cross compilation and floats (again)

Article View: gnu.gcc.bug

Article #639

#639

From: Peter R. Ham
Date: Wed, 27 Sep 1989 00:41

26 lines
1427 bytes

It seems to me that none of the routines that
build immediate floating point values should
ever be called when doing cross compilation
except when REAL_ARITHMETIC is defined.
There are some calls to immed_real_const() and
related routines that should be ifdef'd out.
For instance, in gcc-1.35;
*** expr.c	Fri Sep 22 17:35:48 1989
--- ../dist/expr.c	Fri Sep 22 17:17:17 1989
***************
*** 2209,2218 ****
      case REAL_CST:
        /* If optimized, generate immediate CONST_DOUBLE
  	 which will be turned into memory by reload if necessary.  */
- #if !defined (REAL_IS_NOT_DOUBLE) || defined (REAL_ARITHMETIC)
        if (!cse_not_expected)
  	return immed_real_const (exp);
- #endif /* not REAL_IS_NOT_DOUBLE, or REAL_ARITHMETIC */
      case COMPLEX_CST:
      case STRING_CST:
        if (! TREE_CST_RTL (exp))
--- 2209,2216 ----
--
Peter Ham			PO Box 3430	(h)(415) 324-9645
MS Computer Science Student	Stanford, CA	ham@polya.stanford.edu
Stanford University 		94309		(o)(415) 723-2067

Technical Details

Message-ID: <HAM.89Sep26174131@Neon.Stanford.EDU>

Path: arc1-ca.pugleaf.net!utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!shelby!polya!Polya.Stanford.EDU!ham

Show Full Headers

Complete Article Headers

Path: arc1-ca.pugleaf.net!utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!shelby!polya!Polya.Stanford.EDU!ham
From: ham@Neon.Stanford.EDU (Peter R. Ham)
Newsgroups: gnu.gcc.bug
Subject: cross compilation and floats (again)
Message-ID: <HAM.89Sep26174131@Neon.Stanford.EDU>
Date: 27 Sep 89 00:41:31 GMT
Sender: USENET News System <news@Polya.Stanford.EDU>
Distribution: gnu
Organization: Stanford University
Lines: 26