This is the mail archive of the
gcc-regression@gcc.gnu.org
mailing list for the GCC project.
Re: [3.3 branch, 6 regressions] Re: 'fix' for 11350 causes bootstraphang on Darwin
On Fri, 23 Jan 2004, Krister Walfridsson wrote:
> The change
>
> 2004-01-18 Jan Hubicka <jh@suse.cz>
>
> * basic-block.h (try_redirect_by_replacing_jump): Declare.
> * cfgcleanup.c (try_optimize_cfg): Use it.
> * cfgrtl.c (try_redirect_by_replacing_jump): Export.
> (rtl_redirect_edge_and_branch, cfg_layout_redirect_edge_and_branch):
> Kill hack.
> (cfg_layout_merge_blocks): Use try_redirect_by_replacing_jump.
>
> Revert:
> 2004-01-16 Geoffrey Keating <geoffk@apple.com>
>
> * cfgrtl.c (try_redirect_by_replacing_jump): Optimize tablejumps
> even after reload, just don't remove the actual jump tables.
>
> causes bootstrap comparison failure for i386-unknown-netbsdelf1.6.
> It looks like all .o files differs...
>
> I'll try to dig up some more info, but I'm not really sure where to
> start...
Running the testsuite on the compiler from the first stage shows
that the change introduces these new failures:
FAIL: gcc.c-torture/compile/goto-1.c (test for excess errors)
FAIL: gcc.c-torture/compile/goto-1.c (test for excess errors)
FAIL: gcc.c-torture/compile/goto-1.c (test for excess errors)
FAIL: gcc.c-torture/compile/goto-1.c (test for excess errors)
FAIL: gcc.c-torture/compile/goto-1.c (test for excess errors)
FAIL: gcc.c-torture/execute/930603-2.c compilation, -O3 -fomit-frame-pointer -funroll-loops
FAIL: gcc.c-torture/execute/930603-2.c compilation, -O3 -fomit-frame-pointer -funroll-all-loops -finline-functions
/Krister