[PowerPC] Adjust the naming and operand order of fnmsub patterns

Summary: This is follow up patch of https://reviews.llvm.org/D67595. Adjust naming and the Commutable operands for additional patterns to make it easier to read. The testcase update also show that we can save some unecessary fmr as well. Reviewers: #powerpc, steven.zhang, hfinkel, nemanjai Reviewed By: #powerpc, nemanjai Subscribers: wuzish, hiraditya, kbarton, MaskRay, shchenz, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D68112 llvm-svn: 373652
llvm · Oct 3, 2019 · 4a6881e · 4a6881e
1 parent 18350af
commit 4a6881e
Show file tree

Hide file tree

Showing 6 changed files with 21 additions and 23 deletions.
diff --git a/llvm/lib/Target/PowerPC/PPCInstrVSX.td b/llvm/lib/Target/PowerPC/PPCInstrVSX.td
@@ -1011,21 +1011,21 @@ def : Pat<(f64 (extractelt v2f64:$S, 1)),
           (f64 (EXTRACT_SUBREG $S, sub_64))>;
 }
 
-// Additional fnmsub patterns: -a*c + b == -(a*c - b)
-def : Pat<(fma (fneg f64:$A), f64:$C, f64:$B),
-          (XSNMSUBADP $B, $C, $A)>;
-def : Pat<(fma f64:$A, (fneg f64:$C), f64:$B),
-          (XSNMSUBADP $B, $C, $A)>;
-
-def : Pat<(fma (fneg v2f64:$A), v2f64:$C, v2f64:$B),
-          (XVNMSUBADP $B, $C, $A)>;
-def : Pat<(fma v2f64:$A, (fneg v2f64:$C), v2f64:$B),
-          (XVNMSUBADP $B, $C, $A)>;
-
-def : Pat<(fma (fneg v4f32:$A), v4f32:$C, v4f32:$B),
-          (XVNMSUBASP $B, $C, $A)>;
-def : Pat<(fma v4f32:$A, (fneg v4f32:$C), v4f32:$B),
-          (XVNMSUBASP $B, $C, $A)>;
+// Additional fnmsub patterns: -a*b + c == -(a*b - c)
+def : Pat<(fma (fneg f64:$A), f64:$B, f64:$C),
+          (XSNMSUBADP $C, $A, $B)>;
+def : Pat<(fma f64:$A, (fneg f64:$B), f64:$C),
+          (XSNMSUBADP $C, $A, $B)>;
+
+def : Pat<(fma (fneg v2f64:$A), v2f64:$B, v2f64:$C),
+          (XVNMSUBADP $C, $A, $B)>;
+def : Pat<(fma v2f64:$A, (fneg v2f64:$B), v2f64:$C),
+          (XVNMSUBADP $C, $A, $B)>;
+
+def : Pat<(fma (fneg v4f32:$A), v4f32:$B, v4f32:$C),
+          (XVNMSUBASP $C, $A, $B)>;
+def : Pat<(fma v4f32:$A, (fneg v4f32:$B), v4f32:$C),
+          (XVNMSUBASP $C, $A, $B)>;
 
 def : Pat<(v2f64 (bitconvert v4f32:$A)),
           (COPY_TO_REGCLASS $A, VSRC)>;
@@ -2656,9 +2656,9 @@ let AddedComplexity = 400, Predicates = [HasP9Vector] in {
                                       (fneg (int_ppc_fmaf128_round_to_odd
                                       f128:$vA, f128:$vB, (fneg f128:$vTi))))]>;
 
-  // Additional fnmsub patterns: -a*c + b == -(a*c - b)
-  def : Pat<(fma (fneg f128:$A), f128:$C, f128:$B), (XSNMSUBQP $B, $C, $A)>;
-  def : Pat<(fma f128:$A, (fneg f128:$C), f128:$B), (XSNMSUBQP $B, $C, $A)>;
+  // Additional fnmsub patterns: -a*b + c == -(a*b - c)
+  def : Pat<(fma (fneg f128:$A), f128:$B, f128:$C), (XSNMSUBQP $C, $A, $B)>;
+  def : Pat<(fma f128:$A, (fneg f128:$B), f128:$C), (XSNMSUBQP $C, $A, $B)>;
 
   //===--------------------------------------------------------------------===//
   // Quad/Double-Precision Compare Instructions:

diff --git a/llvm/test/CodeGen/PowerPC/combine-fneg.ll b/llvm/test/CodeGen/PowerPC/combine-fneg.ll
@@ -14,7 +14,7 @@ define <4 x double> @fneg_fdiv_splat(double %a0, <4 x double> %a1) {
 ; CHECK-NEXT:    xvredp 2, 0
 ; CHECK-NEXT:    addi 3, 3, .LCPI0_1@toc@l
 ; CHECK-NEXT:    xxswapd 1, 1
-; CHECK-NEXT:    xvnmsubadp 1, 2, 0
+; CHECK-NEXT:    xvnmsubadp 1, 0, 2
 ; CHECK-NEXT:    xvmaddadp 2, 2, 1
 ; CHECK-NEXT:    lxvd2x 1, 0, 3
 ; CHECK-NEXT:    xxswapd 1, 1

diff --git a/llvm/test/CodeGen/PowerPC/f128-fma.ll b/llvm/test/CodeGen/PowerPC/f128-fma.ll
@@ -129,7 +129,7 @@ entry:
 ; CHECK-DAG: lxv v[[REG3:[0-9]+]], 0(r3)
 ; CHECK-DAG: lxv v[[REG4:[0-9]+]], 0(r4)
 ; CHECK-DAG: lxv v[[REG5:[0-9]+]], 0(r5)
-; CHECK: xsnmsubqp v[[REG3]], v[[REG5]], v[[REG4]]
+; CHECK: xsnmsubqp v[[REG3]], v[[REG4]], v[[REG5]]
 ; CHECK-NEXT: stxv v[[REG3]], 0(r6)
 ; CHECK-NEXT: blr
 }

diff --git a/llvm/test/CodeGen/PowerPC/fma-assoc.ll b/llvm/test/CodeGen/PowerPC/fma-assoc.ll
@@ -260,7 +260,6 @@ define double @test_FMSUB_ASSOC_EXT3(float %A, float %B, double %C,
 ; CHECK-VSX-LABEL: test_FMSUB_ASSOC_EXT3:
 ; CHECK-VSX: xsnmsubmdp
 ; CHECK-VSX-NEXT: xsnmsubadp
-; CHECK-VSX-NEXT: fmr
 ; CHECK-VSX-NEXT: blr
 }
 

diff --git a/llvm/test/CodeGen/PowerPC/fma-ext.ll b/llvm/test/CodeGen/PowerPC/fma-ext.ll
@@ -54,7 +54,6 @@ define double @test_FMSUB_EXT2(float %A, float %B, double %C) {
 
 ; CHECK-VSX-LABEL: test_FMSUB_EXT2:
 ; CHECK-VSX: xsnmsubmdp
-; CHECK-VSX-NEXT: fmr
 ; CHECK-VSX-NEXT: blr
 }
 

diff --git a/llvm/test/CodeGen/PowerPC/repeated-fp-divisors.ll b/llvm/test/CodeGen/PowerPC/repeated-fp-divisors.ll
@@ -13,7 +13,7 @@ define <4 x float> @repeated_fp_divisor(float %a, <4 x float> %b) {
 ; CHECK-NEXT:    lvx 4, 0, 3
 ; CHECK-NEXT:    xxspltw 0, 0, 0
 ; CHECK-NEXT:    xvresp 1, 0
-; CHECK-NEXT:    xvnmsubasp 35, 1, 0
+; CHECK-NEXT:    xvnmsubasp 35, 0, 1
 ; CHECK-NEXT:    xvmulsp 0, 34, 36
 ; CHECK-NEXT:    xvmaddasp 1, 1, 35
 ; CHECK-NEXT:    xvmulsp 34, 0, 1