A lightweight WebAssembly runtime that is fast, secure, and standards-compliant
winch(aarch64): v128 add and sub ops (#13928)
Daniel Macovei19d4993828db
diff --git a/tests/disas/winch/aarch64/v128_ops/f32x4_add.wat b/tests/disas/winch/aarch64/v128_ops/f32x4_add.watnew file mode 100644index 0000000000..23a24a9cb3--- /dev/null+++ b/tests/disas/winch/aarch64/v128_ops/f32x4_add.wat@@ -0,0 +1,51 @@+;;! target = "aarch64"+;;! test = "winch"++(module+ (memory 1)+ (func (export "run")+ (v128.store (i32.const 0)+ (f32x4.add+ (v128.const f32x4 1.5 2.5 -3.25 100)+ (v128.const f32x4 0.25 -0.5 3.25 0.125)))))+;; wasm[0]::function[0]:+;; stp x29, x30, [sp, #-0x10]!+;; mov x29, sp+;; str x28, [sp, #-0x10]!+;; mov x28, sp+;; ldur x16, [x0, #8]+;; ldur x16, [x16, #0x18]+;; mov x17, #0+;; movk x17, #0x10+;; add x16, x16, x17+;; cmp sp, x16+;; b.lo #0x74+;; 2c: mov x9, x0+;; sub x28, x28, #0x10+;; mov sp, x28+;; stur x0, [x28, #8]+;; stur x1, [x28]+;; ldr q0, #0x80+;; ldr q1, #0x90+;; fadd v1.4s, v1.4s, v0.4s+;; mov x0, #0+;; ldur x1, [x9, #0x38]+;; add x1, x1, w0, uxtw+;; stur q1, [x1]+;; add x28, x28, #0x10+;; mov sp, x28+;; mov sp, x28+;; ldr x28, [sp], #0x10+;; ldp x29, x30, [sp], #0x10+;; ret+;; 74: udf #0xc11f+;; 78: udf #0+;; 7c: udf #0+;; 80: .byte 0x00, 0x00, 0x80, 0x3e+;; 84: .byte 0x00, 0x00, 0x00, 0xbf+;; 88: .byte 0x00, 0x00, 0x50, 0x40+;; 8c: .byte 0x00, 0x00, 0x00, 0x3e+;; 90: .byte 0x00, 0x00, 0xc0, 0x3f+;; 94: .byte 0x00, 0x00, 0x20, 0x40+;; 98: .byte 0x00, 0x00, 0x50, 0xc0+;; 9c: .byte 0x00, 0x00, 0xc8, 0x42diff --git a/tests/disas/winch/aarch64/v128_ops/f32x4_sub.wat b/tests/disas/winch/aarch64/v128_ops/f32x4_sub.watnew file mode 100644index 0000000000..6218af2c46--- /dev/null+++ b/tests/disas/winch/aarch64/v128_ops/f32x4_sub.wat@@ -0,0 +1,51 @@+;;! target = "aarch64"+;;! test = "winch"++(module+ (memory 1)+ (func (export "run")+ (v128.store (i32.const 0)+ (f32x4.sub+ (v128.const f32x4 1.5 2.5 -3.25 100)+ (v128.const f32x4 0.25 -0.5 3.25 0.125)))))+;; wasm[0]::function[0]:+;; stp x29, x30, [sp, #-0x10]!+;; mov x29, sp+;; str x28, [sp, #-0x10]!+;; mov x28, sp+;; ldur x16, [x0, #8]+;; ldur x16, [x16, #0x18]+;; mov x17, #0+;; movk x17, #0x10+;; add x16, x16, x17+;; cmp sp, x16+;; b.lo #0x74+;; 2c: mov x9, x0+;; sub x28, x28, #0x10+;; mov sp, x28+;; stur x0, [x28, #8]+;; stur x1, [x28]+;; ldr q0, #0x80+;; ldr q1, #0x90+;; fsub v1.4s, v1.4s, v0.4s+;; mov x0, #0+;; ldur x1, [x9, #0x38]+;; add x1, x1, w0, uxtw+;; stur q1, [x1]+;; add x28, x28, #0x10+;; mov sp, x28+;; mov sp, x28+;; ldr x28, [sp], #0x10+;; ldp x29, x30, [sp], #0x10+;; ret+;; 74: udf #0xc11f+;; 78: udf #0+;; 7c: udf #0+;; 80: .byte 0x00, 0x00, 0x80, 0x3e+;; 84: .byte 0x00, 0x00, 0x00, 0xbf+;; 88: .byte 0x00, 0x00, 0x50, 0x40+;; 8c: .byte 0x00, 0x00, 0x00, 0x3e+;; 90: .byte 0x00, 0x00, 0xc0, 0x3f+;; 94: .byte 0x00, 0x00, 0x20, 0x40+;; 98: .byte 0x00, 0x00, 0x50, 0xc0+;; 9c: .byte 0x00, 0x00, 0xc8, 0x42diff --git a/tests/disas/winch/aarch64/v128_ops/f64x2_add.wat b/tests/disas/winch/aarch64/v128_ops/f64x2_add.watnew file mode 100644index 0000000000..21bcd3df7b--- /dev/null+++ b/tests/disas/winch/aarch64/v128_ops/f64x2_add.wat@@ -0,0 +1,51 @@+;;! target = "aarch64"+;;! test = "winch"++(module+ (memory 1)+ (func (export "run")+ (v128.store (i32.const 0)+ (f64x2.add+ (v128.const f64x2 1.5 -2.25)+ (v128.const f64x2 0.75 2.5)))))+;; wasm[0]::function[0]:+;; stp x29, x30, [sp, #-0x10]!+;; mov x29, sp+;; str x28, [sp, #-0x10]!+;; mov x28, sp+;; ldur x16, [x0, #8]+;; ldur x16, [x16, #0x18]+;; mov x17, #0+;; movk x17, #0x10+;; add x16, x16, x17+;; cmp sp, x16+;; b.lo #0x74+;; 2c: mov x9, x0+;; sub x28, x28, #0x10+;; mov sp, x28+;; stur x0, [x28, #8]+;; stur x1, [x28]+;; ldr q0, #0x80+;; ldr q1, #0x90+;; fadd v1.2d, v1.2d, v0.2d+;; mov x0, #0+;; ldur x1, [x9, #0x38]+;; add x1, x1, w0, uxtw+;; stur q1, [x1]+;; add x28, x28, #0x10+;; mov sp, x28+;; mov sp, x28+;; ldr x28, [sp], #0x10+;; ldp x29, x30, [sp], #0x10+;; ret+;; 74: udf #0xc11f+;; 78: udf #0+;; 7c: udf #0+;; 80: udf #0+;; 84: .byte 0x00, 0x00, 0xe8, 0x3f+;; 88: udf #0+;; 8c: .byte 0x00, 0x00, 0x04, 0x40+;; 90: udf #0+;; 94: .byte 0x00, 0x00, 0xf8, 0x3f+;; 98: udf #0+;; 9c: mov z0.b, p0/m, za0h.b[w12, 0]diff --git a/tests/disas/winch/aarch64/v128_ops/f64x2_sub.wat b/tests/disas/winch/aarch64/v128_ops/f64x2_sub.watnew file mode 100644index 0000000000..da9e79bded--- /dev/null+++ b/tests/disas/winch/aarch64/v128_ops/f64x2_sub.wat@@ -0,0 +1,51 @@+;;! target = "aarch64"+;;! test = "winch"++(module+ (memory 1)+ (func (export "run")+ (v128.store (i32.const 0)+ (f64x2.sub+ (v128.const f64x2 1.5 -2.25)+ (v128.const f64x2 0.75 2.5)))))+;; wasm[0]::function[0]:+;; stp x29, x30, [sp, #-0x10]!+;; mov x29, sp+;; str x28, [sp, #-0x10]!+;; mov x28, sp+;; ldur x16, [x0, #8]+;; ldur x16, [x16, #0x18]+;; mov x17, #0+;; movk x17, #0x10+;; add x16, x16, x17+;; cmp sp, x16+;; b.lo #0x74+;; 2c: mov x9, x0+;; sub x28, x28, #0x10+;; mov sp, x28+;; stur x0, [x28, #8]+;; stur x1, [x28]+;; ldr q0, #0x80+;; ldr q1, #0x90+;; fsub v1.2d, v1.2d, v0.2d+;; mov x0, #0+;; ldur x1, [x9, #0x38]+;; add x1, x1, w0, uxtw+;; stur q1, [x1]+;; add x28, x28, #0x10+;; mov sp, x28+;; mov sp, x28+;; ldr x28, [sp], #0x10+;; ldp x29, x30, [sp], #0x10+;; ret+;; 74: udf #0xc11f+;; 78: udf #0+;; 7c: udf #0+;; 80: udf #0+;; 84: .byte 0x00, 0x00, 0xe8, 0x3f+;; 88: udf #0+;; 8c: .byte 0x00, 0x00, 0x04, 0x40+;; 90: udf #0+;; 94: .byte 0x00, 0x00, 0xf8, 0x3f+;; 98: udf #0+;; 9c: mov z0.b, p0/m, za0h.b[w12, 0]diff --git a/tests/disas/winch/aarch64/v128_ops/i16x8_add.wat b/tests/disas/winch/aarch64/v128_ops/i16x8_add.watnew file mode 100644index 0000000000..ec00729d01--- /dev/null+++ b/tests/disas/winch/aarch64/v128_ops/i16x8_add.wat@@ -0,0 +1,51 @@+;;! target = "aarch64"+;;! test = "winch"++(module+ (memory 1)+ (func (export "run")+ (v128.store (i32.const 0)+ (i16x8.add+ (v128.const i16x8 100 200 300 400 500 600 700 800)+ (v128.const i16x8 1 2 3 4 5 6 7 8)))))+;; wasm[0]::function[0]:+;; stp x29, x30, [sp, #-0x10]!+;; mov x29, sp+;; str x28, [sp, #-0x10]!+;; mov x28, sp+;; ldur x16, [x0, #8]+;; ldur x16, [x16, #0x18]+;; mov x17, #0+;; movk x17, #0x10+;; add x16, x16, x17+;; cmp sp, x16+;; b.lo #0x74+;; 2c: mov x9, x0+;; sub x28, x28, #0x10+;; mov sp, x28+;; stur x0, [x28, #8]+;; stur x1, [x28]+;; ldr q0, #0x80+;; ldr q1, #0x90+;; add v1.8h, v1.8h, v0.8h+;; mov x0, #0+;; ldur x1, [x9, #0x38]+;; add x1, x1, w0, uxtw+;; stur q1, [x1]+;; add x28, x28, #0x10+;; mov sp, x28+;; mov sp, x28+;; ldr x28, [sp], #0x10+;; ldp x29, x30, [sp], #0x10+;; ret+;; 74: udf #0xc11f+;; 78: udf #0+;; 7c: udf #0+;; 80: .byte 0x01, 0x00, 0x02, 0x00+;; 84: .byte 0x03, 0x00, 0x04, 0x00+;; 88: .byte 0x05, 0x00, 0x06, 0x00+;; 8c: .byte 0x07, 0x00, 0x08, 0x00+;; 90: .byte 0x64, 0x00, 0xc8, 0x00+;; 94: .byte 0x2c, 0x01, 0x90, 0x01+;; 98: .byte 0xf4, 0x01, 0x58, 0x02+;; 9c: .byte 0xbc, 0x02, 0x20, 0x03diff --git a/tests/disas/winch/aarch64/v128_ops/i16x8_add_sat_s.wat b/tests/disas/winch/aarch64/v128_ops/i16x8_add_sat_s.watnew file mode 100644index 0000000000..2fbc0248ec--- /dev/null+++ b/tests/disas/winch/aarch64/v128_ops/i16x8_add_sat_s.wat@@ -0,0 +1,51 @@+;;! target = "aarch64"+;;! test = "winch"++(module+ (memory 1)+ (func (export "run")+ (v128.store (i32.const 0)+ (i16x8.add_sat_s+ (v128.const i16x8 32767 -32768 1000 -1000 0 1 32767 -1)+ (v128.const i16x8 1 -1 1000 -1000 0 1 -1 32767)))))+;; wasm[0]::function[0]:+;; stp x29, x30, [sp, #-0x10]!+;; mov x29, sp+;; str x28, [sp, #-0x10]!+;; mov x28, sp+;; ldur x16, [x0, #8]+;; ldur x16, [x16, #0x18]+;; mov x17, #0+;; movk x17, #0x10+;; add x16, x16, x17+;; cmp sp, x16+;; b.lo #0x74+;; 2c: mov x9, x0+;; sub x28, x28, #0x10+;; mov sp, x28+;; stur x0, [x28, #8]+;; stur x1, [x28]+;; ldr q0, #0x80+;; ldr q1, #0x90+;; sqadd v1.8h, v1.8h, v0.8h+;; mov x0, #0+;; ldur x1, [x9, #0x38]+;; add x1, x1, w0, uxtw+;; stur q1, [x1]+;; add x28, x28, #0x10+;; mov sp, x28+;; mov sp, x28+;; ldr x28, [sp], #0x10+;; ldp x29, x30, [sp], #0x10+;; ret+;; 74: udf #0xc11f+;; 78: udf #0+;; 7c: udf #0+;; 80: .byte 0x01, 0x00, 0xff, 0xff+;; 84: stur d8, [sp, #-0x80]+;; 88: .byte 0x00, 0x00, 0x01, 0x00+;; 8c: .byte 0xff, 0xff, 0xff, 0x7f+;; 90: .byte 0xff, 0x7f, 0x00, 0x80+;; 94: stur d8, [sp, #-0x80]+;; 98: .byte 0x00, 0x00, 0x01, 0x00+;; 9c: .byte 0xff, 0x7f, 0xff, 0xffdiff --git a/tests/disas/winch/aarch64/v128_ops/i16x8_add_sat_u.wat b/tests/disas/winch/aarch64/v128_ops/i16x8_add_sat_u.watnew file mode 100644index 0000000000..7b631e6378--- /dev/null+++ b/tests/disas/winch/aarch64/v128_ops/i16x8_add_sat_u.wat@@ -0,0 +1,51 @@+;;! target = "aarch64"+;;! test = "winch"++(module+ (memory 1)+ (func (export "run")+ (v128.store (i32.const 0)+ (i16x8.add_sat_u+ (v128.const i16x8 65535 60000 1000 0 1 32768 65535 2)+ (v128.const i16x8 1 10000 1000 0 1 32768 0 65535)))))+;; wasm[0]::function[0]:+;; stp x29, x30, [sp, #-0x10]!+;; mov x29, sp+;; str x28, [sp, #-0x10]!+;; mov x28, sp+;; ldur x16, [x0, #8]+;; ldur x16, [x16, #0x18]+;; mov x17, #0+;; movk x17, #0x10+;; add x16, x16, x17+;; cmp sp, x16+;; b.lo #0x74+;; 2c: mov x9, x0+;; sub x28, x28, #0x10+;; mov sp, x28+;; stur x0, [x28, #8]+;; stur x1, [x28]+;; ldr q0, #0x80+;; ldr q1, #0x90+;; uqadd v1.8h, v1.8h, v0.8h+;; mov x0, #0+;; ldur x1, [x9, #0x38]+;; add x1, x1, w0, uxtw+;; stur q1, [x1]+;; add x28, x28, #0x10+;; mov sp, x28+;; mov sp, x28+;; ldr x28, [sp], #0x10+;; ldp x29, x30, [sp], #0x10+;; ret+;; 74: udf #0xc11f+;; 78: udf #0+;; 7c: udf #0+;; 80: .byte 0x01, 0x00, 0x10, 0x27+;; 84: udf #0x3e8+;; 88: .byte 0x01, 0x00, 0x00, 0x80+;; 8c: .byte 0x00, 0x00, 0xff, 0xff+;; 90: bics xzr, xzr, x0, lsr #63+;; 94: udf #0x3e8+;; 98: .byte 0x01, 0x00, 0x00, 0x80+;; 9c: .byte 0xff, 0xff, 0x02, 0x00diff --git a/tests/disas/winch/aarch64/v128_ops/i16x8_sub.wat b/tests/disas/winch/aarch64/v128_ops/i16x8_sub.watnew file mode 100644index 0000000000..adbb5795fe--- /dev/null+++ b/tests/disas/winch/aarch64/v128_ops/i16x8_sub.wat@@ -0,0 +1,51 @@+;;! target = "aarch64"+;;! test = "winch"++(module+ (memory 1)+ (func (export "run")+ (v128.store (i32.const 0)+ (i16x8.sub+ (v128.const i16x8 1000 2000 3000 4000 5 6 7 8)+ (v128.const i16x8 1 2 3 4 500 600 700 800)))))+;; wasm[0]::function[0]:+;; stp x29, x30, [sp, #-0x10]!+;; mov x29, sp+;; str x28, [sp, #-0x10]!+;; mov x28, sp+;; ldur x16, [x0, #8]+;; ldur x16, [x16, #0x18]+;; mov x17, #0+;; movk x17, #0x10+;; add x16, x16, x17+;; cmp sp, x16+;; b.lo #0x74+;; 2c: mov x9, x0+;; sub x28, x28, #0x10+;; mov sp, x28+;; stur x0, [x28, #8]+;; stur x1, [x28]+;; ldr q0, #0x80+;; ldr q1, #0x90+;; sub v1.8h, v1.8h, v0.8h+;; mov x0, #0+;; ldur x1, [x9, #0x38]+;; add x1, x1, w0, uxtw+;; stur q1, [x1]+;; add x28, x28, #0x10+;; mov sp, x28+;; mov sp, x28+;; ldr x28, [sp], #0x10+;; ldp x29, x30, [sp], #0x10+;; ret+;; 74: udf #0xc11f+;; 78: udf #0+;; 7c: udf #0+;; 80: .byte 0x01, 0x00, 0x02, 0x00+;; 84: .byte 0x03, 0x00, 0x04, 0x00+;; 88: .byte 0xf4, 0x01, 0x58, 0x02+;; 8c: .byte 0xbc, 0x02, 0x20, 0x03+;; 90: .byte 0xe8, 0x03, 0xd0, 0x07+;; 94: fmlal v24.2s, v29.2h, v0.h[6]+;; 98: .byte 0x05, 0x00, 0x06, 0x00+;; 9c: .byte 0x07, 0x00, 0x08, 0x00diff --git a/tests/disas/winch/aarch64/v128_ops/i16x8_sub_sat_s.wat b/tests/disas/winch/aarch64/v128_ops/i16x8_sub_sat_s.watnew file mode 100644index 0000000000..9065fffab3--- /dev/null+++ b/tests/disas/winch/aarch64/v128_ops/i16x8_sub_sat_s.wat@@ -0,0 +1,51 @@+;;! target = "aarch64"+;;! test = "winch"++(module+ (memory 1)+ (func (export "run")+ (v128.store (i32.const 0)+ (i16x8.sub_sat_s+ (v128.const i16x8 -32768 32767 1000 -1000 0 1 -32768 32767)+ (v128.const i16x8 1 -1 -1000 1000 0 1 32767 -32768)))))+;; wasm[0]::function[0]:+;; stp x29, x30, [sp, #-0x10]!+;; mov x29, sp+;; str x28, [sp, #-0x10]!+;; mov x28, sp+;; ldur x16, [x0, #8]+;; ldur x16, [x16, #0x18]+;; mov x17, #0+;; movk x17, #0x10+;; add x16, x16, x17+;; cmp sp, x16+;; b.lo #0x74+;; 2c: mov x9, x0+;; sub x28, x28, #0x10+;; mov sp, x28+;; stur x0, [x28, #8]+;; stur x1, [x28]+;; ldr q0, #0x80+;; ldr q1, #0x90+;; sqsub v1.8h, v1.8h, v0.8h+;; mov x0, #0+;; ldur x1, [x9, #0x38]+;; add x1, x1, w0, uxtw+;; stur q1, [x1]+;; add x28, x28, #0x10+;; mov sp, x28+;; mov sp, x28+;; ldr x28, [sp], #0x10+;; ldp x29, x30, [sp], #0x10+;; ret+;; 74: udf #0xc11f+;; 78: udf #0+;; 7c: udf #0+;; 80: .byte 0x01, 0x00, 0xff, 0xff+;; 84: .byte 0x18, 0xfc, 0xe8, 0x03+;; 88: .byte 0x00, 0x00, 0x01, 0x00+;; 8c: .byte 0xff, 0x7f, 0x00, 0x80+;; 90: .byte 0x00, 0x80, 0xff, 0x7f+;; 94: stur d8, [sp, #-0x80]+;; 98: .byte 0x00, 0x00, 0x01, 0x00+;; 9c: .byte 0x00, 0x80, 0xff, 0x7fdiff --git a/tests/disas/winch/aarch64/v128_ops/i16x8_sub_sat_u.wat b/tests/disas/winch/aarch64/v128_ops/i16x8_sub_sat_u.watnew file mode 100644index 0000000000..3bda2469b2--- /dev/null+++ b/tests/disas/winch/aarch64/v128_ops/i16x8_sub_sat_u.wat@@ -0,0 +1,51 @@+;;! target = "aarch64"+;;! test = "winch"++(module+ (memory 1)+ (func (export "run")+ (v128.store (i32.const 0)+ (i16x8.sub_sat_u+ (v128.const i16x8 3 65535 1000 0 1 32768 65535 2)+ (v128.const i16x8 10 1 1000 1 1 32769 0 65535)))))+;; wasm[0]::function[0]:+;; stp x29, x30, [sp, #-0x10]!+;; mov x29, sp+;; str x28, [sp, #-0x10]!+;; mov x28, sp+;; ldur x16, [x0, #8]+;; ldur x16, [x16, #0x18]+;; mov x17, #0+;; movk x17, #0x10+;; add x16, x16, x17+;; cmp sp, x16+;; b.lo #0x74+;; 2c: mov x9, x0+;; sub x28, x28, #0x10+;; mov sp, x28+;; stur x0, [x28, #8]+;; stur x1, [x28]+;; ldr q0, #0x80+;; ldr q1, #0x90+;; uqsub v1.8h, v1.8h, v0.8h+;; mov x0, #0+;; ldur x1, [x9, #0x38]+;; add x1, x1, w0, uxtw+;; stur q1, [x1]+;; add x28, x28, #0x10+;; mov sp, x28+;; mov sp, x28+;; ldr x28, [sp], #0x10+;; ldp x29, x30, [sp], #0x10+;; ret+;; 74: udf #0xc11f+;; 78: udf #0+;; 7c: udf #0+;; 80: .byte 0x0a, 0x00, 0x01, 0x00+;; 84: .byte 0xe8, 0x03, 0x01, 0x00+;; 88: .byte 0x01, 0x00, 0x01, 0x80+;; 8c: .byte 0x00, 0x00, 0xff, 0xff+;; 90: .byte 0x03, 0x00, 0xff, 0xff+;; 94: udf #0x3e8+;; 98: .byte 0x01, 0x00, 0x00, 0x80+;; 9c: .byte 0xff, 0xff, 0x02, 0x00diff --git a/tests/disas/winch/aarch64/v128_ops/i32x4_add.wat b/tests/disas/winch/aarch64/v128_ops/i32x4_add.watnew file mode 100644index 0000000000..44d7a775cb--- /dev/null+++ b/tests/disas/winch/aarch64/v128_ops/i32x4_add.wat@@ -0,0 +1,51 @@+;;! target = "aarch64"+;;! test = "winch"++(module+ (memory 1)+ (func (export "run")+ (v128.store (i32.const 0)+ (i32x4.add+ (v128.const i32x4 1 2 3 4)+ (v128.const i32x4 10 20 30 40)))))+;; wasm[0]::function[0]:+;; stp x29, x30, [sp, #-0x10]!+;; mov x29, sp+;; str x28, [sp, #-0x10]!+;; mov x28, sp+;; ldur x16, [x0, #8]+;; ldur x16, [x16, #0x18]+;; mov x17, #0+;; movk x17, #0x10+;; add x16, x16, x17+;; cmp sp, x16+;; b.lo #0x74+;; 2c: mov x9, x0+;; sub x28, x28, #0x10+;; mov sp, x28+;; stur x0, [x28, #8]+;; stur x1, [x28]+;; ldr q0, #0x80+;; ldr q1, #0x90+;; add v1.4s, v1.4s, v0.4s+;; mov x0, #0+;; ldur x1, [x9, #0x38]+;; add x1, x1, w0, uxtw+;; stur q1, [x1]+;; add x28, x28, #0x10+;; mov sp, x28+;; mov sp, x28+;; ldr x28, [sp], #0x10+;; ldp x29, x30, [sp], #0x10+;; ret+;; 74: udf #0xc11f+;; 78: udf #0+;; 7c: udf #0+;; 80: udf #0xa+;; 84: udf #0x14+;; 88: udf #0x1e+;; 8c: udf #0x28+;; 90: udf #1+;; 94: udf #2+;; 98: udf #3+;; 9c: udf #4diff --git a/tests/disas/winch/aarch64/v128_ops/i32x4_sub.wat b/tests/disas/winch/aarch64/v128_ops/i32x4_sub.watnew file mode 100644index 0000000000..5e1151c9b9--- /dev/null+++ b/tests/disas/winch/aarch64/v128_ops/i32x4_sub.wat@@ -0,0 +1,51 @@+;;! target = "aarch64"+;;! test = "winch"++(module+ (memory 1)+ (func (export "run")+ (v128.store (i32.const 0)+ (i32x4.sub+ (v128.const i32x4 100 200 300 400)+ (v128.const i32x4 1 2 3 4)))))+;; wasm[0]::function[0]:+;; stp x29, x30, [sp, #-0x10]!+;; mov x29, sp+;; str x28, [sp, #-0x10]!+;; mov x28, sp+;; ldur x16, [x0, #8]+;; ldur x16, [x16, #0x18]+;; mov x17, #0+;; movk x17, #0x10+;; add x16, x16, x17+;; cmp sp, x16+;; b.lo #0x74+;; 2c: mov x9, x0+;; sub x28, x28, #0x10+;; mov sp, x28+;; stur x0, [x28, #8]+;; stur x1, [x28]+;; ldr q0, #0x80+;; ldr q1, #0x90+;; sub v1.4s, v1.4s, v0.4s+;; mov x0, #0+;; ldur x1, [x9, #0x38]+;; add x1, x1, w0, uxtw+;; stur q1, [x1]+;; add x28, x28, #0x10+;; mov sp, x28+;; mov sp, x28+;; ldr x28, [sp], #0x10+;; ldp x29, x30, [sp], #0x10+;; ret+;; 74: udf #0xc11f+;; 78: udf #0+;; 7c: udf #0+;; 80: udf #1+;; 84: udf #2+;; 88: udf #3+;; 8c: udf #4+;; 90: udf #0x64+;; 94: udf #0xc8+;; 98: udf #0x12c+;; 9c: udf #0x190diff --git a/tests/disas/winch/aarch64/v128_ops/i64x2_add.wat b/tests/disas/winch/aarch64/v128_ops/i64x2_add.watnew file mode 100644index 0000000000..db135a5d67--- /dev/null+++ b/tests/disas/winch/aarch64/v128_ops/i64x2_add.wat@@ -0,0 +1,51 @@+;;! target = "aarch64"+;;! test = "winch"++(module+ (memory 1)+ (func (export "run")+ (v128.store (i32.const 0)+ (i64x2.add+ (v128.const i64x2 100 200)+ (v128.const i64x2 7 9)))))+;; wasm[0]::function[0]:+;; stp x29, x30, [sp, #-0x10]!+;; mov x29, sp+;; str x28, [sp, #-0x10]!+;; mov x28, sp+;; ldur x16, [x0, #8]+;; ldur x16, [x16, #0x18]+;; mov x17, #0+;; movk x17, #0x10+;; add x16, x16, x17+;; cmp sp, x16+;; b.lo #0x74+;; 2c: mov x9, x0+;; sub x28, x28, #0x10+;; mov sp, x28+;; stur x0, [x28, #8]+;; stur x1, [x28]+;; ldr q0, #0x80+;; ldr q1, #0x90+;; add v1.2d, v1.2d, v0.2d+;; mov x0, #0+;; ldur x1, [x9, #0x38]+;; add x1, x1, w0, uxtw+;; stur q1, [x1]+;; add x28, x28, #0x10+;; mov sp, x28+;; mov sp, x28+;; ldr x28, [sp], #0x10+;; ldp x29, x30, [sp], #0x10+;; ret+;; 74: udf #0xc11f+;; 78: udf #0+;; 7c: udf #0+;; 80: udf #7+;; 84: udf #0+;; 88: udf #9+;; 8c: udf #0+;; 90: udf #0x64+;; 94: udf #0+;; 98: udf #0xc8+;; 9c: udf #0diff --git a/tests/disas/winch/aarch64/v128_ops/i64x2_sub.wat b/tests/disas/winch/aarch64/v128_ops/i64x2_sub.watnew file mode 100644index 0000000000..e597b2e815--- /dev/null+++ b/tests/disas/winch/aarch64/v128_ops/i64x2_sub.wat@@ -0,0 +1,51 @@+;;! target = "aarch64"+;;! test = "winch"++(module+ (memory 1)+ (func (export "run")+ (v128.store (i32.const 0)+ (i64x2.sub+ (v128.const i64x2 100 200)+ (v128.const i64x2 7 9)))))+;; wasm[0]::function[0]:+;; stp x29, x30, [sp, #-0x10]!+;; mov x29, sp+;; str x28, [sp, #-0x10]!+;; mov x28, sp+;; ldur x16, [x0, #8]+;; ldur x16, [x16, #0x18]+;; mov x17, #0+;; movk x17, #0x10+;; add x16, x16, x17+;; cmp sp, x16+;; b.lo #0x74+;; 2c: mov x9, x0+;; sub x28, x28, #0x10+;; mov sp, x28+;; stur x0, [x28, #8]+;; stur x1, [x28]+;; ldr q0, #0x80+;; ldr q1, #0x90+;; sub v1.2d, v1.2d, v0.2d+;; mov x0, #0+;; ldur x1, [x9, #0x38]+;; add x1, x1, w0, uxtw+;; stur q1, [x1]+;; add x28, x28, #0x10+;; mov sp, x28+;; mov sp, x28+;; ldr x28, [sp], #0x10+;; ldp x29, x30, [sp], #0x10+;; ret+;; 74: udf #0xc11f+;; 78: udf #0+;; 7c: udf #0+;; 80: udf #7+;; 84: udf #0+;; 88: udf #9+;; 8c: udf #0+;; 90: udf #0x64+;; 94: udf #0+;; 98: udf #0xc8+;; 9c: udf #0diff --git a/tests/disas/winch/aarch64/v128_ops/i8x16_add.wat b/tests/disas/winch/aarch64/v128_ops/i8x16_add.watnew file mode 100644index 0000000000..55fb7cc96a--- /dev/null+++ b/tests/disas/winch/aarch64/v128_ops/i8x16_add.wat@@ -0,0 +1,51 @@+;;! target = "aarch64"+;;! test = "winch"++(module+ (memory 1)+ (func (export "run")+ (v128.store (i32.const 0)+ (i8x16.add+ (v128.const i8x16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16)+ (v128.const i8x16 10 10 10 10 10 10 10 10 20 20 20 20 20 20 20 20)))))+;; wasm[0]::function[0]:+;; stp x29, x30, [sp, #-0x10]!+;; mov x29, sp+;; str x28, [sp, #-0x10]!+;; mov x28, sp+;; ldur x16, [x0, #8]+;; ldur x16, [x16, #0x18]+;; mov x17, #0+;; movk x17, #0x10+;; add x16, x16, x17+;; cmp sp, x16+;; b.lo #0x74+;; 2c: mov x9, x0+;; sub x28, x28, #0x10+;; mov sp, x28+;; stur x0, [x28, #8]+;; stur x1, [x28]+;; ldr q0, #0x80+;; ldr q1, #0x90+;; add v1.16b, v1.16b, v0.16b+;; mov x0, #0+;; ldur x1, [x9, #0x38]+;; add x1, x1, w0, uxtw+;; stur q1, [x1]+;; add x28, x28, #0x10+;; mov sp, x28+;; mov sp, x28+;; ldr x28, [sp], #0x10+;; ldp x29, x30, [sp], #0x10+;; ret+;; 74: udf #0xc11f+;; 78: udf #0+;; 7c: udf #0+;; 80: and w10, w16, w10, lsl #2+;; 84: and w10, w16, w10, lsl #2+;; 88: b #0x5050d8+;; 8c: b #0x5050dc+;; 90: subr z1.b, p0/m, z1.b, z16.b+;; 94: stxrb w7, w5, [x16]+;; 98: .byte 0x09, 0x0a, 0x0b, 0x0c+;; 9c: adr x13, #0x1e25cdiff --git a/tests/disas/winch/aarch64/v128_ops/i8x16_add_sat_s.wat b/tests/disas/winch/aarch64/v128_ops/i8x16_add_sat_s.watnew file mode 100644index 0000000000..c23dcec734--- /dev/null+++ b/tests/disas/winch/aarch64/v128_ops/i8x16_add_sat_s.wat@@ -0,0 +1,51 @@+;;! target = "aarch64"+;;! test = "winch"++(module+ (memory 1)+ (func (export "run")+ (v128.store (i32.const 0)+ (i8x16.add_sat_s+ (v128.const i8x16 127 -128 100 -100 0 1 -1 50 127 -128 5 -5 64 -64 127 -1)+ (v128.const i8x16 1 -1 100 -100 0 1 -1 50 127 -128 5 -5 64 -64 -1 127)))))+;; wasm[0]::function[0]:+;; stp x29, x30, [sp, #-0x10]!+;; mov x29, sp+;; str x28, [sp, #-0x10]!+;; mov x28, sp+;; ldur x16, [x0, #8]+;; ldur x16, [x16, #0x18]+;; mov x17, #0+;; movk x17, #0x10+;; add x16, x16, x17+;; cmp sp, x16+;; b.lo #0x74+;; 2c: mov x9, x0+;; sub x28, x28, #0x10+;; mov sp, x28+;; stur x0, [x28, #8]+;; stur x1, [x28]+;; ldr q0, #0x80+;; ldr q1, #0x90+;; sqadd v1.16b, v1.16b, v0.16b+;; mov x0, #0+;; ldur x1, [x9, #0x38]+;; add x1, x1, w0, uxtw+;; stur q1, [x1]+;; add x28, x28, #0x10+;; mov sp, x28+;; mov sp, x28+;; ldr x28, [sp], #0x10+;; ldp x29, x30, [sp], #0x10+;; ret+;; 74: udf #0xc11f+;; 78: udf #0+;; 7c: udf #0+;; 80: ldr q1, #0xca060+;; 84: .byte 0x00, 0x01, 0xff, 0x32+;; 88: .byte 0x7f, 0x80, 0x05, 0xfb+;; 8c: .byte 0x40, 0xc0, 0xff, 0x7f+;; 90: ldr q31, #0xc909c+;; 94: .byte 0x00, 0x01, 0xff, 0x32+;; 98: .byte 0x7f, 0x80, 0x05, 0xfb+;; 9c: .byte 0x40, 0xc0, 0x7f, 0xffdiff --git a/tests/disas/winch/aarch64/v128_ops/i8x16_add_sat_u.wat b/tests/disas/winch/aarch64/v128_ops/i8x16_add_sat_u.watnew file mode 100644index 0000000000..e7348d0ed4--- /dev/null+++ b/tests/disas/winch/aarch64/v128_ops/i8x16_add_sat_u.wat@@ -0,0 +1,51 @@+;;! target = "aarch64"+;;! test = "winch"++(module+ (memory 1)+ (func (export "run")+ (v128.store (i32.const 0)+ (i8x16.add_sat_u+ (v128.const i8x16 255 200 100 0 1 128 254 255 3 5 250 250 64 64 255 0)+ (v128.const i8x16 1 100 100 0 1 128 1 255 3 5 10 6 64 64 0 255)))))+;; wasm[0]::function[0]:+;; stp x29, x30, [sp, #-0x10]!+;; mov x29, sp+;; str x28, [sp, #-0x10]!+;; mov x28, sp+;; ldur x16, [x0, #8]+;; ldur x16, [x16, #0x18]+;; mov x17, #0+;; movk x17, #0x10+;; add x16, x16, x17+;; cmp sp, x16+;; b.lo #0x74+;; 2c: mov x9, x0+;; sub x28, x28, #0x10+;; mov sp, x28+;; stur x0, [x28, #8]+;; stur x1, [x28]+;; ldr q0, #0x80+;; ldr q1, #0x90+;; uqadd v1.16b, v1.16b, v0.16b+;; mov x0, #0+;; ldur x1, [x9, #0x38]+;; add x1, x1, w0, uxtw+;; stur q1, [x1]+;; add x28, x28, #0x10+;; mov sp, x28+;; mov sp, x28+;; ldr x28, [sp], #0x10+;; ldp x29, x30, [sp], #0x10+;; ret+;; 74: udf #0xc11f+;; 78: udf #0+;; 7c: udf #0+;; 80: .byte 0x01, 0x64, 0x64, 0x00+;; 84: .byte 0x01, 0x80, 0x01, 0xff+;; 88: .byte 0x03, 0x05, 0x0a, 0x06+;; 8c: .byte 0x40, 0x40, 0x00, 0xff+;; 90: .byte 0xff, 0xc8, 0x64, 0x00+;; 94: .byte 0x01, 0x80, 0xfe, 0xff+;; 98: .byte 0x03, 0x05, 0xfa, 0xfa+;; 9c: .byte 0x40, 0x40, 0xff, 0x00diff --git a/tests/disas/winch/aarch64/v128_ops/i8x16_sub.wat b/tests/disas/winch/aarch64/v128_ops/i8x16_sub.watnew file mode 100644index 0000000000..19a0445068--- /dev/null+++ b/tests/disas/winch/aarch64/v128_ops/i8x16_sub.wat@@ -0,0 +1,51 @@+;;! target = "aarch64"+;;! test = "winch"++(module+ (memory 1)+ (func (export "run")+ (v128.store (i32.const 0)+ (i8x16.sub+ (v128.const i8x16 50 60 70 80 90 100 110 120 5 10 15 20 25 30 35 40)+ (v128.const i8x16 10 10 10 10 10 10 10 10 10 20 30 40 50 60 70 80)))))+;; wasm[0]::function[0]:+;; stp x29, x30, [sp, #-0x10]!+;; mov x29, sp+;; str x28, [sp, #-0x10]!+;; mov x28, sp+;; ldur x16, [x0, #8]+;; ldur x16, [x16, #0x18]+;; mov x17, #0+;; movk x17, #0x10+;; add x16, x16, x17+;; cmp sp, x16+;; b.lo #0x74+;; 2c: mov x9, x0+;; sub x28, x28, #0x10+;; mov sp, x28+;; stur x0, [x28, #8]+;; stur x1, [x28]+;; ldr q0, #0x80+;; ldr q1, #0x90+;; sub v1.16b, v1.16b, v0.16b+;; mov x0, #0+;; ldur x1, [x9, #0x38]+;; add x1, x1, w0, uxtw+;; stur q1, [x1]+;; add x28, x28, #0x10+;; mov sp, x28+;; mov sp, x28+;; ldr x28, [sp], #0x10+;; ldp x29, x30, [sp], #0x10+;; ret+;; 74: udf #0xc11f+;; 78: udf #0+;; 7c: udf #0+;; 80: and w10, w16, w10, lsl #2+;; 84: and w10, w16, w10, lsl #2+;; 88: stnp w10, w5, [x0, #0xf0]+;; 8c: adr x18, #0x8c812+;; 90: adr x18, #0x8c816+;; 94: .byte 0x5a, 0x64, 0x6e, 0x78+;; 98: b #0x3c28ac+;; 9c: stnp w25, w7, [x16, #-0xe8]diff --git a/tests/disas/winch/aarch64/v128_ops/i8x16_sub_sat_s.wat b/tests/disas/winch/aarch64/v128_ops/i8x16_sub_sat_s.watnew file mode 100644index 0000000000..9ea51b6454--- /dev/null+++ b/tests/disas/winch/aarch64/v128_ops/i8x16_sub_sat_s.wat@@ -0,0 +1,51 @@+;;! target = "aarch64"+;;! test = "winch"++(module+ (memory 1)+ (func (export "run")+ (v128.store (i32.const 0)+ (i8x16.sub_sat_s+ (v128.const i8x16 -128 127 -100 100 0 1 -1 50 -128 127 5 -5 -64 64 -128 0)+ (v128.const i8x16 1 -1 100 -100 0 1 -1 -50 127 -128 5 -5 64 -64 0 -128)))))+;; wasm[0]::function[0]:+;; stp x29, x30, [sp, #-0x10]!+;; mov x29, sp+;; str x28, [sp, #-0x10]!+;; mov x28, sp+;; ldur x16, [x0, #8]+;; ldur x16, [x16, #0x18]+;; mov x17, #0+;; movk x17, #0x10+;; add x16, x16, x17+;; cmp sp, x16+;; b.lo #0x74+;; 2c: mov x9, x0+;; sub x28, x28, #0x10+;; mov sp, x28+;; stur x0, [x28, #8]+;; stur x1, [x28]+;; ldr q0, #0x80+;; ldr q1, #0x90+;; sqsub v1.16b, v1.16b, v0.16b+;; mov x0, #0+;; ldur x1, [x9, #0x38]+;; add x1, x1, w0, uxtw+;; stur q1, [x1]+;; add x28, x28, #0x10+;; mov sp, x28+;; mov sp, x28+;; ldr x28, [sp], #0x10+;; ldp x29, x30, [sp], #0x10+;; ret+;; 74: udf #0xc11f+;; 78: udf #0+;; 7c: udf #0+;; 80: ldr q1, #0xca060+;; 84: .byte 0x00, 0x01, 0xff, 0xce+;; 88: .byte 0x7f, 0x80, 0x05, 0xfb+;; 8c: .byte 0x40, 0xc0, 0x00, 0x80+;; 90: fcmla z0.s, p7/m, z28.s, z28.s, #0x10e+;; 94: .byte 0x00, 0x01, 0xff, 0x32+;; 98: .byte 0x80, 0x7f, 0x05, 0xfb+;; 9c: .byte 0xc0, 0x40, 0x80, 0x00diff --git a/tests/disas/winch/aarch64/v128_ops/i8x16_sub_sat_u.wat b/tests/disas/winch/aarch64/v128_ops/i8x16_sub_sat_u.watnew file mode 100644index 0000000000..c04504c988--- /dev/null+++ b/tests/disas/winch/aarch64/v128_ops/i8x16_sub_sat_u.wat@@ -0,0 +1,51 @@+;;! target = "aarch64"+;;! test = "winch"++(module+ (memory 1)+ (func (export "run")+ (v128.store (i32.const 0)+ (i8x16.sub_sat_u+ (v128.const i8x16 3 255 100 0 1 128 254 255 3 5 250 6 64 64 255 0)+ (v128.const i8x16 10 1 100 1 1 129 255 254 3 5 10 250 65 63 0 255)))))+;; wasm[0]::function[0]:+;; stp x29, x30, [sp, #-0x10]!+;; mov x29, sp+;; str x28, [sp, #-0x10]!+;; mov x28, sp+;; ldur x16, [x0, #8]+;; ldur x16, [x16, #0x18]+;; mov x17, #0+;; movk x17, #0x10+;; add x16, x16, x17+;; cmp sp, x16+;; b.lo #0x74+;; 2c: mov x9, x0+;; sub x28, x28, #0x10+;; mov sp, x28+;; stur x0, [x28, #8]+;; stur x1, [x28]+;; ldr q0, #0x80+;; ldr q1, #0x90+;; uqsub v1.16b, v1.16b, v0.16b+;; mov x0, #0+;; ldur x1, [x9, #0x38]+;; add x1, x1, w0, uxtw+;; stur q1, [x1]+;; add x28, x28, #0x10+;; mov sp, x28+;; mov sp, x28+;; ldr x28, [sp], #0x10+;; ldp x29, x30, [sp], #0x10+;; ret+;; 74: udf #0xc11f+;; 78: udf #0+;; 7c: udf #0+;; 80: .byte 0x0a, 0x01, 0x64, 0x01+;; 84: .byte 0x01, 0x81, 0xff, 0xfe+;; 88: .byte 0x03, 0x05, 0x0a, 0xfa+;; 8c: .byte 0x41, 0x3f, 0x00, 0xff+;; 90: .byte 0x03, 0xff, 0x64, 0x00+;; 94: .byte 0x01, 0x80, 0xfe, 0xff+;; 98: .byte 0x03, 0x05, 0xfa, 0x06+;; 9c: .byte 0x40, 0x40, 0xff, 0x00diff --git a/winch/codegen/src/isa/aarch64/masm.rs b/winch/codegen/src/isa/aarch64/masm.rsindex b4e83b385b..83a2420144 100644--- a/winch/codegen/src/isa/aarch64/masm.rs+++ b/winch/codegen/src/isa/aarch64/masm.rs@@ -774,6 +774,9 @@ impl Masm for MacroAssembler {_reg: WritableReg,_lane_size: OperandSize,) -> Result<()> {+ if !self.shared_flags.enable_nan_canonicalization() {+ return Ok(());+ }bail!(CodeGenError::unimplemented_masm_instruction())}@@ -1518,24 +1521,96 @@ impl Masm for MacroAssembler {bail!(CodeGenError::unimplemented_masm_instruction())}- fn v128_add(- &mut self,- _lhs: Reg,- _rhs: Reg,- _dst: WritableReg,- _kind: V128AddKind,- ) -> Result<()> {- Err(format_err!(CodeGenError::unimplemented_masm_instruction()))+ fn v128_add(&mut self, lhs: Reg, rhs: Reg, dst: WritableReg, kind: V128AddKind) -> Result<()> {+ match kind {+ V128AddKind::F32x4 => {+ self.asm+ .vec_rrr(VecALUOp::Fadd, lhs, rhs, dst, VectorSize::Size32x4);+ }+ V128AddKind::F64x2 => {+ self.asm+ .vec_rrr(VecALUOp::Fadd, lhs, rhs, dst, VectorSize::Size64x2);+ }+ V128AddKind::I8x16 => {+ self.asm+ .vec_rrr(VecALUOp::Add, lhs, rhs, dst, VectorSize::Size8x16);+ }+ V128AddKind::I8x16SatS => {+ self.asm+ .vec_rrr(VecALUOp::Sqadd, lhs, rhs, dst, VectorSize::Size8x16);+ }+ V128AddKind::I8x16SatU => {+ self.asm+ .vec_rrr(VecALUOp::Uqadd, lhs, rhs, dst, VectorSize::Size8x16);+ }+ V128AddKind::I16x8 => {+ self.asm+ .vec_rrr(VecALUOp::Add, lhs, rhs, dst, VectorSize::Size16x8);+ }+ V128AddKind::I16x8SatS => {+ self.asm+ .vec_rrr(VecALUOp::Sqadd, lhs, rhs, dst, VectorSize::Size16x8);+ }+ V128AddKind::I16x8SatU => {+ self.asm+ .vec_rrr(VecALUOp::Uqadd, lhs, rhs, dst, VectorSize::Size16x8);+ }+ V128AddKind::I32x4 => {+ self.asm+ .vec_rrr(VecALUOp::Add, lhs, rhs, dst, VectorSize::Size32x4);+ }+ V128AddKind::I64x2 => {+ self.asm+ .vec_rrr(VecALUOp::Add, lhs, rhs, dst, VectorSize::Size64x2);+ }+ }+ Ok(())}- fn v128_sub(- &mut self,- _lhs: Reg,- _rhs: Reg,- _dst: WritableReg,- _kind: V128SubKind,- ) -> Result<()> {- Err(format_err!(CodeGenError::unimplemented_masm_instruction()))+ fn v128_sub(&mut self, lhs: Reg, rhs: Reg, dst: WritableReg, kind: V128SubKind) -> Result<()> {+ match kind {+ V128SubKind::F32x4 => {+ self.asm+ .vec_rrr(VecALUOp::Fsub, lhs, rhs, dst, VectorSize::Size32x4);+ }+ V128SubKind::F64x2 => {+ self.asm+ .vec_rrr(VecALUOp::Fsub, lhs, rhs, dst, VectorSize::Size64x2);+ }+ V128SubKind::I8x16 => {+ self.asm+ .vec_rrr(VecALUOp::Sub, lhs, rhs, dst, VectorSize::Size8x16);+ }+ V128SubKind::I8x16SatS => {+ self.asm+ .vec_rrr(VecALUOp::Sqsub, lhs, rhs, dst, VectorSize::Size8x16);+ }+ V128SubKind::I8x16SatU => {+ self.asm+ .vec_rrr(VecALUOp::Uqsub, lhs, rhs, dst, VectorSize::Size8x16);+ }+ V128SubKind::I16x8 => {+ self.asm+ .vec_rrr(VecALUOp::Sub, lhs, rhs, dst, VectorSize::Size16x8);+ }+ V128SubKind::I16x8SatS => {+ self.asm+ .vec_rrr(VecALUOp::Sqsub, lhs, rhs, dst, VectorSize::Size16x8);+ }+ V128SubKind::I16x8SatU => {+ self.asm+ .vec_rrr(VecALUOp::Uqsub, lhs, rhs, dst, VectorSize::Size16x8);+ }+ V128SubKind::I32x4 => {+ self.asm+ .vec_rrr(VecALUOp::Sub, lhs, rhs, dst, VectorSize::Size32x4);+ }+ V128SubKind::I64x2 => {+ self.asm+ .vec_rrr(VecALUOp::Sub, lhs, rhs, dst, VectorSize::Size64x2);+ }+ }+ Ok(())}fn v128_mul(