[llvm] a33f60d - [RISCV] Add test case showing suboptimal BUILD_VECTOR lowering

[Fraser Cormack via llvm-commits]

Author: Fraser Cormack
Date: 2021-07-29T09:21:26+01:00
New Revision: a33f60db39836d94a79557b147515d42d30ecba3

URL: https://github.com/llvm/llvm-project/commit/a33f60db39836d94a79557b147515d42d30ecba3
DIFF: https://github.com/llvm/llvm-project/commit/a33f60db39836d94a79557b147515d42d30ecba3.diff

LOG: [RISCV] Add test case showing suboptimal BUILD_VECTOR lowering

The second test case added here was pointed out to me by @craig.topper
and shows how we "optimize" a two-element BUILD_VECTOR from being one
load from the constant pool to two loads from the constant pool.

The first test case shows that since materialization for the
floating-point +0.0 value is cheap and doesn't involve a load, the
optimization is more clearly beneficial here.

Reviewed By: craig.topper

Differential Revision: https://reviews.llvm.org/D106962

Added: 
    

Modified: 
    llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp-buildvec.ll

Removed: 
    


################################################################################
diff  --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp-buildvec.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp-buildvec.ll
index bcc6a3de50480..203c5fca447f5 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp-buildvec.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp-buildvec.ll
@@ -84,12 +84,50 @@ define <4 x float> @hang_when_merging_stores_after_legalization(<8 x float> %x,
   ret <4 x float> %z
 }
 
+define void @buildvec_dominant0_v2f32(<2 x float>* %x) {
+; CHECK-LABEL: buildvec_dominant0_v2f32:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    lui a1, %hi(.LCPI2_0)
+; CHECK-NEXT:    addi a1, a1, %lo(.LCPI2_0)
+; CHECK-NEXT:    vsetivli zero, 2, e32, mf2, ta, mu
+; CHECK-NEXT:    vlse32.v v25, (a1), zero
+; CHECK-NEXT:    fmv.w.x ft0, zero
+; CHECK-NEXT:    vsetvli zero, zero, e32, mf2, tu, mu
+; CHECK-NEXT:    vfmv.s.f v25, ft0
+; CHECK-NEXT:    vsetvli zero, zero, e32, mf2, ta, mu
+; CHECK-NEXT:    vse32.v v25, (a0)
+; CHECK-NEXT:    ret
+  store <2 x float> <float 0.0, float 1.0>, <2 x float>* %x
+  ret void
+}
+
+; FIXME: We "optimize" this one 2-element load from the constant pool to two
+; loads from the constant pool.
+
+define void @buildvec_dominant1_v2f32(<2 x float>* %x) {
+; CHECK-LABEL: buildvec_dominant1_v2f32:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    lui a1, %hi(.LCPI3_0)
+; CHECK-NEXT:    addi a1, a1, %lo(.LCPI3_0)
+; CHECK-NEXT:    vsetivli zero, 2, e32, mf2, ta, mu
+; CHECK-NEXT:    vlse32.v v25, (a1), zero
+; CHECK-NEXT:    lui a1, %hi(.LCPI3_1)
+; CHECK-NEXT:    flw ft0, %lo(.LCPI3_1)(a1)
+; CHECK-NEXT:    vsetvli zero, zero, e32, mf2, tu, mu
+; CHECK-NEXT:    vfmv.s.f v25, ft0
+; CHECK-NEXT:    vsetvli zero, zero, e32, mf2, ta, mu
+; CHECK-NEXT:    vse32.v v25, (a0)
+; CHECK-NEXT:    ret
+  store <2 x float> <float 1.0, float 2.0>, <2 x float>* %x
+  ret void
+}
+
 define void @buildvec_dominant0_v4f32(<4 x float>* %x) {
 ; CHECK-LABEL: buildvec_dominant0_v4f32:
 ; CHECK:       # %bb.0:
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, mu
-; CHECK-NEXT:    lui a1, %hi(.LCPI2_0)
-; CHECK-NEXT:    addi a1, a1, %lo(.LCPI2_0)
+; CHECK-NEXT:    lui a1, %hi(.LCPI4_0)
+; CHECK-NEXT:    addi a1, a1, %lo(.LCPI4_0)
 ; CHECK-NEXT:    vlse32.v v25, (a1), zero
 ; CHECK-NEXT:    fmv.w.x ft0, zero
 ; CHECK-NEXT:    vfmv.s.f v26, ft0
@@ -125,8 +163,8 @@ define void @buildvec_dominant1_v4f32(<4 x float>* %x, float %f) {
 define void @buildvec_dominant2_v4f32(<4 x float>* %x, float %f) {
 ; CHECK-LABEL: buildvec_dominant2_v4f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    lui a1, %hi(.LCPI4_0)
-; CHECK-NEXT:    flw ft0, %lo(.LCPI4_0)(a1)
+; CHECK-NEXT:    lui a1, %hi(.LCPI6_0)
+; CHECK-NEXT:    flw ft0, %lo(.LCPI6_0)(a1)
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, mu
 ; CHECK-NEXT:    vfmv.s.f v25, ft0
 ; CHECK-NEXT:    vfmv.v.f v26, fa0