i386.md (*floatuns<SWI48:mode><MODEF:mode>2_avx512): New insn pattern.

	* config/i386/i386.md (*floatuns<SWI48:mode><MODEF:mode>2_avx512):
	New insn pattern.
	(floatunssi<mode>2): Also enable for AVX512F and TARGET_SSE_MATH.
	Rewrite expander pattern.  Emit gen_floatunssi<mode>2_i387_with_xmm
	for non-SSE modes.
	(floatunsdisf2): Rewrite expander pattern.  Hanlde TARGET_AVX512F.
	(floatunsdidf2): Ditto.

	* config/i386/i386.md (fixuns_trunc<mode>di2): New insn pattern.
	(fixuns_trunc<mode>si2_avx512f): Ditto.
	(*fixuns_trunc<mode>si2_avx512f_zext): Ditto.
	(fixuns_trunc<mode>si2): Also enable for AVX512F and TARGET_SSE_MATH.
	Emit fixuns_trunc<mode>si2_avx512f for AVX512F targets.

testsuite/ChangeLog:

	* gcc.target/i386/cvt-2.c: New test.

	* gcc.target/i386/cvt-3.c: New test.

From-SVN: r260614

diff --git a/gcc/ChangeLog b/gcc/ChangeLog
index 270c979e84ed3cf29f28aa4f3c6660de55b5933f..e9268d64c78d0be80d27901344ad1c2b05318e59 100644 (file)
--- a/gcc/ChangeLog
+++ b/gcc/ChangeLog
@@ -1,3 +1,21 @@
+2018-05-23  Uros Bizjak  <ubizjak@gmail.com>
+
+       * config/i386/i386.md (*floatuns<SWI48:mode><MODEF:mode>2_avx512):
+       New insn pattern.
+       (floatunssi<mode>2): Also enable for AVX512F and TARGET_SSE_MATH.
+       Rewrite expander pattern.  Emit gen_floatunssi<mode>2_i387_with_xmm
+       for non-SSE modes.
+       (floatunsdisf2): Rewrite expander pattern.  Hanlde TARGET_AVX512F.
+       (floatunsdidf2): Ditto.
+
+2018-05-23  Uros Bizjak  <ubizjak@gmail.com>
+
+       * config/i386/i386.md (fixuns_trunc<mode>di2): New insn pattern.
+       (fixuns_trunc<mode>si2_avx512f): Ditto.
+       (*fixuns_trunc<mode>si2_avx512f_zext): Ditto.
+       (fixuns_trunc<mode>si2): Also enable for AVX512F and TARGET_SSE_MATH.
+       Emit fixuns_trunc<mode>si2_avx512f for AVX512F targets.
+
 2018-05-23  Alexander Monakov  <amonakov@ispras.ru>
 
        PR rtl-optimization/79985

diff --git a/gcc/config/i386/i386.md b/gcc/config/i386/i386.md
index cc993b3aad8f806142d5da48a6e775f8e2e49a4b..bde3c34ee63d14cb328a8ab5e352dac79e4350aa 100644 (file)
--- a/gcc/config/i386/i386.md
+++ b/gcc/config/i386/i386.md
@@ -5017,6 +5017,18 @@
    }
 })
 
+;; Unsigned conversion to DImode
+
+(define_insn "fixuns_trunc<mode>di2"
+  [(set (match_operand:DI 0 "register_operand" "=r")
+       (unsigned_fix:DI
+         (match_operand:MODEF 1 "nonimmediate_operand" "vm")))]
+  "TARGET_64BIT && TARGET_AVX512F && TARGET_SSE_MATH"
+  "vcvtt<ssemodesuffix>2usi\t{%1, %0|%0, %1}"
+  [(set_attr "type" "sseicvt")
+   (set_attr "prefix" "evex")
+   (set_attr "mode" "DI")])
+
 ;; Unsigned conversion to SImode.
 
 (define_expand "fixuns_trunc<mode>si2"
@@ -5027,13 +5039,19 @@
      (use (match_dup 2))
      (clobber (match_scratch:<ssevecmode> 3))
      (clobber (match_scratch:<ssevecmode> 4))])]
-  "!TARGET_64BIT && TARGET_SSE2 && TARGET_SSE_MATH"
+  "(!TARGET_64BIT || TARGET_AVX512F) && TARGET_SSE2 && TARGET_SSE_MATH"
 {
   machine_mode mode = <MODE>mode;
   machine_mode vecmode = <ssevecmode>mode;
   REAL_VALUE_TYPE TWO31r;
   rtx two31;
 
+  if (TARGET_AVX512F)
+    {
+      emit_insn (gen_fixuns_trunc<mode>si2_avx512f (operands[0], operands[1]));
+      DONE;
+    }
+
   if (optimize_insn_for_size_p ())
     FAIL;
 
@@ -5043,6 +5061,27 @@
   operands[2] = force_reg (vecmode, two31);
 })
 
+(define_insn "fixuns_trunc<mode>si2_avx512f"
+  [(set (match_operand:SI 0 "register_operand" "=r")
+       (unsigned_fix:SI
+         (match_operand:MODEF 1 "nonimmediate_operand" "vm")))]
+  "TARGET_AVX512F && TARGET_SSE_MATH"
+  "vcvtt<ssemodesuffix>2usi\t{%1, %0|%0, %1}"
+  [(set_attr "type" "sseicvt")
+   (set_attr "prefix" "evex")
+   (set_attr "mode" "SI")])
+
+(define_insn "*fixuns_trunc<mode>si2_avx512f_zext"
+  [(set (match_operand:DI 0 "register_operand" "=r")
+       (zero_extend:DI
+         (unsigned_fix:SI
+           (match_operand:MODEF 1 "nonimmediate_operand" "vm"))))]
+  "TARGET_64BIT && TARGET_AVX512F"
+  "vcvtt<ssemodesuffix>2usi\t{%1, %k0|%k0, %1}"
+  [(set_attr "type" "sseicvt")
+   (set_attr "prefix" "evex")
+   (set_attr "mode" "SI")])
+
 (define_insn_and_split "*fixuns_trunc<mode>_1"
   [(set (match_operand:SI 0 "register_operand" "=&x,&x")
        (unsigned_fix:SI
@@ -5615,16 +5654,26 @@
   DONE;
 })
 
+(define_insn "*floatuns<SWI48:mode><MODEF:mode>2_avx512"
+  [(set (match_operand:MODEF 0 "register_operand" "=v")
+       (unsigned_float:MODEF
+         (match_operand:SWI48 1 "nonimmediate_operand" "rm")))]
+  "TARGET_AVX512F && TARGET_SSE_MATH"
+  "vcvtusi2<MODEF:ssemodesuffix><SWI48:rex64suffix>\t{%1, %0, %0|%0, %0, %1}"
+  [(set_attr "type" "sseicvt")
+   (set_attr "prefix" "evex")
+   (set_attr "mode" "<MODEF:MODE>")])
+
 ;; Avoid store forwarding (partial memory) stall penalty by extending
 ;; SImode value to DImode through XMM register instead of pushing two
 ;; SImode values to stack. Also note that fild loads from memory only.
 
-(define_insn_and_split "*floatunssi<mode>2_i387_with_xmm"
+(define_insn_and_split "floatunssi<mode>2_i387_with_xmm"
   [(set (match_operand:X87MODEF 0 "register_operand" "=f")
        (unsigned_float:X87MODEF
          (match_operand:SI 1 "nonimmediate_operand" "rm")))
-   (clobber (match_scratch:DI 3 "=x"))
-   (clobber (match_operand:DI 2 "memory_operand" "=m"))]
+   (clobber (match_operand:DI 2 "memory_operand" "=m"))
+   (clobber (match_scratch:DI 3 "=x"))]
   "!TARGET_64BIT
    && TARGET_80387 && X87_ENABLE_FLOAT (<X87MODEF:MODE>mode, DImode)
    && TARGET_SSE2 && TARGET_INTER_UNIT_MOVES_TO_VEC"
@@ -5639,43 +5688,59 @@
    (set_attr "mode" "<MODE>")])
 
 (define_expand "floatunssi<mode>2"
-  [(parallel
-     [(set (match_operand:X87MODEF 0 "register_operand")
-          (unsigned_float:X87MODEF
-            (match_operand:SI 1 "nonimmediate_operand")))
-      (clobber (match_scratch:DI 3))
-      (clobber (match_dup 2))])]
-  "!TARGET_64BIT
-   && ((TARGET_80387 && X87_ENABLE_FLOAT (<X87MODEF:MODE>mode, DImode)
-       && TARGET_SSE2 && TARGET_INTER_UNIT_MOVES_TO_VEC)
-       || (SSE_FLOAT_MODE_P (<MODE>mode) && TARGET_SSE_MATH))"
+  [(set (match_operand:X87MODEF 0 "register_operand")
+       (unsigned_float:X87MODEF
+         (match_operand:SI 1 "nonimmediate_operand")))]
+  "(!TARGET_64BIT
+    && TARGET_80387 && X87_ENABLE_FLOAT (<X87MODEF:MODE>mode, DImode)
+    && TARGET_SSE2 && TARGET_INTER_UNIT_MOVES_TO_VEC)
+   || ((!TARGET_64BIT || TARGET_AVX512F)
+       && SSE_FLOAT_MODE_P (<MODE>mode) && TARGET_SSE_MATH)"
 {
-  if (SSE_FLOAT_MODE_P (<MODE>mode) && TARGET_SSE_MATH)
+  if (!(SSE_FLOAT_MODE_P (<MODE>mode) && TARGET_SSE_MATH))
+    {
+      emit_insn (gen_floatunssi<mode>2_i387_with_xmm
+                 (operands[0], operands[1],
+                  assign_386_stack_local (DImode, SLOT_TEMP)));
+      DONE;
+    }
+  if (!TARGET_AVX512F)
     {
       ix86_expand_convert_uns_si<mode>_sse (operands[0], operands[1]);
       DONE;
     }
-  else
-    operands[2] = assign_386_stack_local (DImode, SLOT_TEMP);
 })
 
 (define_expand "floatunsdisf2"
-  [(use (match_operand:SF 0 "register_operand"))
-   (use (match_operand:DI 1 "nonimmediate_operand"))]
+  [(set (match_operand:SF 0 "register_operand")
+       (unsigned_float:SF
+         (match_operand:DI 1 "nonimmediate_operand")))]
   "TARGET_64BIT && TARGET_SSE && TARGET_SSE_MATH"
-  "x86_emit_floatuns (operands); DONE;")
+{
+  if (!TARGET_AVX512F)
+    {
+      x86_emit_floatuns (operands);
+      DONE;
+    }
+})
 
 (define_expand "floatunsdidf2"
-  [(use (match_operand:DF 0 "register_operand"))
-   (use (match_operand:DI 1 "nonimmediate_operand"))]
-  "(TARGET_64BIT || TARGET_KEEPS_VECTOR_ALIGNED_STACK)
+  [(set (match_operand:DF 0 "register_operand")
+       (unsigned_float:DF
+         (match_operand:DI 1 "nonimmediate_operand")))]
+  "(TARGET_KEEPS_VECTOR_ALIGNED_STACK || TARGET_AVX512F)
    && TARGET_SSE2 && TARGET_SSE_MATH"
 {
-  if (TARGET_64BIT)
-    x86_emit_floatuns (operands);
-  else
-    ix86_expand_convert_uns_didf_sse (operands[0], operands[1]);
-  DONE;
+  if (!TARGET_64BIT)
+    {
+      ix86_expand_convert_uns_didf_sse (operands[0], operands[1]);
+      DONE;
+    }
+  if (!TARGET_AVX512F)
+    {
+      x86_emit_floatuns (operands);
+      DONE;
+    }
 })
 \f
 ;; Load effective address instructions

diff --git a/gcc/testsuite/ChangeLog b/gcc/testsuite/ChangeLog
index 5b16f44d8ff9f1d02b900947d9061e4b57342b02..f45de0973056d3f2b1a72d883ac75143dc3184cc 100644 (file)
--- a/gcc/testsuite/ChangeLog
+++ b/gcc/testsuite/ChangeLog
@@ -1,3 +1,11 @@
+2018-05-23  Uros Bizjak  <ubizjak@gmail.com>
+
+       * gcc.target/i386/cvt-3.c: New test.
+
+2018-05-23  Uros Bizjak  <ubizjak@gmail.com>
+
+       * gcc.target/i386/cvt-2.c: New test.
+
 2018-05-23  Alexander Monakov  <amonakov@ispras.ru>
 
        * gcc.dg/pr79985.c: New testcase.

diff --git a/gcc/testsuite/gcc.target/i386/cvt-2.c b/gcc/testsuite/gcc.target/i386/cvt-2.c
new file mode 100644 (file)
index 0000000..aa0fd86
--- /dev/null
+++ b/gcc/testsuite/gcc.target/i386/cvt-2.c
@@ -0,0 +1,15 @@
+/* { dg-do compile } */
+/* { dg-options "-O2 -mavx512f -mfpmath=sse" } */
+
+unsigned int f2ui (float x) { return x; }
+unsigned int d2ui (double x) { return x; }
+
+#ifdef __x86_64__
+unsigned long f2ul (float x) { return x; }
+unsigned long d2ul (double x) { return x; }
+#endif
+  
+/* { dg-final { scan-assembler-times "vcvttss2usi" 1 { target ia32 } } } */
+/* { dg-final { scan-assembler-times "vcvttsd2usi" 1 { target ia32 } } } */
+/* { dg-final { scan-assembler-times "vcvttss2usi" 2 { target { ! ia32 } } } } */
+/* { dg-final { scan-assembler-times "vcvttsd2usi" 2 { target { ! ia32 } } } } */

diff --git a/gcc/testsuite/gcc.target/i386/cvt-3.c b/gcc/testsuite/gcc.target/i386/cvt-3.c
new file mode 100644 (file)
index 0000000..132ea6e
--- /dev/null
+++ b/gcc/testsuite/gcc.target/i386/cvt-3.c
@@ -0,0 +1,15 @@
+/* { dg-do compile } */
+/* { dg-options "-O2 -mavx512f -mfpmath=sse" } */
+
+float ui2f (unsigned int x) { return x; }
+double ui2d (unsigned int x) { return x; }
+
+#ifdef __x86_64__
+float ul2f (unsigned long x) { return x; }
+double ul2d (unsigned long x) { return x; }
+#endif
+  
+/* { dg-final { scan-assembler-times "vcvtusi2ss" 1 { target ia32 } } } */
+/* { dg-final { scan-assembler-times "vcvtusi2sd" 1 { target ia32 } } } */
+/* { dg-final { scan-assembler-times "vcvtusi2ss" 2 { target { ! ia32 } } } } */
+/* { dg-final { scan-assembler-times "vcvtusi2sd" 2 { target { ! ia32 } } } } */