File gcc-amdfam10-suse-3.patch of Package gcc-amd-barcelona-4.1.2

Overview Repositories Revisions Requests Users Attributes Meta

File gcc-amdfam10-suse-3.patch of Package gcc-amd-barcelona-4.1.2

Index: gcc/config/i386/i386.c
===================================================================
--- gcc/config/i386/i386.c.orig
+++ gcc/config/i386/i386.c
@@ -851,12 +851,71 @@ const int x86_arch_always_fancy_math_387
    this option on P4 brings over 20% SPECfp regression, while enabling it on
    K8 brings roughly 2.4% regression that can be partly masked by careful scheduling
    of moves.  */
+
+/* 
+Code generation for scalar reg-reg moves:
+   if (x86_sse_partial_reg_dependency == true | x86_sse_split_regs == true)
+     movaps reg, reg
+   else
+     movss reg, reg
+  
+   if (x86_sse_partial_reg_dependency == true)
+     movapd reg, reg
+   else
+     movsd reg, reg
+
+Code generation for scalar loads of double precision data:
+   if (x86_sse_split_regs == true)
+     movlpd mem, reg      (gas syntax)
+   else
+     movsd mem, reg
+
+Code generation for unaligned packed loads of single precision data:
+   if (x86_sse_partial_reg_dependency == true)
+     {
+       if (x86_sse_unaligned_move_optimal)
+         {
+           movups mem, reg
+         }
+       else
+         {
+           xorps  reg3, reg3
+           movaps reg3, reg2
+           movlps mem, reg2
+           movhps mem+8, reg2
+         }
+     }
+   else
+     {
+       movlps mem, reg
+       movhps mem+8, reg
+     }
+Code generation for unaligned packed loads of double precision data:
+   if (x86_sse_split_regs == true)
+     {
+       movlpd mem, reg
+       movhpd mem+8, reg
+     }
+   else
+     {
+       if (x86_sse_unaligned_move_optimal)
+         {
+           movupd mem, reg
+         }
+       else
+         {
+           movsd  mem, reg2
+           movhpd mem+8, reg2
+         }
+     }
+*/
 const int x86_sse_partial_reg_dependency = m_PENT4 | m_NOCONA | m_PPRO | m_GENERIC | m_AMDFAM10;
 /* Set for machines where the type and dependencies are resolved on SSE
    register parts instead of whole registers, so we may maintain just
    lower part of scalar values in proper format leaving the upper part
    undefined.  */
 const int x86_sse_split_regs = m_ATHLON_K8;
+const int x86_sse_unaligned_move_optimal = m_AMDFAM10;
 const int x86_sse_typeless_stores = m_ATHLON_K8 | m_AMDFAM10;
 const int x86_sse_load0_by_pxor = m_PPRO | m_PENT4 | m_NOCONA;
 const int x86_use_ffreep = m_ATHLON_K8 | m_AMDFAM10;
@@ -8923,6 +8982,13 @@ ix86_expand_vector_move_misalign (enum m
 	    }
 	  else
 	    {
+	      if (TARGET_SSE_UNALIGNED_MOVE_OPTIMAL)
+                {
+                  op0 = gen_lowpart (V2DFmode, op0);
+                  op1 = gen_lowpart (V2DFmode, op1);
+                  emit_insn (gen_sse2_movupd (op0, op1));
+                  return;
+                }
 	      /* ??? Not sure about the best option for the Intel chips.
 		 The following would seem to satisfy; the register is
 		 entirely cleared, breaking the dependency chain.  We
@@ -8942,7 +9008,16 @@ ix86_expand_vector_move_misalign (enum m
       else
 	{
 	  if (TARGET_SSE_PARTIAL_REG_DEPENDENCY)
+	    {
+	      if (TARGET_SSE_UNALIGNED_MOVE_OPTIMAL)
+                {
+		  op0 = gen_lowpart (V4SFmode, op0);
+                  op1 = gen_lowpart (V4SFmode, op1);
+		  emit_insn (gen_sse_movups (op0, op1));
+		  return;
+		}
 	    emit_move_insn (op0, CONST0_RTX (mode));
+	    }
 	  else
 	    emit_insn (gen_rtx_CLOBBER (VOIDmode, op0));
 
Index: gcc/config/i386/i386.h
===================================================================
--- gcc/config/i386/i386.h.orig
+++ gcc/config/i386/i386.h
@@ -158,6 +158,7 @@ extern const int x86_accumulate_outgoing
 extern const int x86_epilogue_using_move, x86_decompose_lea;
 extern const int x86_arch_always_fancy_math_387, x86_shift1;
 extern const int x86_sse_partial_reg_dependency, x86_sse_split_regs;
+extern const int x86_sse_unaligned_move_optimal;
 extern const int x86_sse_typeless_stores, x86_sse_load0_by_pxor;
 extern const int x86_use_ffreep;
 extern const int x86_inter_unit_moves, x86_schedule;
@@ -207,6 +208,8 @@ extern int x86_prefetch_sse;
 #define TARGET_PARTIAL_REG_DEPENDENCY (x86_partial_reg_dependency & TUNEMASK)
 #define TARGET_SSE_PARTIAL_REG_DEPENDENCY \
 				      (x86_sse_partial_reg_dependency & TUNEMASK)
+#define TARGET_SSE_UNALIGNED_MOVE_OPTIMAL \
+				      (x86_sse_unaligned_move_optimal & TUNEMASK)
 #define TARGET_SSE_SPLIT_REGS (x86_sse_split_regs & TUNEMASK)
 #define TARGET_SSE_TYPELESS_STORES (x86_sse_typeless_stores & TUNEMASK)
 #define TARGET_SSE_LOAD0_BY_PXOR (x86_sse_load0_by_pxor & TUNEMASK)

Places

File gcc-amdfam10-suse-3.patch of Package gcc-amd-barcelona-4.1.2

Places