This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: Support for Xilinx PowerPC single-precision FPU
- From: "Joseph S. Myers" <joseph at codesourcery dot com>
- To: Michael Eager <eager at eagercon dot com>
- Cc: gcc-patches at gcc dot gnu dot org
- Date: Fri, 23 May 2008 21:45:47 +0000 (UTC)
- Subject: Re: Support for Xilinx PowerPC single-precision FPU
- References: <48372CD3.4080301@eagercon.com>
On Fri, 23 May 2008, Michael Eager wrote:
> There had been a question about ABI for floating point
> supported with a mix of hardware and software. I recently
> joined the Power.Org ABI committee. I've submitted proposed
> changes to the 32-bit ABI to support this configuration.
I didn't see anything in those changes regarding va_list layout (which
admittedly wasn't specified properly in the last draft, because my chapter
6 changes have yet to be implemented). That certainly needs to be
specified.
A .gnu_attribute value for single-float needs to be assigned, and
rs6000_file_start needs to use it, so these objects can be properly
diagnosed as incompatible with double-float objects. These assignments
are documented in the assembler manual, so you need to get the binutils
patch with an assignment and compatibility checks and testcases in first.
--
Joseph S. Myers
joseph@codesourcery.com