eh merge complete

Franz Sirl Franz.Sirl-kernel@lauterbach.com
Sun May 13 13:56:00 GMT 2001


On Sunday 13 May 2001 09:20, Richard Henderson wrote:
> Tested on i686 and alphaev6 linux, which should be good enough for a basic
> sanity check.  IA-64 doesn't bootstrap at all on the branch, and my other
> hosts are slow enough to not be done testing until tomorrow.
>
> There are a couple of regressions, but this patch was large enough
> that I'd rather address them separately.  On i686 I see
>
> < + 21_strings/ctor_copy_dtor.cc
>
> > -r 21_strings/ctor_copy_dtor.cc
>
> < + 21_strings/insert.cc
>
> > -r 21_strings/insert.cc
>
> These pass on alpha.  Not sure off-hand how this compares with mainline.
>
> Also, g++.oliva/delete3.C fails everywhere.  This is a regression from
> mainline.  It appears that __cxa_vec_delete isn't being invoked at all,
> so I assume there's some ABI patch on mainline that is missing.
>
> For the record, the final patch is attached.

You just broke bootstrap on powerpc-linux-gnu :-). I  guess you want this one 
too:

2001-03-28  Richard Henderson  <rth@redhat.com>
 
        * config/rs6000/rs6000.h (EPILOGUE_USES): Use TARGET_AIX,
        not TARGET_TOC for special toc restore handling.
        * config/rs6000/rs6000.md (eh_return): Likewise.

Franz.



More information about the Gcc-patches mailing list