Skip to content

[RISCV] Remove unecessary early exit in transferBefore #74040

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

Jump to bottom
Merged
merged 3 commits into from
Dec 12, 2023
Merged

[RISCV] Remove unecessary early exit in transferBefore #74040

merged 3 commits into from
Dec 12, 2023

Conversation

lukel97
Copy link
Contributor

@lukel97 lukel97 commented Dec 1, 2023

Previously we bailed if we encountered a pseudo without a VL op, i.e. vmv.x.s,
which prevented us from preserving VL and VTYPE. It looks like this was copied
over from a time whenever this code was operating on the MachineInstrs in
place, see https://reviews.llvm.org/D127870

However because we no longer mutate the MIs, we can just get rid of this early
exit which allows us to preserve VL and VTYPE when dealing with vmv.x.s.

@lukel97 lukel97 requested review from BeMg, preames and topperc December 1, 2023 05:55
@llvmbot
Copy link
Member

llvmbot commented Dec 1, 2023

@llvm/pr-subscribers-backend-risc-v

Author: Luke Lau (lukel97)

Changes

Previously we bailed if we encountered a pseudo without a VL op, i.e. vmv.x.s,
which prevented us from preserving VL and VTYPE. It looks like this was copied
over from a time whenever this code was operating on the MachineInstrs in
place, see https://reviews.llvm.org/D127870

However because we no longer mutate the MIs, we can just get rid of this early
exit which allows us to preserve VL and VTYPE when dealing with vmv.x.s.

Patch is 90.60 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/74040.diff

12 Files Affected:

  • (modified) llvm/lib/Target/RISCV/RISCVInsertVSETVLI.cpp (-5)
  • (modified) llvm/test/CodeGen/RISCV/rvv/fixed-vectors-extract-i1.ll (+10-10)
  • (modified) llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-gather.ll (+180-316)
  • (modified) llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-scatter.ll (+42-42)
  • (modified) llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-fp.ll (+17-17)
  • (modified) llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-int.ll (+24-24)
  • (modified) llvm/test/CodeGen/RISCV/rvv/fixed-vectors-unaligned.ll (+10-10)
  • (modified) llvm/test/CodeGen/RISCV/rvv/vreductions-fp-sdnode.ll (+8-8)
  • (modified) llvm/test/CodeGen/RISCV/rvv/vreductions-int-vp.ll (+4-4)
  • (modified) llvm/test/CodeGen/RISCV/rvv/vreductions-int.ll (+12-12)
  • (modified) llvm/test/CodeGen/RISCV/rvv/vsetvli-insert.ll (+1-2)
  • (modified) llvm/test/CodeGen/RISCV/rvv/vsetvli-regression.ll (+1-1)
diff --git a/llvm/lib/Target/RISCV/RISCVInsertVSETVLI.cpp b/llvm/lib/Target/RISCV/RISCVInsertVSETVLI.cpp
index 323a92cfb8c83d3..721d623e3653468 100644
--- a/llvm/lib/Target/RISCV/RISCVInsertVSETVLI.cpp
+++ b/llvm/lib/Target/RISCV/RISCVInsertVSETVLI.cpp
@@ -1064,11 +1064,6 @@ void RISCVInsertVSETVLI::transferBefore(VSETVLIInfo &Info,
   if (Info.hasSEWLMULRatioOnly() || !Info.isValid() || Info.isUnknown())
     Info = NewInfo;
 
-  if (!RISCVII::hasVLOp(TSFlags)) {
-    Info = NewInfo;
-    return;
-  }
-
   DemandedFields Demanded = getDemanded(MI, MRI, ST);
   const VSETVLIInfo IncomingInfo =
       adjustIncoming(PrevInfo, NewInfo, Demanded, MRI);
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-extract-i1.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-extract-i1.ll
index 9d689c732d7999f..dccb62877af3cce 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-extract-i1.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-extract-i1.ll
@@ -106,7 +106,7 @@ define i1 @extractelt_v16i1(ptr %x, i64 %idx) nounwind {
 ; RV32-NEXT:    vsetivli zero, 16, e8, m1, ta, ma
 ; RV32-NEXT:    vle8.v v8, (a0)
 ; RV32-NEXT:    vmseq.vi v8, v8, 0
-; RV32-NEXT:    vsetivli zero, 1, e16, mf4, ta, ma
+; RV32-NEXT:    vsetvli zero, zero, e16, m2, ta, ma
 ; RV32-NEXT:    vmv.x.s a0, v8
 ; RV32-NEXT:    srl a0, a0, a1
 ; RV32-NEXT:    andi a0, a0, 1
@@ -117,7 +117,7 @@ define i1 @extractelt_v16i1(ptr %x, i64 %idx) nounwind {
 ; RV64-NEXT:    vsetivli zero, 16, e8, m1, ta, ma
 ; RV64-NEXT:    vle8.v v8, (a0)
 ; RV64-NEXT:    vmseq.vi v8, v8, 0
-; RV64-NEXT:    vsetivli zero, 1, e16, mf4, ta, ma
+; RV64-NEXT:    vsetvli zero, zero, e16, m2, ta, ma
 ; RV64-NEXT:    vmv.x.s a0, v8
 ; RV64-NEXT:    srl a0, a0, a1
 ; RV64-NEXT:    andi a0, a0, 1
@@ -128,7 +128,7 @@ define i1 @extractelt_v16i1(ptr %x, i64 %idx) nounwind {
 ; RV32ZBS-NEXT:    vsetivli zero, 16, e8, m1, ta, ma
 ; RV32ZBS-NEXT:    vle8.v v8, (a0)
 ; RV32ZBS-NEXT:    vmseq.vi v8, v8, 0
-; RV32ZBS-NEXT:    vsetivli zero, 1, e16, mf4, ta, ma
+; RV32ZBS-NEXT:    vsetvli zero, zero, e16, m2, ta, ma
 ; RV32ZBS-NEXT:    vmv.x.s a0, v8
 ; RV32ZBS-NEXT:    bext a0, a0, a1
 ; RV32ZBS-NEXT:    ret
@@ -138,7 +138,7 @@ define i1 @extractelt_v16i1(ptr %x, i64 %idx) nounwind {
 ; RV64ZBS-NEXT:    vsetivli zero, 16, e8, m1, ta, ma
 ; RV64ZBS-NEXT:    vle8.v v8, (a0)
 ; RV64ZBS-NEXT:    vmseq.vi v8, v8, 0
-; RV64ZBS-NEXT:    vsetivli zero, 1, e16, mf4, ta, ma
+; RV64ZBS-NEXT:    vsetvli zero, zero, e16, m2, ta, ma
 ; RV64ZBS-NEXT:    vmv.x.s a0, v8
 ; RV64ZBS-NEXT:    bext a0, a0, a1
 ; RV64ZBS-NEXT:    ret
@@ -155,7 +155,7 @@ define i1 @extractelt_v32i1(ptr %x, i64 %idx) nounwind {
 ; RV32-NEXT:    vsetvli zero, a2, e8, m2, ta, ma
 ; RV32-NEXT:    vle8.v v8, (a0)
 ; RV32-NEXT:    vmseq.vi v10, v8, 0
-; RV32-NEXT:    vsetivli zero, 1, e32, mf2, ta, ma
+; RV32-NEXT:    vsetvli zero, zero, e32, m8, ta, ma
 ; RV32-NEXT:    vmv.x.s a0, v10
 ; RV32-NEXT:    srl a0, a0, a1
 ; RV32-NEXT:    andi a0, a0, 1
@@ -167,7 +167,7 @@ define i1 @extractelt_v32i1(ptr %x, i64 %idx) nounwind {
 ; RV64-NEXT:    vsetvli zero, a2, e8, m2, ta, ma
 ; RV64-NEXT:    vle8.v v8, (a0)
 ; RV64-NEXT:    vmseq.vi v10, v8, 0
-; RV64-NEXT:    vsetivli zero, 1, e32, mf2, ta, ma
+; RV64-NEXT:    vsetvli zero, zero, e32, m8, ta, ma
 ; RV64-NEXT:    vmv.x.s a0, v10
 ; RV64-NEXT:    srl a0, a0, a1
 ; RV64-NEXT:    andi a0, a0, 1
@@ -179,7 +179,7 @@ define i1 @extractelt_v32i1(ptr %x, i64 %idx) nounwind {
 ; RV32ZBS-NEXT:    vsetvli zero, a2, e8, m2, ta, ma
 ; RV32ZBS-NEXT:    vle8.v v8, (a0)
 ; RV32ZBS-NEXT:    vmseq.vi v10, v8, 0
-; RV32ZBS-NEXT:    vsetivli zero, 1, e32, mf2, ta, ma
+; RV32ZBS-NEXT:    vsetvli zero, zero, e32, m8, ta, ma
 ; RV32ZBS-NEXT:    vmv.x.s a0, v10
 ; RV32ZBS-NEXT:    bext a0, a0, a1
 ; RV32ZBS-NEXT:    ret
@@ -190,7 +190,7 @@ define i1 @extractelt_v32i1(ptr %x, i64 %idx) nounwind {
 ; RV64ZBS-NEXT:    vsetvli zero, a2, e8, m2, ta, ma
 ; RV64ZBS-NEXT:    vle8.v v8, (a0)
 ; RV64ZBS-NEXT:    vmseq.vi v10, v8, 0
-; RV64ZBS-NEXT:    vsetivli zero, 1, e32, mf2, ta, ma
+; RV64ZBS-NEXT:    vsetvli zero, zero, e32, m8, ta, ma
 ; RV64ZBS-NEXT:    vmv.x.s a0, v10
 ; RV64ZBS-NEXT:    bext a0, a0, a1
 ; RV64ZBS-NEXT:    ret
@@ -221,7 +221,7 @@ define i1 @extractelt_v64i1(ptr %x, i64 %idx) nounwind {
 ; RV64-NEXT:    vsetvli zero, a2, e8, m4, ta, ma
 ; RV64-NEXT:    vle8.v v8, (a0)
 ; RV64-NEXT:    vmseq.vi v12, v8, 0
-; RV64-NEXT:    vsetivli zero, 1, e64, m1, ta, ma
+; RV64-NEXT:    vsetvli zero, a2, e64, m1, ta, ma
 ; RV64-NEXT:    vmv.x.s a0, v12
 ; RV64-NEXT:    srl a0, a0, a1
 ; RV64-NEXT:    andi a0, a0, 1
@@ -246,7 +246,7 @@ define i1 @extractelt_v64i1(ptr %x, i64 %idx) nounwind {
 ; RV64ZBS-NEXT:    vsetvli zero, a2, e8, m4, ta, ma
 ; RV64ZBS-NEXT:    vle8.v v8, (a0)
 ; RV64ZBS-NEXT:    vmseq.vi v12, v8, 0
-; RV64ZBS-NEXT:    vsetivli zero, 1, e64, m1, ta, ma
+; RV64ZBS-NEXT:    vsetvli zero, a2, e64, m1, ta, ma
 ; RV64ZBS-NEXT:    vmv.x.s a0, v12
 ; RV64ZBS-NEXT:    bext a0, a0, a1
 ; RV64ZBS-NEXT:    ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-gather.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-gather.ll
index d74fd6cd3f03470..ac3bf0d89b5ed9f 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-gather.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-gather.ll
@@ -3844,7 +3844,7 @@ define <4 x i64> @mgather_truemask_v4i64(<4 x ptr> %ptrs, <4 x i64> %passthru) {
 ; RV32ZVE32F-NEXT:    vmv.x.s a6, v9
 ; RV32ZVE32F-NEXT:    bnez zero, .LBB45_5
 ; RV32ZVE32F-NEXT:  # %bb.1: # %cond.load
-; RV32ZVE32F-NEXT:    vsetivli zero, 1, e32, m1, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s a3, v8
 ; RV32ZVE32F-NEXT:    lw a2, 4(a3)
 ; RV32ZVE32F-NEXT:    lw a3, 0(a3)
@@ -4265,12 +4265,12 @@ define <8 x i64> @mgather_baseidx_v8i8_v8i64(ptr %base, <8 x i8> %idxs, <8 x i1>
 ; RV32ZVE32F-NEXT:    vsext.vf4 v10, v8
 ; RV32ZVE32F-NEXT:    vsll.vi v8, v10, 3
 ; RV32ZVE32F-NEXT:    vadd.vx v8, v8, a1
-; RV32ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e8, mf2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s t0, v0
 ; RV32ZVE32F-NEXT:    andi a1, t0, 1
 ; RV32ZVE32F-NEXT:    beqz a1, .LBB48_7
 ; RV32ZVE32F-NEXT:  # %bb.1: # %cond.load
-; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s a3, v8
 ; RV32ZVE32F-NEXT:    lw a1, 4(a3)
 ; RV32ZVE32F-NEXT:    lw a3, 0(a3)
@@ -4539,12 +4539,12 @@ define <8 x i64> @mgather_baseidx_sext_v8i8_v8i64(ptr %base, <8 x i8> %idxs, <8
 ; RV32ZVE32F-NEXT:    vsext.vf4 v10, v8
 ; RV32ZVE32F-NEXT:    vsll.vi v8, v10, 3
 ; RV32ZVE32F-NEXT:    vadd.vx v8, v8, a1
-; RV32ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e8, mf2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s t0, v0
 ; RV32ZVE32F-NEXT:    andi a1, t0, 1
 ; RV32ZVE32F-NEXT:    beqz a1, .LBB49_7
 ; RV32ZVE32F-NEXT:  # %bb.1: # %cond.load
-; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s a3, v8
 ; RV32ZVE32F-NEXT:    lw a1, 4(a3)
 ; RV32ZVE32F-NEXT:    lw a3, 0(a3)
@@ -4815,12 +4815,12 @@ define <8 x i64> @mgather_baseidx_zext_v8i8_v8i64(ptr %base, <8 x i8> %idxs, <8
 ; RV32ZVE32F-NEXT:    vzext.vf4 v10, v8
 ; RV32ZVE32F-NEXT:    vsll.vi v8, v10, 3
 ; RV32ZVE32F-NEXT:    vadd.vx v8, v8, a1
-; RV32ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e8, mf2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s t0, v0
 ; RV32ZVE32F-NEXT:    andi a1, t0, 1
 ; RV32ZVE32F-NEXT:    beqz a1, .LBB50_7
 ; RV32ZVE32F-NEXT:  # %bb.1: # %cond.load
-; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s a3, v8
 ; RV32ZVE32F-NEXT:    lw a1, 4(a3)
 ; RV32ZVE32F-NEXT:    lw a3, 0(a3)
@@ -5098,12 +5098,12 @@ define <8 x i64> @mgather_baseidx_v8i16_v8i64(ptr %base, <8 x i16> %idxs, <8 x i
 ; RV32ZVE32F-NEXT:    vsext.vf2 v10, v8
 ; RV32ZVE32F-NEXT:    vsll.vi v8, v10, 3
 ; RV32ZVE32F-NEXT:    vadd.vx v8, v8, a1
-; RV32ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e8, mf2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s t0, v0
 ; RV32ZVE32F-NEXT:    andi a1, t0, 1
 ; RV32ZVE32F-NEXT:    beqz a1, .LBB51_7
 ; RV32ZVE32F-NEXT:  # %bb.1: # %cond.load
-; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s a3, v8
 ; RV32ZVE32F-NEXT:    lw a1, 4(a3)
 ; RV32ZVE32F-NEXT:    lw a3, 0(a3)
@@ -5373,12 +5373,12 @@ define <8 x i64> @mgather_baseidx_sext_v8i16_v8i64(ptr %base, <8 x i16> %idxs, <
 ; RV32ZVE32F-NEXT:    vsext.vf2 v10, v8
 ; RV32ZVE32F-NEXT:    vsll.vi v8, v10, 3
 ; RV32ZVE32F-NEXT:    vadd.vx v8, v8, a1
-; RV32ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e8, mf2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s t0, v0
 ; RV32ZVE32F-NEXT:    andi a1, t0, 1
 ; RV32ZVE32F-NEXT:    beqz a1, .LBB52_7
 ; RV32ZVE32F-NEXT:  # %bb.1: # %cond.load
-; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s a3, v8
 ; RV32ZVE32F-NEXT:    lw a1, 4(a3)
 ; RV32ZVE32F-NEXT:    lw a3, 0(a3)
@@ -5650,12 +5650,12 @@ define <8 x i64> @mgather_baseidx_zext_v8i16_v8i64(ptr %base, <8 x i16> %idxs, <
 ; RV32ZVE32F-NEXT:    vzext.vf2 v10, v8
 ; RV32ZVE32F-NEXT:    vsll.vi v8, v10, 3
 ; RV32ZVE32F-NEXT:    vadd.vx v8, v8, a1
-; RV32ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e8, mf2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s t0, v0
 ; RV32ZVE32F-NEXT:    andi a1, t0, 1
 ; RV32ZVE32F-NEXT:    beqz a1, .LBB53_7
 ; RV32ZVE32F-NEXT:  # %bb.1: # %cond.load
-; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s a3, v8
 ; RV32ZVE32F-NEXT:    lw a1, 4(a3)
 ; RV32ZVE32F-NEXT:    lw a3, 0(a3)
@@ -5934,12 +5934,12 @@ define <8 x i64> @mgather_baseidx_v8i32_v8i64(ptr %base, <8 x i32> %idxs, <8 x i
 ; RV32ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
 ; RV32ZVE32F-NEXT:    vsll.vi v8, v8, 3
 ; RV32ZVE32F-NEXT:    vadd.vx v8, v8, a1
-; RV32ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e8, mf2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s t0, v0
 ; RV32ZVE32F-NEXT:    andi a1, t0, 1
 ; RV32ZVE32F-NEXT:    beqz a1, .LBB54_7
 ; RV32ZVE32F-NEXT:  # %bb.1: # %cond.load
-; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s a3, v8
 ; RV32ZVE32F-NEXT:    lw a1, 4(a3)
 ; RV32ZVE32F-NEXT:    lw a3, 0(a3)
@@ -6207,12 +6207,12 @@ define <8 x i64> @mgather_baseidx_sext_v8i32_v8i64(ptr %base, <8 x i32> %idxs, <
 ; RV32ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
 ; RV32ZVE32F-NEXT:    vsll.vi v8, v8, 3
 ; RV32ZVE32F-NEXT:    vadd.vx v8, v8, a1
-; RV32ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e8, mf2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s t0, v0
 ; RV32ZVE32F-NEXT:    andi a1, t0, 1
 ; RV32ZVE32F-NEXT:    beqz a1, .LBB55_7
 ; RV32ZVE32F-NEXT:  # %bb.1: # %cond.load
-; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s a3, v8
 ; RV32ZVE32F-NEXT:    lw a1, 4(a3)
 ; RV32ZVE32F-NEXT:    lw a3, 0(a3)
@@ -6481,12 +6481,12 @@ define <8 x i64> @mgather_baseidx_zext_v8i32_v8i64(ptr %base, <8 x i32> %idxs, <
 ; RV32ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
 ; RV32ZVE32F-NEXT:    vsll.vi v8, v8, 3
 ; RV32ZVE32F-NEXT:    vadd.vx v8, v8, a1
-; RV32ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e8, mf2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s t0, v0
 ; RV32ZVE32F-NEXT:    andi a1, t0, 1
 ; RV32ZVE32F-NEXT:    beqz a1, .LBB56_7
 ; RV32ZVE32F-NEXT:  # %bb.1: # %cond.load
-; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s a3, v8
 ; RV32ZVE32F-NEXT:    lw a1, 4(a3)
 ; RV32ZVE32F-NEXT:    lw a3, 0(a3)
@@ -6778,12 +6778,12 @@ define <8 x i64> @mgather_baseidx_v8i64(ptr %base, <8 x i64> %idxs, <8 x i1> %m,
 ; RV32ZVE32F-NEXT:    vslide1down.vx v8, v8, a4
 ; RV32ZVE32F-NEXT:    vsll.vi v8, v8, 3
 ; RV32ZVE32F-NEXT:    vadd.vx v8, v8, a1
-; RV32ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e8, mf2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s t0, v0
 ; RV32ZVE32F-NEXT:    andi a1, t0, 1
 ; RV32ZVE32F-NEXT:    beqz a1, .LBB57_7
 ; RV32ZVE32F-NEXT:  # %bb.1: # %cond.load
-; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV32ZVE32F-NEXT:    lw a1, 4(a2)
 ; RV32ZVE32F-NEXT:    lw a2, 0(a2)
@@ -9652,7 +9652,7 @@ define <4 x double> @mgather_truemask_v4f64(<4 x ptr> %ptrs, <4 x double> %passt
 ; RV32ZVE32F-NEXT:    fsd fa3, 24(a0)
 ; RV32ZVE32F-NEXT:    ret
 ; RV32ZVE32F-NEXT:  .LBB84_6: # %cond.load
-; RV32ZVE32F-NEXT:    vsetivli zero, 1, e32, m1, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV32ZVE32F-NEXT:    fld fa0, 0(a2)
 ; RV32ZVE32F-NEXT:    andi a2, a1, 2
@@ -9964,7 +9964,7 @@ define <8 x double> @mgather_baseidx_v8i8_v8f64(ptr %base, <8 x i8> %idxs, <8 x
 ; RV32ZVE32F-NEXT:    vsext.vf4 v10, v8
 ; RV32ZVE32F-NEXT:    vsll.vi v8, v10, 3
 ; RV32ZVE32F-NEXT:    vadd.vx v8, v8, a1
-; RV32ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e8, mf2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s a1, v0
 ; RV32ZVE32F-NEXT:    andi a2, a1, 1
 ; RV32ZVE32F-NEXT:    bnez a2, .LBB87_10
@@ -10005,7 +10005,7 @@ define <8 x double> @mgather_baseidx_v8i8_v8f64(ptr %base, <8 x i8> %idxs, <8 x
 ; RV32ZVE32F-NEXT:    fsd fa7, 56(a0)
 ; RV32ZVE32F-NEXT:    ret
 ; RV32ZVE32F-NEXT:  .LBB87_10: # %cond.load
-; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV32ZVE32F-NEXT:    fld fa0, 0(a2)
 ; RV32ZVE32F-NEXT:    andi a2, a1, 2
@@ -10179,7 +10179,7 @@ define <8 x double> @mgather_baseidx_sext_v8i8_v8f64(ptr %base, <8 x i8> %idxs,
 ; RV32ZVE32F-NEXT:    vsext.vf4 v10, v8
 ; RV32ZVE32F-NEXT:    vsll.vi v8, v10, 3
 ; RV32ZVE32F-NEXT:    vadd.vx v8, v8, a1
-; RV32ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e8, mf2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s a1, v0
 ; RV32ZVE32F-NEXT:    andi a2, a1, 1
 ; RV32ZVE32F-NEXT:    bnez a2, .LBB88_10
@@ -10220,7 +10220,7 @@ define <8 x double> @mgather_baseidx_sext_v8i8_v8f64(ptr %base, <8 x i8> %idxs,
 ; RV32ZVE32F-NEXT:    fsd fa7, 56(a0)
 ; RV32ZVE32F-NEXT:    ret
 ; RV32ZVE32F-NEXT:  .LBB88_10: # %cond.load
-; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV32ZVE32F-NEXT:    fld fa0, 0(a2)
 ; RV32ZVE32F-NEXT:    andi a2, a1, 2
@@ -10396,7 +10396,7 @@ define <8 x double> @mgather_baseidx_zext_v8i8_v8f64(ptr %base, <8 x i8> %idxs,
 ; RV32ZVE32F-NEXT:    vzext.vf4 v10, v8
 ; RV32ZVE32F-NEXT:    vsll.vi v8, v10, 3
 ; RV32ZVE32F-NEXT:    vadd.vx v8, v8, a1
-; RV32ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e8, mf2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s a1, v0
 ; RV32ZVE32F-NEXT:    andi a2, a1, 1
 ; RV32ZVE32F-NEXT:    bnez a2, .LBB89_10
@@ -10437,7 +10437,7 @@ define <8 x double> @mgather_baseidx_zext_v8i8_v8f64(ptr %base, <8 x i8> %idxs,
 ; RV32ZVE32F-NEXT:    fsd fa7, 56(a0)
 ; RV32ZVE32F-NEXT:    ret
 ; RV32ZVE32F-NEXT:  .LBB89_10: # %cond.load
-; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV32ZVE32F-NEXT:    fld fa0, 0(a2)
 ; RV32ZVE32F-NEXT:    andi a2, a1, 2
@@ -10620,7 +10620,7 @@ define <8 x double> @mgather_baseidx_v8i16_v8f64(ptr %base, <8 x i16> %idxs, <8
 ; RV32ZVE32F-NEXT:    vsext.vf2 v10, v8
 ; RV32ZVE32F-NEXT:    vsll.vi v8, v10, 3
 ; RV32ZVE32F-NEXT:    vadd.vx v8, v8, a1
-; RV32ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e8, mf2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s a1, v0
 ; RV32ZVE32F-NEXT:    andi a2, a1, 1
 ; RV32ZVE32F-NEXT:    bnez a2, .LBB90_10
@@ -10661,7 +10661,7 @@ define <8 x double> @mgather_baseidx_v8i16_v8f64(ptr %base, <8 x i16> %idxs, <8
 ; RV32ZVE32F-NEXT:    fsd fa7, 56(a0)
 ; RV32ZVE32F-NEXT:    ret
 ; RV32ZVE32F-NEXT:  .LBB90_10: # %cond.load
-; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV32ZVE32F-NEXT:    fld fa0, 0(a2)
 ; RV32ZVE32F-NEXT:    andi a2, a1, 2
@@ -10836,7 +10836,7 @@ define <8 x double> @mgather_baseidx_sext_v8i16_v8f64(ptr %base, <8 x i16> %idxs
 ; RV32ZVE32F-NEXT:    vsext.vf2 v10, v8
 ; RV32ZVE32F-NEXT:    vsll.vi v8, v10, 3
 ; RV32ZVE32F-NEXT:    vadd.vx v8, v8, a1
-; RV32ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e8, mf2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s a1, v0
 ; RV32ZVE32F-NEXT:    andi a2, a1, 1
 ; RV32ZVE32F-NEXT:    bnez a2, .LBB91_10
@@ -10877,7 +10877,7 @@ define <8 x double> @mgather_baseidx_sext_v8i16_v8f64(ptr %base, <8 x i16> %idxs
 ; RV32ZVE32F-NEXT:    fsd fa7, 56(a0)
 ; RV32ZVE32F-NEXT:    ret
 ; RV32ZVE32F-NEXT:  .LBB91_10: # %cond.load
-; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV32ZVE32F-NEXT:    fld fa0, 0(a2)
 ; RV32ZVE32F-NEXT:    andi a2, a1, 2
@@ -11054,7 +11054,7 @@ define <8 x double> @mgather_baseidx_zext_v8i16_v8f64(ptr %base, <8 x i16> %idxs
 ; RV32ZVE32F-NEXT:    vzext.vf2 v10, v8
 ; RV32ZVE32F-NEXT:    vsll.vi v8, v10, 3
 ; RV32ZVE32F-NEXT:    vadd.vx v8, v8, a1
-; RV32ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e8, mf2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s a1, v0
 ; RV32ZVE32F-NEXT:    andi a2, a1, 1
 ; RV32ZVE32F-NEXT:    bnez a2, .LBB92_10
@@ -11095,7 +11095,7 @@ define <8 x double> @mgather_baseidx_zext_v8i16_v8f64(ptr %base, <8 x i16> %idxs
 ; RV32ZVE32F-NEXT:    fsd fa7, 56(a0)
 ; RV32ZVE32F-NEXT:    ret
 ; RV32ZVE32F-NEXT:  .LBB92_10: # %cond.load
-; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV32ZVE32F-NEXT:    fld fa0, 0(a2)
 ; RV32ZVE32F-NEXT:    andi a2, a1, 2
@@ -11279,7 +11279,7 @@ define <8 x double> @mgather_baseidx_v8i32_v8f64(ptr %base, <8 x i32> %idxs, <8
 ; RV32ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
 ; RV32ZVE32F-NEXT:    vsll.vi v8, v8, 3
 ; RV32ZVE32F-NEXT:    vadd.vx v8, v8, a1
-; RV32ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e8, mf2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s a1, v0
 ; RV32ZVE32F-NEXT:    andi a2, a1, 1
 ; RV32ZVE32F-NEXT:    bnez a2, .LBB93_10
@@ -11320,7 +11320,7 @@ define <8 x double> @mgather_baseidx_v8i32_v8f64(ptr %base, <8 x i32> %idxs, <8
 ; RV32ZVE32F-NEXT:    fsd fa7, 56(a0)
 ; RV32ZVE32F-NEXT:    ret
 ; RV32ZVE32F-NEXT:  .LBB93_10: # %cond.load
-; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e32, m2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV32ZVE32F-NEXT:    fld fa0, 0(a2)
 ; RV32ZVE32F-NEXT:    andi a2, a1, 2
@@ -11493,7 +11493,7 @@ define <8 x double> @mgather_baseidx_sext_v8i32_v8f64(ptr %base, <8 x i32> %idxs
 ; RV32ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
 ; RV32ZVE32F-NEXT:    vsll.vi v8, v8, 3
 ; RV32ZVE32F-NEXT:    vadd.vx v8, v8, a1
-; RV32ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV32ZVE32F-NEXT:    vsetvli zero, zero, e8, mf2, ta, ma
 ; RV32ZVE32F-NEXT:    vmv.x.s a1, v0
 ; RV32ZVE32F-NEXT:    andi a2, a1, 1
 ; RV32ZVE32F-NEXT:    bnez a2, .LBB94_10
@@ -11534,7 +1...
[truncated]

lukel97
lukel97 commented Dec 1, 2023
View reviewed changes
llvm/test/CodeGen/RISCV/rvv/fixed-vectors-extract-i1.ll
@@ -221,7 +221,7 @@ define i1 @extractelt_v64i1(ptr %x, i64 %idx) nounwind {
; RV64-NEXT: vsetvli zero, a2, e8, m4, ta, ma
; RV64-NEXT: vle8.v v8, (a0)
; RV64-NEXT: vmseq.vi v12, v8, 0
; RV64-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; RV64-NEXT: vsetvli zero, a2, e64, m1, ta, ma
Copy link
Contributor Author

@lukel97 lukel97 Dec 1, 2023 •
edited
Loading

Choose a reason for hiding this comment

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

We're preserving the AVL here but changing the VTYPE, i.e. performing the minimal change needed to get the correct state, which is what transferBefore should be doing. This looks like a case where the postpass could be replacing it with 1 to reduce the live range

Copy link
Collaborator

Choose a reason for hiding this comment

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

Another way to look at this is that transferBefore should chose to use the next AVL (1) since it has the choice here, and using the previous VL extends the live range. I'm fine with this landing as is, but this might be a good follow up.

@lukel97 lukel97 mentioned this pull request Dec 6, 2023
Merged
preames
preames approved these changes Dec 6, 2023
View reviewed changes
Copy link
Collaborator

@preames preames left a comment

Choose a reason for hiding this comment

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

LGTM

llvm/test/CodeGen/RISCV/rvv/fixed-vectors-extract-i1.ll
@@ -221,7 +221,7 @@ define i1 @extractelt_v64i1(ptr %x, i64 %idx) nounwind {
; RV64-NEXT: vsetvli zero, a2, e8, m4, ta, ma
; RV64-NEXT: vle8.v v8, (a0)
; RV64-NEXT: vmseq.vi v12, v8, 0
; RV64-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; RV64-NEXT: vsetvli zero, a2, e64, m1, ta, ma
Copy link
Collaborator

Choose a reason for hiding this comment

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

Another way to look at this is that transferBefore should chose to use the next AVL (1) since it has the choice here, and using the previous VL extends the live range. I'm fine with this landing as is, but this might be a good follow up.

@lukel97
[RISCV] Remove unecessary early exit in transferBefore
a60c334 
Previously we bailed if we encountered a pseudo without a VL op, i.e. vmv.x.s,
which prevented us from preserving VL and VTYPE. It looks like this was copied
over from a time whenever this code was operating on the MachineInstrs in
place, see https://reviews.llvm.org/D127870

However because we no longer mutate the MIs, we can just get rid of this early
exit which allows us to preserve VL and VTYPE when dealing with vmv.x.s.
@lukel97
Update test outside of rvv directory
079473d
@lukel97
Rebase and update test
e7b4ea5
@lukel97 lukel97 force-pushed the transferBefore-remove-vlop-exit branch from 5a40e7c to e7b4ea5 Compare December 12, 2023 08:04
@lukel97 lukel97 merged commit 3944504 into llvm:main Dec 12, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@preames preames preames approved these changes

@BeMg BeMg Awaiting requested review from BeMg

@topperc topperc Awaiting requested review from topperc

Assignees
No one assigned
Labels
backend:RISC-V
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
3 participants
@lukel97 @llvmbot @preames