[Mesa-dev] [PATCH] radeonsi: Bitcast result of packf16 intrinsic to float for export intrinsic.

[Tom Stellard]

On Tue, Nov 27, 2012 at 07:55:05PM +0100, Michel Dänzer wrote:
> From: Michel Dänzer <michel.daenzer at amd.com>
> 
> Fixes 4 piglit tests and prevents many more from crashing.
> 
> Signed-off-by: Michel Dänzer <michel.daenzer at amd.com>
> ---
>  src/gallium/drivers/radeonsi/radeonsi_shader.c |    6 +++++-
>  1 file changed, 5 insertions(+), 1 deletion(-)
> 
> diff --git a/src/gallium/drivers/radeonsi/radeonsi_shader.c b/src/gallium/drivers/radeonsi/radeonsi_shader.c
> index d6e37ac..d5d56c4 100644
> --- a/src/gallium/drivers/radeonsi/radeonsi_shader.c
> +++ b/src/gallium/drivers/radeonsi/radeonsi_shader.c
> @@ -496,7 +496,11 @@ static void si_llvm_init_export_args(struct lp_build_tgsi_context *bld_base,
>  						LLVMInt32TypeInContext(base->gallivm->context),
>  						args, 2,
>  						LLVMReadNoneAttribute);
> -			args[chan + 7] = args[chan + 5];
> +			args[chan + 7] = args[chan + 5] =
> +				LLVMBuildBitCast(base->gallivm->builder,
> +						 args[chan + 5],
> +						 LLVMFloatTypeInContext(base->gallivm->context),
> +						 "");

After reading the "Fix lp_build_float_to_half" thread and reviewing the
LLVM IR docs, I realized that we can't use fptrunc instead of SI_packf16,
becuase it produces undefined results when the value won't fit into
the smaller type.

-Tom 

>  		}
>  
>  		/* Set COMPR flag */
> -- 
> 1.7.10.4
> 
> _______________________________________________
> mesa-dev mailing list
> mesa-dev at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/mesa-dev