arm.c (arm_tune_cortex_a9): Define.

2008-09-02  Paul Brook  <paul@codesourcery.com>

	* config/arm/arm.c (arm_tune_cortex_a9): Define.
	(arm_override_options): Set arm_tune_cortex_a9.
	(arm_rtx_costs_1): Make register shifts more expensive on Cortex-A9.
	(arm_issue_rate): Handle cortexa9.
	* config/arm/arm.h (arm_tune_cortex_a9): Declare.
	* config/arm/arm-cores.def: Add cortex-a9.
	* config/arm/arm-tune.md: Regenerate.
	* config/arm/arm.md: Include cortex-a9.md.
	(generic_sched, generic_vfp): Add cortexa9.
	* config/arm/cortex-a9.md: New file.

From-SVN: r139891

6 files changed, 89 insertions, 5 deletions

diff --git a/gcc/config/arm/arm-cores.def b/gcc/config/arm/arm-cores.def
index c7e425b..fe2f2b5 100644
--- a/gcc/config/arm/arm-cores.def
+++ b/gcc/config/arm/arm-cores.def
@@ -116,6 +116,7 @@ ARM_CORE("mpcorenovfp",	  mpcorenovfp,	6K,				 FL_LDSCHED, 9e)
 ARM_CORE("mpcore",	  mpcore,	6K,				 FL_LDSCHED | FL_VFPV2, 9e)
 ARM_CORE("arm1156t2-s",	  arm1156t2s,	6T2,				 FL_LDSCHED, 9e)
 ARM_CORE("cortex-a8",	  cortexa8,	7A,				 FL_LDSCHED, 9e)
+ARM_CORE("cortex-a9",	  cortexa9,	7A,				 FL_LDSCHED, 9e)
 ARM_CORE("cortex-r4",	  cortexr4,	7R,				 FL_LDSCHED, 9e)
 ARM_CORE("cortex-r4f",	  cortexr4f,	7R,				 FL_LDSCHED, 9e)
 ARM_CORE("cortex-m3",	  cortexm3,	7M,				 FL_LDSCHED, 9e)
diff --git a/gcc/config/arm/arm-tune.md b/gcc/config/arm/arm-tune.md
index ee5606b..beb8f9f 100644
--- a/gcc/config/arm/arm-tune.md
+++ b/gcc/config/arm/arm-tune.md
@@ -1,5 +1,5 @@
 ;; -*- buffer-read-only: t -*-
 ;; Generated automatically by gentune.sh from arm-cores.def
 (define_attr "tune"
-	"arm2,arm250,arm3,arm6,arm60,arm600,arm610,arm620,arm7,arm7d,arm7di,arm70,arm700,arm700i,arm710,arm720,arm710c,arm7100,arm7500,arm7500fe,arm7m,arm7dm,arm7dmi,arm8,arm810,strongarm,strongarm110,strongarm1100,strongarm1110,arm7tdmi,arm7tdmis,arm710t,arm720t,arm740t,arm9,arm9tdmi,arm920,arm920t,arm922t,arm940t,ep9312,arm10tdmi,arm1020t,arm9e,arm946es,arm966es,arm968es,arm10e,arm1020e,arm1022e,xscale,iwmmxt,arm926ejs,arm1026ejs,arm1136js,arm1136jfs,arm1176jzs,arm1176jzfs,mpcorenovfp,mpcore,arm1156t2s,cortexa8,cortexr4,cortexr4f,cortexm3,cortexm1"
+	"arm2,arm250,arm3,arm6,arm60,arm600,arm610,arm620,arm7,arm7d,arm7di,arm70,arm700,arm700i,arm710,arm720,arm710c,arm7100,arm7500,arm7500fe,arm7m,arm7dm,arm7dmi,arm8,arm810,strongarm,strongarm110,strongarm1100,strongarm1110,arm7tdmi,arm7tdmis,arm710t,arm720t,arm740t,arm9,arm9tdmi,arm920,arm920t,arm922t,arm940t,ep9312,arm10tdmi,arm1020t,arm9e,arm946es,arm966es,arm968es,arm10e,arm1020e,arm1022e,xscale,iwmmxt,arm926ejs,arm1026ejs,arm1136js,arm1136jfs,arm1176jzs,arm1176jzfs,mpcorenovfp,mpcore,arm1156t2s,cortexa8,cortexa9,cortexr4,cortexr4f,cortexm3,cortexm1"
 	(const (symbol_ref "arm_tune")))
diff --git a/gcc/config/arm/arm.c b/gcc/config/arm/arm.c
index 724f1ac..ec28f79 100644
--- a/gcc/config/arm/arm.c
+++ b/gcc/config/arm/arm.c
@@ -544,6 +544,9 @@ int arm_tune_xscale = 0;
    This typically means an ARM6 or ARM7 with MMU or MPU.  */
 int arm_tune_wbuf = 0;
 
+/* Nonzero if tuning for Cortex-A9.  */
+int arm_tune_cortex_a9 = 0;
+
 /* Nonzero if generating Thumb instructions.  */
 int thumb_code = 0;
 
@@ -1289,6 +1292,7 @@ arm_override_options (void)
   arm_tune_xscale = (tune_flags & FL_XSCALE) != 0;
   arm_arch_iwmmxt = (insn_flags & FL_IWMMXT) != 0;
   arm_arch_hwdiv = (insn_flags & FL_DIV) != 0;
+  arm_tune_cortex_a9 = (arm_tune == cortexa9) != 0;
 
   /* If we are not using the default (ARM mode) section anchor offset
      ranges, then set the correct ranges now.  */
@@ -4913,7 +4917,15 @@ arm_rtx_costs_1 (rtx x, enum rtx_code code, enum rtx_code outer)
 		    || (GET_CODE (XEXP (x, 0)) == SUBREG
 			&& GET_CODE (SUBREG_REG (XEXP (x, 0))) == REG))
 		   ? 0 : 8));
-      return (1 + ((GET_CODE (XEXP (x, 0)) == REG
+
+      extra_cost = 1;
+      /* Increase the cost of complex shifts because they aren't any faster,
+         and reduce dual issue opportunities.  */
+      if (arm_tune_cortex_a9
+	  && outer != SET && GET_CODE (XEXP (x, 1)) != CONST_INT)
+	extra_cost++;
+
+      return (extra_cost + ((GET_CODE (XEXP (x, 0)) == REG
 		    || (GET_CODE (XEXP (x, 0)) == SUBREG
 			&& GET_CODE (SUBREG_REG (XEXP (x, 0))) == REG))
 		   ? 0 : 4)
@@ -5028,7 +5040,8 @@ arm_rtx_costs_1 (rtx x, enum rtx_code code, enum rtx_code outer)
 			 && ((INTVAL (XEXP (XEXP (x, 0), 1)) &
 			      (INTVAL (XEXP (XEXP (x, 0), 1)) - 1)) == 0)))
 		    && (REG_OR_SUBREG_REG (XEXP (XEXP (x, 0), 0)))
-		    && ((REG_OR_SUBREG_REG (XEXP (XEXP (x, 0), 1)))
+		    && ((REG_OR_SUBREG_REG (XEXP (XEXP (x, 0), 1))
+			 && !arm_tune_cortex_a9)
 			|| GET_CODE (XEXP (XEXP (x, 0), 1)) == CONST_INT))
 		   ? 0 : 4));
 
@@ -19066,6 +19079,7 @@ arm_issue_rate (void)
     case cortexr4:
     case cortexr4f:
     case cortexa8:
+    case cortexa9:
       return 2;
 
     default:
diff --git a/gcc/config/arm/arm.h b/gcc/config/arm/arm.h
index e12c3f9..f83aaba 100644
--- a/gcc/config/arm/arm.h
+++ b/gcc/config/arm/arm.h
@@ -407,6 +407,9 @@ extern int arm_tune_xscale;
 /* Nonzero if tuning for stores via the write buffer.  */
 extern int arm_tune_wbuf;
 
+/* Nonzero if tuning for Cortex-A9.  */
+extern int arm_tune_cortex_a9;
+
 /* Nonzero if we should define __THUMB_INTERWORK__ in the
    preprocessor.
    XXX This is a bit of a hack, it's intended to help work around
diff --git a/gcc/config/arm/arm.md b/gcc/config/arm/arm.md
index c5708d6..1c27909 100644
--- a/gcc/config/arm/arm.md
+++ b/gcc/config/arm/arm.md
@@ -341,7 +341,7 @@
 
 (define_attr "generic_sched" "yes,no"
   (const (if_then_else 
-          (ior (eq_attr "tune" "arm926ejs,arm1020e,arm1026ejs,arm1136js,arm1136jfs,cortexa8")
+          (ior (eq_attr "tune" "arm926ejs,arm1020e,arm1026ejs,arm1136js,arm1136jfs,cortexa8,cortexa9")
 	      (eq_attr "tune_cortexr4" "yes"))
           (const_string "no")
           (const_string "yes"))))
@@ -349,7 +349,7 @@
 (define_attr "generic_vfp" "yes,no"
   (const (if_then_else
 	  (and (eq_attr "fpu" "vfp")
-	       (eq_attr "tune" "!arm1020e,arm1022e,cortexa8")
+	       (eq_attr "tune" "!arm1020e,arm1022e,cortexa8,cortexa9")
 	       (eq_attr "tune_cortexr4" "no"))
 	  (const_string "yes")
 	  (const_string "no"))))
@@ -360,6 +360,7 @@
 (include "arm1026ejs.md")
 (include "arm1136jfs.md")
 (include "cortex-a8.md")
+(include "cortex-a9.md")
 (include "cortex-r4.md")
 (include "cortex-r4f.md")
 (include "vfp11.md")
diff --git a/gcc/config/arm/cortex-a9.md b/gcc/config/arm/cortex-a9.md
new file mode 100644
index 0000000..121fd2d
--- /dev/null
+++ b/gcc/config/arm/cortex-a9.md
@@ -0,0 +1,65 @@
+;; ARM Cortex-A9 VFP pipeline description
+;; Copyright (C) 2008 Free Software Foundation, Inc.
+;; Written by CodeSourcery.
+;;
+;; This file is part of GCC.
+;;
+;; GCC is free software; you can redistribute it and/or modify it
+;; under the terms of the GNU General Public License as published by
+;; the Free Software Foundation; either version 3, or (at your option)
+;; any later version.
+;;
+;; GCC is distributed in the hope that it will be useful, but
+;; WITHOUT ANY WARRANTY; without even the implied warranty of
+;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+;; General Public License for more details.
+;;
+;; You should have received a copy of the GNU General Public License
+;; along with GCC; see the file COPYING3.  If not see
+;; <http://www.gnu.org/licenses/>.
+
+(define_automaton "cortex_a9")
+
+;; FIXME: We model a single pipeline for all instructions.
+;; Is dual-issue possible, and do we have other pipelines?
+(define_cpu_unit "cortex_a9_vfp" "cortex_a9")
+
+(define_insn_reservation "cortex_a9_ffarith" 1
+ (and (eq_attr "tune" "cortexa9")
+      (eq_attr "type" "fcpys,ffariths,ffarithd,fcmps,fcmpd,fconsts,fconstd"))
+ "cortex_a9_vfp")
+
+(define_insn_reservation "cortex_a9_fadd" 4
+ (and (eq_attr "tune" "cortexa9")
+      (eq_attr "type" "fadds,faddd,f_cvt"))
+ "cortex_a9_vfp")
+
+(define_insn_reservation "cortex_a9_fmuls" 5
+ (and (eq_attr "tune" "cortexa9")
+      (eq_attr "type" "fmuls"))
+ "cortex_a9_vfp")
+
+(define_insn_reservation "cortex_a9_fmuld" 6
+ (and (eq_attr "tune" "cortexa9")
+      (eq_attr "type" "fmuld"))
+ "cortex_a9_vfp*2")
+
+(define_insn_reservation "cortex_a9_fmacs" 8
+ (and (eq_attr "tune" "cortexa9")
+      (eq_attr "type" "fmacs"))
+ "cortex_a9_vfp")
+
+(define_insn_reservation "cortex_a9_fmacd" 8
+ (and (eq_attr "tune" "cortexa9")
+      (eq_attr "type" "fmacd"))
+ "cortex_a9_vfp*2")
+
+(define_insn_reservation "cortex_a9_fdivs" 15
+ (and (eq_attr "tune" "cortexa9")
+      (eq_attr "type" "fdivs"))
+ "cortex_a9_vfp*10")
+
+(define_insn_reservation "cortex_a9_fdivd" 25
+ (and (eq_attr "tune" "cortexa9")
+      (eq_attr "type" "fdivd"))
+ "cortex_a9_vfp*20")