Re: [Mesa-dev] [PATCH 06/12] glsl: Add lowering pass for ir_triop_vector_insert

2013-05-07 Thread Eric Anholt
Ian Romanick writes: > On 05/07/2013 09:53 AM, Eric Anholt wrote: >> Ian Romanick writes: >> >>> From: Ian Romanick >>> >>> This will eventually replace do_vec_index_to_cond_assign. This lowering >>> pass is called in all the places where do_vec_index_to_cond_assign or >>> do_vec_index_to_swiz

Re: [Mesa-dev] [PATCH 06/12] glsl: Add lowering pass for ir_triop_vector_insert

2013-05-07 Thread Ian Romanick
On 05/07/2013 09:53 AM, Eric Anholt wrote: Ian Romanick writes: From: Ian Romanick This will eventually replace do_vec_index_to_cond_assign. This lowering pass is called in all the places where do_vec_index_to_cond_assign or do_vec_index_to_swizzle is called. v2: Use WRITEMASK_* instead of

Re: [Mesa-dev] [PATCH 06/12] glsl: Add lowering pass for ir_triop_vector_insert

2013-05-07 Thread Eric Anholt
Ian Romanick writes: > From: Ian Romanick > > This will eventually replace do_vec_index_to_cond_assign. This lowering > pass is called in all the places where do_vec_index_to_cond_assign or > do_vec_index_to_swizzle is called. > > v2: Use WRITEMASK_* instead of integer literals. Use a more con

Re: [Mesa-dev] [PATCH 06/12] glsl: Add lowering pass for ir_triop_vector_insert

2013-05-06 Thread Kenneth Graunke
On 05/03/2013 04:07 PM, Ian Romanick wrote: From: Ian Romanick This will eventually replace do_vec_index_to_cond_assign. This lowering pass is called in all the places where do_vec_index_to_cond_assign or do_vec_index_to_swizzle is called. v2: Use WRITEMASK_* instead of integer literals. Use

[Mesa-dev] [PATCH 06/12] glsl: Add lowering pass for ir_triop_vector_insert

2013-05-03 Thread Ian Romanick
From: Ian Romanick This will eventually replace do_vec_index_to_cond_assign. This lowering pass is called in all the places where do_vec_index_to_cond_assign or do_vec_index_to_swizzle is called. v2: Use WRITEMASK_* instead of integer literals. Use a more concise method of generating broadcast

Re: [Mesa-dev] [PATCH 06/12] glsl: Add lowering pass for ir_triop_vector_insert

2013-04-10 Thread Eric Anholt
Ian Romanick writes: > On 04/09/2013 10:29 AM, Eric Anholt wrote: >> Ian Romanick writes: >> I know our hardware doesn't like the swizzling of that bvec compare >> result and we'd rather just see individual compares as the condition of >> each if statement. (we basically have to emit a compare

Re: [Mesa-dev] [PATCH 06/12] glsl: Add lowering pass for ir_triop_vector_insert

2013-04-09 Thread Ian Romanick
On 04/09/2013 10:29 AM, Eric Anholt wrote: Ian Romanick writes: From: Ian Romanick This will eventually replace do_vec_index_to_cond_assign. This lowering pass is called in all the places where do_vec_index_to_cond_assign or do_vec_index_to_swizzle is called. + + ir_constant *const

Re: [Mesa-dev] [PATCH 06/12] glsl: Add lowering pass for ir_triop_vector_insert

2013-04-09 Thread Eric Anholt
Ian Romanick writes: > From: Ian Romanick > > This will eventually replace do_vec_index_to_cond_assign. This lowering > pass is called in all the places where do_vec_index_to_cond_assign or > do_vec_index_to_swizzle is called. > + > + ir_constant *const cmp_indices = > + new(facto

[Mesa-dev] [PATCH 06/12] glsl: Add lowering pass for ir_triop_vector_insert

2013-04-08 Thread Ian Romanick
From: Ian Romanick This will eventually replace do_vec_index_to_cond_assign. This lowering pass is called in all the places where do_vec_index_to_cond_assign or do_vec_index_to_swizzle is called. Signed-off-by: Ian Romanick --- src/glsl/Makefile.sources | 1 + src/glsl/glsl