[RISCV] Optimize getVLENFactoredAmount function.

[craig.topper]

Summary

If the local variable NumOfVReg isPowerOf2_32(NumOfVReg - 1) or isPowerOf2_32(NumOfVReg + 1), the ADDI and MUL instructions can be replaced with SLLI and ADD(or SUB) instructions.

[StephenFan]

Add Kill flag

[craig.topper]

LGTM but please add a description of the optimization so that once it is committed it can be understood from git log without needing to read the patch.

[frasercrmck]

LGTM but I agree with Craig.

[StephenFan]

Hi @craig.topper , @frasercrmck , Sorry, I missed it. And I will add it right now.

[StephenFan]

Add git log info.

[craig.topper]

@StephenFan Is something holding up committing this?

[craig.topper]

Commandeering so I can update this. It no longer applies cleanly.

[craig.topper]

Minimize number of virtual registers used to match current version of getVLENFactoredAmount

[StephenFan]

In D100577#2774378, @craig.topper wrote:

@StephenFan Is something holding up committing this?

Sorry, I’m writing my undergraduate thesis recently, so I don’t have time to deal with this patch, thank you for helping me deal with it.

[rogfer01]

llvm/lib/Target/RISCV/RISCVInstrInfo.cpp

1394

I saw a test failing downstream due to this. This should be addImm(ShiftAmount)

[frasercrmck]

llvm/test/CodeGen/RISCV/rvv/allocate-lmul-2-4-8.ll

307

@rogfer01 I believe this line shows the issue you're talking about?

[frasercrmck]

llvm/lib/Target/RISCV/RISCVInstrInfo.cpp

1394

@craig.topper has fixed it in https://github.com/llvm/llvm-project/commit/020df692d801c4fa9a67eb32e923927e33f9e4b5

llvm/lib/Target/RISCV/RISCVInstrInfo.cpp

assert(isInt<12>(NumOfVReg) &&
          "Expect the number of vector registers within 12-bits.");
   if (isPowerOf2_32(NumOfVReg)) {
     uint32_t ShiftAmount = Log2_32(NumOfVReg);
     if (ShiftAmount == 0)
       return VL;
     BuildMI(MBB, II, DL, TII->get(RISCV::SLLI), VL)
         .addReg(VL, RegState::Kill)
         .addImm(ShiftAmount);
+  } else if (isPowerOf2_32(NumOfVReg - 1)) {
+    Register ScaledRegister = MRI.createVirtualRegister(&RISCV::GPRRegClass);
+    uint32_t ShiftAmount = Log2_32(NumOfVReg - 1);
+    BuildMI(MBB, II, DL, TII->get(RISCV::SLLI), ScaledRegister)
+        .addReg(VL)
+        .addImm(ShiftAmount);
+    BuildMI(MBB, II, DL, TII->get(RISCV::ADD), VL)
+        .addReg(ScaledRegister, RegState::Kill)
+        .addReg(VL, RegState::Kill);
+  } else if (isPowerOf2_32(NumOfVReg + 1)) {
+    Register ScaledRegister = MRI.createVirtualRegister(&RISCV::GPRRegClass);
+    uint32_t ShiftAmount = Log2_32(NumOfVReg + 1);
+    BuildMI(MBB, II, DL, TII->get(RISCV::SLLI), ScaledRegister)
+        .addReg(VL)
+        .addReg(ShiftAmount);
+    BuildMI(MBB, II, DL, TII->get(RISCV::SUB), VL)
+        .addReg(ScaledRegister, RegState::Kill)
+        .addReg(VL, RegState::Kill);
   } else {
     Register N = MRI.createVirtualRegister(&RISCV::GPRRegClass);
     BuildMI(MBB, II, DL, TII->get(RISCV::ADDI), N)
         .addReg(RISCV::X0)
         .addImm(NumOfVReg);
     if (!MF.getSubtarget<RISCVSubtarget>().hasStdExtM())
       MF.getFunction().getContext().diagnose(DiagnosticInfoUnsupported{
           MF.getFunction(),

llvm/test/CodeGen/RISCV/rvv/allocate-lmul-2-4-8.ll

; CHECK-NEXT:    ret
   ret void
 }
 
 define void @lmul1_and_2() nounwind {
 ; CHECK-LABEL: lmul1_and_2:
 ; CHECK:       # %bb.0:
 ; CHECK-NEXT:    addi sp, sp, -16
 ; CHECK-NEXT:    csrr a0, vlenb
-; CHECK-NEXT:    addi a1, zero, 3
-; CHECK-NEXT:    mul a0, a0, a1
+; CHECK-NEXT:    slli a1, a0, 1
+; CHECK-NEXT:    add a0, a1, a0
 ; CHECK-NEXT:    sub sp, sp, a0
 ; CHECK-NEXT:    csrr a0, vlenb
-; CHECK-NEXT:    addi a1, zero, 3
-; CHECK-NEXT:    mul a0, a0, a1
+; CHECK-NEXT:    slli a1, a0, 1
+; CHECK-NEXT:    add a0, a1, a0
 ; CHECK-NEXT:    add sp, sp, a0
 ; CHECK-NEXT:    addi sp, sp, 16
 ; CHECK-NEXT:    ret
   %v1 = alloca <vscale x 1 x i64>
   %v2 = alloca <vscale x 2 x i64>
   ret void
 }
 

 define void @lmul1_and_4() nounwind {
 ; CHECK-LABEL: lmul1_and_4:
 ; CHECK:       # %bb.0:
 ; CHECK-NEXT:    addi sp, sp, -32
 ; CHECK-NEXT:    sd ra, 24(sp) # 8-byte Folded Spill
 ; CHECK-NEXT:    sd s0, 16(sp) # 8-byte Folded Spill
 ; CHECK-NEXT:    addi s0, sp, 32
 ; CHECK-NEXT:    csrr a0, vlenb
-; CHECK-NEXT:    addi a1, zero, 5
-; CHECK-NEXT:    mul a0, a0, a1
+; CHECK-NEXT:    slli a1, a0, 2
+; CHECK-NEXT:    add a0, a1, a0
 ; CHECK-NEXT:    sub sp, sp, a0
 ; CHECK-NEXT:    andi sp, sp, -32
 ; CHECK-NEXT:    addi sp, s0, -32
 ; CHECK-NEXT:    ld s0, 16(sp) # 8-byte Folded Reload
 ; CHECK-NEXT:    ld ra, 24(sp) # 8-byte Folded Reload
 ; CHECK-NEXT:    addi sp, sp, 32
 ; CHECK-NEXT:    ret
   %v1 = alloca <vscale x 1 x i64>
   %v2 = alloca <vscale x 4 x i64>
   ret void
 }
 
 define void @lmul2_and_1() nounwind {
 ; CHECK-LABEL: lmul2_and_1:
 ; CHECK:       # %bb.0:
 ; CHECK-NEXT:    addi sp, sp, -16
 ; CHECK-NEXT:    csrr a0, vlenb
-; CHECK-NEXT:    addi a1, zero, 3
-; CHECK-NEXT:    mul a0, a0, a1
+; CHECK-NEXT:    slli a1, a0, 1
+; CHECK-NEXT:    add a0, a1, a0
 ; CHECK-NEXT:    sub sp, sp, a0
 ; CHECK-NEXT:    csrr a0, vlenb
-; CHECK-NEXT:    addi a1, zero, 3
-; CHECK-NEXT:    mul a0, a0, a1
+; CHECK-NEXT:    slli a1, a0, 1
+; CHECK-NEXT:    add a0, a1, a0
 ; CHECK-NEXT:    add sp, sp, a0
 ; CHECK-NEXT:    addi sp, sp, 16
 ; CHECK-NEXT:    ret
   %v1 = alloca <vscale x 2 x i64>
   %v2 = alloca <vscale x 1 x i64>
   ret void
 }
 
 define void @lmul4_and_1() nounwind {
 ; CHECK-LABEL: lmul4_and_1:
 ; CHECK:       # %bb.0:
 ; CHECK-NEXT:    addi sp, sp, -32
 ; CHECK-NEXT:    sd ra, 24(sp) # 8-byte Folded Spill
 ; CHECK-NEXT:    sd s0, 16(sp) # 8-byte Folded Spill
 ; CHECK-NEXT:    addi s0, sp, 32
 ; CHECK-NEXT:    csrr a0, vlenb
-; CHECK-NEXT:    addi a1, zero, 5
-; CHECK-NEXT:    mul a0, a0, a1
+; CHECK-NEXT:    slli a1, a0, 2
+; CHECK-NEXT:    add a0, a1, a0
 ; CHECK-NEXT:    sub sp, sp, a0
 ; CHECK-NEXT:    andi sp, sp, -32
 ; CHECK-NEXT:    addi sp, s0, -32
 ; CHECK-NEXT:    ld s0, 16(sp) # 8-byte Folded Reload
 ; CHECK-NEXT:    ld ra, 24(sp) # 8-byte Folded Reload
 ; CHECK-NEXT:    addi sp, sp, 32
 ; CHECK-NEXT:    ret
   %v1 = alloca <vscale x 4 x i64>

 }
 
 
 define void @gpr_and_lmul1_and_2() nounwind {
 ; CHECK-LABEL: gpr_and_lmul1_and_2:
 ; CHECK:       # %bb.0:
 ; CHECK-NEXT:    addi sp, sp, -32
 ; CHECK-NEXT:    csrr a0, vlenb
-; CHECK-NEXT:    addi a1, zero, 3
-; CHECK-NEXT:    mul a0, a0, a1
+; CHECK-NEXT:    slli a1, a0, 1
+; CHECK-NEXT:    add a0, a1, a0
 ; CHECK-NEXT:    sub sp, sp, a0
 ; CHECK-NEXT:    addi a0, zero, 3
 ; CHECK-NEXT:    sd a0, 24(sp)
 ; CHECK-NEXT:    csrr a0, vlenb
-; CHECK-NEXT:    addi a1, zero, 3
-; CHECK-NEXT:    mul a0, a0, a1
+; CHECK-NEXT:    slli a1, a0, 1
+; CHECK-NEXT:    add a0, a1, a0
 ; CHECK-NEXT:    add sp, sp, a0
 ; CHECK-NEXT:    addi sp, sp, 32
 ; CHECK-NEXT:    ret
   %x1 = alloca i64
   %v1 = alloca <vscale x 1 x i64>
   %v2 = alloca <vscale x 2 x i64>
   store volatile i64 3, i64* %x1
   ret void
 }
 
 define void @gpr_and_lmul1_and_4() nounwind {
 ; CHECK-LABEL: gpr_and_lmul1_and_4:
 ; CHECK:       # %bb.0:
 ; CHECK-NEXT:    addi sp, sp, -64
 ; CHECK-NEXT:    sd ra, 56(sp) # 8-byte Folded Spill
 ; CHECK-NEXT:    sd s0, 48(sp) # 8-byte Folded Spill
 ; CHECK-NEXT:    addi s0, sp, 64
 ; CHECK-NEXT:    csrr a0, vlenb
-; CHECK-NEXT:    addi a1, zero, 5
-; CHECK-NEXT:    mul a0, a0, a1
+; CHECK-NEXT:    slli a1, a0, 2
+; CHECK-NEXT:    add a0, a1, a0
 ; CHECK-NEXT:    sub sp, sp, a0
 ; CHECK-NEXT:    andi sp, sp, -32
 ; CHECK-NEXT:    addi a0, zero, 3
 ; CHECK-NEXT:    sd a0, 40(sp)
 ; CHECK-NEXT:    addi sp, s0, -64
 ; CHECK-NEXT:    ld s0, 48(sp) # 8-byte Folded Reload
 ; CHECK-NEXT:    ld ra, 56(sp) # 8-byte Folded Reload
 ; CHECK-NEXT:    addi sp, sp, 64
 ; CHECK-NEXT:    ret
   %x1 = alloca i64
   %v1 = alloca <vscale x 1 x i64>
   %v2 = alloca <vscale x 4 x i64>
   store volatile i64 3, i64* %x1
   ret void
 }
 
 define void @lmul_1_2_4_8() nounwind {
 ; CHECK-LABEL: lmul_1_2_4_8:
 ; CHECK:       # %bb.0:
 ; CHECK-NEXT:    addi sp, sp, -64
 ; CHECK-NEXT:    sd ra, 56(sp) # 8-byte Folded Spill
 ; CHECK-NEXT:    sd s0, 48(sp) # 8-byte Folded Spill
 ; CHECK-NEXT:    addi s0, sp, 64
 ; CHECK-NEXT:    csrr a0, vlenb
-; CHECK-NEXT:    addi a1, zero, 15
-; CHECK-NEXT:    mul a0, a0, a1
+; CHECK-NEXT:    slli a1, a0, vl
+; CHECK-NEXT:    sub a0, a1, a0
 ; CHECK-NEXT:    sub sp, sp, a0
 ; CHECK-NEXT:    andi sp, sp, -64
 ; CHECK-NEXT:    addi sp, s0, -64
 ; CHECK-NEXT:    ld s0, 48(sp) # 8-byte Folded Reload
 ; CHECK-NEXT:    ld ra, 56(sp) # 8-byte Folded Reload
 ; CHECK-NEXT:    addi sp, sp, 64
 ; CHECK-NEXT:    ret
   %v1 = alloca <vscale x 1 x i64>

llvm/test/CodeGen/RISCV/rvv/rvv-framelayout.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -mtriple=riscv64 -mattr=+experimental-v,+m -verify-machineinstrs < %s | FileCheck %s
 
 define void @rvv_vla(i64 %n, i64 %i) nounwind {
 ; CHECK-LABEL: rvv_vla:
 ; CHECK:       # %bb.0:
 ; CHECK-NEXT:    addi sp, sp, -32
 ; CHECK-NEXT:    sd ra, 24(sp) # 8-byte Folded Spill
 ; CHECK-NEXT:    sd s0, 16(sp) # 8-byte Folded Spill
 ; CHECK-NEXT:    addi s0, sp, 32
 ; CHECK-NEXT:    csrr a2, vlenb
-; CHECK-NEXT:    addi a3, zero, 3
-; CHECK-NEXT:    mul a2, a2, a3
+; CHECK-NEXT:    slli a3, a2, 1
+; CHECK-NEXT:    add a2, a3, a2
 ; CHECK-NEXT:    sub sp, sp, a2
 ; CHECK-NEXT:    slli a0, a0, 2
 ; CHECK-NEXT:    addi a0, a0, 15
 ; CHECK-NEXT:    andi a0, a0, -16
 ; CHECK-NEXT:    sub a0, sp, a0
 ; CHECK-NEXT:    mv sp, a0
 ; CHECK-NEXT:    csrr a2, vlenb
 ; CHECK-NEXT:    sub a2, s0, a2
 ; CHECK-NEXT:    addi a2, a2, -32
 ; CHECK-NEXT:    vl1re64.v v25, (a2)
 ; CHECK-NEXT:    csrr a2, vlenb
-; CHECK-NEXT:    addi a3, zero, 3
-; CHECK-NEXT:    mul a2, a2, a3
+; CHECK-NEXT:    slli a3, a2, 1
+; CHECK-NEXT:    add a2, a3, a2
 ; CHECK-NEXT:    sub a2, s0, a2
 ; CHECK-NEXT:    addi a2, a2, -32
 ; CHECK-NEXT:    vl2re64.v v26, (a2)
 ; CHECK-NEXT:    slli a1, a1, 2
 ; CHECK-NEXT:    add a0, a0, a1
 ; CHECK-NEXT:    lw a0, 0(a0)
 ; CHECK-NEXT:    addi sp, s0, -32
 ; CHECK-NEXT:    ld s0, 16(sp) # 8-byte Folded Reload

 define void @rvv_overaligned() nounwind {
 ; CHECK-LABEL: rvv_overaligned:
 ; CHECK:       # %bb.0:
 ; CHECK-NEXT:    addi sp, sp, -128
 ; CHECK-NEXT:    sd ra, 120(sp) # 8-byte Folded Spill
 ; CHECK-NEXT:    sd s0, 112(sp) # 8-byte Folded Spill
 ; CHECK-NEXT:    addi s0, sp, 128
 ; CHECK-NEXT:    csrr a0, vlenb
-; CHECK-NEXT:    addi a1, zero, 3
-; CHECK-NEXT:    mul a0, a0, a1
+; CHECK-NEXT:    slli a1, a0, 1
+; CHECK-NEXT:    add a0, a1, a0
 ; CHECK-NEXT:    sub sp, sp, a0
 ; CHECK-NEXT:    andi sp, sp, -64
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    slli a0, a0, 1
 ; CHECK-NEXT:    add a0, sp, a0
 ; CHECK-NEXT:    addi a0, a0, 112
 ; CHECK-NEXT:    vl1re64.v v25, (a0)
 ; CHECK-NEXT:    addi a0, sp, 112

 ; CHECK-LABEL: rvv_vla_and_overaligned:
 ; CHECK:       # %bb.0:
 ; CHECK-NEXT:    addi sp, sp, -128
 ; CHECK-NEXT:    sd ra, 120(sp) # 8-byte Folded Spill
 ; CHECK-NEXT:    sd s0, 112(sp) # 8-byte Folded Spill
 ; CHECK-NEXT:    sd s1, 104(sp) # 8-byte Folded Spill
 ; CHECK-NEXT:    addi s0, sp, 128
 ; CHECK-NEXT:    csrr a2, vlenb
-; CHECK-NEXT:    addi a3, zero, 3
-; CHECK-NEXT:    mul a2, a2, a3
+; CHECK-NEXT:    slli a3, a2, 1
+; CHECK-NEXT:    add a2, a3, a2
 ; CHECK-NEXT:    sub sp, sp, a2
 ; CHECK-NEXT:    andi sp, sp, -64
 ; CHECK-NEXT:    mv s1, sp
 ; CHECK-NEXT:    slli a0, a0, 2
 ; CHECK-NEXT:    addi a0, a0, 15
 ; CHECK-NEXT:    andi a0, a0, -16
 ; CHECK-NEXT:    sub a0, sp, a0
 ; CHECK-NEXT:    mv sp, a0