[36/43] i386: Allow MMX vector expanders with TARGET_MMX_WITH_SSE

Message ID 20190209132352.1828-37-hjl.tools@gmail.com
State Superseded
Headers show
Series
  • V2: Emulate MMX intrinsics with SSE
Related show

Commit Message

H.J. Lu Feb. 9, 2019, 1:23 p.m.
PR target/89021
	* config/i386/i386.c (ix86_expand_vector_init_duplicate): Set
	mmx_ok to true if TARGET_MMX_WITH_SSE is true.
	(ix86_expand_vector_init_one_nonzero): Likewise.
	(ix86_expand_vector_init_one_var): Likewise.
	(ix86_expand_vector_init_general): Likewise.
	(ix86_expand_vector_init): Likewise.
	(ix86_expand_vector_set): Likewise.
	(ix86_expand_vector_extract): Likewise.
	* config/i386/mmx.md (*vec_dupv2sf): Changed to
	define_insn_and_split to support SSE emulation.
	(vec_setv2sf): Also allow TARGET_MMX_WITH_SSE.
	(vec_extractv2sf_1 splitter): Likewise.
	(vec_extractv2sfsf): Likewise.
	(vec_setv2si): Likewise.
	(vec_extractv2si_1 splitter): Likewise.
	(vec_extractv2sisi): Likewise.
	(vec_setv4hi): Likewise.
	(vec_extractv4hihi): Likewise.
	(vec_setv8qi): Likewise.
	(vec_extractv8qiqi): Likewise.
	(*vec_extractv2sf_0): Don't allow TARGET_MMX_WITH_SSE.
	(*vec_extractv2sf_1): Likewise.
	(*vec_extractv2si_0): Likewise.
	(*vec_extractv2si_1): Likewise.
	(*vec_extractv2sf_0_sse): New.
	(*vec_extractv2sf_1_sse): Likewise.
	(*vec_extractv2si_0_sse): Likewise.
	(*vec_extractv2si_1_sse): Likewise.
---
 gcc/config/i386/i386.c |   8 +++
 gcc/config/i386/mmx.md | 133 +++++++++++++++++++++++++++++++++--------
 2 files changed, 117 insertions(+), 24 deletions(-)

-- 
2.20.1

Patch

diff --git a/gcc/config/i386/i386.c b/gcc/config/i386/i386.c
index 6e67ac346dd..3770bb882d4 100644
--- a/gcc/config/i386/i386.c
+++ b/gcc/config/i386/i386.c
@@ -42364,6 +42364,7 @@  ix86_expand_vector_init_duplicate (bool mmx_ok, machine_mode mode,
 {
   bool ok;
 
+  mmx_ok |= TARGET_MMX_WITH_SSE;
   switch (mode)
     {
     case E_V2SImode:
@@ -42523,6 +42524,7 @@  ix86_expand_vector_init_one_nonzero (bool mmx_ok, machine_mode mode,
   bool use_vector_set = false;
   rtx (*gen_vec_set_0) (rtx, rtx, rtx) = NULL;
 
+  mmx_ok |= TARGET_MMX_WITH_SSE;
   switch (mode)
     {
     case E_V2DImode:
@@ -42716,6 +42718,7 @@  ix86_expand_vector_init_one_var (bool mmx_ok, machine_mode mode,
   XVECEXP (const_vec, 0, one_var) = CONST0_RTX (GET_MODE_INNER (mode));
   const_vec = gen_rtx_CONST_VECTOR (mode, XVEC (const_vec, 0));
 
+  mmx_ok |= TARGET_MMX_WITH_SSE;
   switch (mode)
     {
     case E_V2DFmode:
@@ -43101,6 +43104,7 @@  ix86_expand_vector_init_general (bool mmx_ok, machine_mode mode,
   machine_mode quarter_mode = VOIDmode;
   int n, i;
 
+  mmx_ok |= TARGET_MMX_WITH_SSE;
   switch (mode)
     {
     case E_V2SFmode:
@@ -43300,6 +43304,8 @@  ix86_expand_vector_init (bool mmx_ok, rtx target, rtx vals)
   int i;
   rtx x;
 
+  mmx_ok |= TARGET_MMX_WITH_SSE;
+
   /* Handle first initialization from vector elts.  */
   if (n_elts != XVECLEN (vals, 0))
     {
@@ -43399,6 +43405,7 @@  ix86_expand_vector_set (bool mmx_ok, rtx target, rtx val, int elt)
   machine_mode mmode = VOIDmode;
   rtx (*gen_blendm) (rtx, rtx, rtx, rtx);
 
+  mmx_ok |= TARGET_MMX_WITH_SSE;
   switch (mode)
     {
     case E_V2SFmode:
@@ -43754,6 +43761,7 @@  ix86_expand_vector_extract (bool mmx_ok, rtx target, rtx vec, int elt)
   bool use_vec_extr = false;
   rtx tmp;
 
+  mmx_ok |= TARGET_MMX_WITH_SSE;
   switch (mode)
     {
     case E_V2SImode:
diff --git a/gcc/config/i386/mmx.md b/gcc/config/i386/mmx.md
index 283661f7887..31db0e8b0c7 100644
--- a/gcc/config/i386/mmx.md
+++ b/gcc/config/i386/mmx.md
@@ -595,14 +595,27 @@ 
    (set_attr "prefix_extra" "1")
    (set_attr "mode" "V2SF")])
 
-(define_insn "*vec_dupv2sf"
-  [(set (match_operand:V2SF 0 "register_operand" "=y")
+(define_insn_and_split "*vec_dupv2sf"
+  [(set (match_operand:V2SF 0 "register_operand" "=y,Yx,Yy")
 	(vec_duplicate:V2SF
-	  (match_operand:SF 1 "register_operand" "0")))]
-  "TARGET_MMX"
-  "punpckldq\t%0, %0"
-  [(set_attr "type" "mmxcvt")
-   (set_attr "mode" "DI")])
+	  (match_operand:SF 1 "register_operand" "0,0,Yy")))]
+  "TARGET_MMX || TARGET_MMX_WITH_SSE"
+  "@
+   punpckldq\t%0, %0
+   #
+   #"
+  "&& reload_completed && TARGET_MMX_WITH_SSE"
+  [(const_int 0)]
+{
+  /* Emulate MMX vec_dupv2sf with SSE vec_dupv4sf.  */
+  rtx op0 = gen_rtx_REG (V4SFmode, REGNO (operands[0]));
+  rtx insn = gen_vec_dupv4sf (op0, operands[1]);
+  emit_insn (insn);
+  DONE;
+}
+  [(set_attr "mmx_isa" "native,x64_noavx,x64_avx")
+   (set_attr "type" "mmxcvt,ssemov,ssemov")
+   (set_attr "mode" "DI,TI,TI")])
 
 (define_insn "*mmx_concatv2sf"
   [(set (match_operand:V2SF 0 "register_operand"     "=y,y")
@@ -620,7 +633,7 @@ 
   [(match_operand:V2SF 0 "register_operand")
    (match_operand:SF 1 "register_operand")
    (match_operand 2 "const_int_operand")]
-  "TARGET_MMX"
+  "TARGET_MMX || TARGET_MMX_WITH_SSE"
 {
   ix86_expand_vector_set (false, operands[0], operands[1],
 			  INTVAL (operands[2]));
@@ -634,7 +647,20 @@ 
 	(vec_select:SF
 	  (match_operand:V2SF 1 "nonimmediate_operand" " xm,x,ym,y,m,m")
 	  (parallel [(const_int 0)])))]
-  "TARGET_MMX && !(MEM_P (operands[0]) && MEM_P (operands[1]))"
+  "TARGET_MMX
+   && !TARGET_MMX_WITH_SSE
+   && !(MEM_P (operands[0]) && MEM_P (operands[1]))"
+  "#"
+  "&& reload_completed"
+  [(set (match_dup 0) (match_dup 1))]
+  "operands[1] = gen_lowpart (SFmode, operands[1]);")
+
+(define_insn_and_split "*vec_extractv2sf_0_sse"
+  [(set (match_operand:SF 0 "nonimmediate_operand"     "=x, m,f,r")
+	(vec_select:SF
+	  (match_operand:V2SF 1 "nonimmediate_operand" " xm,x,m,m")
+	  (parallel [(const_int 0)])))]
+  "TARGET_MMX_WITH_SSE && !(MEM_P (operands[0]) && MEM_P (operands[1]))"
   "#"
   "&& reload_completed"
   [(set (match_dup 0) (match_dup 1))]
@@ -647,7 +673,9 @@ 
 	(vec_select:SF
 	  (match_operand:V2SF 1 "nonimmediate_operand" " 0,x,x,o,o,o,o")
 	  (parallel [(const_int 1)])))]
-  "TARGET_MMX && !(MEM_P (operands[0]) && MEM_P (operands[1]))"
+  "TARGET_MMX
+   && !TARGET_MMX_WITH_SSE
+   && !(MEM_P (operands[0]) && MEM_P (operands[1]))"
   "@
    punpckhdq\t%0, %0
    %vmovshdup\t{%1, %0|%0, %1}
@@ -669,12 +697,33 @@ 
    (set_attr "prefix" "orig,maybe_vex,orig,orig,orig,orig,orig")
    (set_attr "mode" "DI,V4SF,V4SF,SF,SF,SF,SF")])
 
+(define_insn "*vec_extractv2sf_1_sse"
+  [(set (match_operand:SF 0 "nonimmediate_operand"     "=x,x,x,f,r")
+	(vec_select:SF
+	  (match_operand:V2SF 1 "nonimmediate_operand" " x,x,o,o,o")
+	  (parallel [(const_int 1)])))]
+  "TARGET_MMX_WITH_SSE && !(MEM_P (operands[0]) && MEM_P (operands[1]))"
+  "@
+   %vmovshdup\t{%1, %0|%0, %1}
+   shufps\t{$0xe5, %1, %0|%0, %1, 0xe5}
+   #
+   #
+   #"
+  [(set_attr "isa" "sse3,noavx,*,*,*")
+   (set_attr "type" "sse,sseshuf1,ssemov,fmov,imov")
+   (set (attr "length_immediate")
+     (if_then_else (eq_attr "alternative" "1")
+		   (const_string "1")
+		   (const_string "*")))
+   (set_attr "prefix" "maybe_vex,orig,orig,orig,orig")
+   (set_attr "mode" "V4SF,V4SF,SF,SF,SF")])
+
 (define_split
   [(set (match_operand:SF 0 "register_operand")
 	(vec_select:SF
 	  (match_operand:V2SF 1 "memory_operand")
 	  (parallel [(const_int 1)])))]
-  "TARGET_MMX && reload_completed"
+  "(TARGET_MMX || TARGET_MMX_WITH_SSE) && reload_completed"
   [(set (match_dup 0) (match_dup 1))]
   "operands[1] = adjust_address (operands[1], SFmode, 4);")
 
@@ -682,7 +731,7 @@ 
   [(match_operand:SF 0 "register_operand")
    (match_operand:V2SF 1 "register_operand")
    (match_operand 2 "const_int_operand")]
-  "TARGET_MMX"
+  "TARGET_MMX || TARGET_MMX_WITH_SSE"
 {
   ix86_expand_vector_extract (false, operands[0], operands[1],
 			      INTVAL (operands[2]));
@@ -1523,7 +1572,7 @@ 
   [(match_operand:V2SI 0 "register_operand")
    (match_operand:SI 1 "register_operand")
    (match_operand 2 "const_int_operand")]
-  "TARGET_MMX"
+  "TARGET_MMX || TARGET_MMX_WITH_SSE"
 {
   ix86_expand_vector_set (false, operands[0], operands[1],
 			  INTVAL (operands[2]));
@@ -1537,7 +1586,20 @@ 
 	(vec_select:SI
 	  (match_operand:V2SI 1 "nonimmediate_operand" "xm,x,ym,y,m")
 	  (parallel [(const_int 0)])))]
-  "TARGET_MMX && !(MEM_P (operands[0]) && MEM_P (operands[1]))"
+  "TARGET_MMX
+   && !TARGET_MMX_WITH_SSE
+   && !(MEM_P (operands[0]) && MEM_P (operands[1]))"
+  "#"
+  "&& reload_completed"
+  [(set (match_dup 0) (match_dup 1))]
+  "operands[1] = gen_lowpart (SImode, operands[1]);")
+
+(define_insn_and_split "*vec_extractv2si_0_sse"
+  [(set (match_operand:SI 0 "nonimmediate_operand"     "=x,m,r")
+	(vec_select:SI
+	  (match_operand:V2SI 1 "nonimmediate_operand" "xm,x,m")
+	  (parallel [(const_int 0)])))]
+  "TARGET_MMX_WITH_SSE && !(MEM_P (operands[0]) && MEM_P (operands[1]))"
   "#"
   "&& reload_completed"
   [(set (match_dup 0) (match_dup 1))]
@@ -1550,7 +1612,9 @@ 
 	(vec_select:SI
 	  (match_operand:V2SI 1 "nonimmediate_operand" " 0,x,x,o,o,o")
 	  (parallel [(const_int 1)])))]
-  "TARGET_MMX && !(MEM_P (operands[0]) && MEM_P (operands[1]))"
+  "TARGET_MMX
+   && !TARGET_MMX_WITH_SSE
+   && !(MEM_P (operands[0]) && MEM_P (operands[1]))"
   "@
    punpckhdq\t%0, %0
    %vpshufd\t{$0xe5, %1, %0|%0, %1, 0xe5}
@@ -1567,22 +1631,43 @@ 
    (set_attr "prefix" "orig,maybe_vex,orig,orig,orig,orig")
    (set_attr "mode" "DI,TI,V4SF,SI,SI,SI")])
 
+(define_insn "*vec_extractv2si_1_sse"
+  [(set (match_operand:SI 0 "nonimmediate_operand"     "=x,x,x,r")
+	(vec_select:SI
+	  (match_operand:V2SI 1 "nonimmediate_operand" " x,x,o,o")
+	  (parallel [(const_int 1)])))]
+  "TARGET_MMX_WITH_SSE
+   && !(MEM_P (operands[0]) && MEM_P (operands[1]))"
+  "@
+   %vpshufd\t{$0xe5, %1, %0|%0, %1, 0xe5}
+   shufps\t{$0xe5, %1, %0|%0, %1, 0xe5}
+   #
+   #"
+  [(set_attr "isa" "sse2,noavx,*,*")
+   (set_attr "type" "sseshuf1,sseshuf1,ssemov,imov")
+   (set (attr "length_immediate")
+     (if_then_else (eq_attr "alternative" "0,1")
+		   (const_string "1")
+		   (const_string "*")))
+   (set_attr "prefix" "maybe_vex,orig,orig,orig")
+   (set_attr "mode" "TI,V4SF,SI,SI")])
+
 (define_split
   [(set (match_operand:SI 0 "register_operand")
 	(vec_select:SI
 	  (match_operand:V2SI 1 "memory_operand")
 	  (parallel [(const_int 1)])))]
-  "TARGET_MMX && reload_completed"
+  "(TARGET_MMX || TARGET_MMX_WITH_SSE) && reload_completed"
   [(set (match_dup 0) (match_dup 1))]
   "operands[1] = adjust_address (operands[1], SImode, 4);")
 
 (define_insn_and_split "*vec_extractv2si_zext_mem"
-  [(set (match_operand:DI 0 "register_operand" "=y,x,r")
+  [(set (match_operand:DI 0 "register_operand" "=x,r")
 	(zero_extend:DI
 	  (vec_select:SI
-	    (match_operand:V2SI 1 "memory_operand" "o,o,o")
+	    (match_operand:V2SI 1 "memory_operand" "o,o")
 	    (parallel [(match_operand:SI 2 "const_0_to_1_operand")]))))]
-  "TARGET_64BIT && TARGET_MMX"
+  "TARGET_64BIT"
   "#"
   "&& reload_completed"
   [(set (match_dup 0) (zero_extend:DI (match_dup 1)))]
@@ -1594,7 +1679,7 @@ 
   [(match_operand:SI 0 "register_operand")
    (match_operand:V2SI 1 "register_operand")
    (match_operand 2 "const_int_operand")]
-  "TARGET_MMX"
+  "TARGET_MMX || TARGET_MMX_WITH_SSE"
 {
   ix86_expand_vector_extract (false, operands[0], operands[1],
 			      INTVAL (operands[2]));
@@ -1614,7 +1699,7 @@ 
   [(match_operand:V4HI 0 "register_operand")
    (match_operand:HI 1 "register_operand")
    (match_operand 2 "const_int_operand")]
-  "TARGET_MMX"
+  "TARGET_MMX || TARGET_MMX_WITH_SSE"
 {
   ix86_expand_vector_set (false, operands[0], operands[1],
 			  INTVAL (operands[2]));
@@ -1625,7 +1710,7 @@ 
   [(match_operand:HI 0 "register_operand")
    (match_operand:V4HI 1 "register_operand")
    (match_operand 2 "const_int_operand")]
-  "TARGET_MMX"
+  "TARGET_MMX || TARGET_MMX_WITH_SSE"
 {
   ix86_expand_vector_extract (false, operands[0], operands[1],
 			      INTVAL (operands[2]));
@@ -1645,7 +1730,7 @@ 
   [(match_operand:V8QI 0 "register_operand")
    (match_operand:QI 1 "register_operand")
    (match_operand 2 "const_int_operand")]
-  "TARGET_MMX"
+  "TARGET_MMX || TARGET_MMX_WITH_SSE"
 {
   ix86_expand_vector_set (false, operands[0], operands[1],
 			  INTVAL (operands[2]));
@@ -1656,7 +1741,7 @@ 
   [(match_operand:QI 0 "register_operand")
    (match_operand:V8QI 1 "register_operand")
    (match_operand 2 "const_int_operand")]
-  "TARGET_MMX"
+  "TARGET_MMX || TARGET_MMX_WITH_SSE"
 {
   ix86_expand_vector_extract (false, operands[0], operands[1],
 			      INTVAL (operands[2]));