[Mesa-dev] [PATCH] gallivm/gs: fix indirect addressing in geometry shaders

[Roland Scheidegger]

Am 17.04.2013 21:17, schrieb Zack Rusin:
> We were always treating the vertex index as a scalar but when the
> shader is using indirect addressing it will be a vector of indices
> for each channel. This was causing some nasty crashes insides
> LLVM.
> 
> Signed-off-by: Zack Rusin <zackr at vmware.com>
> ---
>  src/gallium/auxiliary/draw/draw_llvm.c          |   34 +++++++++++++++++++----
>  src/gallium/auxiliary/gallivm/lp_bld_tgsi.h     |    1 +
>  src/gallium/auxiliary/gallivm/lp_bld_tgsi_soa.c |    1 +
>  3 files changed, 30 insertions(+), 6 deletions(-)
> 
> diff --git a/src/gallium/auxiliary/draw/draw_llvm.c b/src/gallium/auxiliary/draw/draw_llvm.c
> index 33fe40d..8e3ea88 100644
> --- a/src/gallium/auxiliary/draw/draw_llvm.c
> +++ b/src/gallium/auxiliary/draw/draw_llvm.c
> @@ -1253,6 +1253,7 @@ clipmask_booli32(struct gallivm_state *gallivm,
>  static LLVMValueRef
>  draw_gs_llvm_fetch_input(const struct lp_build_tgsi_gs_iface *gs_iface,
>                           struct lp_build_tgsi_context * bld_base,
> +                         boolean is_indirect,
>                           LLVMValueRef vertex_index,
>                           LLVMValueRef attrib_index,
>                           LLVMValueRef swizzle_index)
> @@ -1262,13 +1263,34 @@ draw_gs_llvm_fetch_input(const struct lp_build_tgsi_gs_iface *gs_iface,
>     LLVMBuilderRef builder = gallivm->builder;
>     LLVMValueRef indices[3];
>     LLVMValueRef res;
> +   struct lp_type type = bld_base->base.type;
> +
> +   if (is_indirect) {
> +      int i;
> +      res = bld_base->base.zero;
> +      for (i = 0; i < type.length; ++i) {
> +         LLVMValueRef idx = lp_build_const_int32(gallivm, i);
> +         LLVMValueRef vert_chan_index = LLVMBuildExtractElement(builder,
> +                                                                vertex_index, idx, "");
> +         LLVMValueRef channel_vec, value;
> +         indices[0] = vert_chan_index;
> +         indices[1] = attrib_index;
> +         indices[2] = swizzle_index;
> +         
> +         channel_vec = LLVMBuildGEP(builder, gs->input, indices, 3, "");
> +         channel_vec = LLVMBuildLoad(builder, channel_vec, "");
> +         value = LLVMBuildExtractElement(builder, channel_vec, idx, "");
> +
> +         res = LLVMBuildInsertElement(builder, res, value, idx, "");
> +      }
> +   } else {
> +      indices[0] = vertex_index;
> +      indices[1] = attrib_index;
> +      indices[2] = swizzle_index;
>  
> -   indices[0] = vertex_index;
> -   indices[1] = attrib_index;
> -   indices[2] = swizzle_index;
> -   
> -   res = LLVMBuildGEP(builder, gs->input, indices, 3, "");
> -   res = LLVMBuildLoad(builder, res, "");
> +      res = LLVMBuildGEP(builder, gs->input, indices, 3, "");
> +      res = LLVMBuildLoad(builder, res, "");
> +   }
>  
>     return res;
>  }
> diff --git a/src/gallium/auxiliary/gallivm/lp_bld_tgsi.h b/src/gallium/auxiliary/gallivm/lp_bld_tgsi.h
> index 0fbb8aa..78a1f0e 100644
> --- a/src/gallium/auxiliary/gallivm/lp_bld_tgsi.h
> +++ b/src/gallium/auxiliary/gallivm/lp_bld_tgsi.h
> @@ -368,6 +368,7 @@ struct lp_build_tgsi_gs_iface
>  {
>     LLVMValueRef (*fetch_input)(const struct lp_build_tgsi_gs_iface *gs_iface,
>                                 struct lp_build_tgsi_context * bld_base,
> +                               boolean is_indirect,
>                                 LLVMValueRef vertex_index,
>                                 LLVMValueRef attrib_index,
>                                 LLVMValueRef swizzle_index);
> diff --git a/src/gallium/auxiliary/gallivm/lp_bld_tgsi_soa.c b/src/gallium/auxiliary/gallivm/lp_bld_tgsi_soa.c
> index 28eb57b..ea7dec7 100644
> --- a/src/gallium/auxiliary/gallivm/lp_bld_tgsi_soa.c
> +++ b/src/gallium/auxiliary/gallivm/lp_bld_tgsi_soa.c
> @@ -829,6 +829,7 @@ emit_fetch_gs_input(
>     }
>  
>     res = bld->gs_iface->fetch_input(bld->gs_iface, bld_base,
> +                                    reg->Dimension.Indirect,
>                                      vertex_index, attrib_index,
>                                      swizzle_index);
>  
> 

Looks good.
Reviewed-by: Roland Scheidegger <sroland at vmware.com>