[llvm-commits] [llvm] r98504 - /llvm/trunk/test/CodeGen/ARM/fp16.ll

Anton Korobeynikov asl at math.spbu.ru
Sun Mar 14 11:42:43 PDT 2010


Author: asl
Date: Sun Mar 14 13:42:43 2010
New Revision: 98504

URL: http://llvm.org/viewvc/llvm-project?rev=98504&view=rev
Log:
Feature test for half precision FP.

Added:
    llvm/trunk/test/CodeGen/ARM/fp16.ll

Added: llvm/trunk/test/CodeGen/ARM/fp16.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/ARM/fp16.ll?rev=98504&view=auto
==============================================================================
--- llvm/trunk/test/CodeGen/ARM/fp16.ll (added)
+++ llvm/trunk/test/CodeGen/ARM/fp16.ll Sun Mar 14 13:42:43 2010
@@ -0,0 +1,32 @@
+; RUN: llc < %s | FileCheck %s
+; RUN: llc -mattr=+fp16,+vfp3 | FileCheck --check-prefix=CHECK-FP16 %s
+target datalayout = "e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-n32"
+target triple = "armv7-eabi"
+
+ at x = global i16 12902
+ at y = global i16 0
+ at z = common global i16 0
+
+define arm_aapcs_vfpcc void @foo() nounwind {
+; CHECK: foo:
+; CHECK-FP6: foo:
+entry:
+  %0 = load i16* @x, align 2
+  %1 = load i16* @y, align 2
+  %2 = tail call float @llvm.convert.from.fp16(i16 %0)
+; CHECK: __gnu_h2f_ieee
+; CHECK-FP16: vcvtb.f16.f32
+  %3 = tail call float @llvm.convert.from.fp16(i16 %1)
+; CHECK: __gnu_h2f_ieee
+; CHECK-FP16: vcvtb.f16.f32
+  %4 = fadd float %2, %3
+  %5 = tail call i16 @llvm.convert.to.fp16(float %4)
+; CHECK: __gnu_f2h_ieee
+; CHECK-FP16: vcvtb.f32.f16
+  store i16 %5, i16* @x, align 2
+  ret void
+}
+
+declare float @llvm.convert.from.fp16(i16) nounwind readnone
+
+declare i16 @llvm.convert.to.fp16(float) nounwind readnone





More information about the llvm-commits mailing list