[PATCH] Set properties of f128 type

[Akira Hatanaka]

Sorry, I forgot to include the changes I made to GetFPLibCall in
LegalizeFloatTypes.cpp in my last email. Please see the updated patch.

With this patch, the attached test case (mips64-f128.ll) passes make check.

On Thu, Feb 21, 2013 at 12:20 PM, Tim Northover <t.p.northover at gmail.com>wrote:

> > This patch sets properties of f128 type in
> > TargetLoweringBase::computeRegisterProperties. This is needed to enable
> > soft-float on architectures which have support for f128. mips64 is an
> > example of such an architecture. aarch64 also supports f128, but it
> doesn't
> > need this because it doesn't have a soft-float ABI.
>
> It looks clearly in line with the surrounding code to me, but is there
> some kind of test you could write for MIPS that would demonstrate it?
> We don't do much merging on trunk, but it could help avoid an obvious
> blunder in some future refactoring effort.
>
> Cheers.
>
> Tim.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20130222/4aefb8a6/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: f128soft2.patch
Type: application/octet-stream
Size: 21493 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20130222/4aefb8a6/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: mips64-f128.ll
Type: application/octet-stream
Size: 13603 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20130222/4aefb8a6/attachment-0001.obj>