14 GCC regressions, 11 new, with your patch on 2003-07-05T14:15:01Z.

[Zack Weinberg]

Andrew Pinski <pinskia@physics.uc.edu> writes:

> On Saturday, Jul 5, 2003, at 15:03 US/Eastern, GCC regression checker
> wrote:
>
>> With your recent patch, GCC has some regression test failures, which
>> used to pass.  There are 11 new failures, and 3
>> failures that existed before and after that patch; 16 failures
>> have been fixed.
>>
>> The new failures are:
>> native g++.sum g++.dg/debug/debug4.C
>> native gcc.sum gcc.c-torture/execute/20010325-1.c
>> native gcc.sum gcc.c-torture/execute/wchar_t-1.c
>> native gcc.sum gcc.c-torture/execute/widechar-2.c
>> native gcc.sum gcc.dg/c99-init-1.c
>> native gcc.sum gcc.dg/cpp/escape-2.c
>> native gcc.sum gcc.dg/cpp/escape.c
>> native gcc.sum gcc.dg/cpp/if-2.c
>> native gcc.sum gcc.dg/cpp/lexstrng.c
>> native gcc.sum gcc.dg/cpp/paste2.c
>> native gcc.sum gcc.dg/cpp/ucs.c
>
> All of these really are caused by Zack's patch which added USC to CPP
> but Darwin does not iconv so they fail.

Either we need to put GNU iconv into the GCC distribution, or all
these tests are going to have to be XFAILed on targets that don't have
iconv.  Personally I lean towards importing GNU iconv, as this will
provide consistent behavior across hosts and let me get rid of some
dodgy fallback logic.

> gcc.c-torture/execute/wchar_t-1.c and
> gcc.c-torture/execute/widechar-2.c: fails on execution.

These may be worth further investigation, as they suggest that the
fallback logic doesn't work.

zw