🚀 go-pugleaf

Return-Path: <MAILER-DAEMON@ai.mit.edu>
Date: Wed, 17 Jan 90 11:03:52 EST
From: MAILER-DAEMON@ai.mit.edu (Mail Delivery Subsystem)
Subject: Returned mail: User unknown
To: owner-bug-gcc

   ----- Transcript of session follows -----
Connected to tut.cis.ohio-state.edu:
>>> RCPT To:<bug-gcc-gate@tut.cis.ohio-state.edu>
<<< 550 <bug-gcc-gate@tut.cis.ohio-state.edu>... User unknown
550 bug-gcc-gate@tut.cis.ohio-state.edu... User unknown

   ----- Unsent message follows -----
Return-Path: <@dg-rtp.rtp.dg.com:wood@DG-RTP.DG.COM>
Received: from aeneas.MIT.EDU by life.ai.mit.edu (4.0/AI-4.10) id AA13567; Wed, 17 Jan 90 11:03:52 EST
Received: from relay.cs.net by aeneas.MIT.EDU (5.61/4.7) id AA13705; Wed, 17 Jan 90 11:03:33 -0500
Received: from dg-rtp.rtp.dg.com by RELAY.CS.NET id aa18148; 17 Jan 90 10:02 EST
Received: from gen-rtx.rtp.dg.com (gen-rtx) by dg-rtp.dg.com (4.12/4.7)
	id AA07605; Wed, 17 Jan 90 11:05:51 est via SMTP
Received: by gen-rtx.rtp.dg.com (4.20/4.7)
	id AA18914; Wed, 17 Jan 90 11:01:55 est
From: Tom Wood <wood@dg-rtp.dg.com>
Message-Id: <9001171601.AA18914@gen-rtx.rtp.dg.com>
Organization: Data General, RTP, NC
Reply-To: wood@dg-rtp.dg.com
Subject: Problem with pp-number in GCC 1.36.92
To: bug-gcc@prep.ai.mit.edu
Date: Wed, 17 Jan 90 11:01:53 EST
X-Mailer: ELM [version 2.2 PL10]

GCC versions 1.35 through 1.36.92 fail the enclosed preprocessor test.

#	This is a shell archive.
#	Remove everything above and including the cut line.
#	Then run the rest of the file through sh.
-----cut here-----cut here-----cut here-----cut here-----
#!/bin/sh
# shar:	Shell Archiver
#	Run the following text with /bin/sh to create:
#	pp-number.c		Test program
#	pp-number-bad.i		Incorrect output
#	pp-number-ok.i		Correct output
# This archive created: Wed Jan 17 11:00:20 EST 1990
echo x - pp-number.c
cat << \SHAR_EOF > pp-number.c
#define N 4
#define EXP 3e-N
double v = 1.2+EXP;
SHAR_EOF
echo x - pp-number-bad.i
cat << \SHAR_EOF > pp-number-bad.i
# 1 "pp-number.c"


double v = 1.2+3e-N ;
SHAR_EOF
echo x - pp-number-ok.i
cat << \SHAR_EOF > pp-number-ok.i
# 1 "pp-number.c"


double v = 1.2+3e-4  ;
SHAR_EOF
#	End of shell archive
exit 0

*** gcc-1.35.00/cccp.c	Wed Aug 16 14:12:06 1989
--- gcc-1.35/cccp.c	Mon Jan 15 10:08:49 1990
***************
*** 1552,1562 ****
  	    break;
  	  }
  	  *obp++ = c;
! 	  /* A sign can be part of a preprocessing number
! 	     if it follows an e.  */
  	  if (c == 'e' || c == 'E') {
! 	    if (ibp < limit && (*ibp == '+' || *ibp == '-'))
  	      *obp++ = *ibp++;
  	  }
  	}
  	break;
--- 1644,1657 ----
  	    break;
  	  }
  	  *obp++ = c;
! 	  /* A sign can be part of a preprocessing number if it follows
! 	     an e.  However, ANSI and PCC treat this as the termination
! 	     of the pp-number.  The exponent token may be rescanned. */
  	  if (c == 'e' || c == 'E') {
! 	    if (ibp < limit && (*ibp == '+' || *ibp == '-')) {
  	      *obp++ = *ibp++;
+ 	      break;
+ 	    }
  	  }
  	}
  	break;
---
			Tom Wood	(919) 248-6067
			Data General, Research Triangle Park, NC
			{the known world}!rti!xyzzy!wood, wood@dg-rtp.dg.com