From Philip.Blundell@pobox.com Sat Oct 2 06:34:05 1999 Date: Thu, 16 Sep 1999 20:11:36 +0100 From: Philip Blundell To: Francois Desloges Cc: linux-arm@vger.rutgers.edu Subject: Re: GCC 2.95 >Now, what about that extra CVS patch? Where is this CVS thing and how do I >apply the patch ? I've appended the patch in question below. See the GCC web site if you are interested in CVS access. >>2.9.5 is not 100% compatible with the older versions. Normal programs >>shouldn't care but gcc and glibc need to know the difference. >> >>I hope it will be fixed in glibc before 2.1.2 is actually released, though >>I can't make any promise. This was indeed fixed in glibc 2.1.2. >>As a side note, I am about to change the mainline egcs sources so that >>a target name of `arm-*-linux-gnu' implies binutils 2.9.5, and a target >>name of `arm-*-linux-gnuoldld' implies an older version. This is true in the GCC mainline now, though config.guess has not yet been updated. It is not in the 2.95 branch but when I get around to making a new 2.95.x patch (hopefully within a couple of days) it will be in there. p. Thu Sep 2 00:06:43 1999 Jeffrey A Law (law@cygnus.com) * fold-const.c (fold_range_test): Do not try to fold the range test if the rhs or lhs has side effects. Index: fold-const.c =================================================================== RCS file: /egcs/carton/cvsfiles/egcs/gcc/fold-const.c,v retrieving revision 1.69 retrieving revision 1.70 diff -u -p -r1.69 -r1.70 --- fold-const.c 1999/08/27 07:37:16 1.69 +++ fold-const.c 1999/09/02 09:37:20 1.70 @@ -3560,6 +3560,10 @@ fold_range_test (exp) tree rhs = make_range (TREE_OPERAND (exp, 1), &in1_p, &low1, &high1); tree tem; + /* Fail if anything is volatile. */ + if (TREE_SIDE_EFFECTS (lhs) || TREE_SIDE_EFFECTS (rhs)) + return 0; + /* If this is an OR operation, invert both sides; we will invert again at the end. */ if (or_op) unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu