Limit multi-byte nop instructions to 10 bytes

There is no performance advantage to use multi-byte nop instructions
greater than 10 bytes.  This patch limits multi-byte nop instructions
to 10 bytes.  Since there is only one way to encode multi-byte nop
instructions now, it also removed redundant nop tests.

gas/

	* config/tc-i386.c (i386_align_code): Limit multi-byte nop
	instructions to 10 bytes.

gas/testsuite/

	* gas/i386/i386.exp: Don't run nops-1-bdver1, nops-1-bdver2,
	nops-1-bdver3, nops-1-bdver4, nops-1-znver1, nops-1-btver1
	nops-1-btver2, x86-64-nops-1-nocona, x86-64-nops-1-bdver1,
	x86-64-nops-1-bdver2, x86-64-nops-1-bdver3, x86-64-nops-1-bdver4,
	x86-64-nops-1-znver1, x86-64-nops-1-btver1 nor
	x86-64-nops-1-btver2.
	* gas/i386/nops-1-core2.d: Updated.
	* gas/i386/nops-1-k8.d: Likewise.
	* gas/i386/nops-4a-i686.d: Likewise.
	* gas/i386/nops-5-i686.d: Likewise.
	* gas/i386/nops-5.d: Likewise.
	* gas/i386/nops-6.d: Likewise.
	* gas/i386/x86-64-nops-1-core2.d: Likewise.
	* gas/i386/x86-64-nops-1-g64.d: Likewise.
	* gas/i386/x86-64-nops-1-k8.d: Likewise.
	* gas/i386/x86-64-nops-1.d: Likewise.
	* gas/i386/x86-64-nops-2.d: Likewise.
	* gas/i386/x86-64-nops-3.d: Likewise.
	* gas/i386/x86-64-nops-4-core2.d: Likewise.
	* gas/i386/x86-64-nops-4-k8.d: Likewise.
	* gas/i386/x86-64-nops-4.d: Likewise.
	* gas/i386/x86-64-nops-5-k8.d: Likewise.
	* gas/i386/x86-64-nops-5.d: Likewise.
	* gas/i386/ilp32/x86-64-nops-1-core2.d: Likewise.
	* gas/i386/ilp32/x86-64-nops-1-k8.d: Likewise.
	* gas/i386/ilp32/x86-64-nops-1.d: Likewise.
	* gas/i386/ilp32/x86-64-nops-2.d: Likewise.
	* gas/i386/ilp32/x86-64-nops-3.d: Likewise.
	* gas/i386/ilp32/x86-64-nops-4-core2.d: Likewise.
	* gas/i386/ilp32/x86-64-nops-4-k8.d: Likewise.
	* gas/i386/ilp32/x86-64-nops-4.d: Likewise.
	* gas/i386/ilp32/x86-64-nops-5-k8.d: Likewise.
	* gas/i386/ilp32/x86-64-nops-5.d: Likewise.
	* gas/i386/nops-1-bdver1.d: Removed.
	* gas/i386/nops-1-bdver2.d: Likewise.
	* gas/i386/nops-1-bdver3.d: Likewise.
	* gas/i386/nops-1-bdver4.d: Likewise.
	* gas/i386/nops-1-btver1.d: Likewise.
	* gas/i386/nops-1-btver2.d: Likewise.
	* gas/i386/nops-1-znver1.d: Likewise.
	* gas/i386/x86-64-nops-1-bdver1.d: Likewise.
	* gas/i386/x86-64-nops-1-bdver2.d: Likewise.
	* gas/i386/x86-64-nops-1-bdver3.d: Likewise.
	* gas/i386/x86-64-nops-1-bdver4.d: Likewise.
	* gas/i386/x86-64-nops-1-btver1.d: Likewise.
	* gas/i386/x86-64-nops-1-btver2.d: Likewise.
	* gas/i386/x86-64-nops-1-nocona.d: Likewise.
	* gas/i386/x86-64-nops-1-znver1.d: Likewise.
	* gas/i386/ilp32/x86-64-nops-1-nocona.d: Likewise.

1 files changed, 15 insertions, 98 deletions

diff --git a/gas/config/tc-i386.c b/gas/config/tc-i386.c
index b6e8f5c..d710216 100644
--- a/gas/config/tc-i386.c
+++ b/gas/config/tc-i386.c
@@ -1138,85 +1138,9 @@ i386_align_code (fragS *fragP, int count)
   /* nopw %cs:0L(%[re]ax,%[re]ax,1) */
   static const char alt_10[] =
     {0x66,0x2e,0x0f,0x1f,0x84,0x00,0x00,0x00,0x00,0x00};
-  /* data16
-     nopw %cs:0L(%[re]ax,%[re]ax,1) */
-  static const char alt_long_11[] =
-    {0x66,
-     0x66,0x2e,0x0f,0x1f,0x84,0x00,0x00,0x00,0x00,0x00};
-  /* data16
-     data16
-     nopw %cs:0L(%[re]ax,%[re]ax,1) */
-  static const char alt_long_12[] =
-    {0x66,
-     0x66,
-     0x66,0x2e,0x0f,0x1f,0x84,0x00,0x00,0x00,0x00,0x00};
-  /* data16
-     data16
-     data16
-     nopw %cs:0L(%[re]ax,%[re]ax,1) */
-  static const char alt_long_13[] =
-    {0x66,
-     0x66,
-     0x66,
-     0x66,0x2e,0x0f,0x1f,0x84,0x00,0x00,0x00,0x00,0x00};
-  /* data16
-     data16
-     data16
-     data16
-     nopw %cs:0L(%[re]ax,%[re]ax,1) */
-  static const char alt_long_14[] =
-    {0x66,
-     0x66,
-     0x66,
-     0x66,
-     0x66,0x2e,0x0f,0x1f,0x84,0x00,0x00,0x00,0x00,0x00};
-  /* data16
-     data16
-     data16
-     data16
-     data16
-     nopw %cs:0L(%[re]ax,%[re]ax,1) */
-  static const char alt_long_15[] =
-    {0x66,
-     0x66,
-     0x66,
-     0x66,
-     0x66,
-     0x66,0x2e,0x0f,0x1f,0x84,0x00,0x00,0x00,0x00,0x00};
-  /* nopl 0(%[re]ax,%[re]ax,1)
-     nopw 0(%[re]ax,%[re]ax,1) */
-  static const char alt_short_11[] =
-    {0x0f,0x1f,0x44,0x00,0x00,
-     0x66,0x0f,0x1f,0x44,0x00,0x00};
-  /* nopw 0(%[re]ax,%[re]ax,1)
-     nopw 0(%[re]ax,%[re]ax,1) */
-  static const char alt_short_12[] =
-    {0x66,0x0f,0x1f,0x44,0x00,0x00,
-     0x66,0x0f,0x1f,0x44,0x00,0x00};
-  /* nopw 0(%[re]ax,%[re]ax,1)
-     nopl 0L(%[re]ax) */
-  static const char alt_short_13[] =
-    {0x66,0x0f,0x1f,0x44,0x00,0x00,
-     0x0f,0x1f,0x80,0x00,0x00,0x00,0x00};
-  /* nopl 0L(%[re]ax)
-     nopl 0L(%[re]ax) */
-  static const char alt_short_14[] =
-    {0x0f,0x1f,0x80,0x00,0x00,0x00,0x00,
-     0x0f,0x1f,0x80,0x00,0x00,0x00,0x00};
-  /* nopl 0L(%[re]ax)
-     nopl 0L(%[re]ax,%[re]ax,1) */
-  static const char alt_short_15[] =
-    {0x0f,0x1f,0x80,0x00,0x00,0x00,0x00,
-     0x0f,0x1f,0x84,0x00,0x00,0x00,0x00,0x00};
-  static const char *const alt_short_patt[] = {
+  static const char *const alt_patt[] = {
     f32_1, f32_2, alt_3, alt_4, alt_5, alt_6, alt_7, alt_8,
-    alt_9, alt_10, alt_short_11, alt_short_12, alt_short_13,
-    alt_short_14, alt_short_15
-  };
-  static const char *const alt_long_patt[] = {
-    f32_1, f32_2, alt_3, alt_4, alt_5, alt_6, alt_7, alt_8,
-    alt_9, alt_10, alt_long_11, alt_long_12, alt_long_13,
-    alt_long_14, alt_long_15
+    alt_9, alt_10
   };
 
   /* Only align for at least a positive non-zero boundary. */
@@ -1228,14 +1152,9 @@ i386_align_code (fragS *fragP, int count)
 
      1. For PROCESSOR_I386, PROCESSOR_I486, PROCESSOR_PENTIUM and
      PROCESSOR_GENERIC32, f32_patt will be used.
-     2. For PROCESSOR_PENTIUMPRO, PROCESSOR_PENTIUM4, PROCESSOR_NOCONA,
-     PROCESSOR_CORE, PROCESSOR_CORE2, PROCESSOR_COREI7, and
-     PROCESSOR_GENERIC64, alt_long_patt will be used.
-     3. For PROCESSOR_ATHLON, PROCESSOR_K6, PROCESSOR_K8 and
-     PROCESSOR_AMDFAM10, PROCESSOR_BD and PROCESSOR_BT, alt_short_patt
-     will be used.
-
-     When -mtune= isn't used, alt_long_patt will be used if
+     2. For the rest, alt_patt will be used.
+
+     When -mtune= isn't used, alt_patt will be used if
      cpu_arch_isa_flags has CpuNop.  Otherwise, f32_patt will
      be used.
 
@@ -1268,7 +1187,7 @@ i386_align_code (fragS *fragP, int count)
 	      /* We use cpu_arch_isa_flags to check if we SHOULD
 		 optimize with nops.  */
 	      if (fragP->tc_frag_data.isa_flags.bitfield.cpunop)
-		patt = alt_long_patt;
+		patt = alt_patt;
 	      else
 		patt = f32_patt;
 	      break;
@@ -1280,8 +1199,6 @@ i386_align_code (fragS *fragP, int count)
 	    case PROCESSOR_L1OM:
 	    case PROCESSOR_K1OM:
 	    case PROCESSOR_GENERIC64:
-	      patt = alt_long_patt;
-	      break;
 	    case PROCESSOR_K6:
 	    case PROCESSOR_ATHLON:
 	    case PROCESSOR_K8:
@@ -1289,7 +1206,7 @@ i386_align_code (fragS *fragP, int count)
 	    case PROCESSOR_BD:
 	    case PROCESSOR_ZNVER:
 	    case PROCESSOR_BT:
-	      patt = alt_short_patt;
+	      patt = alt_patt;
 	      break;
 	    case PROCESSOR_I386:
 	    case PROCESSOR_I486:
@@ -1324,7 +1241,7 @@ i386_align_code (fragS *fragP, int count)
 	      /* We use cpu_arch_isa_flags to check if we CAN optimize
 		 with nops.  */
 	      if (fragP->tc_frag_data.isa_flags.bitfield.cpunop)
-		patt = alt_short_patt;
+		patt = alt_patt;
 	      else
 		patt = f32_patt;
 	      break;
@@ -1337,12 +1254,12 @@ i386_align_code (fragS *fragP, int count)
 	    case PROCESSOR_L1OM:
 	    case PROCESSOR_K1OM:
 	      if (fragP->tc_frag_data.isa_flags.bitfield.cpunop)
-		patt = alt_long_patt;
+		patt = alt_patt;
 	      else
 		patt = f32_patt;
 	      break;
 	    case PROCESSOR_GENERIC64:
-	      patt = alt_long_patt;
+	      patt = alt_patt;
 	      break;
 	    }
 	}
@@ -1373,15 +1290,15 @@ i386_align_code (fragS *fragP, int count)
 	}
       else
 	{
-	  /* Maximum length of an instruction is 15 byte.  If the
-	     padding is greater than 15 bytes and we don't use jump,
+	  /* Maximum length of an instruction is 10 byte.  If the
+	     padding is greater than 10 bytes and we don't use jump,
 	     we have to break it into smaller pieces.  */
 	  int padding = count;
-	  while (padding > 15)
+	  while (padding > 10)
 	    {
-	      padding -= 15;
+	      padding -= 10;
 	      memcpy (fragP->fr_literal + fragP->fr_fix + padding,
-		      patt [14], 15);
+		      patt [9], 10);
 	    }
 
 	  if (padding)