KonstantinSeurer
/
mesa
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

st/glsl_to_tgsi: make use of driver_cache_blob with the disk cache 

driver_cache_blob was introduced with the i965 disk cache, it allows
us to simplify the cache a little and possibly offers some minor
speed improvements since we load the GLSL metadata and TGSI from
disk in one pass.

Using driver_cache_blob should also make it straight forward to
implement binary support for ARB_get_program_binary in gallium.

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
This commit is contained in:
Timothy Arceri 2017-11-27 16:25:11 +11:00
parent 4cb27047c8
commit 3e789026ca
4 changed files with 139 additions and 260 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
src/compiler/glsl/shader_cache.cpp
View File

@ -1299,7 +1299,7 @@ shader_cache_read_program_metadata(struct gl_context *ctx,
return false;
struct disk_cache *cache = ctx->Cache;
if (!cache || prog->data->skip_cache)
if (!cache)
return false;
/* Include bindings when creating sha1. These bindings change the resulting

7
src/mesa/main/mtypes.h
View File

@ -2887,13 +2887,6 @@ struct gl_shader_program_data
unsigned NumUniformDataSlots;
union gl_constant_value *UniformDataSlots;
/* TODO: This used by Gallium drivers to skip the cache on tgsi fallback.
* All structures (gl_program, uniform storage, etc) will get recreated
* even though we have already loaded them from cache. We should instead
* switch to storing the GLSL metadata and TGSI IR in a single cache item
* like the i965 driver does with NIR.
*/
bool skip_cache;
GLboolean Validated;
/** List of all active resources after linking. */

3
src/mesa/state_tracker/st_program.h
View File

@ -150,9 +150,6 @@ struct st_fragment_program
struct gl_shader_program *shader_program;
struct st_fp_variant *variants;
/** SHA1 hash of linked tgsi shader program, used for on-disk cache */
unsigned char sha1[20];
};

387
src/mesa/state_tracker/st_shader_cache.c
View File

@ -40,14 +40,15 @@ write_stream_out_to_cache(struct blob *blob,
static void
write_tgsi_to_cache(struct blob *blob, struct pipe_shader_state *tgsi,
struct st_context *st, unsigned char *sha1,
unsigned num_tokens)
struct gl_program *prog, unsigned num_tokens)
{
blob_write_uint32(blob, num_tokens);
blob_write_bytes(blob, tgsi->tokens,
num_tokens * sizeof(struct tgsi_token));
disk_cache_put(st->ctx->Cache, sha1, blob->data, blob->size, NULL);
prog->driver_cache_blob = ralloc_size(NULL, blob->size);
memcpy(prog->driver_cache_blob, blob->data, blob->size);
prog->driver_cache_blob_size = blob->size;
}
/**
@ -68,14 +69,12 @@ st_store_tgsi_in_disk_cache(struct st_context *st, struct gl_program *prog,
if (memcmp(prog->sh.data->sha1, zero, sizeof(prog->sh.data->sha1)) == 0)
return;
unsigned char *sha1;
struct blob blob;
blob_init(&blob);
switch (prog->info.stage) {
case MESA_SHADER_VERTEX: {
struct st_vertex_program *stvp = (struct st_vertex_program *) prog;
sha1 = stvp->sha1;
blob_write_uint32(&blob, stvp->num_inputs);
blob_write_bytes(&blob, stvp->index_to_input,
@ -84,31 +83,24 @@ st_store_tgsi_in_disk_cache(struct st_context *st, struct gl_program *prog,
sizeof(stvp->result_to_output));
write_stream_out_to_cache(&blob, &stvp->tgsi);
write_tgsi_to_cache(&blob, &stvp->tgsi, st, sha1, num_tokens);
write_tgsi_to_cache(&blob, &stvp->tgsi, prog, num_tokens);
break;
}
case MESA_SHADER_TESS_CTRL:
case MESA_SHADER_TESS_EVAL:
case MESA_SHADER_GEOMETRY: {
struct st_common_program *p = st_common_program(prog);
sha1 = p->sha1;
write_stream_out_to_cache(&blob, out_state);
write_tgsi_to_cache(&blob, out_state, st, sha1, num_tokens);
write_tgsi_to_cache(&blob, out_state, prog, num_tokens);
break;
}
case MESA_SHADER_FRAGMENT: {
struct st_fragment_program *stfp = (struct st_fragment_program *) prog;
sha1 = stfp->sha1;
write_tgsi_to_cache(&blob, &stfp->tgsi, st, sha1, num_tokens);
write_tgsi_to_cache(&blob, &stfp->tgsi, prog, num_tokens);
break;
}
case MESA_SHADER_COMPUTE: {
struct st_compute_program *stcp = (struct st_compute_program *) prog;
sha1 = stcp->sha1;
write_tgsi_to_cache(&blob, out_state, st, sha1, num_tokens);
write_tgsi_to_cache(&blob, out_state, prog, num_tokens);
break;
}
default:
@ -116,10 +108,8 @@ st_store_tgsi_in_disk_cache(struct st_context *st, struct gl_program *prog,
}
if (st->ctx->_Shader->Flags & GLSL_CACHE_INFO) {
char sha1_buf[41];
_mesa_sha1_format(sha1_buf, sha1);
fprintf(stderr, "putting %s tgsi_tokens in cache: %s\n",
_mesa_shader_stage_to_string(prog->info.stage), sha1_buf);
fprintf(stderr, "putting %s tgsi_tokens in cache\n",
_mesa_shader_stage_to_string(prog->info.stage));
}
blob_finish(&blob);
@ -150,254 +140,153 @@ st_load_tgsi_from_disk_cache(struct gl_context *ctx,
if (!ctx->Cache)
return false;
unsigned char *stage_sha1[MESA_SHADER_STAGES];
char sha1_buf[41];
/* Compute and store sha1 for each stage. These will be reused by the
* cache store pass if we fail to find the cached tgsi.
*/
for (unsigned i = 0; i < MESA_SHADER_STAGES; i++) {
if (prog->_LinkedShaders[i] == NULL)
continue;
char *buf = ralloc_strdup(NULL, "tgsi_tokens ");
_mesa_sha1_format(sha1_buf,
prog->_LinkedShaders[i]->Program->sh.data->sha1);
ralloc_strcat(&buf, sha1_buf);
struct gl_program *glprog = prog->_LinkedShaders[i]->Program;
switch (glprog->info.stage) {
case MESA_SHADER_VERTEX: {
struct st_vertex_program *stvp = (struct st_vertex_program *) glprog;
stage_sha1[i] = stvp->sha1;
ralloc_strcat(&buf, " vs");
disk_cache_compute_key(ctx->Cache, buf, strlen(buf), stage_sha1[i]);
break;
}
case MESA_SHADER_TESS_CTRL: {
struct st_common_program *stcp = st_common_program(glprog);
stage_sha1[i] = stcp->sha1;
ralloc_strcat(&buf, " tcs");
disk_cache_compute_key(ctx->Cache, buf, strlen(buf), stage_sha1[i]);
break;
}
case MESA_SHADER_TESS_EVAL: {
struct st_common_program *step = st_common_program(glprog);
stage_sha1[i] = step->sha1;
ralloc_strcat(&buf, " tes");
disk_cache_compute_key(ctx->Cache, buf, strlen(buf), stage_sha1[i]);
break;
}
case MESA_SHADER_GEOMETRY: {
struct st_common_program *stgp = st_common_program(glprog);
stage_sha1[i] = stgp->sha1;
ralloc_strcat(&buf, " gs");
disk_cache_compute_key(ctx->Cache, buf, strlen(buf), stage_sha1[i]);
break;
}
case MESA_SHADER_FRAGMENT: {
struct st_fragment_program *stfp =
(struct st_fragment_program *) glprog;
stage_sha1[i] = stfp->sha1;
ralloc_strcat(&buf, " fs");
disk_cache_compute_key(ctx->Cache, buf, strlen(buf), stage_sha1[i]);
break;
}
case MESA_SHADER_COMPUTE: {
struct st_compute_program *stcp =
(struct st_compute_program *) glprog;
stage_sha1[i] = stcp->sha1;
ralloc_strcat(&buf, " cs");
disk_cache_compute_key(ctx->Cache, buf, strlen(buf), stage_sha1[i]);
break;
}
default:
unreachable("Unsupported stage");
}
ralloc_free(buf);
}
/* Now that we have created the sha1 keys that will be used for writting to
* the tgsi cache fallback to the regular glsl to tgsi path if we didn't
* load the GLSL IR from cache. We do this as glsl to tgsi can alter things
* such as gl_program_parameter_list which holds things like uniforms.
/* If we didn't load the GLSL metadata from cache then we could not have
* loaded the tgsi either.
*/
if (prog->data->LinkStatus != linking_skipped)
return false;
uint8_t *buffer = NULL;
if (ctx->_Shader->Flags & GLSL_CACHE_FALLBACK) {
goto fallback_recompile;
}
struct st_context *st = st_context(ctx);
for (unsigned i = 0; i < MESA_SHADER_STAGES; i++) {
if (prog->_LinkedShaders[i] == NULL)
continue;
unsigned char *sha1 = stage_sha1[i];
size_t size;
buffer = (uint8_t *) disk_cache_get(ctx->Cache, sha1, &size);
if (buffer) {
struct blob_reader blob_reader;
blob_reader_init(&blob_reader, buffer, size);
struct gl_program *glprog = prog->_LinkedShaders[i]->Program;
struct gl_program *glprog = prog->_LinkedShaders[i]->Program;
switch (glprog->info.stage) {
case MESA_SHADER_VERTEX: {
struct st_vertex_program *stvp =
(struct st_vertex_program *) glprog;
size_t size = glprog->driver_cache_blob_size;
uint8_t *buffer = (uint8_t *) glprog->driver_cache_blob;
st_release_vp_variants(st, stvp);
struct blob_reader blob_reader;
blob_reader_init(&blob_reader, buffer, size);
stvp->num_inputs = blob_read_uint32(&blob_reader);
blob_copy_bytes(&blob_reader, (uint8_t *) stvp->index_to_input,
sizeof(stvp->index_to_input));
blob_copy_bytes(&blob_reader, (uint8_t *) stvp->result_to_output,
sizeof(stvp->result_to_output));
switch (glprog->info.stage) {
case MESA_SHADER_VERTEX: {
struct st_vertex_program *stvp = (struct st_vertex_program *) glprog;
read_stream_out_from_cache(&blob_reader, &stvp->tgsi);
read_tgsi_from_cache(&blob_reader, &stvp->tgsi.tokens);
st_release_vp_variants(st, stvp);
if (st->vp == stvp)
st->dirty |= ST_NEW_VERTEX_PROGRAM(st, stvp);
stvp->num_inputs = blob_read_uint32(&blob_reader);
blob_copy_bytes(&blob_reader, (uint8_t *) stvp->index_to_input,
sizeof(stvp->index_to_input));
blob_copy_bytes(&blob_reader, (uint8_t *) stvp->result_to_output,
sizeof(stvp->result_to_output));
break;
}
case MESA_SHADER_TESS_CTRL: {
struct st_common_program *sttcp = st_common_program(glprog);
read_stream_out_from_cache(&blob_reader, &stvp->tgsi);
read_tgsi_from_cache(&blob_reader, &stvp->tgsi.tokens);
st_release_basic_variants(st, sttcp->Base.Target,
&sttcp->variants, &sttcp->tgsi);
if (st->vp == stvp)
st->dirty |= ST_NEW_VERTEX_PROGRAM(st, stvp);
read_stream_out_from_cache(&blob_reader, &sttcp->tgsi);
read_tgsi_from_cache(&blob_reader, &sttcp->tgsi.tokens);
if (st->tcp == sttcp)
st->dirty |= sttcp->affected_states;
break;
}
case MESA_SHADER_TESS_EVAL: {
struct st_common_program *sttep = st_common_program(glprog);
st_release_basic_variants(st, sttep->Base.Target,
&sttep->variants, &sttep->tgsi);
read_stream_out_from_cache(&blob_reader, &sttep->tgsi);
read_tgsi_from_cache(&blob_reader, &sttep->tgsi.tokens);
if (st->tep == sttep)
st->dirty |= sttep->affected_states;
break;
}
case MESA_SHADER_GEOMETRY: {
struct st_common_program *stgp = st_common_program(glprog);
st_release_basic_variants(st, stgp->Base.Target, &stgp->variants,
&stgp->tgsi);
read_stream_out_from_cache(&blob_reader, &stgp->tgsi);
read_tgsi_from_cache(&blob_reader, &stgp->tgsi.tokens);
if (st->gp == stgp)
st->dirty |= stgp->affected_states;
break;
}
case MESA_SHADER_FRAGMENT: {
struct st_fragment_program *stfp =
(struct st_fragment_program *) glprog;
st_release_fp_variants(st, stfp);
read_tgsi_from_cache(&blob_reader, &stfp->tgsi.tokens);
if (st->fp == stfp)
st->dirty |= stfp->affected_states;
break;
}
case MESA_SHADER_COMPUTE: {
struct st_compute_program *stcp =
(struct st_compute_program *) glprog;
st_release_cp_variants(st, stcp);
read_tgsi_from_cache(&blob_reader,
(const struct tgsi_token**) &stcp->tgsi.prog);
stcp->tgsi.req_local_mem = stcp->Base.info.cs.shared_size;
stcp->tgsi.req_private_mem = 0;
stcp->tgsi.req_input_mem = 0;
if (st->cp == stcp)
st->dirty |= stcp->affected_states;
break;
}
default:
unreachable("Unsupported stage");
}
if (blob_reader.current != blob_reader.end || blob_reader.overrun) {
/* Something very bad has gone wrong discard the item from the
* cache and rebuild/link from source.
*/
assert(!"Invalid TGSI shader disk cache item!");
if (ctx->_Shader->Flags & GLSL_CACHE_INFO) {
fprintf(stderr, "Error reading program from cache (invalid "
"TGSI cache item)\n");
}
disk_cache_remove(ctx->Cache, sha1);
goto fallback_recompile;
}
if (ctx->_Shader->Flags & GLSL_CACHE_INFO) {
_mesa_sha1_format(sha1_buf, sha1);
fprintf(stderr, "%s tgsi_tokens retrieved from cache: %s\n",
_mesa_shader_stage_to_string(i), sha1_buf);
}
st_set_prog_affected_state_flags(glprog);
_mesa_associate_uniform_storage(ctx, prog, glprog, false);
/* Create Gallium shaders now instead of on demand. */
if (ST_DEBUG & DEBUG_PRECOMPILE ||
st->shader_has_one_variant[glprog->info.stage])
st_precompile_shader_variant(st, glprog);
free(buffer);
} else {
/* Failed to find a matching cached shader so fallback to recompile.
*/
if (ctx->_Shader->Flags & GLSL_CACHE_INFO) {
fprintf(stderr, "TGSI cache item not found.\n");
}
goto fallback_recompile;
break;
}
case MESA_SHADER_TESS_CTRL: {
struct st_common_program *sttcp = st_common_program(glprog);
st_release_basic_variants(st, sttcp->Base.Target,
&sttcp->variants, &sttcp->tgsi);
read_stream_out_from_cache(&blob_reader, &sttcp->tgsi);
read_tgsi_from_cache(&blob_reader, &sttcp->tgsi.tokens);
if (st->tcp == sttcp)
st->dirty |= sttcp->affected_states;
break;
}
case MESA_SHADER_TESS_EVAL: {
struct st_common_program *sttep = st_common_program(glprog);
st_release_basic_variants(st, sttep->Base.Target,
&sttep->variants, &sttep->tgsi);
read_stream_out_from_cache(&blob_reader, &sttep->tgsi);
read_tgsi_from_cache(&blob_reader, &sttep->tgsi.tokens);
if (st->tep == sttep)
st->dirty |= sttep->affected_states;
break;
}
case MESA_SHADER_GEOMETRY: {
struct st_common_program *stgp = st_common_program(glprog);
st_release_basic_variants(st, stgp->Base.Target, &stgp->variants,
&stgp->tgsi);
read_stream_out_from_cache(&blob_reader, &stgp->tgsi);
read_tgsi_from_cache(&blob_reader, &stgp->tgsi.tokens);
if (st->gp == stgp)
st->dirty |= stgp->affected_states;
break;
}
case MESA_SHADER_FRAGMENT: {
struct st_fragment_program *stfp =
(struct st_fragment_program *) glprog;
st_release_fp_variants(st, stfp);
read_tgsi_from_cache(&blob_reader, &stfp->tgsi.tokens);
if (st->fp == stfp)
st->dirty |= stfp->affected_states;
break;
}
case MESA_SHADER_COMPUTE: {
struct st_compute_program *stcp =
(struct st_compute_program *) glprog;
st_release_cp_variants(st, stcp);
read_tgsi_from_cache(&blob_reader,
(const struct tgsi_token**) &stcp->tgsi.prog);
stcp->tgsi.req_local_mem = stcp->Base.info.cs.shared_size;
stcp->tgsi.req_private_mem = 0;
stcp->tgsi.req_input_mem = 0;
if (st->cp == stcp)
st->dirty |= stcp->affected_states;
break;
}
default:
unreachable("Unsupported stage");
}
/* Make sure we don't try to read more data than we wrote. This should
* never happen in release builds but its useful to have this check to
* catch development bugs.
*/
if (blob_reader.current != blob_reader.end || blob_reader.overrun) {
assert(!"Invalid TGSI shader disk cache item!");
if (ctx->_Shader->Flags & GLSL_CACHE_INFO) {
fprintf(stderr, "Error reading program from cache (invalid "
"TGSI cache item)\n");
}
}
if (ctx->_Shader->Flags & GLSL_CACHE_INFO) {
fprintf(stderr, "%s tgsi_tokens retrieved from cache\n",
_mesa_shader_stage_to_string(i));
}
st_set_prog_affected_state_flags(glprog);
_mesa_associate_uniform_storage(ctx, prog, glprog, false);
/* Create Gallium shaders now instead of on demand. */
if (ST_DEBUG & DEBUG_PRECOMPILE ||
st->shader_has_one_variant[glprog->info.stage])
st_precompile_shader_variant(st, glprog);
/* We don't need the cached blob anymore so free it */
ralloc_free(glprog->driver_cache_blob);
glprog->driver_cache_blob = NULL;
glprog->driver_cache_blob_size = 0;
}
return true;
fallback_recompile:
free(buffer);
if (ctx->_Shader->Flags & GLSL_CACHE_INFO)
fprintf(stderr, "TGSI cache falling back to recompile.\n");
for (unsigned i = 0; i < prog->NumShaders; i++) {
_mesa_glsl_compile_shader(ctx, prog->Shaders[i], false, false, true);
}
prog->data->skip_cache = true;
_mesa_glsl_link_shader(ctx, prog);
return true;
}