[Patch] Avoid conversion to float when creating ConstantDataArray/ConstantDataVector

[Raoux, Thomas F]

Hi Rafael,

I added the same test for the ConstantDataArray path. If this is patch looks ok to you, could you commit it as I don't have write access?

Thank you

-----Original Message-----
From: Rafael Espíndola [mailto:rafael.espindola at gmail.com] 
Sent: Thursday, February 05, 2015 1:50 PM
To: Raoux, Thomas F
Cc: llvm-commits at cs.uiuc.edu; Nick Lewycky
Subject: Re: [Patch] Avoid conversion to float when creating ConstantDataArray/ConstantDataVector

On 5 February 2015 at 16:34, Raoux, Thomas F <thomas.f.raoux at intel.com> wrote:
> There is a test as part of the patch. Let me know if any other tests are needed.

Duh, I applied the patch to make it easier to read and didn't notice the test. Sorry about that. The test only checks the vector code path, no?

> The spec already specify that bitcast doesn't change any bits. Although here it is not the bitcast changing the bits but the insertElement. I would expect that the insertElement shouldn't change the bits either.
>
> "The ‘bitcast‘ instruction converts value to type ty2 without changing any bits."

Cool. I think the patch is fine. Nick, would you mind double checking?

Cheers,
Rafael
-------------- next part --------------
A non-text attachment was scrubbed...
Name: ConstantDataVector.patch
Type: application/octet-stream
Size: 8035 bytes
Desc: ConstantDataVector.patch
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20150206/9a2de94d/attachment.obj>