This is the mail archive of the
gcc-regression@gcc.gnu.org
mailing list for the GCC project.
2 GCC regressions, 2 new, with your patch on 2001-07-23T09:42:34Z.
- To: gcc-regression at gcc dot gnu dot org
- Subject: 2 GCC regressions, 2 new, with your patch on 2001-07-23T09:42:34Z.
- From: "GCC regression checker" <regress at maat dot cygnus dot com>
- Date: Mon, 23 Jul 2001 11:48:59 +0000
- CC: rth at redhat dot com
- Reply-To: gcc-regression at gcc dot gnu dot org
With your recent patch, GCC has some regression test failures, which
used to pass. There are 2 new failures, and 0
failures that existed before and after that patch; 0 failures
have been fixed.
The new failures are:
powerpc-eabisim gcc.sum gcc.c-torture/compile/921013-1.c,
powerpc-eabisim gcc.sum gcc.c-torture/unsorted/storecc.c,
For more information, see <http://www.cygnus.com/~geoffk/gcc-regression/>.
ChangeLog entries since last run on 2001-07-23T07:00:46Z:
--- /maat/heart/tbox/changelog_mail/gcc/gcc/ChangeLog Mon Jul 23 00:01:01 2001
+++ gcc/gcc/ChangeLog Mon Jul 23 02:42:49 2001
@@ -1,5 +1,21 @@
2001-07-22 Richard Henderson <rth@redhat.com>
+ * flow.c: Grammar check and clarify a lot of comments.
+ (try_simplify_condjump): Rename variables to be clearer.
+ (try_forward_edges): Skip complex and fallthru edges.
+ Rearrange tests to avoid duplicate checks.
+ (flow_find_cross_jump): Likewise.
+ (outgoing_edges_match): Allow match if neither branch has
+ probability data. Loosen probability match to 5%.
+ (try_crossjump_to_edge): Hoist repeated indirection into
+ local variables.
+ (try_crossjump_bb): Don't check complex edges. Eliminate
+ redundant crossjump tests.
+ (try_optimize_cfg): Fix use of bool. Reorganize cheaper
+ checks before more expensive checks.
+
+2001-07-22 Richard Henderson <rth@redhat.com>
+
* fold-const.c (fold): Test vs FLOAT_TYPE_P instead of
INTEGRAL_TYPE_P when folding comparisons with operand_equal_p
arguments.
ChangeLog entries between 2001-07-23T07:00:46Z and the last time everything
passed, on 2001-07-23T04:00:03Z:
--- /maat/heart/tbox/changelog_pass/gcc/gcc/ChangeLog Sun Jul 22 16:06:37 2001
+++ gcc/gcc/ChangeLog Mon Jul 23 00:01:01 2001
@@ -1,5 +1,22 @@
2001-07-22 Richard Henderson <rth@redhat.com>
+ * fold-const.c (fold): Test vs FLOAT_TYPE_P instead of
+ INTEGRAL_TYPE_P when folding comparisons with operand_equal_p
+ arguments.
+
+2001-07-22 Richard Henderson <rth@redhat.com>
+
+ * m68k.c (const_uint32_operand): Abort if mode is <= 32 bits.
+ (const_sint32_operand): Likewise.
+ * m68k.md (anon mulsi pattern): Use const_int_operand not
+ const_sint32_operand.
+ (umulsi3_highpart): Zero extend a constant input.
+ (smulsi3_highpart): Don't bother checking SImode constant.
+ (const_umulsi3_highpart): Give op3 DImode.
+ (const_smulsi3_highpart): Likewise.
+
+2001-07-22 Richard Henderson <rth@redhat.com>
+
* flow.c (split_block): Make sure bb_note is included in the
new block when splitting before a label.
--- /maat/heart/tbox/changelog_pass/gcc/gcc/cp/ChangeLog Sat Jul 21 05:25:35 2001
+++ gcc/gcc/cp/ChangeLog Mon Jul 23 00:01:03 2001
@@ -1,3 +1,8 @@
+2001-07-23 Graham Stott <grahams@redhat.com>
+
+ * cp/class.c (type_requires_array_cookie): Fix use of uninitialised
+ variable has_two_argument_delete_p.
+
2001-07-21 Nathan Sidwell <nathan@codesourcery.com>
Remove flag_vtable_thunk. It is always on for the 3.0 ABI.
--- /maat/heart/tbox/changelog_pass/gcc/gcc/testsuite/ChangeLog Sun Jul 22 05:04:16 2001
+++ gcc/gcc/testsuite/ChangeLog Mon Jul 23 00:01:04 2001
@@ -1,3 +1,7 @@
+2001-07-23 Ben Elliston <bje@redhat.com>
+
+ * g++.old-deja/g++.other/enum5.C: Mark as XFAIL.
+
2001-07-22 Franz Sirl <Franz.Sirl-kernel@lauterbach.com>
* gcc.c-torture/execute/ieee/ieee.exp: Only use -ffloat-store only on