[llvm-commits] [patch] fp128 sint_to_fp folding fix

[Tim Northover]

On Monday 07 Jan 2013 16:06:28 Renato Golin wrote:
> On 7 January 2013 14:50, Tim Northover
> I don't get what IEEE has to do with it...

APFloat can (allegedly) model both IEEE and PPC 128-bit floats, which have
completely different formats (PPC ones are just 2 doubles glued together). You
have to make sure you get the right one or constant folding will just get the
wrong bitpattern.

> Nothing wrong with your commit, but looks like a special case to me to
> force PPC types for non-IEEE... Even weirder is that this is the
> default...

I presume that's because PPC was the first Target to support any kind of 128-
bit float in some reasonably complete form.

Tim.

-- IMPORTANT NOTICE: The contents of this email and any attachments are confidential and may also be privileged. If you are not the intended recipient, please notify the sender immediately and do not disclose the contents to any other person, use it for any purpose, or store or copy the information in any medium.  Thank you.