Re: movstrti use or loose

Re: movstrti use or loose

[Jeffrey A Law]

  In message <199801182204.XAA01982@scar.axis.se>you write:
  > Sun Jan 18 21:23:15 1998  Hans-Peter Nilsson  <hp@axis.se>
  > 
  > 	* expr.c (emit_push_insn): Add code to use movstrti if present.
Thanks.  I've installed this patch into egcs.

jeff

movstrti use or loose

[Hans-Peter Nilsson]

 The pattern "movstrti" is mentioned in *one* place in expr.c (at MOVE_RATIO
in expr.c; c:a line 234), but not defined in any any port in the distribution,
and not explicitly supported by other code, as the explicit
"HAVE_movstr.."-tests in expr.c (but may be use implicitly via movstr_optab).
 So I think support for movstrti as for movstrdi should be added, or movstrti
not be mentioned at all, but I think it's better to include it than exclude it.
 Sent to gcc2 as well.

Sun Jan 18 21:23:15 1998  Hans-Peter Nilsson  <hp@axis.se>

	* expr.c (emit_push_insn): Add code to use movstrti if present.

*** expr.c.original	Sun Jan 18 21:16:49 1998
--- expr.c	Sun Jan 18 21:17:27 1998
*************** emit_push_insn (x, mode, type, size, ali
*** 2704,2709 ****
--- 2704,2721 ----
		}
	    }
  #endif
+ #ifdef HAVE_movstrti
+	  if (HAVE_movstrti)
+	    {
+	      rtx pat = gen_movstrti (gen_rtx_MEM (BLKmode, temp),
+				      xinner, size, GEN_INT (align));
+	      if (pat != 0)
+		{
+		  emit_insn (pat);
+		  goto ret;
+		}
+	    }
+ #endif
  
  #ifndef ACCUMULATE_OUTGOING_ARGS
	  /* If the source is referenced relative to the stack pointer,

Best regards,
-- 
Hans-Peter Nilsson, Axis Communications AB, S - 223 70 LUND, SWEDEN
Hans-Peter.Nilsson@axis.se | http://www.axis.se/ | Tel +46 46191867,191800
Fax +46 46136130 | RFC 1855 compliance implemented; report loss of brain.