i386-c.c (ix86_target_macros_internal): New case.

        * config/i386/i386-c.c (ix86_target_macros_internal): New case.
        (ix86_target_macros_internal): Likewise.

        * config/i386/i386.c (m_CORE2I7): Removed.
        (m_CORE_HASWELL): New macro.
        (m_CORE_ALL): Likewise.
        (initial_ix86_tune_features): m_CORE2I7 is replaced by m_CORE_ALL.
        (initial_ix86_arch_features): Likewise.
        (processor_target_table): Initializations for Core avx2.
        (cpu_names): New names "core-avx2".
        (ix86_option_override_internal): Changed PROCESSOR_COREI7 by
        PROCESSOR_CORE_HASWELL.
        (ix86_issue_rate): New case.
        (ia32_multipass_dfa_lookahead): Likewise.
        (ix86_sched_init_global): Likewise.

        * config/i386/i386.h (TARGET_HASWELL): New macro.
        (target_cpu_default): New TARGET_CPU_DEFAULT_haswell.
        (processor_type): New PROCESSOR_HASWELL.

From-SVN: r195191

4 files changed, 67 insertions, 28 deletions

diff --git a/gcc/ChangeLog b/gcc/ChangeLog
index ea0e4fa..efe553b 100644
--- a/gcc/ChangeLog
+++ b/gcc/ChangeLog
@@ -1,3 +1,25 @@
+2013-01-15  Vladimir Yakovlev  <vladimir.b.yakovlev@intel.com>
+
+	* config/i386/i386-c.c (ix86_target_macros_internal): New case.
+	(ix86_target_macros_internal): Likewise.
+
+	* config/i386/i386.c (m_CORE2I7): Removed.
+	(m_CORE_HASWELL): New macro.
+	(m_CORE_ALL): Likewise.
+	(initial_ix86_tune_features): m_CORE2I7 is replaced by m_CORE_ALL.
+	(initial_ix86_arch_features): Likewise.
+	(processor_target_table): Initializations for Core avx2.
+	(cpu_names): New names "core-avx2".
+	(ix86_option_override_internal): Changed PROCESSOR_COREI7 by
+	PROCESSOR_CORE_HASWELL.
+	(ix86_issue_rate): New case.
+	(ia32_multipass_dfa_lookahead): Likewise.
+	(ix86_sched_init_global): Likewise.
+
+	* config/i386/i386.h (TARGET_HASWELL): New macro.
+	(target_cpu_default): New TARGET_CPU_DEFAULT_haswell.
+	(processor_type): New PROCESSOR_HASWELL.
+
 2013-01-15  Jakub Jelinek  <jakub@redhat.com>
 
 	PR tree-optimization/55955
diff --git a/gcc/config/i386/i386-c.c b/gcc/config/i386/i386-c.c
index 49545d1..51fec84 100644
--- a/gcc/config/i386/i386-c.c
+++ b/gcc/config/i386/i386-c.c
@@ -141,6 +141,10 @@ ix86_target_macros_internal (HOST_WIDE_INT isa_flag,
       def_or_undef (parse_in, "__corei7");
       def_or_undef (parse_in, "__corei7__");
       break;
+    case PROCESSOR_HASWELL:
+      def_or_undef (parse_in, "__core_avx2");
+      def_or_undef (parse_in, "__core_avx2__");
+      break;
     case PROCESSOR_ATOM:
       def_or_undef (parse_in, "__atom");
       def_or_undef (parse_in, "__atom__");
@@ -231,6 +235,9 @@ ix86_target_macros_internal (HOST_WIDE_INT isa_flag,
     case PROCESSOR_COREI7:
       def_or_undef (parse_in, "__tune_corei7__");
       break;
+    case PROCESSOR_HASWELL:
+      def_or_undef (parse_in, "__tune_core_avx2__");
+      break;
     case PROCESSOR_ATOM:
       def_or_undef (parse_in, "__tune_atom__");
       break;
diff --git a/gcc/config/i386/i386.c b/gcc/config/i386/i386.c
index 4f778c1..0e98a1b 100644
--- a/gcc/config/i386/i386.c
+++ b/gcc/config/i386/i386.c
@@ -1730,7 +1730,8 @@ const struct processor_costs *ix86_cost = &pentium_cost;
 #define m_P4_NOCONA (m_PENT4 | m_NOCONA)
 #define m_CORE2 (1<<PROCESSOR_CORE2)
 #define m_COREI7 (1<<PROCESSOR_COREI7)
-#define m_CORE2I7 (m_CORE2 | m_COREI7)
+#define m_HASWELL (1<<PROCESSOR_HASWELL)
+#define m_CORE_ALL (m_CORE2 | m_COREI7  | m_HASWELL)
 #define m_ATOM (1<<PROCESSOR_ATOM)
 
 #define m_GEODE (1<<PROCESSOR_GEODE)
@@ -1766,16 +1767,16 @@ static unsigned int initial_ix86_tune_features[X86_TUNE_LAST] = {
      negatively, so enabling for Generic64 seems like good code size
      tradeoff.  We can't enable it for 32bit generic because it does not
      work well with PPro base chips.  */
-  m_386 | m_CORE2I7 | m_K6_GEODE | m_AMD_MULTIPLE | m_GENERIC64,
+  m_386 | m_CORE_ALL | m_K6_GEODE | m_AMD_MULTIPLE | m_GENERIC64,
 
   /* X86_TUNE_PUSH_MEMORY */
-  m_386 | m_P4_NOCONA | m_CORE2I7 | m_K6_GEODE | m_AMD_MULTIPLE | m_GENERIC,
+  m_386 | m_P4_NOCONA | m_CORE_ALL | m_K6_GEODE | m_AMD_MULTIPLE | m_GENERIC,
 
   /* X86_TUNE_ZERO_EXTEND_WITH_AND */
   m_486 | m_PENT,
 
   /* X86_TUNE_UNROLL_STRLEN */
-  m_486 | m_PENT | m_PPRO | m_ATOM | m_CORE2I7 | m_K6 | m_AMD_MULTIPLE | m_GENERIC,
+  m_486 | m_PENT | m_PPRO | m_ATOM | m_CORE_ALL | m_K6 | m_AMD_MULTIPLE | m_GENERIC,
 
   /* X86_TUNE_BRANCH_PREDICTION_HINTS: Branch hints were put in P4 based
      on simulation result. But after P4 was made, no performance benefit
@@ -1787,11 +1788,11 @@ static unsigned int initial_ix86_tune_features[X86_TUNE_LAST] = {
   ~m_386,
 
   /* X86_TUNE_USE_SAHF */
-  m_PPRO | m_P4_NOCONA | m_CORE2I7 | m_ATOM | m_K6_GEODE | m_K8 | m_AMDFAM10 | m_BDVER | m_BTVER | m_GENERIC,
+  m_PPRO | m_P4_NOCONA | m_CORE_ALL | m_ATOM | m_K6_GEODE | m_K8 | m_AMDFAM10 | m_BDVER | m_BTVER | m_GENERIC,
 
   /* X86_TUNE_MOVX: Enable to zero extend integer registers to avoid
      partial dependencies.  */
-  m_PPRO | m_P4_NOCONA | m_CORE2I7 | m_ATOM | m_GEODE | m_AMD_MULTIPLE  | m_GENERIC,
+  m_PPRO | m_P4_NOCONA | m_CORE_ALL | m_ATOM | m_GEODE | m_AMD_MULTIPLE  | m_GENERIC,
 
   /* X86_TUNE_PARTIAL_REG_STALL: We probably ought to watch for partial
      register stalls on Generic32 compilation setting as well.  However
@@ -1804,17 +1805,17 @@ static unsigned int initial_ix86_tune_features[X86_TUNE_LAST] = {
   m_PPRO,
 
   /* X86_TUNE_PARTIAL_FLAG_REG_STALL */
-  m_CORE2I7 | m_GENERIC,
+  m_CORE_ALL | m_GENERIC,
 
   /* X86_TUNE_LCP_STALL: Avoid an expensive length-changing prefix stall
    * on 16-bit immediate moves into memory on Core2 and Corei7.  */
-  m_CORE2I7 | m_GENERIC,
+  m_CORE_ALL | m_GENERIC,
 
   /* X86_TUNE_USE_HIMODE_FIOP */
   m_386 | m_486 | m_K6_GEODE,
 
   /* X86_TUNE_USE_SIMODE_FIOP */
-  ~(m_PENT | m_PPRO | m_CORE2I7 | m_ATOM | m_AMD_MULTIPLE | m_GENERIC),
+  ~(m_PENT | m_PPRO | m_CORE_ALL | m_ATOM | m_AMD_MULTIPLE | m_GENERIC),
 
   /* X86_TUNE_USE_MOV0 */
   m_K6,
@@ -1835,7 +1836,7 @@ static unsigned int initial_ix86_tune_features[X86_TUNE_LAST] = {
   ~(m_PENT | m_PPRO),
 
   /* X86_TUNE_PROMOTE_QIMODE */
-  m_386 | m_486 | m_PENT | m_CORE2I7 | m_ATOM | m_K6_GEODE | m_AMD_MULTIPLE | m_GENERIC,
+  m_386 | m_486 | m_PENT | m_CORE_ALL | m_ATOM | m_K6_GEODE | m_AMD_MULTIPLE | m_GENERIC,
 
   /* X86_TUNE_FAST_PREFIX */
   ~(m_386 | m_486 | m_PENT),
@@ -1876,10 +1877,10 @@ static unsigned int initial_ix86_tune_features[X86_TUNE_LAST] = {
 
   /* X86_TUNE_INTEGER_DFMODE_MOVES: Enable if integer moves are preferred
      for DFmode copies */
-  ~(m_PPRO | m_P4_NOCONA | m_CORE2I7 | m_ATOM | m_GEODE | m_AMD_MULTIPLE | m_ATOM | m_GENERIC),
+  ~(m_PPRO | m_P4_NOCONA | m_CORE_ALL | m_ATOM | m_GEODE | m_AMD_MULTIPLE | m_ATOM | m_GENERIC),
 
   /* X86_TUNE_PARTIAL_REG_DEPENDENCY */
-  m_P4_NOCONA | m_CORE2I7 | m_ATOM | m_AMD_MULTIPLE | m_GENERIC,
+  m_P4_NOCONA | m_CORE_ALL | m_ATOM | m_AMD_MULTIPLE | m_GENERIC,
 
   /* X86_TUNE_SSE_PARTIAL_REG_DEPENDENCY: In the Generic model we have a
      conflict here in between PPro/Pentium4 based chips that thread 128bit
@@ -1890,7 +1891,7 @@ static unsigned int initial_ix86_tune_features[X86_TUNE_LAST] = {
      shows that disabling 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.  */
-  m_PPRO | m_P4_NOCONA | m_CORE2I7 | m_ATOM  | m_AMDFAM10 | m_BDVER | m_GENERIC,
+  m_PPRO | m_P4_NOCONA | m_CORE_ALL | m_ATOM  | m_AMDFAM10 | m_BDVER | m_GENERIC,
 
   /* X86_TUNE_SSE_UNALIGNED_LOAD_OPTIMAL */
   m_COREI7 | m_AMDFAM10 | m_BDVER | m_BTVER,
@@ -1914,7 +1915,7 @@ static unsigned int initial_ix86_tune_features[X86_TUNE_LAST] = {
   m_PPRO | m_P4_NOCONA,
 
   /* X86_TUNE_MEMORY_MISMATCH_STALL */
-  m_P4_NOCONA | m_CORE2I7 | m_ATOM | m_AMD_MULTIPLE | m_GENERIC,
+  m_P4_NOCONA | m_CORE_ALL | m_ATOM | m_AMD_MULTIPLE | m_GENERIC,
 
   /* X86_TUNE_PROLOGUE_USING_MOVE */
   m_PPRO | m_ATHLON_K8,
@@ -1936,28 +1937,28 @@ static unsigned int initial_ix86_tune_features[X86_TUNE_LAST] = {
 
   /* X86_TUNE_FOUR_JUMP_LIMIT: Some CPU cores are not able to predict more
      than 4 branch instructions in the 16 byte window.  */
-  m_PPRO | m_P4_NOCONA | m_CORE2I7 | m_ATOM | m_AMD_MULTIPLE | m_GENERIC,
+  m_PPRO | m_P4_NOCONA | m_CORE_ALL | m_ATOM | m_AMD_MULTIPLE | m_GENERIC,
 
   /* X86_TUNE_SCHEDULE */
-  m_PENT | m_PPRO | m_CORE2I7 | m_ATOM | m_K6_GEODE | m_AMD_MULTIPLE | m_GENERIC,
+  m_PENT | m_PPRO | m_CORE_ALL | m_ATOM | m_K6_GEODE | m_AMD_MULTIPLE | m_GENERIC,
 
   /* X86_TUNE_USE_BT */
-  m_CORE2I7 | m_ATOM | m_AMD_MULTIPLE | m_GENERIC,
+  m_CORE_ALL | m_ATOM | m_AMD_MULTIPLE | m_GENERIC,
 
   /* X86_TUNE_USE_INCDEC */
-  ~(m_P4_NOCONA | m_CORE2I7 | m_ATOM | m_GENERIC),
+  ~(m_P4_NOCONA | m_CORE_ALL | m_ATOM | m_GENERIC),
 
   /* X86_TUNE_PAD_RETURNS */
-  m_CORE2I7 | m_AMD_MULTIPLE | m_GENERIC,
+  m_CORE_ALL | m_AMD_MULTIPLE | m_GENERIC,
 
   /* X86_TUNE_PAD_SHORT_FUNCTION: Pad short funtion.  */
   m_ATOM,
 
   /* X86_TUNE_EXT_80387_CONSTANTS */
-  m_PPRO | m_P4_NOCONA | m_CORE2I7 | m_ATOM | m_K6_GEODE | m_ATHLON_K8 | m_GENERIC,
+  m_PPRO | m_P4_NOCONA | m_CORE_ALL | m_ATOM | m_K6_GEODE | m_ATHLON_K8 | m_GENERIC,
 
   /* X86_TUNE_AVOID_VECTOR_DECODE */
-  m_CORE2I7 | m_K8 | m_GENERIC64,
+  m_CORE_ALL | m_K8 | m_GENERIC64,
 
   /* X86_TUNE_PROMOTE_HIMODE_IMUL: Modern CPUs have same latency for HImode
      and SImode multiply, but 386 and 486 do HImode multiply faster.  */
@@ -1965,11 +1966,11 @@ static unsigned int initial_ix86_tune_features[X86_TUNE_LAST] = {
 
   /* X86_TUNE_SLOW_IMUL_IMM32_MEM: Imul of 32-bit constant and memory is
      vector path on AMD machines.  */
-  m_CORE2I7 | m_K8 | m_AMDFAM10 | m_BDVER | m_BTVER | m_GENERIC64,
+  m_CORE_ALL | m_K8 | m_AMDFAM10 | m_BDVER | m_BTVER | m_GENERIC64,
 
   /* X86_TUNE_SLOW_IMUL_IMM8: Imul of 8-bit constant is vector path on AMD
      machines.  */
-  m_CORE2I7 | m_K8 | m_AMDFAM10 | m_BDVER | m_BTVER | m_GENERIC64,
+  m_CORE_ALL | m_K8 | m_AMDFAM10 | m_BDVER | m_BTVER | m_GENERIC64,
 
   /* X86_TUNE_MOVE_M1_VIA_OR: On pentiums, it is faster to load -1 via OR
      than a MOV.  */
@@ -1986,7 +1987,7 @@ static unsigned int initial_ix86_tune_features[X86_TUNE_LAST] = {
 
   /* X86_TUNE_USE_VECTOR_FP_CONVERTS: Prefer vector packed SSE conversion
      from FP to FP. */
-  m_CORE2I7 | m_AMDFAM10 | m_GENERIC,
+  m_CORE_ALL | m_AMDFAM10 | m_GENERIC,
 
   /* X86_TUNE_USE_VECTOR_CONVERTS: Prefer vector packed SSE conversion
      from integer to FP. */
@@ -2024,7 +2025,7 @@ static unsigned int initial_ix86_tune_features[X86_TUNE_LAST] = {
 
   /* X86_TUNE_GENERAL_REGS_SSE_SPILL: Try to spill general regs to SSE
      regs instead of memory.  */
-  m_COREI7 | m_CORE2I7,
+  m_CORE_ALL,
 
   /* X86_TUNE_AVOID_MEM_OPND_FOR_CMOVE: Try to avoid memory operands for
      a conditional move.  */
@@ -2054,10 +2055,10 @@ static unsigned int initial_ix86_arch_features[X86_ARCH_LAST] = {
 };
 
 static const unsigned int x86_accumulate_outgoing_args
-  = m_PPRO | m_P4_NOCONA | m_ATOM | m_CORE2I7 | m_AMD_MULTIPLE | m_GENERIC;
+  = m_PPRO | m_P4_NOCONA | m_ATOM | m_CORE_ALL | m_AMD_MULTIPLE | m_GENERIC;
 
 static const unsigned int x86_arch_always_fancy_math_387
-  = m_PENT | m_PPRO | m_P4_NOCONA | m_CORE2I7 | m_ATOM | m_AMD_MULTIPLE | m_GENERIC;
+  = m_PENT | m_PPRO | m_P4_NOCONA | m_CORE_ALL | m_ATOM | m_AMD_MULTIPLE | m_GENERIC;
 
 static const unsigned int x86_avx256_split_unaligned_load
   = m_COREI7 | m_GENERIC;
@@ -2432,6 +2433,8 @@ static const struct ptt processor_target_table[PROCESSOR_max] =
   {&core_cost, 16, 10, 16, 10, 16},
   /* Core i7  */
   {&core_cost, 16, 10, 16, 10, 16},
+  /* Core avx2  */
+  {&core_cost, 16, 10, 16, 10, 16},
   {&generic32_cost, 16, 7, 16, 7, 16},
   {&generic64_cost, 16, 10, 16, 10, 16},
   {&amdfam10_cost, 32, 24, 32, 7, 32},
@@ -2459,6 +2462,7 @@ static const char *const cpu_names[TARGET_CPU_DEFAULT_max] =
   "nocona",
   "core2",
   "corei7",
+  "core-avx2",
   "atom",
   "geode",
   "k6",
@@ -2910,7 +2914,7 @@ ix86_option_override_internal (bool main_args_p)
 	| PTA_SSSE3 | PTA_SSE4_1 | PTA_SSE4_2 | PTA_AVX
 	| PTA_CX16 | PTA_POPCNT | PTA_AES | PTA_PCLMUL | PTA_FSGSBASE
 	| PTA_RDRND | PTA_F16C | PTA_FXSR | PTA_XSAVE | PTA_XSAVEOPT},
-      {"core-avx2", PROCESSOR_COREI7, CPU_COREI7,
+      {"core-avx2", PROCESSOR_HASWELL, CPU_COREI7,
 	PTA_64BIT | PTA_MMX | PTA_SSE | PTA_SSE2 | PTA_SSE3
 	| PTA_SSSE3 | PTA_SSE4_1 | PTA_SSE4_2 | PTA_AVX | PTA_AVX2
 	| PTA_CX16 | PTA_POPCNT | PTA_AES | PTA_PCLMUL | PTA_FSGSBASE
@@ -24048,6 +24052,7 @@ ix86_issue_rate (void)
     case PROCESSOR_PENTIUM4:
     case PROCESSOR_CORE2:
     case PROCESSOR_COREI7:
+    case PROCESSOR_HASWELL:
     case PROCESSOR_ATHLON:
     case PROCESSOR_K8:
     case PROCESSOR_AMDFAM10:
@@ -24304,6 +24309,7 @@ ia32_multipass_dfa_lookahead (void)
 
     case PROCESSOR_CORE2:
     case PROCESSOR_COREI7:
+    case PROCESSOR_HASWELL:
     case PROCESSOR_ATOM:
       /* Generally, we want haifa-sched:max_issue() to look ahead as far
 	 as many instructions can be executed on a cycle, i.e.,
@@ -24848,6 +24854,7 @@ ix86_sched_init_global (FILE *dump ATTRIBUTE_UNUSED,
     {
     case PROCESSOR_CORE2:
     case PROCESSOR_COREI7:
+    case PROCESSOR_HASWELL:
       /* Do not perform multipass scheduling for pre-reload schedule
          to save compile time.  */
       if (reload_completed)
diff --git a/gcc/config/i386/i386.h b/gcc/config/i386/i386.h
index f86e346..af293b4 100644
--- a/gcc/config/i386/i386.h
+++ b/gcc/config/i386/i386.h
@@ -246,6 +246,7 @@ extern const struct processor_costs ix86_size_cost;
 #define TARGET_NOCONA (ix86_tune == PROCESSOR_NOCONA)
 #define TARGET_CORE2 (ix86_tune == PROCESSOR_CORE2)
 #define TARGET_COREI7 (ix86_tune == PROCESSOR_COREI7)
+#define TARGET_HASWELL (ix86_tune == PROCESSOR_HASWELL)
 #define TARGET_GENERIC32 (ix86_tune == PROCESSOR_GENERIC32)
 #define TARGET_GENERIC64 (ix86_tune == PROCESSOR_GENERIC64)
 #define TARGET_GENERIC (TARGET_GENERIC32 || TARGET_GENERIC64)
@@ -604,6 +605,7 @@ enum target_cpu_default
   TARGET_CPU_DEFAULT_nocona,
   TARGET_CPU_DEFAULT_core2,
   TARGET_CPU_DEFAULT_corei7,
+  TARGET_CPU_DEFAULT_haswell,
   TARGET_CPU_DEFAULT_atom,
 
   TARGET_CPU_DEFAULT_geode,
@@ -2096,6 +2098,7 @@ enum processor_type
   PROCESSOR_NOCONA,
   PROCESSOR_CORE2,
   PROCESSOR_COREI7,
+  PROCESSOR_HASWELL,
   PROCESSOR_GENERIC32,
   PROCESSOR_GENERIC64,
   PROCESSOR_AMDFAM10,