[multiple changes]

2012-11-20  Uros Bizjak  <ubizjak@gmail.com>

	* config/i386/i386.md (fix_trunc<MODEF:mode><SWI48:mode>_sse): Macroize
	insn from fix_trunc<mode>{si,di}_sse using SWI48 mode iterator.
	(peephole2 to avoid vector decoded forms): Macroize peephole2
	using MODEF mode iterator.  Use SWI48 mode iterator instead of SWI48x.

2012-11-20  Uros Bizjak  <ubizjak@gmail.com>

	PR target/19398
	* config/i386/i386.md
	(peephole2 to shorten x87->SSE reload sequences): Remove peephole2.
	* config/i386/i386.h (enum ix86_tune_indices)
	<IX86_TUNE_SHORTEN_X87_SSE>: Remove.
	* config/i386/i386.h (initial_ix86_tune_features): Update.

2012-11-20  Vladimir Makarov  <vmakarov@redhat.com>

	PR target/19398
	* lra-constraints.c (process_alt_operands): Discourage reloads
	through secodnary memory.

testsuite/ChangeLog:

2012-11-20  Uros Bizjak  <ubizjak@gmail.com>

	PR target/19398
	* gcc.target/i386/pr19398.c: New test.

From-SVN: r193671

7 files changed, 68 insertions, 51 deletions

diff --git a/gcc/ChangeLog b/gcc/ChangeLog
index e022c8a..cc6441c 100644
--- a/gcc/ChangeLog
+++ b/gcc/ChangeLog
@@ -1,3 +1,25 @@
+2012-11-20  Uros Bizjak  <ubizjak@gmail.com>
+
+	* config/i386/i386.md (fix_trunc<MODEF:mode><SWI48:mode>_sse): Macroize
+	insn from fix_trunc<mode>{si,di}_sse using SWI48 mode iterator.
+	(peephole2 to avoid vector decoded forms): Macroize peephole2
+	using MODEF mode iterator.  Use SWI48 mode iterator instead of SWI48x.
+
+2012-11-20  Uros Bizjak  <ubizjak@gmail.com>
+
+	PR target/19398
+	* config/i386/i386.md
+	(peephole2 to shorten x87->SSE reload sequences): Remove peephole2.
+	* config/i386/i386.h (enum ix86_tune_indices)
+	<IX86_TUNE_SHORTEN_X87_SSE>: Remove.
+	* config/i386/i386.h (initial_ix86_tune_features): Update.
+
+2012-11-20  Vladimir Makarov  <vmakarov@redhat.com>
+
+	PR target/19398
+	* lra-constraints.c (process_alt_operands): Discourage reloads
+	through secodnary memory.
+
 2012-11-20  David Edelsohn  <dje.gcc@gmail.com>
 
 	* config/rs6000/rs6000.md (largetoc_low): Revert.
@@ -48,8 +70,7 @@
 	(v850_function_arg_advance): Likewise.
 	(v850_print_operand): Handle CONST_INT and CONST_DOUBLE.
 	(compute_register_save_size): Use df_regs_ever_live_p.
-	(increment_stack): Mark prologue adjustments as being frame
-	related.
+	(increment_stack): Mark prologue adjustments as being frame related.
 	(expand_prologue): Handle pretend args.  Mark insns generated as
 	being frame related.
 	(expand_epilogue): Likewise.
diff --git a/gcc/config/i386/i386.c b/gcc/config/i386/i386.c
index 8ce32be..fc75771 100644
--- a/gcc/config/i386/i386.c
+++ b/gcc/config/i386/i386.c
@@ -1855,9 +1855,6 @@ static unsigned int initial_ix86_tune_features[X86_TUNE_LAST] = {
   /* X86_TUNE_EXT_80387_CONSTANTS */
   m_PPRO | m_P4_NOCONA | m_CORE2I7 | m_ATOM | m_K6_GEODE | m_ATHLON_K8 | m_GENERIC,
 
-  /* X86_TUNE_SHORTEN_X87_SSE */
-  ~m_K8,
-
   /* X86_TUNE_AVOID_VECTOR_DECODE */
   m_CORE2I7_64 | m_K8 | m_GENERIC64,
 
diff --git a/gcc/config/i386/i386.h b/gcc/config/i386/i386.h
index 0cdbee1..ef62683 100644
--- a/gcc/config/i386/i386.h
+++ b/gcc/config/i386/i386.h
@@ -314,7 +314,6 @@ enum ix86_tune_indices {
   X86_TUNE_PAD_RETURNS,
   X86_TUNE_PAD_SHORT_FUNCTION,
   X86_TUNE_EXT_80387_CONSTANTS,
-  X86_TUNE_SHORTEN_X87_SSE,
   X86_TUNE_AVOID_VECTOR_DECODE,
   X86_TUNE_PROMOTE_HIMODE_IMUL,
   X86_TUNE_SLOW_IMUL_IMM32_MEM,
@@ -408,7 +407,6 @@ extern unsigned char ix86_tune_features[X86_TUNE_LAST];
 	ix86_tune_features[X86_TUNE_PAD_SHORT_FUNCTION]
 #define TARGET_EXT_80387_CONSTANTS \
 	ix86_tune_features[X86_TUNE_EXT_80387_CONSTANTS]
-#define TARGET_SHORTEN_X87_SSE	ix86_tune_features[X86_TUNE_SHORTEN_X87_SSE]
 #define TARGET_AVOID_VECTOR_DECODE \
 	ix86_tune_features[X86_TUNE_AVOID_VECTOR_DECODE]
 #define TARGET_TUNE_PROMOTE_HIMODE_IMUL \
diff --git a/gcc/config/i386/i386.md b/gcc/config/i386/i386.md
index aa75d6b..05449ba 100644
--- a/gcc/config/i386/i386.md
+++ b/gcc/config/i386/i386.md
@@ -4465,60 +4465,34 @@
   "operands[2] = gen_reg_rtx (SImode);")
 
 ;; When SSE is available, it is always faster to use it!
-(define_insn "fix_trunc<mode>di_sse"
-  [(set (match_operand:DI 0 "register_operand" "=r,r")
-	(fix:DI (match_operand:MODEF 1 "nonimmediate_operand" "x,m")))]
-  "TARGET_64BIT && SSE_FLOAT_MODE_P (<MODE>mode)
+(define_insn "fix_trunc<MODEF:mode><SWI48:mode>_sse"
+  [(set (match_operand:SWI48 0 "register_operand" "=r,r")
+	(fix:SWI48 (match_operand:MODEF 1 "nonimmediate_operand" "x,m")))]
+  "SSE_FLOAT_MODE_P (<MODEF:MODE>mode)
    && (!TARGET_FISTTP || TARGET_SSE_MATH)"
-  "%vcvtt<ssemodesuffix>2si{q}\t{%1, %0|%0, %1}"
+  "%vcvtt<MODEF:ssemodesuffix>2si<SWI48:rex64suffix>\t{%1, %0|%0, %1}"
   [(set_attr "type" "sseicvt")
    (set_attr "prefix" "maybe_vex")
-   (set_attr "prefix_rex" "1")
-   (set_attr "mode" "<MODE>")
-   (set_attr "athlon_decode" "double,vector")
-   (set_attr "amdfam10_decode" "double,double")
-   (set_attr "bdver1_decode" "double,double")])
-
-(define_insn "fix_trunc<mode>si_sse"
-  [(set (match_operand:SI 0 "register_operand" "=r,r")
-	(fix:SI (match_operand:MODEF 1 "nonimmediate_operand" "x,m")))]
-  "SSE_FLOAT_MODE_P (<MODE>mode)
-   && (!TARGET_FISTTP || TARGET_SSE_MATH)"
-  "%vcvtt<ssemodesuffix>2si\t{%1, %0|%0, %1}"
-  [(set_attr "type" "sseicvt")
-   (set_attr "prefix" "maybe_vex")
-   (set_attr "mode" "<MODE>")
+   (set (attr "prefix_rex")
+	(if_then_else
+	  (match_test "<SWI48:MODE>mode == DImode")
+	  (const_string "1")
+	  (const_string "*")))
+   (set_attr "mode" "<MODEF:MODE>")
    (set_attr "athlon_decode" "double,vector")
    (set_attr "amdfam10_decode" "double,double")
    (set_attr "bdver1_decode" "double,double")])
 
-;; Shorten x87->SSE reload sequences of fix_trunc?f?i_sse patterns.
-(define_peephole2
-  [(set (match_operand:MODEF 0 "register_operand")
-	(match_operand:MODEF 1 "memory_operand"))
-   (set (match_operand:SWI48x 2 "register_operand")
-	(fix:SWI48x (match_dup 0)))]
-  "TARGET_SHORTEN_X87_SSE
-   && !(TARGET_AVOID_VECTOR_DECODE && optimize_insn_for_speed_p ())
-   && peep2_reg_dead_p (2, operands[0])"
-  [(set (match_dup 2) (fix:SWI48x (match_dup 1)))])
-
 ;; Avoid vector decoded forms of the instruction.
 (define_peephole2
-  [(match_scratch:DF 2 "x")
-   (set (match_operand:SWI48x 0 "register_operand")
-	(fix:SWI48x (match_operand:DF 1 "memory_operand")))]
-  "TARGET_SSE2 && TARGET_AVOID_VECTOR_DECODE && optimize_insn_for_speed_p ()"
-  [(set (match_dup 2) (match_dup 1))
-   (set (match_dup 0) (fix:SWI48x (match_dup 2)))])
-
-(define_peephole2
-  [(match_scratch:SF 2 "x")
-   (set (match_operand:SWI48x 0 "register_operand")
-	(fix:SWI48x (match_operand:SF 1 "memory_operand")))]
-  "TARGET_AVOID_VECTOR_DECODE && optimize_insn_for_speed_p ()"
+  [(match_scratch:MODEF 2 "x")
+   (set (match_operand:SWI48 0 "register_operand")
+	(fix:SWI48 (match_operand:MODEF 1 "memory_operand")))]
+  "TARGET_AVOID_VECTOR_DECODE
+   && SSE_FLOAT_MODE_P (<MODEF:MODE>mode)
+   && optimize_insn_for_speed_p ()"
   [(set (match_dup 2) (match_dup 1))
-   (set (match_dup 0) (fix:SWI48x (match_dup 2)))])
+   (set (match_dup 0) (fix:SWI48 (match_dup 2)))])
 
 (define_insn_and_split "fix_trunc<mode>_fisttp_i387_1"
   [(set (match_operand:SWI248x 0 "nonimmediate_operand")
diff --git a/gcc/lra-constraints.c b/gcc/lra-constraints.c
index 6f19c18..9df7b97 100644
--- a/gcc/lra-constraints.c
+++ b/gcc/lra-constraints.c
@@ -1942,6 +1942,19 @@ process_alt_operands (int only_alternative)
 	      if (no_regs_p && REG_P (op))
 		reject++;
 
+#ifdef SECONDARY_MEMORY_NEEDED
+	      /* If reload requires moving value through secondary
+		 memory, it will need one more insn at least.  */
+	      if (this_alternative != NO_REGS 
+		  && REG_P (op) && (cl = get_reg_class (REGNO (op))) != NO_REGS
+		  && ((curr_static_id->operand[nop].type != OP_OUT
+		       && SECONDARY_MEMORY_NEEDED (cl, this_alternative,
+						   GET_MODE (op)))
+		      || (curr_static_id->operand[nop].type != OP_IN
+			  && SECONDARY_MEMORY_NEEDED (this_alternative, cl,
+						      GET_MODE (op)))))
+		losers++;
+#endif
 	      /* Input reloads can be inherited more often than output
 		 reloads can be removed, so penalize output
 		 reloads.  */
diff --git a/gcc/testsuite/ChangeLog b/gcc/testsuite/ChangeLog
index cb65187..0cff709 100644
--- a/gcc/testsuite/ChangeLog
+++ b/gcc/testsuite/ChangeLog
@@ -1,3 +1,8 @@
+2012-11-20  Uros Bizjak  <ubizjak@gmail.com>
+
+	PR target/19398
+	* gcc.target/i386/pr19398.c: New test.
+
 2012-11-20  Martin Jambor  <mjambor@suse.cz>
 
 	PR tree-optimization/55260
diff --git a/gcc/testsuite/gcc.target/i386/pr19398.c b/gcc/testsuite/gcc.target/i386/pr19398.c
new file mode 100644
index 0000000..60931c0
--- /dev/null
+++ b/gcc/testsuite/gcc.target/i386/pr19398.c
@@ -0,0 +1,9 @@
+/* { dg-do compile } */
+/* { dg-options "-Os -msse -mno-sse3 -mfpmath=387" } */
+
+int test (float a)
+{
+  return (a * a);
+}
+
+/* { dg-final { scan-assembler-not "cvttss2si\[^\\n\]*%xmm" } } */