This is the mail archive of the
mailing list for the GCC project.
Re: C++ ABI Issues
- From: Mike Stump <mrs at apple dot com>
- To: Ziemowit Laski <zlaski at apple dot com>
- Cc: gcc at gcc dot gnu dot org
- Date: Tue, 27 Aug 2002 11:56:45 -0700
- Subject: Re: C++ ABI Issues
On Tuesday, August 27, 2002, at 11:27 AM, Ziemowit Laski wrote:
Yes, but that's a pipe dream. On cannot have any abi velocity of any
sort for multivendor abi compatibility. One cannot have any abi bugs
either. If both of these were true, we would not be talking about
-fabi-correctly and friends.
On Tuesday, Aug 27, 2002, at 11:15 US/Pacific, Mike Stump wrote:
If the C++ ABI is to be a multi-vendor ABI, then I think you _would_
need some sort of multi-vendor numbering scheme. (Or is source-level
compatibility not desirable?) You're right about the difficulty of
ensuring bug-for-bug compatibility, though.
On Tuesday, August 27, 2002, at 11:07 AM, Ziemowit Laski wrote:
Another question is whether proprietary compilers (HP aC++, etc.)
No. It is impossible for someone else to have exactly the same bugs
as we have without very deliberate engineering, also they would have
to be perfect in their abi to not require their own abi number.