[Mesa-dev] [PATCH 53/61] spirv/cfg: Make the builder fully capable for both walks

Jason Ekstrand jason at jlekstrand.net
Fri Mar 23 21:42:59 UTC 2018


We were only initializing vtn_builder::func for the pre-walk where we
build the CFG.  We were only initializing the nir_builder for the later
walk through the instructions even though were were setting b->cursor
for the pre-walk.  Let's set both both places so that everything is
consistent.  This useful because we handle OpFunctionParameter in the
pre-walk and we're going to need to be able to emit instructions.
---
 src/compiler/spirv/vtn_cfg.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/src/compiler/spirv/vtn_cfg.c b/src/compiler/spirv/vtn_cfg.c
index 692e3b6..6e80ff5 100644
--- a/src/compiler/spirv/vtn_cfg.c
+++ b/src/compiler/spirv/vtn_cfg.c
@@ -95,6 +95,7 @@ vtn_cfg_handle_prepass_instruction(struct vtn_builder *b, SpvOp opcode,
       func->return_type = func_type->return_type->type;
 
       b->func->impl = nir_function_impl_create(func);
+      nir_builder_init(&b->nb, func->impl);
       b->nb.cursor = nir_before_cf_list(&b->func->impl->body);
 
       b->func_param_idx = 0;
@@ -878,6 +879,7 @@ vtn_function_emit(struct vtn_builder *b, struct vtn_function *func,
                   vtn_instruction_handler instruction_handler)
 {
    nir_builder_init(&b->nb, func->impl);
+   b->func = func;
    b->nb.cursor = nir_after_cf_list(&func->impl->body);
    b->has_loop_continue = false;
    b->phi_table = _mesa_hash_table_create(b, _mesa_hash_pointer,
-- 
2.5.0.400.gff86faf



More information about the mesa-dev mailing list