Re: pr83 random failure

[Andrew Haley <aph at redhat dot com>]

Michael Ritzert writes:
 > On Sat, Sep 25, 2004 at 10:04:13PM +0200, Michael Ritzert wrote:
 > > On Sat, Sep 25, 2004@10:51:52AM +0100, Andrew Haley wrote:
 > > > The file i686-pc-linux-gnu/libjava/testsuite/libjava.log that I
 > > > downloaded from
 > > > http://www.mailsys.de/~gcc/logs-HEAD-i686-1096068427.tar.bz2 contains
 > > > a number of failures that indicate gcj is pretty seriously broken on
 > > > that platform.
 > > > 
 > > > FAIL: Array_3 execution - source compiled test
 > > > FAIL: Array_3 -O3 execution - source compiled test
 > > > FAIL: Invoke_1 execution - source compiled test
 > > > FAIL: Invoke_1 execution - bytecode->native test
 > > > FAIL: Invoke_1 -O3 execution - source compiled test
 > > > FAIL: Invoke_1 -O3 execution - bytecode->native test
 > > > 
 > > > These indicate something very broken with optimization and exception
 > > > handling.
 > > 
 > > See http://gcc.gnu.org/ml/gcc-regression/2004-09/msg00184.html .
 > 
 > The second machine caught this in the testing phase.  Taking the intersection
 > of the ChangeLog diffs and stripping the testcase-only, Sparc and Makefile
 > dependency commits, I can further narrow it down to
 > 
 > 2004-09-22  Diego Novillo  <dnovillo@redhat.com>
 > 
 >         PR tree-optimization/16721
 >         * tree-dfa.c (dump_variable): Show TREE_THIS_VOLATILE.
 >         * tree-ssa-alias.c (create_memory_tag): Move setting of
 >         TREE_THIS_VOLATILE ...
 >         (get_tmt_for): ... here.
 > 
 > 2004-09-22  Joseph S. Myers  <jsm@polyomino.org.uk>
 > 
 >         PR c/16566
 >         * c-typeck.c (build_component_ref): Don't special-case
 >         COMPOUND_EXPR.
 > 
 > 2004-09-22  Andreas Tobler  <a.tobler@schweiz.ch>
 > 
 >         * configure.ac: Introduce AC_C_BIGENDIAN_CROSS for WORDS_BIGENDIAN.
 >         * configure: Regenerate.
 >         * include/config.h.in: Likewise.
 >         * jni/gtk-peer/gtkpeer.h (SWAPU32): Introduce macro to swap pixels.
 >         * jni/gtk-peer/gnu_java_awt_peer_gtk_GtkImagePainter.c: Moved SWAPU32
 >         macro to gtkpeer.h.
 >         * jni/gtk-peer/gnu_java_awt_peer_gtk_GdkGraphics2D.c
 >         (Java_gnu_java_awt_peer_gtk_GdkGraphics2D_getImagePixels): Convert
 >         pixels from  0xBBGGRRAA to 0xAARRGGBB only on Little Endian
 >         architectures.
 >         * jni/gtk-peer/gnu_java_awt_peer_gtk_GdkPixbufDecoder.c (area_updated):
 >         Likewise.

Great, thanks.  I think we can assume it's not c-typeck.c, so either
Diego's or Andreas's patch is in the headlights.

Andrew.