[llvm] r257470 - The isel pattern that selects the memory-register form of VCVTPH2PS

[Robert Lougher via llvm-commits]

Author: rlougher
Date: Tue Jan 12 05:48:25 2016
New Revision: 257470

URL: http://llvm.org/viewvc/llvm-project?rev=257470&view=rev
Log:
The isel pattern that selects the memory-register form of VCVTPH2PS
(64 to 128-bit) matches against the pattern fragment 'vzmovl_v2i64'
(a zero-extended 64-bit load).

However, a change in r248784 teaches the instruction combiner that only
the lower 64 bits of the input to a 128-bit vcvtph2ps are used.  This means
the instruction combiner will ordinarily optimize away the upper 64-bit
insertelement instruction in the zero-extension and so we no longer select
the memory-register form.  To fix this a new pattern has been added.

Differential Revision: http://reviews.llvm.org/D16067

Modified:
    llvm/trunk/lib/Target/X86/X86InstrSSE.td
    llvm/trunk/test/CodeGen/X86/f16c-intrinsics.ll

Modified: llvm/trunk/lib/Target/X86/X86InstrSSE.td
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/X86/X86InstrSSE.td?rev=257470&r1=257469&r2=257470&view=diff
==============================================================================
--- llvm/trunk/lib/Target/X86/X86InstrSSE.td (original)
+++ llvm/trunk/lib/Target/X86/X86InstrSSE.td Tue Jan 12 05:48:25 2016
@@ -8257,6 +8257,9 @@ let Predicates = [HasF16C] in {
             (VCVTPH2PSrm addr:$src)>;
   def : Pat<(int_x86_vcvtph2ps_128 (vzload_v2i64 addr:$src)),
             (VCVTPH2PSrm addr:$src)>;
+  def : Pat<(int_x86_vcvtph2ps_128 (bitconvert
+              (v2i64 (scalar_to_vector (loadi64 addr:$src))))),
+            (VCVTPH2PSrm addr:$src)>;
 
   def : Pat<(store (f64 (extractelt (bc_v2f64 (v8i16
                   (int_x86_vcvtps2ph_128 VR128:$src1, i32:$src2))), (iPTR 0))),

Modified: llvm/trunk/test/CodeGen/X86/f16c-intrinsics.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/X86/f16c-intrinsics.ll?rev=257470&r1=257469&r2=257470&view=diff
==============================================================================
--- llvm/trunk/test/CodeGen/X86/f16c-intrinsics.ll (original)
+++ llvm/trunk/test/CodeGen/X86/f16c-intrinsics.ll Tue Jan 12 05:48:25 2016
@@ -61,6 +61,18 @@ define <4 x float> @test_x86_vcvtps2ph_1
   ret <4 x float> %res
 }
 
+define <4 x float> @test_x86_vcvtps2ph_128_scalar2(i64* %ptr) {
+; CHECK-LABEL: test_x86_vcvtps2ph_128_scalar2:
+; CHECK-NOT: vmov
+; CHECK: vcvtph2ps (%
+
+  %load = load i64, i64* %ptr
+  %ins = insertelement <2 x i64> undef, i64 %load, i32 0
+  %bc = bitcast <2 x i64> %ins to <8 x i16>
+  %res = tail call <4 x float> @llvm.x86.vcvtph2ps.128(<8 x i16> %bc)
+  ret <4 x float> %res
+}
+
 define void @test_x86_vcvtps2ph_256_m(<8 x i16>* nocapture %d, <8 x float> %a) nounwind {
 entry:
   ; CHECK-LABEL: test_x86_vcvtps2ph_256_m: