[llvm-commits] PATCH: llvm-gcc in SPARC

Venkatraman Govindaraju venkatra at cs.wisc.edu
Thu Aug 27 09:35:03 PDT 2009

Can I apply the patch then?


On Wed, Aug 26, 2009 at 10:10 PM, Chris Lattner<clattner at apple.com> wrote:
> On Aug 26, 2009, at 2:53 PM, Venkatraman Govindaraju wrote:
>> On Wed, Aug 26, 2009 at 1:29 PM, Anton
>> Korobeynikov<anton at korobeynikov.info> wrote:
>>> Hello Venkatraman,
>>>> Although sparc's long double is 128 bit, neither llvm-gcc (see
>>>> llvm-types.cpp)
>>> It does support. I just committed fixes there. You only need to enable
>>> it for sparc.
>> I enabled it for sparc now. Although llvm-gcc now supports 128 bit
>> long double, sparc back end  does not support long doubles yet (it
>> does not have quad fp registers or instructions yet). Hence xgcc fails
>> to compile __powitf2.o which belongs to libgcc.
>> I added a FIXME comment.
>>>> This script generates the options.h which is included in
>>>> gcc/config/sparc/gmon-sol2.h through tm.h. However when gmon-sol2.h
>>>> compiles as part of gcc, HOST_BITS_PER_INT is not defined. However
>>>> HOST_BITS_PER_INT is used in options.h to explicitly pad a struct.
>>>> If there is a better way to fix this, let me know.
>>> Well, the proper solution is to find where this define is added and
>>> try to figure out how to include the file in question, I suppose.
>> When I commented out tm.h, gmon-sol2.c compiles fine. Can I just
>> commented out tm.h from gmon-sol2.c?
> I don't have a problem with this if it works!
> -Chris

More information about the llvm-commits mailing list