x4_u (local.get $0))) + (func (export "v128.load8x8_u") (param $0 i32) (result v128) (v128.load8x8_u (local.get $0))) + (func (export "v128.load8x8_s") (param $0 i32) (result v128) (v128.load8x8_s (local.get $0))) + (func (export "v128.load16x4_u") (param $0 i32) (result v128) (v128.load16x4_u (local.get $0))) + (func (export "v128.load16x4_s") (param $0 i32) (result v128) (v128.load16x4_s (local.get $0))) + (func (export "v128.load32x2_u") (param $0 i32) (result v128) (v128.load32x2_u (local.get $0))) + (func (export "v128.load32x2_s") (param $0 i32) (result v128) (v128.load32x2_s (local.get $0))) (func (export "v128.load32_zero") (param $0 i32) (result v128) (v128.load32_zero (local.get $0))) (func (export "v128.load64_zero") (param $0 i32) (result v128) (v128.load64_zero (local.get $0))) - (func (export "v8x16.swizzle") (param $0 v128) (param $1 v128) (result v128) (v8x16.swizzle (local.get $0) (local.get $1))) + (func (export "i8x16.swizzle") (param $0 v128) (param $1 v128) (result v128) (i8x16.swizzle (local.get $0) (local.get $1))) ) ;; TODO: Additional f64x2 conversions if specified ;; Basic v128 manipulation (assert_return (invoke "v128.load" (i32.const 128)) (v128.const i8x16 87 65 83 77 83 73 77 68 71 79 69 83 70 65 83 84)) (assert_return (invoke "v128.store" (i32.const 16) (v128.const i32x4 1 2 3 4)) (v128.const i32x4 1 2 3 4)) -(assert_return (invoke "v8x16.load_splat" (i32.const 128)) (v128.const i8x16 87 87 87 87 87 87 87 87 87 87 87 87 87 87 87 87)) -(assert_return (invoke "v16x8.load_splat" (i32.const 128)) (v128.const i8x16 87 65 87 65 87 65 87 65 87 65 87 65 87 65 87 65)) -(assert_return (invoke "v32x4.load_splat" (i32.const 128)) (v128.const i8x16 87 65 83 77 87 65 83 77 87 65 83 77 87 65 83 77)) -(assert_return (invoke "v64x2.load_splat" (i32.const 128)) (v128.const i8x16 87 65 83 77 83 73 77 68 87 65 83 77 83 73 77 68)) +(assert_return (invoke "v128.load8_splat" (i32.const 128)) (v128.const i8x16 87 87 87 87 87 87 87 87 87 87 87 87 87 87 87 87)) +(assert_return (invoke "v128.load16_splat" (i32.const 128)) (v128.const i8x16 87 65 87 65 87 65 87 65 87 65 87 65 87 65 87 65)) +(assert_return (invoke "v128.load32_splat" (i32.const 128)) (v128.const i8x16 87 65 83 77 87 65 83 77 87 65 83 77 87 65 83 77)) +(assert_return (invoke "v128.load64_splat" (i32.const 128)) (v128.const i8x16 87 65 83 77 83 73 77 68 87 65 83 77 83 73 77 68)) (assert_return (invoke "v128.const.i8x16") (v128.const i32x4 0x04030201 0x08070605 0x0c0b0a09 0x100f0e0d)) (assert_return (invoke "v128.const.i16x8") (v128.const i8x16 01 00 02 00 03 00 04 00 05 00 06 00 07 00 08 00)) (assert_return (invoke "v128.const.i32x4") (v128.const i8x16 01 00 00 00 02 00 00 00 03 00 00 00 04 00 00 00)) @@ -281,13 +257,13 @@ (assert_return (invoke "v128.const.f32x4") (v128.const f32x4 1 2 3 4)) (assert_return (invoke "v128.const.f64x2") (v128.const f64x2 1 2)) (assert_return - (invoke "v128.shuffle_interleave_bytes" + (invoke "i8x16.shuffle_interleave_bytes" (v128.const i8x16 1 0 3 0 5 0 7 0 9 0 11 0 13 0 15 0) (v128.const i8x16 0 2 0 4 0 6 0 8 0 10 0 12 0 14 0 16) ) (v128.const i8x16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16) ) -(assert_return (invoke "v128.shuffle_reverse_i32s" (v128.const i32x4 1 2 3 4)) (v128.const i32x4 4 3 2 1)) +(assert_return (invoke "i8x16.shuffle_reverse_i32s" (v128.const i32x4 1 2 3 4)) (v128.const i32x4 4 3 2 1)) ;; i8x16 lane accesses (assert_return (invoke "i8x16.splat" (i32.const 5)) (v128.const i8x16 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5)) @@ -604,10 +580,6 @@ (assert_return (invoke "i8x16.neg" (v128.const i32x4 0 1 42 -3 -56 127 -128 -126 0 -1 -42 3 56 -127 -128 126)) (v128.const i32x4 0 -1 -42 3 56 -127 -128 126 0 1 42 -3 -56 127 -128 -126) ) -(assert_return (invoke "i8x16.any_true" (v128.const i32x4 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0)) (i32.const 0)) -(assert_return (invoke "i8x16.any_true" (v128.const i32x4 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0)) (i32.const 1)) -(assert_return (invoke "i8x16.any_true" (v128.const i32x4 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1)) (i32.const 1)) -(assert_return (invoke "i8x16.any_true" (v128.const i32x4 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1)) (i32.const 1)) (assert_return (invoke "i8x16.all_true" (v128.const i32x4 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0)) (i32.const 0)) (assert_return (invoke "i8x16.all_true" (v128.const i32x4 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0)) (i32.const 0)) (assert_return (invoke "i8x16.all_true" (v128.const i32x4 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1)) (i32.const 0)) @@ -639,14 +611,14 @@ (v128.const i32x4 3 17 0 0 0 135 109 46 145 225 48 184 17 249 128 215) ) (assert_return - (invoke "i8x16.add_saturate_s" + (invoke "i8x16.add_sat_s" (v128.const i32x4 0 42 255 128 127 129 6 29 103 196 231 142 17 250 1 73) (v128.const i32x4 3 231 1 128 129 6 103 17 42 29 73 42 0 255 127 142) ) (v128.const i32x4 3 17 0 128 0 135 109 46 127 225 48 184 17 249 127 215) ) (assert_return - (invoke "i8x16.add_saturate_u" + (invoke "i8x16.add_sat_u" (v128.const i32x4 0 42 255 128 127 129 6 29 103 196 231 142 17 250 1 73) (v128.const i32x4 3 231 1 128 129 6 103 17 42 29 73 42 0 255 127 142) ) @@ -660,27 +632,20 @@ (v128.const i32x4 253 67 254 0 254 123 159 12 61 167 158 100 17 251 130 187) ) (assert_return - (invoke "i8x16.sub_saturate_s" + (invoke "i8x16.sub_sat_s" (v128.const i32x4 0 42 255 128 127 129 6 29 103 196 231 142 17 250 1 73) (v128.const i32x4 3 231 1 128 129 6 103 17 42 29 73 42 0 255 127 142) ) (v128.const i32x4 253 67 254 0 127 128 159 12 61 167 158 128 17 251 130 127) ) (assert_return - (invoke "i8x16.sub_saturate_u" + (invoke "i8x16.sub_sat_u" (v128.const i32x4 0 42 255 128 127 129 6 29 103 196 231 142 17 250 1 73) (v128.const i32x4 3 231 1 128 129 6 103 17 42 29 73 42 0 255 127 142) ) (v128.const i32x4 0 0 254 0 0 123 0 12 61 167 158 100 17 0 0 0) ) (assert_return - (invoke "i8x16.mul" - (v128.const i32x4 0 42 255 128 127 129 6 29 103 196 231 142 17 250 1 73) - (v128.const i32x4 3 231 1 128 129 6 103 17 42 29 73 42 0 255 127 142) - ) - (v128.const i32x4 0 230 255 0 255 6 106 237 230 52 223 76 0 6 127 126) -) -(assert_return (invoke "i8x16.min_s" (v128.const i32x4 0 42 255 128 127 129 6 29 103 196 231 142 17 250 1 73) (v128.const i32x4 3 231 1 128 129 6 103 17 42 29 73 42 0 255 127 142) @@ -723,10 +688,6 @@ (assert_return (invoke "i16x8.neg" (v128.const i32x4 0 1 42 -3 -56 32767 -32768 32766)) (v128.const i32x4 0 -1 -42 3 56 -32767 -32768 -32766) ) -(assert_return (invoke "i16x8.any_true" (v128.const i32x4 0 0 0 0 0 0 0 0)) (i32.const 0)) -(assert_return (invoke "i16x8.any_true" (v128.const i32x4 0 0 1 0 0 0 0 0)) (i32.const 1)) -(assert_return (invoke "i16x8.any_true" (v128.const i32x4 1 1 1 1 1 0 1 1)) (i32.const 1)) -(assert_return (invoke "i16x8.any_true" (v128.const i32x4 1 1 1 1 1 1 1 1)) (i32.const 1)) (assert_return (invoke "i16x8.all_true" (v128.const i32x4 0 0 0 0 0 0 0 0)) (i32.const 0)) (assert_return (invoke "i16x8.all_true" (v128.const i32x4 0 0 1 0 0 0 0 0)) (i32.const 0)) (assert_return (invoke "i16x8.all_true" (v128.const i32x4 1 1 1 1 1 0 1 1)) (i32.const 0)) @@ -746,14 +707,14 @@ (v128.const i32x4 768 65281 0 0 34560 12288 63744 32768) ) (assert_return - (invoke "i16x8.add_saturate_s" + (invoke "i16x8.add_sat_s" (v128.const i32x4 0 65280 32768 32512 33024 59136 64000 32766) (v128.const i32x4 768 1 32768 33024 1536 18688 65280 2) ) (v128.const i32x4 768 65281 32768 0 34560 12288 63744 32767) ) (assert_return - (invoke "i16x8.add_saturate_u" + (invoke "i16x8.add_sat_u" (v128.const i32x4 0 65280 32768 32512 33024 59136 64000 32766) (v128.const i32x4 768 1 32768 33024 1536 18688 65280 2) ) @@ -767,14 +728,14 @@ (v128.const i32x4 64768 65279 0 65024 31488 40448 64256 32764) ) (assert_return - (invoke "i16x8.sub_saturate_s" + (invoke "i16x8.sub_sat_s" (v128.const i32x4 0 65280 32768 32512 33024 59136 64000 32766) (v128.const i32x4 768 1 32768 33024 1536 18688 65280 2) ) (v128.const i32x4 64768 65279 0 32767 32768 40448 64256 32764) ) (assert_return - (invoke "i16x8.sub_saturate_u" + (invoke "i16x8.sub_sat_u" (v128.const i32x4 0 65280 32768 32512 33024 59136 64000 32766) (v128.const i32x4 768 1 32768 33024 1536 18688 65280 2) ) @@ -826,10 +787,6 @@ ;; i32x4 arithmetic (assert_return (invoke "i32x4.abs" (v128.const i32x4 0 1 0x80000000 0x80000001)) (v128.const i32x4 0 1 0x80000000 0x7fffffff)) (assert_return (invoke "i32x4.neg" (v128.const i32x4 0 1 0x80000000 0x80000001)) (v128.const i32x4 0 -1 0x80000000 0x7fffffff)) -(assert_return (invoke "i32x4.any_true" (v128.const i32x4 0 0 0 0)) (i32.const 0)) -(assert_return (invoke "i32x4.any_true" (v128.const i32x4 0 0 1 0)) (i32.const 1)) -(assert_return (invoke "i32x4.any_true" (v128.const i32x4 1 0 1 1)) (i32.const 1)) -(assert_return (invoke "i32x4.any_true" (v128.const i32x4 1 1 1 1)) (i32.const 1)) (assert_return (invoke "i32x4.all_true" (v128.const i32x4 0 0 0 0)) (i32.const 0)) (assert_return (invoke "i32x4.all_true" (v128.const i32x4 0 0 1 0)) (i32.const 0)) (assert_return (invoke "i32x4.all_true" (v128.const i32x4 1 0 1 1)) (i32.const 0)) @@ -882,7 +839,6 @@ (assert_return (invoke "f32x4.abs" (v128.const f32x4 -0 nan -infinity 5)) (v128.const f32x4 0 nan infinity 5)) (assert_return (invoke "f32x4.neg" (v128.const f32x4 -0 nan -infinity 5)) (v128.const f32x4 0 -nan infinity -5)) (assert_return (invoke "f32x4.sqrt" (v128.const f32x4 -0 nan infinity 4)) (v128.const f32x4 -0 nan infinity 2)) -;; TODO: qfma/qfms tests (assert_return (invoke "f32x4.add" (v128.const f32x4 nan -nan infinity 42) (v128.const f32x4 42 infinity infinity 1)) (v128.const f32x4 nan nan infinity 43)) (assert_return (invoke "f32x4.sub" (v128.const f32x4 nan -nan infinity 42) (v128.const f32x4 42 infinity -infinity 1)) (v128.const f32x4 nan nan infinity 41)) (assert_return (invoke "f32x4.mul" (v128.const f32x4 nan -nan infinity 42) (v128.const f32x4 42 infinity infinity 2)) (v128.const f32x4 nan nan infinity 84)) @@ -911,7 +867,6 @@ (assert_return (invoke "f64x2.neg" (v128.const f64x2 -infinity 5)) (v128.const f64x2 infinity -5)) (assert_return (invoke "f64x2.sqrt" (v128.const f64x2 -0 nan)) (v128.const f64x2 -0 nan)) (assert_return (invoke "f64x2.sqrt" (v128.const f64x2 infinity 4)) (v128.const f64x2 infinity 2)) -;; TODO: qfma/qfms tests (assert_return (invoke "f64x2.add" (v128.const f64x2 nan -nan) (v128.const f64x2 42 infinity)) (v128.const f64x2 nan nan)) (assert_return (invoke "f64x2.add" (v128.const f64x2 infinity 42) (v128.const f64x2 infinity 1)) (v128.const f64x2 infinity 43)) (assert_return (invoke "f64x2.sub" (v128.const f64x2 nan -nan) (v128.const f64x2 42 infinity)) (v128.const f64x2 nan nan)) @@ -956,16 +911,8 @@ ;; conversions (assert_return (invoke "i32x4.trunc_sat_f32x4_s" (v128.const f32x4 42 nan infinity -infinity)) (v128.const i32x4 42 0 2147483647 -2147483648)) (assert_return (invoke "i32x4.trunc_sat_f32x4_u" (v128.const f32x4 42 nan infinity -infinity)) (v128.const i32x4 42 0 4294967295 0)) -(assert_return (invoke "i64x2.trunc_sat_f64x2_s" (v128.const f64x2 42 nan)) (v128.const i64x2 42 0)) -(assert_return (invoke "i64x2.trunc_sat_f64x2_s" (v128.const f64x2 infinity -infinity)) (v128.const i64x2 9223372036854775807 -9223372036854775808)) -(assert_return (invoke "i64x2.trunc_sat_f64x2_u" (v128.const f64x2 42 nan)) (v128.const i64x2 42 0)) -(assert_return (invoke "i64x2.trunc_sat_f64x2_u" (v128.const f64x2 infinity -infinity)) (v128.const i64x2 18446744073709551615 0)) (assert_return (invoke "f32x4.convert_i32x4_s" (v128.const i32x4 0 -1 2147483647 -2147483648)) (v128.const f32x4 0 -1 2147483648 -2147483648)) (assert_return (invoke "f32x4.convert_i32x4_u" (v128.const i32x4 0 -1 2147483647 -2147483648)) (v128.const f32x4 0 4294967296 2147483648 2147483648)) -(assert_return (invoke "f64x2.convert_i64x2_s" (v128.const i64x2 0 -1)) (v128.const f64x2 0 -1)) -(assert_return (invoke "f64x2.convert_i64x2_s" (v128.const i64x2 9223372036854775807 -9223372036854775808)) (v128.const f64x2 9223372036854775807 -9223372036854775808)) -(assert_return (invoke "f64x2.convert_i64x2_u" (v128.const i64x2 0 -1)) (v128.const f64x2 0 18446744073709551616)) -(assert_return (invoke "f64x2.convert_i64x2_u" (v128.const i64x2 9223372036854775807 -9223372036854775808)) (v128.const f64x2 9223372036854775807 9223372036854775808)) (assert_return (invoke "i8x16.narrow_i16x8_s" (v128.const i16x8 129 127 -32767 32767 -32768 -1 1 0) @@ -995,44 +942,43 @@ (v128.const i16x8 32769 32767 0 65535 0 1 0 0) ) (assert_return - (invoke "i16x8.widen_low_i8x16_s" + (invoke "i16x8.extend_low_i8x16_s" (v128.const i8x16 0 1 -1 -128 127 129 64 -64 -64 64 129 127 -128 -1 1 0) ) (v128.const i16x8 0 1 -1 -128 127 -127 64 -64) ) (assert_return - (invoke "i16x8.widen_high_i8x16_s" + (invoke "i16x8.extend_high_i8x16_s" (v128.const i8x16 0 1 -1 -128 127 129 64 -64 -64 64 129 127 -128 -1 1 0) ) (v128.const i16x8 -64 64 -127 127 -128 -1 1 0) ) (assert_return - (invoke "i16x8.widen_low_i8x16_u" + (invoke "i16x8.extend_low_i8x16_u" (v128.const i8x16 0 1 -1 -128 127 129 64 -64 -64 64 129 127 -128 -1 1 0) ) (v128.const i16x8 0 1 255 128 127 129 64 192) ) (assert_return - (invoke "i16x8.widen_high_i8x16_u" + (invoke "i16x8.extend_high_i8x16_u" (v128.const i8x16 0 1 -1 -128 127 129 64 -64 -64 64 129 127 -128 -1 1 0) ) (v128.const i16x8 192 64 129 127 128 255 1 0) ) -(assert_return (invoke "i32x4.widen_low_i16x8_s" (v128.const i16x8 0 1 -1 32768 32767 32769 16384 -16384)) (v128.const i32x4 0 1 -1 -32768)) -(assert_return (invoke "i32x4.widen_high_i16x8_s" (v128.const i16x8 0 1 -1 32768 32767 32769 16384 -16384)) (v128.const i32x4 32767 -32767 16384 -16384)) -(assert_return (invoke "i32x4.widen_low_i16x8_u" (v128.const i16x8 0 1 -1 32768 32767 32769 16384 -16384)) (v128.const i32x4 0 1 65535 32768)) -(assert_return (invoke "i32x4.widen_high_i16x8_u" (v128.const i16x8 0 1 -1 32768 32767 32769 16384 -16384)) (v128.const i32x4 32767 32769 16384 49152)) -;; TODO: test i64x2 widens -(assert_return (invoke "i16x8.load8x8_s" (i32.const 256)) (v128.const i16x8 0xff80 0xff90 0xffa0 0xffb0 0xffc0 0xffd0 0xffe0 0xfff0)) -(assert_return (invoke "i16x8.load8x8_u" (i32.const 256)) (v128.const i16x8 0x0080 0x0090 0x00a0 0x00b0 0x00c0 0x00d0 0x00e0 0x00f0)) -(assert_return (invoke "i32x4.load16x4_s" (i32.const 256)) (v128.const i32x4 0xffff9080 0xffffb0a0 0xffffd0c0 0xfffff0e0)) -(assert_return (invoke "i32x4.load16x4_u" (i32.const 256)) (v128.const i32x4 0x00009080 0x0000b0a0 0x0000d0c0 0x0000f0e0)) -(assert_return (invoke "i64x2.load32x2_s" (i32.const 256)) (v128.const i64x2 0xffffffffb0a09080 0xfffffffff0e0d0c0)) -(assert_return (invoke "i64x2.load32x2_u" (i32.const 256)) (v128.const i64x2 0x00000000b0a09080 0x00000000f0e0d0c0)) +(assert_return (invoke "i32x4.extend_low_i16x8_s" (v128.const i16x8 0 1 -1 32768 32767 32769 16384 -16384)) (v128.const i32x4 0 1 -1 -32768)) +(assert_return (invoke "i32x4.extend_high_i16x8_s" (v128.const i16x8 0 1 -1 32768 32767 32769 16384 -16384)) (v128.const i32x4 32767 -32767 16384 -16384)) +(assert_return (invoke "i32x4.extend_low_i16x8_u" (v128.const i16x8 0 1 -1 32768 32767 32769 16384 -16384)) (v128.const i32x4 0 1 65535 32768)) +(assert_return (invoke "i32x4.extend_high_i16x8_u" (v128.const i16x8 0 1 -1 32768 32767 32769 16384 -16384)) (v128.const i32x4 32767 32769 16384 49152)) +(assert_return (invoke "v128.load8x8_s" (i32.const 256)) (v128.const i16x8 0xff80 0xff90 0xffa0 0xffb0 0xffc0 0xffd0 0xffe0 0xfff0)) +(assert_return (invoke "v128.load8x8_u" (i32.const 256)) (v128.const i16x8 0x0080 0x0090 0x00a0 0x00b0 0x00c0 0x00d0 0x00e0 0x00f0)) +(assert_return (invoke "v128.load16x4_s" (i32.const 256)) (v128.const i32x4 0xffff9080 0xffffb0a0 0xffffd0c0 0xfffff0e0)) +(assert_return (invoke "v128.load16x4_u" (i32.const 256)) (v128.const i32x4 0x00009080 0x0000b0a0 0x0000d0c0 0x0000f0e0)) +(assert_return (invoke "v128.load32x2_s" (i32.const 256)) (v128.const i64x2 0xffffffffb0a09080 0xfffffffff0e0d0c0)) +(assert_return (invoke "v128.load32x2_u" (i32.const 256)) (v128.const i64x2 0x00000000b0a09080 0x00000000f0e0d0c0)) (assert_return (invoke "v128.load32_zero" (i32.const 256)) (v128.const i32x4 0xb0a09080 0 0 0)) (assert_return (invoke "v128.load64_zero" (i32.const 256)) (v128.const i64x2 0xf0e0d0c0b0a09080 0)) (assert_return - (invoke "v8x16.swizzle" + (invoke "i8x16.swizzle" (v128.const i8x16 0xf0 0xf1 0xf2 0xf3 0xf4 0xf5 0xf6 0xf7 0xf8 0xf9 0xfa 0xfb 0xfc 0xfd 0xfe 0xff) (v128.const i8x16 0 4 8 12 16 255 129 128 127 17 15 13 12 8 4 0) ) |