From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 13417 invoked by alias); 23 Jan 2009 20:04:15 -0000 Received: (qmail 13220 invoked by uid 48); 23 Jan 2009 20:04:02 -0000 Date: Fri, 23 Jan 2009 20:04:00 -0000 Message-ID: <20090123200402.13219.qmail@sourceware.org> X-Bugzilla-Reason: CC References: Subject: [Bug fortran/38946] gcc trunk 143562 - Testsuite - gfortran failing tests that worked previously In-Reply-To: Reply-To: gcc-bugzilla@gcc.gnu.org To: gcc-bugs@gcc.gnu.org From: "mikael at gcc dot gnu dot org" Mailing-List: contact gcc-bugs-help@gcc.gnu.org; run by ezmlm Precedence: bulk List-Id: List-Archive: List-Post: List-Help: Sender: gcc-bugs-owner@gcc.gnu.org X-SW-Source: 2009-01/txt/msg02475.txt.bz2 ------- Comment #1 from mikael at gcc dot gnu dot org 2009-01-23 20:04 ------- (In reply to comment #0) > === gfortran tests === > FAIL: gfortran.dg/array_constructor_23.f -O3 -fomit-frame-pointer execution > test > FAIL: gfortran.dg/array_constructor_23.f -O3 -fomit-frame-pointer > -funroll-loops execution test > FAIL: gfortran.dg/array_constructor_23.f -O3 -fomit-frame-pointer > -funroll-all-loops -finline-functions execution test > FAIL: gfortran.dg/array_constructor_23.f -O3 -g execution test Those were failing in your previous report too. ;) > When I view the log I see this (dupe blank lines removed): > > Executing on host: > /usr/share/src/gcc_build/gcc/testsuite/gfortran/../../gfortran > -B/usr/share/src/gcc_build/gcc/testsui$ > /usr/share/src/gcc_trunk/gcc/testsuite/gfortran.dg/array_constructor_26.f03:13.26: > > CHARACTER (MAX_FLD_HED, 1) :: DWFdHd(MAXFLD) = [(" ", i = 1, MAXFLD)] > > 1 > > Error: Symbol 'max_fld_hed' at (1) has no IMPLICIT type > This error (like the others) is expected: we test that the compiler gives the expected output. Look for lines starting with FAIL for real failures. But as they are run-time, it won't help probably. Quite many of the newly-failing tests are about I/O; I suspect the following to be the culprit. Could you try before/after this? ------------------------------------------------------------------------ r143462 | pault | 2009-01-17 12:32:02 +0100 (sam. 17 janv. 2009) | 17 lines 2009-01-17 Paul Thomas PR fortran/34955 * trans-intrinsic.c (gfc_conv_intrinsic_array_transfer): Note that some of the tests require specific features (such as denormalized long doubles) and are x-failed(which means: known to fail) on some targets or have some dg-require-effective-targets conditions on them. Thus, maybe the correct behaviour for these tests is to fail on solaris. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38946