diff options
| author | Ilia Mirkin <[email protected]> | 2016-11-05 09:31:21 -0400 |
|---|---|---|
| committer | Ilia Mirkin <[email protected]> | 2016-11-09 20:26:48 -0500 |
| commit | 73f53c097a6bbd07572d5e46af4893d17ff8af93 (patch) | |
| tree | c224c022da56f96956f2c898ece75d2fc66f0c2f /src/compiler/glsl/lower_packed_varyings.cpp | |
| parent | 885c78801785701fdca0217a7b0f992f26115ee4 (diff) | |
glsl: record number of components used in each slot for varying packing
Instead of packing varyings into vec4's, keep track of how many
components each slot uses and create varyings with matching types. This
ensures that we don't end up using more components than the orginal
shader, which is especially important for geometry shader output limits.
This comes up for NVIDIA hw, where the limit is 1024 output components
for a GS, and the hardware complains *loudly* if you even think about
going over.
Signed-off-by: Ilia Mirkin <[email protected]>
Reviewed-by: Timothy Arceri <[email protected]>
Diffstat (limited to 'src/compiler/glsl/lower_packed_varyings.cpp')
| -rw-r--r-- | src/compiler/glsl/lower_packed_varyings.cpp | 21 |
1 files changed, 16 insertions, 5 deletions
diff --git a/src/compiler/glsl/lower_packed_varyings.cpp b/src/compiler/glsl/lower_packed_varyings.cpp index 19bbe576a69..b16f25fc93d 100644 --- a/src/compiler/glsl/lower_packed_varyings.cpp +++ b/src/compiler/glsl/lower_packed_varyings.cpp @@ -164,7 +164,9 @@ namespace { class lower_packed_varyings_visitor { public: - lower_packed_varyings_visitor(void *mem_ctx, unsigned locations_used, + lower_packed_varyings_visitor(void *mem_ctx, + unsigned locations_used, + const uint8_t *components, ir_variable_mode mode, unsigned gs_input_vertices, exec_list *out_instructions, @@ -203,6 +205,8 @@ private: */ const unsigned locations_used; + const uint8_t* components; + /** * Array of pointers to the packed varyings that have been created for each * generic varying slot. NULL entries in this array indicate varying slots @@ -241,12 +245,14 @@ private: } /* anonymous namespace */ lower_packed_varyings_visitor::lower_packed_varyings_visitor( - void *mem_ctx, unsigned locations_used, ir_variable_mode mode, + void *mem_ctx, unsigned locations_used, const uint8_t *components, + ir_variable_mode mode, unsigned gs_input_vertices, exec_list *out_instructions, exec_list *out_variables, bool disable_varying_packing, bool xfb_enabled) : mem_ctx(mem_ctx), locations_used(locations_used), + components(components), packed_varyings((ir_variable **) rzalloc_array_size(mem_ctx, sizeof(*packed_varyings), locations_used)), @@ -607,10 +613,11 @@ lower_packed_varyings_visitor::get_packed_varying_deref( if (this->packed_varyings[slot] == NULL) { char *packed_name = ralloc_asprintf(this->mem_ctx, "packed:%s", name); const glsl_type *packed_type; + assert(components[slot] != 0); if (unpacked_var->is_interpolation_flat()) - packed_type = glsl_type::ivec4_type; + packed_type = glsl_type::get_instance(GLSL_TYPE_INT, components[slot], 1); else - packed_type = glsl_type::vec4_type; + packed_type = glsl_type::get_instance(GLSL_TYPE_FLOAT, components[slot], 1); if (this->gs_input_vertices != 0) { packed_type = glsl_type::get_array_instance(packed_type, @@ -771,6 +778,7 @@ lower_packed_varyings_return_splicer::visit_leave(ir_return *ret) void lower_packed_varyings(void *mem_ctx, unsigned locations_used, + const uint8_t *components, ir_variable_mode mode, unsigned gs_input_vertices, gl_linked_shader *shader, bool disable_varying_packing, bool xfb_enabled) @@ -781,7 +789,10 @@ lower_packed_varyings(void *mem_ctx, unsigned locations_used, ir_function_signature *main_func_sig = main_func->matching_signature(NULL, &void_parameters, false); exec_list new_instructions, new_variables; - lower_packed_varyings_visitor visitor(mem_ctx, locations_used, mode, + lower_packed_varyings_visitor visitor(mem_ctx, + locations_used, + components, + mode, gs_input_vertices, &new_instructions, &new_variables, |