Skip to content

[AArch64] Update zero latency instructions in Neoverse scheduling tables #165690

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 8 commits into from
Nov 19, 2025
Merged

[AArch64] Update zero latency instructions in Neoverse scheduling tables #165690

Show file tree
Hide file tree
Changes from 6 commits
Commits
Show all changes
8 commits
Select commit Hold shift + click to select a range
90fdc60
[AArch64] Update zero latency instructions in scheduling tables for N…
simonwallis2 Oct 30, 2025
695b27d
[AArch64] Update zero latency instructions in scheduling tables for N…
simonwallis2 Oct 30, 2025
9cebca2
[AArch64] Update zero latency instructions in Neoverse scheduling tables
simonwallis2 Nov 7, 2025
824b640
[AArch64] Update zero latency instructions in Neoverse scheduling tables
simonwallis2 Nov 13, 2025
f2d0a29
[AArch64] Update zero latency instructions in Neoverse scheduling tables
simonwallis2 Nov 13, 2025
684ad3e
[AArch64] Update zero latency instructions in Neoverse scheduling tables
simonwallis2 Nov 14, 2025
54fc6b5
[AArch64] Update zero latency instructions in Neoverse scheduling tables
simonwallis2 Nov 18, 2025
f0c2a71
Merge branch 'main' into sched-zero-latency
simonwallis2 Nov 19, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
28 changes: 25 additions & 3 deletions llvm/lib/Target/AArch64/AArch64SchedNeoverseN2.td
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -72,6 +72,10 @@ def : WriteRes<WriteLDHi, []> { let Latency = 4; }
// Define customized scheduler read/write types specific to the Neoverse N2.

//===----------------------------------------------------------------------===//

// Define generic 0 micro-op types
def N2Write_0c : SchedWriteRes<[]> { let Latency = 0; }

// Define generic 1 micro-op types

def N2Write_1c_1B : SchedWriteRes<[N2UnitB]> { let Latency = 1; }
Expand Down Expand Up @@ -645,6 +649,21 @@ def N2Write_11c_9L01_9S_9V : SchedWriteRes<[N2UnitL01, N2UnitL01, N2UnitL01,
let NumMicroOps = 27;
}

//===----------------------------------------------------------------------===//
// Define predicate-controlled types

def N2Write_0or1c_1I : SchedWriteVariant<[
SchedVar<NeoverseZeroMove, [N2Write_0c]>,
SchedVar<NoSchedPred, [N2Write_1c_1I]>]>;

def N2Write_0or2c_1V : SchedWriteVariant<[
SchedVar<NeoverseZeroMove, [N2Write_0c]>,
SchedVar<NoSchedPred, [N2Write_2c_1V]>]>;

def N2Write_0or3c_1M0 : SchedWriteVariant<[
SchedVar<NeoverseZeroMove, [N2Write_0c]>,
SchedVar<NoSchedPred, [N2Write_3c_1M0]>]>;

//===----------------------------------------------------------------------===//
// Define types for arithmetic and logical ops with short shifts
def N2Write_Arith : SchedWriteVariant<[
Expand Down Expand Up @@ -680,6 +699,7 @@ def : InstRW<[N2Write_1c_1B_1S], (instrs BL, BLR)>;
// ALU, basic
// ALU, basic, flagset
def : SchedAlias<WriteI, N2Write_1c_1I>;
def : InstRW<[N2Write_0or1c_1I], (instregex "^MOVZ[WX]i$")>;

// ALU, extend and shift
def : SchedAlias<WriteIEReg, N2Write_2c_1M>;
Expand All @@ -691,7 +711,8 @@ def : SchedAlias<WriteISReg, N2Write_Arith>;

// Logical, shift, no flagset
def : InstRW<[N2Write_1c_1I],
(instregex "^(AND|BIC|EON|EOR|ORN|ORR)[WX]rs$")>;
(instregex "^(AND|BIC|EON|EOR|ORN)[WX]rs$")>;
def : InstRW<[N2Write_0or1c_1I], (instregex "^ORR[WX]rs$")>;

// Logical, shift, flagset
def : InstRW<[N2Write_Logical], (instregex "^(AND|BIC)S[WX]rs$")>;
Expand Down Expand Up @@ -882,8 +903,7 @@ def : SchedAlias<WriteFImm, N2Write_2c_1V>;
def : InstRW<[N2Write_2c_1V], (instrs FMOVHr, FMOVSr, FMOVDr)>;

// FP transfer, from gen to low half of vec reg
def : InstRW<[N2Write_3c_1M0], (instrs FMOVWHr, FMOVXHr, FMOVWSr, FMOVXDr,
FMOVHWr, FMOVHXr, FMOVSWr, FMOVDXr)>;
def : InstRW<[N2Write_0or3c_1M0], (instrs FMOVWHr, FMOVXHr, FMOVWSr, FMOVXDr)>;

// FP transfer, from gen to high half of vec reg
def : InstRW<[N2Write_5c_1M0_1V], (instrs FMOVXDHighr)>;
Expand Down Expand Up @@ -1225,6 +1245,8 @@ def : InstRW<[N2Write_3c_1V0], (instrs BFCVT)>;
// ASIMD unzip/zip
// Handled by SchedAlias<WriteV[dq], ...>

def : InstRW<[N2Write_0or2c_1V], (instrs MOVID, MOVIv2d_ns)>;

// ASIMD duplicate, gen reg
def : InstRW<[N2Write_3c_1M0], (instregex "^DUPv.+gpr")>;

Expand Down
42 changes: 35 additions & 7 deletions llvm/lib/Target/AArch64/AArch64SchedNeoverseN3.td
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -75,7 +75,7 @@ def : WriteRes<WriteHint, []> { let Latency = 1; }

def N3Write_0c : SchedWriteRes<[]> {
let Latency = 0;
let NumMicroOps = 0;
let NumMicroOps = 1;
}

def N3Write_4c : SchedWriteRes<[]> {
Expand Down Expand Up @@ -553,6 +553,26 @@ def N3Write_16c_16V0 : SchedWriteRes<[N3UnitV0, N3UnitV0, N3UnitV0, N3UnitV0,
let NumMicroOps = 16;
}


//===----------------------------------------------------------------------===//
// Define predicate-controlled types

def N3Write_0or1c_1I : SchedWriteVariant<[
SchedVar<NeoverseZeroMove, [N3Write_0c]>,
SchedVar<NoSchedPred, [N3Write_1c_1I]>]>;

def N3Write_0or2c_1V : SchedWriteVariant<[
SchedVar<NeoverseZeroMove, [N3Write_0c]>,
SchedVar<NoSchedPred, [N3Write_2c_1V]>]>;

def N3Write_0or2c_1M : SchedWriteVariant<[
SchedVar<NeoverseAllElement, [N3Write_0c]>,
SchedVar<NoSchedPred, [N3Write_2c_1M]>]>;

def N3Write_0or3c_1M0 : SchedWriteVariant<[
SchedVar<NeoverseZeroMove, [N3Write_0c]>,
SchedVar<NoSchedPred, [N3Write_3c_1M0]>]>;

// Miscellaneous
// -----------------------------------------------------------------------------

Expand Down Expand Up @@ -581,6 +601,7 @@ def : InstRW<[N3Write_1c_1B_1S], (instrs BL, BLR)>;
// Conditional compare
// Conditional select
def : SchedAlias<WriteI, N3Write_1c_1I>;
def : InstRW<[N3Write_0or1c_1I], (instregex "^MOVZ[WX]i$")>;

// ALU, extend and shift
def : SchedAlias<WriteIEReg, N3Write_2c_1M>;
Expand Down Expand Up @@ -610,7 +631,8 @@ def : InstRW<[N3Write_1c_1I], (instrs GMI, SUBP, SUBPS)>;

// Logical, shift, no flagset
def : InstRW<[N3Write_1c_1I],
(instregex "^(AND|BIC|EON|EOR|ORN|ORR)[WX]rs$")>;
(instregex "^(AND|BIC|EON|EOR|ORN)[WX]rs$")>;
def : InstRW<[N3Write_0or1c_1I], (instregex "^ORR[WX]rs$")>;

// Logical, shift, flagset
def : InstRW<[N3Write_2c_1M], (instregex "^(AND|BIC)S[WX]rs$")>;
Expand Down Expand Up @@ -855,10 +877,11 @@ def : SchedAlias<WriteFCvt, N3Write_3c_1V0>;
def : SchedAlias<WriteFImm, N3Write_2c_1V>;

// FP move, register
def : InstRW<[N3Write_2c_1V], (instrs FMOVHr, FMOVSr, FMOVDr)>;
def : InstRW<[N3Write_2c_1V], (instrs FMOVHr)>;
def : InstRW<[N3Write_0c], (instrs FMOVSr, FMOVDr)>;

// FP transfer, from gen to low half of vec reg
def : InstRW<[N3Write_3c_1M0], (instrs FMOVWHr, FMOVXHr, FMOVWSr, FMOVXDr)>;
def : InstRW<[N3Write_0or3c_1M0], (instrs FMOVWHr, FMOVXHr, FMOVWSr, FMOVXDr)>;

// FP transfer, from gen to high half of vec reg
def : InstRW<[N3Write_5c_1M0_1V], (instrs FMOVXDHighr)>;
Expand Down Expand Up @@ -962,6 +985,8 @@ def : InstRW<[WriteAdr, N3Write_2c_1L01_1V_1I], (instregex "^STP[SDQ](post|pre)$
// ASIMD compare
// ASIMD logical
// ASIMD max/min, basic and pair-wise
def : InstRW<[N3Write_0or2c_1V], (instregex "^ORRv16i8", "^ORRv8i8")>;

def : SchedAlias<WriteVd, N3Write_2c_1V>;
def : SchedAlias<WriteVq, N3Write_2c_1V>;

Expand Down Expand Up @@ -1186,6 +1211,7 @@ def : InstRW<[N3Write_3c_1V0], (instrs BFCVT)>;
// ASIMD transpose
// ASIMD unzip/zip
// Covered by WriteV[dq]
def : InstRW<[N3Write_0or2c_1V], (instrs MOVID, MOVIv2d_ns)>;

// ASIMD duplicate, gen reg
def : InstRW<[N3Write_3c_1M0], (instregex "^DUPv.+gpr")>;
Expand Down Expand Up @@ -1564,10 +1590,11 @@ def : InstRW<[N3Write_2c_1M], (instregex "^REV_PP_[BHSD]")>;
def : InstRW<[N3Write_1c_1M], (instrs SEL_PPPP)>;

// Predicate set
def : InstRW<[N3Write_2c_1M], (instregex "^PFALSE", "^PTRUE_[BHSD]")>;
def : InstRW<[N3Write_0c], (instregex "^PFALSE")>;
def : InstRW<[N3Write_0or2c_1M], (instregex "^PTRUE_[BHSD]")>;

// Predicate set/initialize, set flags
def : InstRW<[N3Write_2c_1M], (instregex "^PTRUES_[BHSD]")>;
def : InstRW<[N3Write_0or2c_1M], (instregex "^PTRUES_[BHSD]")>;
Copy link
Contributor

@rj-jesus rj-jesus Nov 14, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

PTRUES isn't listed in Section 4.11, but it is mentioned in Table 2-23, so I'll assume the latter to be correct (matching what you've implemented).


// Predicate find first/next
def : InstRW<[N3Write_2c_1M], (instregex "^PFIRST_B$", "^PNEXT_[BHSD]$")>;
Expand Down Expand Up @@ -1790,10 +1817,11 @@ def : InstRW<[N3Write_5c_1M0_1V], (instregex "^INDEX_(IR|RI|RR)_D$")>;
// Logical
def : InstRW<[N3Write_2c_1V],
(instregex "^(AND|EOR|ORR)_ZI",
"^(AND|BIC|EOR|ORR)_ZZZ",
"^(AND|BIC|EOR)_ZZZ",
"^EOR(BT|TB)_ZZZ_[BHSD]",
"^(AND|BIC|EOR|NOT|ORR)_(ZPmZ|ZPZZ)_[BHSD]",
"^NOT_ZPmZ_[BHSD]")>;
def : InstRW<[N3Write_0or2c_1V], (instregex "^ORR_ZZZ")>;

// Max/min, basic and pairwise
def : InstRW<[N3Write_2c_1V], (instregex "^[SU](MAX|MIN)_ZI_[BHSD]",
Expand Down
17 changes: 15 additions & 2 deletions llvm/lib/Target/AArch64/AArch64SchedNeoverseV1.td
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -91,7 +91,7 @@ def : WriteRes<WriteHint, []> { let Latency = 1; }
//===----------------------------------------------------------------------===//
// Define generic 0 micro-op types

let Latency = 0, NumMicroOps = 0 in
let Latency = 0, NumMicroOps = 1 in
def V1Write_0c_0Z : SchedWriteRes<[]>;


Expand Down Expand Up @@ -472,6 +472,17 @@ def V1Write_11c_9L01_9S_9V : SchedWriteRes<[V1UnitL01, V1UnitL01, V1UnitL01,
V1UnitV, V1UnitV, V1UnitV,
V1UnitV, V1UnitV, V1UnitV]>;

//===----------------------------------------------------------------------===//
// Define predicate-controlled types

def V1Write_0or1c_1I : SchedWriteVariant<[
SchedVar<NeoverseZeroMove, [V1Write_0c_0Z]>,
SchedVar<NoSchedPred, [V1Write_1c_1I]>]>;

def V1Write_0or3c_1M0 : SchedWriteVariant<[
SchedVar<NeoverseZeroMove, [V1Write_0c_0Z]>,
SchedVar<NoSchedPred, [V1Write_3c_1M0]>]>;

//===----------------------------------------------------------------------===//
// Define forwarded types

Expand Down Expand Up @@ -603,6 +614,7 @@ def : InstRW<[V1Write_1c_1I_1Flg],
"^(ADC|SBC)S[WX]r$",
"^ANDS[WX]ri$",
"^(AND|BIC)S[WX]rr$")>;
def : InstRW<[V1Write_0or1c_1I], (instregex "^MOVZ[WX]i$")>;

// ALU, extend and shift
def : SchedAlias<WriteIEReg, V1Write_2c_1M>;
Expand All @@ -623,7 +635,8 @@ def : InstRW<[V1WriteISRegS],
(instregex "^(ADD|SUB)S(([WX]r[sx])|Xrx64)$")>;

// Logical, shift, no flagset
def : InstRW<[V1Write_1c_1I], (instregex "^(AND|BIC|EON|EOR|ORN|ORR)[WX]rs$")>;
def : InstRW<[V1Write_1c_1I], (instregex "^(AND|BIC|EON|EOR|ORN)[WX]rs$")>;
def : InstRW<[V1Write_0or1c_1I], (instregex "^ORR[WX]rs$")>;

// Logical, shift, flagset
def : InstRW<[V1Write_2c_1M_1Flg], (instregex "^(AND|BIC)S[WX]rs$")>;
Expand Down
19 changes: 18 additions & 1 deletion llvm/lib/Target/AArch64/AArch64SchedPredNeoverse.td
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -80,5 +80,22 @@ def NeoverseZeroMove : MCSchedPredicate<
// MOVI Dd, #0
// MOVI Vd.2D, #0
CheckAll<[CheckOpcode<[MOVID, MOVIv2d_ns]>,
CheckImmOperand<1, 0>]>
CheckImmOperand<1, 0>]>,
// MOV Zd, Zn
CheckAll<[CheckOpcode<[ORR_ZZZ]>,
CheckSameRegOperand<1, 2>]>,
// MOV Vd, Vn
CheckAll<[CheckOpcode<[ORRv16i8, ORRv8i8]>,
CheckSameRegOperand<1, 2>]>,
]>>;

def NeoverseAllElement : MCSchedPredicate<
CheckAny<[
// PTRUE Pd, ALL
// PTRUES Pd, ALL
CheckAll<[CheckOpcode<[
PTRUE_B, PTRUE_H, PTRUE_S, PTRUE_D,
PTRUES_B, PTRUES_H, PTRUES_S, PTRUES_D]>,
CheckIsImmOperand<1>,
CheckImmOperand<1, 31>]>,
]>>;
30 changes: 15 additions & 15 deletions llvm/test/tools/llvm-mca/AArch64/Neoverse/N2-basic-instructions.s
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2086,9 +2086,9 @@ drps
# CHECK-NEXT: 1 3 0.50 fcvtas x27, d28
# CHECK-NEXT: 1 3 0.50 fcvtau w29, d30
# CHECK-NEXT: 1 3 0.50 fcvtau xzr, d0
# CHECK-NEXT: 1 3 3.00 fmov w3, s9
# CHECK-NEXT: 1 2 0.50 fmov w3, s9
# CHECK-NEXT: 1 3 3.00 fmov s9, w3
# CHECK-NEXT: 1 3 3.00 fmov x20, d31
# CHECK-NEXT: 1 2 0.50 fmov x20, d31
# CHECK-NEXT: 1 3 3.00 fmov d1, x15
# CHECK-NEXT: 1 2 0.50 fmov x3, v12.d[1]
# CHECK-NEXT: 2 5 1.00 fmov v1.d[1], x19
Expand Down Expand Up @@ -2508,14 +2508,14 @@ drps
# CHECK-NEXT: 1 2 0.50 bics x3, xzr, x3, lsl #1
# CHECK-NEXT: 1 2 0.50 tst w3, w7, lsl #31
# CHECK-NEXT: 1 2 0.50 tst x2, x20, asr #2
# CHECK-NEXT: 1 1 0.25 mov x3, x6
# CHECK-NEXT: 1 1 0.25 mov x3, xzr
# CHECK-NEXT: 1 1 0.25 mov wzr, w2
# CHECK-NEXT: 1 1 0.25 mov w3, w5
# CHECK-NEXT: 1 0 0.20 mov x3, x6
# CHECK-NEXT: 1 0 0.20 mov x3, xzr
# CHECK-NEXT: 1 0 0.20 mov wzr, w2
# CHECK-NEXT: 1 0 0.20 mov w3, w5
# CHECK-NEXT: 1 1 0.25 movz w2, #0, lsl #16
# CHECK-NEXT: 1 1 0.25 mov w2, #-1235
# CHECK-NEXT: 1 1 0.25 mov x2, #5299989643264
# CHECK-NEXT: 1 1 0.25 mov x2, #0
# CHECK-NEXT: 1 0 0.20 mov x2, #0
# CHECK-NEXT: 1 1 0.25 movk w3, #0
# CHECK-NEXT: 1 1 0.25 movz x4, #0, lsl #16
# CHECK-NEXT: 1 1 0.25 movk w5, #0, lsl #16
Expand Down Expand Up @@ -2557,7 +2557,7 @@ drps

# CHECK: Resource pressure per iteration:
# CHECK-NEXT: [0.0] [0.1] [1.0] [1.1] [2] [3.0] [3.1] [4] [5] [6.0] [6.1] [7] [8]
# CHECK-NEXT: 11.00 11.00 33.00 33.00 87.33 151.33 151.33 517.00 251.00 162.50 162.50 215.50 85.50
# CHECK-NEXT: 11.00 11.00 33.00 33.00 87.33 151.33 151.33 509.75 249.75 161.25 161.25 216.50 86.50

# CHECK: Resource pressure by instruction:
# CHECK-NEXT: [0.0] [0.1] [1.0] [1.1] [2] [3.0] [3.1] [4] [5] [6.0] [6.1] [7] [8] Instructions:
Expand Down Expand Up @@ -3270,9 +3270,9 @@ drps
# CHECK-NEXT: - - - - - - - - - - - 0.50 0.50 fcvtas x27, d28
# CHECK-NEXT: - - - - - - - - - - - 0.50 0.50 fcvtau w29, d30
# CHECK-NEXT: - - - - - - - - - - - 0.50 0.50 fcvtau xzr, d0
# CHECK-NEXT: - - - - - - - 3.00 - - - - - fmov w3, s9
# CHECK-NEXT: - - - - - - - - - - - 0.50 0.50 fmov w3, s9
# CHECK-NEXT: - - - - - - - 3.00 - - - - - fmov s9, w3
# CHECK-NEXT: - - - - - - - 3.00 - - - - - fmov x20, d31
# CHECK-NEXT: - - - - - - - - - - - 0.50 0.50 fmov x20, d31
# CHECK-NEXT: - - - - - - - 3.00 - - - - - fmov d1, x15
# CHECK-NEXT: - - - - - - - - - - - 0.50 0.50 fmov x3, v12.d[1]
# CHECK-NEXT: - - - - - - - 1.00 - - - 0.50 0.50 fmov v1.d[1], x19
Expand Down Expand Up @@ -3692,14 +3692,14 @@ drps
# CHECK-NEXT: - - - - - - - 0.50 0.50 - - - - bics x3, xzr, x3, lsl #1
# CHECK-NEXT: - - - - - - - 0.50 0.50 - - - - tst w3, w7, lsl #31
# CHECK-NEXT: - - - - - - - 0.50 0.50 - - - - tst x2, x20, asr #2
# CHECK-NEXT: - - - - - - - 0.25 0.25 0.25 0.25 - - mov x3, x6
# CHECK-NEXT: - - - - - - - 0.25 0.25 0.25 0.25 - - mov x3, xzr
# CHECK-NEXT: - - - - - - - 0.25 0.25 0.25 0.25 - - mov wzr, w2
# CHECK-NEXT: - - - - - - - 0.25 0.25 0.25 0.25 - - mov w3, w5
# CHECK-NEXT: - - - - - - - - - - - - - mov x3, x6
# CHECK-NEXT: - - - - - - - - - - - - - mov x3, xzr
# CHECK-NEXT: - - - - - - - - - - - - - mov wzr, w2
# CHECK-NEXT: - - - - - - - - - - - - - mov w3, w5
# CHECK-NEXT: - - - - - - - 0.25 0.25 0.25 0.25 - - movz w2, #0, lsl #16
# CHECK-NEXT: - - - - - - - 0.25 0.25 0.25 0.25 - - mov w2, #-1235
# CHECK-NEXT: - - - - - - - 0.25 0.25 0.25 0.25 - - mov x2, #5299989643264
# CHECK-NEXT: - - - - - - - 0.25 0.25 0.25 0.25 - - mov x2, #0
# CHECK-NEXT: - - - - - - - - - - - - - mov x2, #0
# CHECK-NEXT: - - - - - - - 0.25 0.25 0.25 0.25 - - movk w3, #0
# CHECK-NEXT: - - - - - - - 0.25 0.25 0.25 0.25 - - movz x4, #0, lsl #16
# CHECK-NEXT: - - - - - - - 0.25 0.25 0.25 0.25 - - movk w5, #0, lsl #16
Expand Down
Loading