From: Andrew Stubbs <ams@codesourcery.com>
Date: Fri, 28 Oct 2022 12:09:20 +0000 (+0100)
Subject: amdgcn: add fmin/fmax patterns
X-Git-Tag: basepoints/gcc-14~3632
X-Git-Url: http://git.ipfire.org/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=10aa0356118f44e5f4d720a2a4c731b173baa298;p=thirdparty%2Fgcc.git

amdgcn: add fmin/fmax patterns

Add fmin/fmax for scalar, vector, and reductions.  The smin/smax patterns are
already using the IEEE compliant hardware instructions anyway, so we can just
expand to use those insns.

gcc/ChangeLog:

	* config/gcn/gcn-valu.md (fminmaxop): New iterator.
	(<fexpander><mode>3): New define_expand.
	(<fexpander><mode>3<exec>): Likewise.
	(reduc_<fexpander>_scal_<mode>): Likewise.
	* config/gcn/gcn.md (fexpander): New attribute.
---

diff --git a/gcc/config/gcn/gcn-valu.md b/gcc/config/gcn/gcn-valu.md
index 6274d2e92280..3b619512e13f 100644
--- a/gcc/config/gcn/gcn-valu.md
+++ b/gcc/config/gcn/gcn-valu.md
@@ -2466,6 +2466,23 @@
   [(set_attr "type" "vop2")
    (set_attr "length" "8,8")])
 
+(define_code_iterator fminmaxop [smin smax])
+(define_expand "<fexpander><mode>3"
+  [(set (match_operand:FP 0 "gcn_valu_dst_operand")
+	(fminmaxop:FP
+	  (match_operand:FP 1 "gcn_valu_src0_operand")
+	  (match_operand:FP 2 "gcn_valu_src1_operand")))]
+  ""
+  {})
+
+(define_expand "<fexpander><mode>3<exec>"
+  [(set (match_operand:V_FP 0 "gcn_valu_dst_operand")
+	(fminmaxop:V_FP
+	  (match_operand:V_FP 1 "gcn_valu_src0_operand")
+	  (match_operand:V_FP 2 "gcn_valu_src1_operand")))]
+  ""
+  {})
+
 ;; }}}
 ;; {{{ FP unops
 
@@ -3522,6 +3539,17 @@
     DONE;
   })
 
+(define_expand "reduc_<fexpander>_scal_<mode>"
+  [(match_operand:<SCALAR_MODE> 0 "register_operand")
+   (fminmaxop:V_FP
+     (match_operand:V_FP 1 "register_operand"))]
+  ""
+  {
+    /* fmin/fmax are identical to smin/smax.  */
+    emit_insn (gen_reduc_<expander>_scal_<mode> (operands[0], operands[1]));
+    DONE;
+  })
+
 ;; Warning: This "-ffast-math" implementation converts in-order reductions
 ;;          into associative reductions. It's also used where OpenMP or
 ;;          OpenACC paralellization has already broken the in-order semantics.
diff --git a/gcc/config/gcn/gcn.md b/gcc/config/gcn/gcn.md
index 6c1a438f9d1a..987b76396cc1 100644
--- a/gcc/config/gcn/gcn.md
+++ b/gcc/config/gcn/gcn.md
@@ -372,6 +372,10 @@
    (sign_extend "extend")
    (zero_extend "zero_extend")])
 
+(define_code_attr fexpander
+  [(smin "fmin")
+   (smax "fmax")])
+
 ;; }}}
 ;; {{{ Miscellaneous instructions