[Libclc-dev] [PATCH 0/2] modf math builtin

[Pavel Ondračka via Libclc-dev]

Jan Vesely píše v Út 19. 01. 2016 v 15:37 -0500:
> On Tue, 2016-01-19 at 20:12 +0000, Arsenault, Matthew via Libclc-dev
> wrote:
> > I tested your patch with OpenCL conformance, and it passes.
> > 
> > The documentation's pseudocode also passes conformance, and emits
> > superior code from using the v_trunc_* instructions.
> 
> I think it'd be nice to have the test in piglit as well.
> 
> Jan
> 
I have used piglit locally to test my implementation. Attached is the
handcrafted test I used. I have no idea if it can be included like
that. Sadly the math builtins generating framework would need some work
to support tests for functions that have two outputs. I'll try to do it
but it will take some time to improve my python skills. 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: builtin-float-modf-1.0.cl
Type: text/x-csrc
Size: 5068 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/libclc-dev/attachments/20160120/7128c9aa/attachment-0001.c>