egcs-980122 results for updated RHL/Intel 5.0 with binutils-2.1.0.19 added

[Mike Stump]

> To: egcs@cygnus.com
> Date: Wed, 28 Jan 1998 08:50:00 +0100

> System type: i586-pc-linux-gnu

> FAIL: g++.mike/p6610a.C  Execution test

If I recall correctly, this is the old `we decided to use thunk, even
though they don't do ctor time vbase conversions correctly' problem.
As I recall this was broken by:

Wed Aug 27 01:24:25 1997  H.J. Lu   (hjl@gnu.ai.mit.edu)

    * config/linux.h (DEFAULT_VTABLE_THUNKS): New. Defined as 1 if
    USE_GNULIBC_1 is not defined.

    * config/rs6000/linux.h (DEFAULT_VTABLE_THUNKS): New.  Defined as 1.

    * config/sparc/linux.h (DEFAULT_VTABLE_THUNKS): New. Defined
    as 1 if USE_GNULIBC_1 is not defined.

Wed Aug 27 16:08:43 1997  Jim Wilson  <wilson@cygnus.com>

    * config/alpha/elf.h (DEFAULT_VTABLE_THUNKS): New. Defined as 1
    if USE_GNULIBC_1 is not defined.

I think it is the responsibility of those that break test cases to
mark them as expected failures, if they know or are told of it, and
they don't plan to fix the problem.

If the target triplet isn't complete enough, then having an unexpected
pass seems better than an unexpected failure.

Mike (thanks for listening, but I really do hate unexpected failures) Stump