[llvm] 8ecbad7 - [AArch64][GlobalISel] Add fpowi fp128 libcall legalization.

[David Green via llvm-commits]

Author: David Green
Date: 2024-10-03T08:48:26+01:00
New Revision: 8ecbad7e416b7c2a8eadf02be553a1caf328a14f

URL: https://github.com/llvm/llvm-project/commit/8ecbad7e416b7c2a8eadf02be553a1caf328a14f
DIFF: https://github.com/llvm/llvm-project/commit/8ecbad7e416b7c2a8eadf02be553a1caf328a14f.diff

LOG: [AArch64][GlobalISel] Add fpowi fp128 libcall legalization.

Added: 
    

Modified: 
    llvm/lib/Target/AArch64/GISel/AArch64LegalizerInfo.cpp
    llvm/test/CodeGen/AArch64/fpowi.ll

Removed: 
    


################################################################################
diff  --git a/llvm/lib/Target/AArch64/GISel/AArch64LegalizerInfo.cpp b/llvm/lib/Target/AArch64/GISel/AArch64LegalizerInfo.cpp
index c00470b4d5c763..9588b49b7a5c2b 100644
--- a/llvm/lib/Target/AArch64/GISel/AArch64LegalizerInfo.cpp
+++ b/llvm/lib/Target/AArch64/GISel/AArch64LegalizerInfo.cpp
@@ -295,7 +295,7 @@ AArch64LegalizerInfo::AArch64LegalizerInfo(const AArch64Subtarget &ST)
   getActionDefinitionsBuilder(G_FPOWI)
       .scalarize(0)
       .minScalar(0, s32)
-      .libcallFor({{s32, s32}, {s64, s32}});
+      .libcallFor({{s32, s32}, {s64, s32}, {s128, s32}});
 
   getActionDefinitionsBuilder(G_INSERT)
       .legalIf(all(typeInSet(0, {s32, s64, p0}),

diff  --git a/llvm/test/CodeGen/AArch64/fpowi.ll b/llvm/test/CodeGen/AArch64/fpowi.ll
index 62fc1c0854ca8b..8948556d1b380a 100644
--- a/llvm/test/CodeGen/AArch64/fpowi.ll
+++ b/llvm/test/CodeGen/AArch64/fpowi.ll
@@ -36,6 +36,29 @@ entry:
   ret half %c
 }
 
+define fp128 @powi_fp128(fp128 %a, i32 %b) {
+; CHECK-LABEL: powi_fp128:
+; CHECK:       // %bb.0: // %entry
+; CHECK-NEXT:    b __powitf2
+entry:
+  %c = call fp128 @llvm.powi.fp128.i32(fp128 %a, i32 %b)
+  ret fp128 %c
+}
+
+define <1 x double> @powi_v1f64(<1 x double> %a, i32 %b) {
+; CHECK-LABEL: powi_v1f64:
+; CHECK:       // %bb.0: // %entry
+; CHECK-NEXT:    str x30, [sp, #-16]! // 8-byte Folded Spill
+; CHECK-NEXT:    .cfi_def_cfa_offset 16
+; CHECK-NEXT:    .cfi_offset w30, -16
+; CHECK-NEXT:    bl __powidf2
+; CHECK-NEXT:    ldr x30, [sp], #16 // 8-byte Folded Reload
+; CHECK-NEXT:    ret
+entry:
+  %c = call <1 x double> @llvm.powi.v1f64.i32(<1 x double> %a, i32 %b)
+  ret <1 x double> %c
+}
+
 define <2 x double> @powi_v2f64(<2 x double> %a, i32 %b) {
 ; CHECK-SD-LABEL: powi_v2f64:
 ; CHECK-SD:       // %bb.0: // %entry
@@ -1406,9 +1429,56 @@ entry:
   ret <16 x half> %c
 }
 
+define <2 x fp128> @powi_v2fp128(<2 x fp128> %a, i32 %b) {
+; CHECK-SD-LABEL: powi_v2fp128:
+; CHECK-SD:       // %bb.0: // %entry
+; CHECK-SD-NEXT:    sub sp, sp, #48
+; CHECK-SD-NEXT:    stp x30, x19, [sp, #32] // 16-byte Folded Spill
+; CHECK-SD-NEXT:    .cfi_def_cfa_offset 48
+; CHECK-SD-NEXT:    .cfi_offset w19, -8
+; CHECK-SD-NEXT:    .cfi_offset w30, -16
+; CHECK-SD-NEXT:    mov w19, w0
+; CHECK-SD-NEXT:    str q1, [sp, #16] // 16-byte Folded Spill
+; CHECK-SD-NEXT:    bl __powitf2
+; CHECK-SD-NEXT:    str q0, [sp] // 16-byte Folded Spill
+; CHECK-SD-NEXT:    ldr q0, [sp, #16] // 16-byte Folded Reload
+; CHECK-SD-NEXT:    mov w0, w19
+; CHECK-SD-NEXT:    bl __powitf2
+; CHECK-SD-NEXT:    ldp x30, x19, [sp, #32] // 16-byte Folded Reload
+; CHECK-SD-NEXT:    mov v1.16b, v0.16b
+; CHECK-SD-NEXT:    ldr q0, [sp] // 16-byte Folded Reload
+; CHECK-SD-NEXT:    add sp, sp, #48
+; CHECK-SD-NEXT:    ret
+;
+; CHECK-GI-LABEL: powi_v2fp128:
+; CHECK-GI:       // %bb.0: // %entry
+; CHECK-GI-NEXT:    sub sp, sp, #48
+; CHECK-GI-NEXT:    stp x30, x19, [sp, #32] // 16-byte Folded Spill
+; CHECK-GI-NEXT:    .cfi_def_cfa_offset 48
+; CHECK-GI-NEXT:    .cfi_offset w19, -8
+; CHECK-GI-NEXT:    .cfi_offset w30, -16
+; CHECK-GI-NEXT:    str q1, [sp, #16] // 16-byte Folded Spill
+; CHECK-GI-NEXT:    mov w19, w0
+; CHECK-GI-NEXT:    bl __powitf2
+; CHECK-GI-NEXT:    str q0, [sp] // 16-byte Folded Spill
+; CHECK-GI-NEXT:    ldr q0, [sp, #16] // 16-byte Folded Reload
+; CHECK-GI-NEXT:    mov w0, w19
+; CHECK-GI-NEXT:    bl __powitf2
+; CHECK-GI-NEXT:    ldp x30, x19, [sp, #32] // 16-byte Folded Reload
+; CHECK-GI-NEXT:    mov v1.16b, v0.16b
+; CHECK-GI-NEXT:    ldr q0, [sp] // 16-byte Folded Reload
+; CHECK-GI-NEXT:    add sp, sp, #48
+; CHECK-GI-NEXT:    ret
+entry:
+  %c = call <2 x fp128> @llvm.powi.v2fp128.i32(<2 x fp128> %a, i32 %b)
+  ret <2 x fp128> %c
+}
+
+declare <1 x double> @llvm.powi.v1f64.i32(<1 x double>, i32)
 declare <16 x half> @llvm.powi.v16f16.i32(<16 x half>, i32)
 declare <2 x double> @llvm.powi.v2f64.i32(<2 x double>, i32)
 declare <2 x float> @llvm.powi.v2f32.i32(<2 x float>, i32)
+declare <2 x fp128> @llvm.powi.v2fp128.i32(<2 x fp128>, i32)
 declare <3 x double> @llvm.powi.v3f64.i32(<3 x double>, i32)
 declare <3 x float> @llvm.powi.v3f32.i32(<3 x float>, i32)
 declare <4 x double> @llvm.powi.v4f64.i32(<4 x double>, i32)
@@ -1419,4 +1489,5 @@ declare <8 x float> @llvm.powi.v8f32.i32(<8 x float>, i32)
 declare <8 x half> @llvm.powi.v8f16.i32(<8 x half>, i32)
 declare double @llvm.powi.f64.i32(double, i32)
 declare float @llvm.powi.f32.i32(float, i32)
+declare fp128 @llvm.powi.fp128.i32(fp128, i32)
 declare half @llvm.powi.f16.i32(half, i32)