FEX/unittests/InstructionCountCI/Secondary_REP.json
2024-06-25 11:37:18 -07:00

637 lines
15 KiB
JSON

{
"Features": {
"Bitness": 64,
"EnabledHostFeatures": [],
"DisabledHostFeatures": [
"SVE128",
"SVE256",
"RPRES",
"AFP",
"FLAGM",
"FLAGM2"
]
},
"Instructions": {
"movss xmm0, xmm1": {
"ExpectedInstructionCount": 1,
"Comment": "0xf3 0x0f 0x10",
"ExpectedArm64ASM": [
"mov v16.s[0], v17.s[0]"
]
},
"movss xmm0, [rax]": {
"ExpectedInstructionCount": 1,
"Comment": "0xf3 0x0f 0x10",
"ExpectedArm64ASM": [
"ldr s16, [x4]"
]
},
"movss [rax], xmm0": {
"ExpectedInstructionCount": 1,
"Comment": "0xf3 0x0f 0x11",
"ExpectedArm64ASM": [
"str s16, [x4]"
]
},
"movsldup xmm0, xmm1": {
"ExpectedInstructionCount": 1,
"Comment": "0xf3 0x0f 0x12",
"ExpectedArm64ASM": [
"trn1 v16.4s, v17.4s, v17.4s"
]
},
"movsldup xmm0, [rax]": {
"ExpectedInstructionCount": 2,
"Comment": "0xf3 0x0f 0x12",
"ExpectedArm64ASM": [
"ldr q2, [x4]",
"trn1 v16.4s, v2.4s, v2.4s"
]
},
"movshdup xmm0, xmm1": {
"ExpectedInstructionCount": 1,
"Comment": "0xf3 0x0f 0x16",
"ExpectedArm64ASM": [
"trn2 v16.4s, v17.4s, v17.4s"
]
},
"movshdup xmm0, [rax]": {
"ExpectedInstructionCount": 2,
"Comment": "0xf3 0x0f 0x16",
"ExpectedArm64ASM": [
"ldr q2, [x4]",
"trn2 v16.4s, v2.4s, v2.4s"
]
},
"cvtsi2ss xmm0, eax": {
"ExpectedInstructionCount": 2,
"Comment": [
"0xf3 0x0f 0x2a"
],
"ExpectedArm64ASM": [
"scvtf s0, w4",
"mov v16.s[0], v0.s[0]"
]
},
"cvtsi2ss xmm0, dword [rax]": {
"ExpectedInstructionCount": 3,
"Comment": [
"0xf3 0x0f 0x2a"
],
"ExpectedArm64ASM": [
"ldr s2, [x4]",
"scvtf s0, s2",
"mov v16.s[0], v0.s[0]"
]
},
"cvtsi2ss xmm0, rax": {
"ExpectedInstructionCount": 2,
"Comment": "0xf3 0x0f 0x2a",
"ExpectedArm64ASM": [
"scvtf s0, x4",
"mov v16.s[0], v0.s[0]"
]
},
"cvtsi2ss xmm0, qword [rax]": {
"ExpectedInstructionCount": 3,
"Comment": [
"0xf3 0x0f 0x2a"
],
"ExpectedArm64ASM": [
"ldr x20, [x4]",
"scvtf s0, x20",
"mov v16.s[0], v0.s[0]"
]
},
"movntss [rax], xmm0": {
"ExpectedInstructionCount": 1,
"Comment": "0xf3 0x0f 0x2b",
"ExpectedArm64ASM": [
"str s16, [x4]"
]
},
"cvttss2si eax, xmm0": {
"ExpectedInstructionCount": 1,
"Comment": "0xf3 0x0f 0x2c",
"ExpectedArm64ASM": [
"fcvtzs w4, s16"
]
},
"cvttss2si eax, dword [rbx]": {
"ExpectedInstructionCount": 2,
"Comment": "0xf3 0x0f 0x2c",
"ExpectedArm64ASM": [
"ldr s2, [x7]",
"fcvtzs w4, s2"
]
},
"cvttss2si rax, xmm0": {
"ExpectedInstructionCount": 1,
"Comment": "0xf3 0x0f 0x2c",
"ExpectedArm64ASM": [
"fcvtzs x4, s16"
]
},
"cvttss2si rax, dword [rbx]": {
"ExpectedInstructionCount": 2,
"Comment": "0xf3 0x0f 0x2c",
"ExpectedArm64ASM": [
"ldr d2, [x7]",
"fcvtzs x4, s2"
]
},
"cvtss2si eax, xmm0": {
"ExpectedInstructionCount": 2,
"Comment": "0xf3 0x0f 0x2d",
"ExpectedArm64ASM": [
"frinti s0, s16",
"fcvtzs w4, s0"
]
},
"cvtss2si eax, dword [rbx]": {
"ExpectedInstructionCount": 3,
"Comment": "0xf3 0x0f 0x2d",
"ExpectedArm64ASM": [
"ldr s2, [x7]",
"frinti s0, s2",
"fcvtzs w4, s0"
]
},
"cvtss2si rax, xmm0": {
"ExpectedInstructionCount": 2,
"Comment": "0xf3 0x0f 0x2d",
"ExpectedArm64ASM": [
"frinti s0, s16",
"fcvtzs x4, s0"
]
},
"cvtss2si rax, dword [rbx]": {
"ExpectedInstructionCount": 3,
"Comment": "0xf3 0x0f 0x2d",
"ExpectedArm64ASM": [
"ldr d2, [x7]",
"frinti s0, s2",
"fcvtzs x4, s0"
]
},
"sqrtss xmm0, xmm1": {
"ExpectedInstructionCount": 2,
"Comment": "0xf3 0x0f 0x51",
"ExpectedArm64ASM": [
"fsqrt s0, s17",
"mov v16.s[0], v0.s[0]"
]
},
"rsqrtss xmm0, xmm1": {
"ExpectedInstructionCount": 4,
"Comment": [
"0xf3 0x0f 0x52"
],
"ExpectedArm64ASM": [
"fmov s0, #0x70 (1.0000)",
"fsqrt s1, s17",
"fdiv s0, s0, s1",
"mov v16.s[0], v0.s[0]"
]
},
"rcpss xmm0, xmm1": {
"ExpectedInstructionCount": 3,
"Comment": [
"0xf3 0x0f 0x53"
],
"ExpectedArm64ASM": [
"fmov s0, #0x70 (1.0000)",
"fdiv s0, s0, s17",
"mov v16.s[0], v0.s[0]"
]
},
"addss xmm0, xmm1": {
"ExpectedInstructionCount": 2,
"Comment": [
"0xf3 0x0f 0x58"
],
"ExpectedArm64ASM": [
"fadd s0, s16, s17",
"mov v16.s[0], v0.s[0]"
]
},
"mulss xmm0, xmm1": {
"ExpectedInstructionCount": 2,
"Comment": [
"0xf3 0x0f 0x59"
],
"ExpectedArm64ASM": [
"fmul s0, s16, s17",
"mov v16.s[0], v0.s[0]"
]
},
"cvtss2sd xmm0, xmm1": {
"ExpectedInstructionCount": 2,
"Comment": "0xf3 0x0f 0x5a",
"ExpectedArm64ASM": [
"fcvt d0, s17",
"mov v16.d[0], v0.d[0]"
]
},
"cvtss2sd xmm0, [rax]": {
"ExpectedInstructionCount": 3,
"Comment": "0xf3 0x0f 0x5a",
"ExpectedArm64ASM": [
"ldr d2, [x4]",
"fcvt d0, s2",
"mov v16.d[0], v0.d[0]"
]
},
"cvttps2dq xmm0, xmm1": {
"ExpectedInstructionCount": 1,
"Comment": "0xf3 0x0f 0x5b",
"ExpectedArm64ASM": [
"fcvtzs v16.4s, v17.4s"
]
},
"subss xmm0, xmm1": {
"ExpectedInstructionCount": 2,
"Comment": [
"0xf3 0x0f 0x5c"
],
"ExpectedArm64ASM": [
"fsub s0, s16, s17",
"mov v16.s[0], v0.s[0]"
]
},
"minss xmm0, xmm1": {
"ExpectedInstructionCount": 5,
"Comment": [
"0xf3 0x0f 0x5d"
],
"ExpectedArm64ASM": [
"mrs x20, nzcv",
"fcmp s16, s17",
"fcsel s0, s16, s17, mi",
"mov v16.s[0], v0.s[0]",
"msr nzcv, x20"
]
},
"divss xmm0, xmm1": {
"ExpectedInstructionCount": 2,
"Comment": [
"0xf3 0x0f 0x5e"
],
"ExpectedArm64ASM": [
"fdiv s0, s16, s17",
"mov v16.s[0], v0.s[0]"
]
},
"maxss xmm0, xmm1": {
"ExpectedInstructionCount": 5,
"Comment": [
"0xf3 0x0f 0x5f"
],
"ExpectedArm64ASM": [
"mrs x20, nzcv",
"fcmp s16, s17",
"fcsel s0, s17, s16, mi",
"mov v16.s[0], v0.s[0]",
"msr nzcv, x20"
]
},
"movdqu xmm0, xmm0": {
"ExpectedInstructionCount": 0,
"Comment": "0xf3 0x0f 0x6f",
"ExpectedArm64ASM": []
},
"movdqu xmm0, xmm1": {
"ExpectedInstructionCount": 1,
"Comment": "0xf3 0x0f 0x6f",
"ExpectedArm64ASM": [
"mov v16.16b, v17.16b"
]
},
"movdqu xmm0, [rax]": {
"ExpectedInstructionCount": 1,
"Comment": "0xf3 0x0f 0x6f",
"ExpectedArm64ASM": [
"ldr q16, [x4]"
]
},
"pshufhw xmm0, xmm1, 0": {
"ExpectedInstructionCount": 2,
"Comment": [
"Broadcast upper-half element 0",
"0xf3 0x0f 0x70"
],
"ExpectedArm64ASM": [
"dup v2.8h, v17.h[4]",
"trn1 v16.2d, v17.2d, v2.2d"
]
},
"pshufhw xmm0, xmm1, 11100100b": {
"ExpectedInstructionCount": 1,
"Comment": [
"Identity copy",
"0xf3 0x0f 0x70"
],
"ExpectedArm64ASM": [
"mov v16.16b, v17.16b"
]
},
"pshufhw xmm0, xmm1, 01010000b": {
"ExpectedInstructionCount": 3,
"Comment": [
"Upper elements Self-zip",
"0xf3 0x0f 0x70"
],
"ExpectedArm64ASM": [
"ldr x0, [x28, #2040]",
"ldr q2, [x0, #1280]",
"tbl v16.16b, {v17.16b}, v2.16b"
]
},
"pshufhw xmm0, xmm1, 1": {
"ExpectedInstructionCount": 3,
"Comment": [
"Broadcast element 0 in the upper-half",
"Upper-half Element 0 gets turned in to element 1",
"0xf3 0x0f 0x70"
],
"ExpectedArm64ASM": [
"ldr x0, [x28, #2040]",
"ldr q2, [x0, #16]",
"tbl v16.16b, {v17.16b}, v2.16b"
]
},
"pshufhw xmm0, xmm1, 0xff": {
"ExpectedInstructionCount": 2,
"Comment": [
"Broadcast upper-half Element 3",
"0xf3 0x0f 0x70"
],
"ExpectedArm64ASM": [
"dup v2.8h, v17.h[7]",
"trn1 v16.2d, v17.2d, v2.2d"
]
},
"movq xmm0, xmm0": {
"ExpectedInstructionCount": 1,
"Comment": "0xf3 0x0f 0x7e",
"ExpectedArm64ASM": [
"mov v16.8b, v16.8b"
]
},
"movq xmm0, xmm1": {
"ExpectedInstructionCount": 1,
"Comment": "0xf3 0x0f 0x7e",
"ExpectedArm64ASM": [
"mov v16.8b, v17.8b"
]
},
"movq xmm0, [rax]": {
"ExpectedInstructionCount": 1,
"Comment": "0xf3 0x0f 0x7e",
"ExpectedArm64ASM": [
"ldr d16, [x4]"
]
},
"movdqu [rax], xmm0": {
"ExpectedInstructionCount": 1,
"Comment": "0xf3 0x0f 0x7f",
"ExpectedArm64ASM": [
"str q16, [x4]"
]
},
"popcnt ax, bx": {
"ExpectedInstructionCount": 9,
"Comment": "0xf3 0x0f 0xb8",
"ExpectedArm64ASM": [
"uxth w20, w7",
"fmov s0, w20",
"cnt v0.8b, v0.8b",
"addp v0.8b, v0.8b, v0.8b",
"umov w20, v0.b[0]",
"bfxil x4, x20, #0, #16",
"tst w20, w20",
"mov w26, #0x1",
"mov w27, #0x0"
]
},
"popcnt eax, ebx": {
"ExpectedInstructionCount": 7,
"Comment": "0xf3 0x0f 0xb8",
"ExpectedArm64ASM": [
"fmov s0, w7",
"cnt v0.8b, v0.8b",
"addv b0, v0.8b",
"umov w4, v0.b[0]",
"tst w4, w4",
"mov w26, #0x1",
"mov w27, #0x0"
]
},
"popcnt rax, rbx": {
"ExpectedInstructionCount": 7,
"Comment": "0xf3 0x0f 0xb8",
"ExpectedArm64ASM": [
"fmov d0, x7",
"cnt v0.8b, v0.8b",
"addv b0, v0.8b",
"umov w4, v0.b[0]",
"tst w4, w4",
"mov w26, #0x1",
"mov w27, #0x0"
]
},
"tzcnt ax, bx": {
"ExpectedInstructionCount": 9,
"Comment": "0xf3 0x0f 0xbc",
"ExpectedArm64ASM": [
"rbit w20, w7",
"orr w20, w20, #0x8000",
"clz w20, w20",
"bfxil x4, x20, #0, #16",
"cmn wzr, w20, lsl #16",
"ubfx x20, x20, #4, #1",
"mrs x21, nzcv",
"orr w20, w21, w20, lsl #29",
"msr nzcv, x20"
]
},
"tzcnt eax, ebx": {
"ExpectedInstructionCount": 7,
"Comment": "0xf3 0x0f 0xbc",
"ExpectedArm64ASM": [
"rbit w4, w7",
"clz w4, w4",
"tst w4, w4",
"ubfx x20, x4, #5, #1",
"mrs x21, nzcv",
"orr w20, w21, w20, lsl #29",
"msr nzcv, x20"
]
},
"tzcnt rax, rbx": {
"ExpectedInstructionCount": 7,
"Comment": "0xf3 0x0f 0xbc",
"ExpectedArm64ASM": [
"rbit x4, x7",
"clz x4, x4",
"tst x4, x4",
"ubfx x20, x4, #6, #1",
"mrs x21, nzcv",
"orr w20, w21, w20, lsl #29",
"msr nzcv, x20"
]
},
"lzcnt ax, bx": {
"ExpectedInstructionCount": 9,
"Comment": "0xf3 0x0f 0xbd",
"ExpectedArm64ASM": [
"lsl w20, w7, #16",
"orr w20, w20, #0x8000",
"clz w20, w20",
"bfxil x4, x20, #0, #16",
"cmn wzr, w20, lsl #16",
"ubfx x20, x20, #4, #1",
"mrs x21, nzcv",
"orr w20, w21, w20, lsl #29",
"msr nzcv, x20"
]
},
"lzcnt eax, ebx": {
"ExpectedInstructionCount": 6,
"Comment": "0xf3 0x0f 0xbd",
"ExpectedArm64ASM": [
"clz w4, w7",
"tst w4, w4",
"ubfx x20, x4, #5, #1",
"mrs x21, nzcv",
"orr w20, w21, w20, lsl #29",
"msr nzcv, x20"
]
},
"lzcnt rax, rbx": {
"ExpectedInstructionCount": 6,
"Comment": "0xf3 0x0f 0xbd",
"ExpectedArm64ASM": [
"clz x4, x7",
"tst x4, x4",
"ubfx x20, x4, #6, #1",
"mrs x21, nzcv",
"orr w20, w21, w20, lsl #29",
"msr nzcv, x20"
]
},
"cmpss xmm0, xmm1, 0": {
"ExpectedInstructionCount": 2,
"Comment": [
"0xf3 0x0f 0xc2"
],
"ExpectedArm64ASM": [
"fcmeq s0, s17, s16",
"mov v16.s[0], v0.s[0]"
]
},
"cmpss xmm0, xmm1, 1": {
"ExpectedInstructionCount": 2,
"Comment": [
"0xf3 0x0f 0xc2"
],
"ExpectedArm64ASM": [
"fcmgt s0, s17, s16",
"mov v16.s[0], v0.s[0]"
]
},
"cmpss xmm0, xmm1, 2": {
"ExpectedInstructionCount": 2,
"Comment": [
"0xf3 0x0f 0xc2"
],
"ExpectedArm64ASM": [
"fcmge s0, s17, s16",
"mov v16.s[0], v0.s[0]"
]
},
"cmpss xmm0, xmm1, 3": {
"ExpectedInstructionCount": 5,
"Comment": [
"0xf3 0x0f 0xc2"
],
"ExpectedArm64ASM": [
"fcmge s0, s16, s17",
"fcmgt s1, s17, s16",
"orr v0.8b, v0.8b, v1.8b",
"mvn v0.8b, v0.8b",
"mov v16.s[0], v0.s[0]"
]
},
"cmpss xmm0, xmm1, 4": {
"ExpectedInstructionCount": 3,
"Comment": [
"0xf3 0x0f 0xc2"
],
"ExpectedArm64ASM": [
"fcmeq s0, s17, s16",
"mvn v0.8b, v0.8b",
"mov v16.s[0], v0.s[0]"
]
},
"cmpss xmm0, xmm1, 5": {
"ExpectedInstructionCount": 3,
"Comment": [
"0xf3 0x0f 0xc2"
],
"ExpectedArm64ASM": [
"fcmgt s2, s17, s16",
"mvn v2.16b, v2.16b",
"mov v16.s[0], v2.s[0]"
]
},
"cmpss xmm0, xmm1, 6": {
"ExpectedInstructionCount": 3,
"Comment": [
"0xf3 0x0f 0xc2"
],
"ExpectedArm64ASM": [
"fcmge s2, s17, s16",
"mvn v2.16b, v2.16b",
"mov v16.s[0], v2.s[0]"
]
},
"cmpss xmm0, xmm1, 7": {
"ExpectedInstructionCount": 4,
"Comment": [
"0xf3 0x0f 0xc2"
],
"ExpectedArm64ASM": [
"fcmge s0, s16, s17",
"fcmgt s1, s17, s16",
"orr v0.8b, v0.8b, v1.8b",
"mov v16.s[0], v0.s[0]"
]
},
"movq2dq xmm0, mm0": {
"ExpectedInstructionCount": 1,
"Comment": "0xf3 0x0f 0xd6",
"ExpectedArm64ASM": [
"ldr d16, [x28, #1040]"
]
},
"cvtdq2pd xmm0, xmm1": {
"ExpectedInstructionCount": 2,
"Comment": "0xf3 0x0f 0xe6",
"ExpectedArm64ASM": [
"sxtl v2.2d, v17.2s",
"scvtf v16.2d, v2.2d"
]
},
"cvtdq2pd xmm0, [rax]": {
"ExpectedInstructionCount": 3,
"Comment": "0xf3 0x0f 0xe6",
"ExpectedArm64ASM": [
"ldr d2, [x4]",
"sxtl v2.2d, v2.2s",
"scvtf v16.2d, v2.2d"
]
}
}
}