[llvm-commits] [llvm] r125393 - in /llvm/trunk: lib/VMCore/ConstantFold.cpp test/Transforms/InstCombine/bitcast-vec-uniform.ll test/Transforms/InstCombine/fold-vector-select.ll

[Rotem, Nadav]

Hi Frits, Nick and Duncan,

I attached a revised patch which supports getAllOnesValue on FP values. I added minor changes to APFloat as well as new tests. 

Thanks, 
Nadav

-----Original Message-----
From: Frits van Bommel [mailto:fvbommel at gmail.com] 
Sent: Tuesday, February 15, 2011 00:25
To: Rotem, Nadav
Cc: Nick Lewycky; Duncan Sands; llvm-commits at cs.uiuc.edu
Subject: Re: [llvm-commits] [llvm] r125393 - in /llvm/trunk: lib/VMCore/ConstantFold.cpp test/Transforms/InstCombine/bitcast-vec-uniform.ll test/Transforms/InstCombine/fold-vector-select.ll

On Mon, Feb 14, 2011 at 10:56 PM, Rotem, Nadav <nadav.rotem at intel.com> wrote:
> I started by looking at adding alone support for constants. I started by adding support for float/double inside the Constant function. A change to APFloat is also possible but more complex. What do you think ?

No need to go messing around with floats and doubles and
reinterpret_casts (or changing APFloat).
APFloat has a constructor that takes an APInt and initializes the
floating-point value with those bits (just be careful with 'fp128' vs
'ppc_fp128' values, that's what the second parameter is for).
---------------------------------------------------------------------
Intel Israel (74) Limited

This e-mail and any attachments may contain confidential material for
the sole use of the intended recipient(s). Any review or distribution
by others is strictly prohibited. If you are not the intended
recipient, please contact the sender and delete all copies.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: fp.diff
Type: application/octet-stream
Size: 4493 bytes
Desc: fp.diff
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20110216/e7616222/attachment.obj>