vrend: grok FBFETCH instruction

The TGSI FBFETCH instruction is used for implementing framebuffer-reads
when doing for instance advanced blends. It's used by both
EXT_shader_framebuffer_fetch_non_coherent and
KHR_blend_equation_advanced.

Sadly, libepoxy doesn't yet support the glFramebufferFetchBarrierEXT
function needed to implement this on top of
EXT_shader_framebuffer_fetch_non_coherent itself, so let's just implement
it on top of EXT_shader_framebuffer_fetch instead for now.

This means that we'll have slightly higher requirements for enabling
this functionality than the level of functionality we provide through it.

Signed-off-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
macos/master
Erik Faye-Lund 6 years ago committed by Dave Airlie
parent 66307104d5
commit 3e0f7bd4ae
  1. 14
      src/vrend_shader.c

@ -60,6 +60,7 @@ extern int vrend_dump_shaders;
#define SHADER_REQ_ES31_COMPAT (1 << 16)
#define SHADER_REQ_IMAGE_SIZE (1 << 17)
#define SHADER_REQ_TXQS (1 << 18)
#define SHADER_REQ_FBFETCH (1 << 19)
struct vrend_shader_io {
unsigned name;
@ -76,6 +77,7 @@ struct vrend_shader_io {
bool glsl_gl_block;
bool override_no_wm;
bool is_int;
bool fbfetch_used;
char glsl_name[64];
unsigned stream;
};
@ -234,6 +236,7 @@ static const struct vrend_shader_table shader_req_table[] = {
{ SHADER_REQ_ES31_COMPAT, "GL_ARB_ES3_1_compatibility" },
{ SHADER_REQ_IMAGE_SIZE, "GL_ARB_shader_image_size" },
{ SHADER_REQ_TXQS, "GL_ARB_shader_texture_image_samples" },
{ SHADER_REQ_FBFETCH, "GL_EXT_shader_framebuffer_fetch" },
};
enum vrend_type_qualifier {
@ -929,6 +932,7 @@ iter_declaration(struct tgsi_iterate_context *iter,
ctx->outputs[i].glsl_no_index = false;
ctx->outputs[i].override_no_wm = false;
ctx->outputs[i].is_int = false;
ctx->outputs[i].fbfetch_used = false;
switch (ctx->outputs[i].name) {
case TGSI_SEMANTIC_POSITION:
@ -2948,6 +2952,11 @@ get_source_info(struct dump_ctx *ctx,
} else if (src->Register.File == TGSI_FILE_OUTPUT) {
for (uint32_t j = 0; j < ctx->num_outputs; j++) {
if (ctx->outputs[j].first == src->Register.Index) {
if (inst->Instruction.Opcode == TGSI_OPCODE_FBFETCH) {
ctx->outputs[j].fbfetch_used = true;
ctx->shader_req_bits |= SHADER_REQ_FBFETCH;
}
enum vrend_type_qualifier srcstypeprefix = stypeprefix;
if (stype == TGSI_TYPE_UNSIGNED && ctx->outputs[j].is_int)
srcstypeprefix = TYPE_CONVERSION_NONE;
@ -3436,6 +3445,7 @@ iter_instruction(struct tgsi_iterate_context *iter,
snprintf(buf, 255, "%s = %s(inversesqrt(%s.x));\n", dsts[0], get_string(dinfo.dstconv), srcs[0]);
EMIT_BUF_WITH_RET(ctx, buf);
break;
case TGSI_OPCODE_FBFETCH:
case TGSI_OPCODE_MOV:
snprintf(buf, 255, "%s = %s(%s(%s%s));\n", dsts[0], get_string(dinfo.dstconv), get_string(dinfo.dtypeprefix), srcs[0], sinfo.override_no_wm[0] ? "" : writemask);
EMIT_BUF_WITH_RET(ctx, buf);
@ -4404,9 +4414,11 @@ static char *emit_ios(struct dump_ctx *ctx, char *glsl_hdr)
ctx->outputs[i].invariant ? "invariant " : "",
ctx->outputs[i].glsl_name);
else
snprintf(buf, 255, "%s%s%sout vec4 %s;\n", prefix,
snprintf(buf, 255, "%s%s%s%s vec4 %s;\n",
prefix,
ctx->outputs[i].precise ? "precise " : "",
ctx->outputs[i].invariant ? "invariant " : "",
ctx->outputs[i].fbfetch_used ? "inout" : "out",
ctx->outputs[i].glsl_name);
STRCAT_WITH_RET(glsl_hdr, buf);
} else if (ctx->outputs[i].invariant || ctx->outputs[i].precise) {

Loading…
Cancel
Save