You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
virglrenderer/src/vrend_renderer.c

7212 lines
232 KiB

/**************************************************************************
*
* Copyright (C) 2014 Red Hat Inc.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and associated documentation files (the "Software"),
* to deal in the Software without restriction, including without limitation
* the rights to use, copy, modify, merge, publish, distribute, sublicense,
* and/or sell copies of the Software, and to permit persons to whom the
* Software is furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included
* in all copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
* OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
* THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR
* OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
* ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
* OTHER DEALINGS IN THE SOFTWARE.
*
**************************************************************************/
renderer: use a thread to block for fences. Instead of polling the fences regularly, have a thread that blocks for a single fence using a separate shared context, then uses eventfd to wake up the main thread when something happens. Inside the guest, glmark2 typicially runs twice as fast with the thread sync. Although in general, the performances seems to be about +30%. The benefits is mostly for CPU-bounds tasks (when main the thread hits 100%) A naive perf stat of the vtest renderer with glmark2 "build" test with a fixed number of frames (500) results in the following stats data: (do not value timing related informations, since the renderer is ran and stopped manually) without thread: 3032.282265 task-clock (msec) # 0.420 CPUs utilized 4,277 context-switches # 0.001 M/sec 102 cpu-migrations # 0.034 K/sec 9,020 page-faults # 0.003 M/sec 7,884,098,254 cycles # 2.600 GHz 4,440,126,451 stalled-cycles-frontend # 56.32% frontend cycles idle <not supported> stalled-cycles-backend 11,024,091,578 instructions # 1.40 insns per cycle # 0.40 stalled # cycles per insn 1,091,831,588 branches # 360.069 M/sec 5,426,846 branch-misses # 0.50% of all branches with thread: 3403.592921 task-clock (msec) # 0.452 CPUs utilized 7,145 context-switches # 0.002 M/sec 410 cpu-migrations # 0.120 K/sec 6,191 page-faults # 0.002 M/sec 7,475,038,064 cycles # 2.196 GHz 4,487,043,071 stalled-cycles-frontend # 60.03% frontend cycles idle <not supported> stalled-cycles-backend 9,925,205,494 instructions # 1.33 insns per cycle # 0.45 stalled # cycles per insn 834,375,503 branches # 245.146 M/sec 4,919,995 branch-misses # 0.59% of all branches Signed-off-by: Marc-André Lureau <marcandre.lureau@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
9 years ago
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
#include <epoxy/gl.h>
#include <unistd.h>
#include <stdio.h>
#include <errno.h>
#include "pipe/p_shader_tokens.h"
#include "pipe/p_context.h"
#include "pipe/p_defines.h"
#include "pipe/p_screen.h"
#include "pipe/p_state.h"
#include "util/u_inlines.h"
#include "util/u_memory.h"
#include "util/u_dual_blend.h"
renderer: use a thread to block for fences. Instead of polling the fences regularly, have a thread that blocks for a single fence using a separate shared context, then uses eventfd to wake up the main thread when something happens. Inside the guest, glmark2 typicially runs twice as fast with the thread sync. Although in general, the performances seems to be about +30%. The benefits is mostly for CPU-bounds tasks (when main the thread hits 100%) A naive perf stat of the vtest renderer with glmark2 "build" test with a fixed number of frames (500) results in the following stats data: (do not value timing related informations, since the renderer is ran and stopped manually) without thread: 3032.282265 task-clock (msec) # 0.420 CPUs utilized 4,277 context-switches # 0.001 M/sec 102 cpu-migrations # 0.034 K/sec 9,020 page-faults # 0.003 M/sec 7,884,098,254 cycles # 2.600 GHz 4,440,126,451 stalled-cycles-frontend # 56.32% frontend cycles idle <not supported> stalled-cycles-backend 11,024,091,578 instructions # 1.40 insns per cycle # 0.40 stalled # cycles per insn 1,091,831,588 branches # 360.069 M/sec 5,426,846 branch-misses # 0.50% of all branches with thread: 3403.592921 task-clock (msec) # 0.452 CPUs utilized 7,145 context-switches # 0.002 M/sec 410 cpu-migrations # 0.120 K/sec 6,191 page-faults # 0.002 M/sec 7,475,038,064 cycles # 2.196 GHz 4,487,043,071 stalled-cycles-frontend # 60.03% frontend cycles idle <not supported> stalled-cycles-backend 9,925,205,494 instructions # 1.33 insns per cycle # 0.45 stalled # cycles per insn 834,375,503 branches # 245.146 M/sec 4,919,995 branch-misses # 0.59% of all branches Signed-off-by: Marc-André Lureau <marcandre.lureau@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
9 years ago
#include "os/os_thread.h"
#include "util/u_double_list.h"
#include "util/u_format.h"
#include "tgsi/tgsi_parse.h"
#include "vrend_object.h"
#include "vrend_shader.h"
#include "vrend_renderer.h"
#include "virgl_hw.h"
#include "tgsi/tgsi_text.h"
renderer: use a thread to block for fences. Instead of polling the fences regularly, have a thread that blocks for a single fence using a separate shared context, then uses eventfd to wake up the main thread when something happens. Inside the guest, glmark2 typicially runs twice as fast with the thread sync. Although in general, the performances seems to be about +30%. The benefits is mostly for CPU-bounds tasks (when main the thread hits 100%) A naive perf stat of the vtest renderer with glmark2 "build" test with a fixed number of frames (500) results in the following stats data: (do not value timing related informations, since the renderer is ran and stopped manually) without thread: 3032.282265 task-clock (msec) # 0.420 CPUs utilized 4,277 context-switches # 0.001 M/sec 102 cpu-migrations # 0.034 K/sec 9,020 page-faults # 0.003 M/sec 7,884,098,254 cycles # 2.600 GHz 4,440,126,451 stalled-cycles-frontend # 56.32% frontend cycles idle <not supported> stalled-cycles-backend 11,024,091,578 instructions # 1.40 insns per cycle # 0.40 stalled # cycles per insn 1,091,831,588 branches # 360.069 M/sec 5,426,846 branch-misses # 0.50% of all branches with thread: 3403.592921 task-clock (msec) # 0.452 CPUs utilized 7,145 context-switches # 0.002 M/sec 410 cpu-migrations # 0.120 K/sec 6,191 page-faults # 0.002 M/sec 7,475,038,064 cycles # 2.196 GHz 4,487,043,071 stalled-cycles-frontend # 60.03% frontend cycles idle <not supported> stalled-cycles-backend 9,925,205,494 instructions # 1.33 insns per cycle # 0.45 stalled # cycles per insn 834,375,503 branches # 245.146 M/sec 4,919,995 branch-misses # 0.59% of all branches Signed-off-by: Marc-André Lureau <marcandre.lureau@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
9 years ago
#ifdef HAVE_EVENTFD
#include <sys/eventfd.h>
#endif
/* debugging aid to dump shaders */
int vrend_dump_shaders;
/* debugging via KHR_debug extension */
int vrend_use_debug_cb = 0;
struct vrend_if_cbs *vrend_clicbs;
struct vrend_fence {
uint32_t fence_id;
uint32_t ctx_id;
GLsync syncobj;
struct list_head fences;
};
struct vrend_query {
struct list_head waiting_queries;
GLuint id;
GLuint type;
GLuint index;
GLuint gltype;
int ctx_id;
struct vrend_resource *res;
uint64_t current_total;
};
struct global_error_state {
enum virgl_errors last_error;
};
struct global_renderer_state {
int gl_major_ver;
int gl_minor_ver;
struct vrend_context *current_ctx;
struct vrend_context *current_hw_ctx;
struct list_head waiting_query_list;
bool inited;
bool use_gles;
bool use_core_profile;
bool have_debug_cb;
bool have_mesa_invert;
bool have_samplers;
bool have_gles_khr_robustness;
bool have_arb_robustness;
bool have_arb_or_gles_ext_texture_buffer;
bool have_multisample;
bool have_ms_scaled_blit;
bool have_nv_prim_restart;
bool have_gl_prim_restart;
bool have_bit_encoding;
bool have_vertex_attrib_binding;
bool have_tf2;
bool have_stencil_texturing;
bool have_sample_shading;
/* these appeared broken on at least one driver */
bool use_explicit_locations;
uint32_t max_uniform_blocks;
struct list_head active_ctx_list;
renderer: use a thread to block for fences. Instead of polling the fences regularly, have a thread that blocks for a single fence using a separate shared context, then uses eventfd to wake up the main thread when something happens. Inside the guest, glmark2 typicially runs twice as fast with the thread sync. Although in general, the performances seems to be about +30%. The benefits is mostly for CPU-bounds tasks (when main the thread hits 100%) A naive perf stat of the vtest renderer with glmark2 "build" test with a fixed number of frames (500) results in the following stats data: (do not value timing related informations, since the renderer is ran and stopped manually) without thread: 3032.282265 task-clock (msec) # 0.420 CPUs utilized 4,277 context-switches # 0.001 M/sec 102 cpu-migrations # 0.034 K/sec 9,020 page-faults # 0.003 M/sec 7,884,098,254 cycles # 2.600 GHz 4,440,126,451 stalled-cycles-frontend # 56.32% frontend cycles idle <not supported> stalled-cycles-backend 11,024,091,578 instructions # 1.40 insns per cycle # 0.40 stalled # cycles per insn 1,091,831,588 branches # 360.069 M/sec 5,426,846 branch-misses # 0.50% of all branches with thread: 3403.592921 task-clock (msec) # 0.452 CPUs utilized 7,145 context-switches # 0.002 M/sec 410 cpu-migrations # 0.120 K/sec 6,191 page-faults # 0.002 M/sec 7,475,038,064 cycles # 2.196 GHz 4,487,043,071 stalled-cycles-frontend # 60.03% frontend cycles idle <not supported> stalled-cycles-backend 9,925,205,494 instructions # 1.33 insns per cycle # 0.45 stalled # cycles per insn 834,375,503 branches # 245.146 M/sec 4,919,995 branch-misses # 0.59% of all branches Signed-off-by: Marc-André Lureau <marcandre.lureau@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
9 years ago
/* threaded sync */
bool stop_sync_thread;
int eventfd;
pipe_mutex fence_mutex;
struct list_head fence_list;
struct list_head fence_wait_list;
pipe_condvar fence_cond;
pipe_thread sync_thread;
virgl_gl_context sync_context;
};
static struct global_renderer_state vrend_state;
struct vrend_linked_shader_program {
struct list_head head;
struct list_head sl[PIPE_SHADER_TYPES];
GLuint id;
bool dual_src_linked;
struct vrend_shader *ss[PIPE_SHADER_TYPES];
uint32_t samplers_used_mask[PIPE_SHADER_TYPES];
GLuint *samp_locs[PIPE_SHADER_TYPES];
GLuint *shadow_samp_mask_locs[PIPE_SHADER_TYPES];
GLuint *shadow_samp_add_locs[PIPE_SHADER_TYPES];
GLuint *const_locs[PIPE_SHADER_TYPES];
GLuint *attrib_locs;
uint32_t shadow_samp_mask[PIPE_SHADER_TYPES];
GLuint *ubo_locs[PIPE_SHADER_TYPES];
GLuint vs_ws_adjust_loc;
GLuint fs_stipple_loc;
GLuint clip_locs[8];
};
struct vrend_shader {
struct vrend_shader *next_variant;
struct vrend_shader_selector *sel;
GLchar *glsl_prog;
GLuint id;
GLuint compiled_fs_id;
struct vrend_shader_key key;
struct list_head programs;
};
struct vrend_shader_selector {
struct pipe_reference reference;
unsigned num_shaders;
unsigned type;
struct vrend_shader_info sinfo;
struct vrend_shader *current;
struct tgsi_token *tokens;
char *tmp_buf;
uint32_t buf_len;
uint32_t buf_offset;
};
struct vrend_texture {
struct vrend_resource base;
struct pipe_sampler_state state;
GLenum cur_swizzle_r;
GLenum cur_swizzle_g;
GLenum cur_swizzle_b;
GLenum cur_swizzle_a;
GLuint srgb_decode;
};
struct vrend_surface {
struct pipe_reference reference;
GLuint id;
GLuint res_handle;
GLuint format;
GLuint val0, val1;
struct vrend_resource *texture;
};
struct vrend_sampler_state {
struct pipe_sampler_state base;
GLuint id;
};
struct vrend_so_target {
struct pipe_reference reference;
GLuint res_handle;
unsigned buffer_offset;
unsigned buffer_size;
struct vrend_resource *buffer;
struct vrend_sub_context *sub_ctx;
};
struct vrend_sampler_view {
struct pipe_reference reference;
GLuint id;
GLuint format;
GLuint val0, val1;
GLuint gl_swizzle_r;
GLuint gl_swizzle_g;
GLuint gl_swizzle_b;
GLuint gl_swizzle_a;
GLuint cur_base, cur_max;
GLenum depth_texture_mode;
GLuint srgb_decode;
struct vrend_resource *texture;
};
struct vrend_vertex_element {
struct pipe_vertex_element base;
GLenum type;
GLboolean norm;
GLuint nr_chan;
};
struct vrend_vertex_element_array {
unsigned count;
struct vrend_vertex_element elements[PIPE_MAX_ATTRIBS];
GLuint id;
};
struct vrend_constants {
unsigned int *consts;
uint32_t num_consts;
};
struct vrend_shader_view {
int num_views;
struct vrend_sampler_view *views[PIPE_MAX_SHADER_SAMPLER_VIEWS];
uint32_t res_id[PIPE_MAX_SHADER_SAMPLER_VIEWS];
uint32_t old_ids[PIPE_MAX_SHADER_SAMPLER_VIEWS];
};
struct vrend_viewport {
GLint cur_x, cur_y;
GLsizei width, height;
GLclampd near_val, far_val;
};
/* create a streamout object to support pause/resume */
struct vrend_streamout_object {
GLuint id;
uint32_t num_targets;
uint32_t handles[16];
struct list_head head;
int xfb_state;
struct vrend_so_target *so_targets[16];
};
#define XFB_STATE_OFF 0
#define XFB_STATE_STARTED_NEED_BEGIN 1
#define XFB_STATE_STARTED 2
#define XFB_STATE_PAUSED 3
struct vrend_sub_context {
struct list_head head;
virgl_gl_context gl_context;
int sub_ctx_id;
GLuint vaoid;
uint32_t enabled_attribs_bitmask;
struct list_head programs;
struct util_hash_table *object_hash;
struct vrend_vertex_element_array *ve;
int num_vbos;
int old_num_vbos; /* for cleaning up */
struct pipe_vertex_buffer vbo[PIPE_MAX_ATTRIBS];
uint32_t vbo_res_ids[PIPE_MAX_ATTRIBS];
struct pipe_index_buffer ib;
uint32_t index_buffer_res_id;
bool vbo_dirty;
bool shader_dirty;
bool sampler_state_dirty;
bool stencil_state_dirty;
uint32_t long_shader_in_progress_handle[PIPE_SHADER_TYPES];
struct vrend_shader_selector *shaders[PIPE_SHADER_TYPES];
struct vrend_linked_shader_program *prog;
int prog_ids[PIPE_SHADER_TYPES];
struct vrend_shader_view views[PIPE_SHADER_TYPES];
struct vrend_constants consts[PIPE_SHADER_TYPES];
bool const_dirty[PIPE_SHADER_TYPES];
struct vrend_sampler_state *sampler_state[PIPE_SHADER_TYPES][PIPE_MAX_SAMPLERS];
struct pipe_constant_buffer cbs[PIPE_SHADER_TYPES][PIPE_MAX_CONSTANT_BUFFERS];
uint32_t const_bufs_used_mask[PIPE_SHADER_TYPES];
int num_sampler_states[PIPE_SHADER_TYPES];
uint32_t fb_id;
int nr_cbufs, old_nr_cbufs;
struct vrend_surface *zsurf;
struct vrend_surface *surf[PIPE_MAX_COLOR_BUFS];
struct vrend_viewport vps[PIPE_MAX_VIEWPORTS];
float depth_transform, depth_scale;
/* viewport is negative */
uint32_t scissor_state_dirty;
uint32_t viewport_state_dirty;
uint32_t fb_height;
struct pipe_scissor_state ss[PIPE_MAX_VIEWPORTS];
struct pipe_blend_state blend_state;
struct pipe_depth_stencil_alpha_state dsa_state;
struct pipe_rasterizer_state rs_state;
uint8_t stencil_refs[2];
bool viewport_is_negative;
/* this is set if the contents of the FBO look upside down when viewed
with 0,0 as the bottom corner */
bool inverted_fbo_content;
GLuint blit_fb_ids[2];
struct pipe_depth_stencil_alpha_state *dsa;
struct pipe_clip_state ucp_state;
bool blend_enabled;
bool depth_test_enabled;
bool alpha_test_enabled;
bool stencil_test_enabled;
GLuint program_id;
int last_shader_idx;
struct pipe_rasterizer_state hw_rs_state;
struct pipe_blend_state hw_blend_state;
struct list_head streamout_list;
struct vrend_streamout_object *current_so;
struct pipe_blend_color blend_color;
uint32_t cond_render_q_id;
GLenum cond_render_gl_mode;
};
struct vrend_context {
char debug_name[64];
struct list_head sub_ctxs;
struct vrend_sub_context *sub;
struct vrend_sub_context *sub0;
int ctx_id;
/* has this ctx gotten an error? */
bool in_error;
bool ctx_switch_pending;
bool pstip_inited;
GLuint pstipple_tex_id;
enum virgl_ctx_errors last_error;
/* resource bounds to this context */
struct util_hash_table *res_hash;
struct list_head active_nontimer_query_list;
struct list_head ctx_entry;
struct vrend_shader_cfg shader_cfg;
};
static struct vrend_resource *vrend_renderer_ctx_res_lookup(struct vrend_context *ctx, int res_handle);
static void vrend_pause_render_condition(struct vrend_context *ctx, bool pause);
static void vrend_update_viewport_state(struct vrend_context *ctx);
static void vrend_update_scissor_state(struct vrend_context *ctx);
static void vrend_destroy_query_object(void *obj_ptr);
static void vrend_finish_context_switch(struct vrend_context *ctx);
static void vrend_patch_blend_state(struct vrend_context *ctx);
static void vrend_update_frontface_state(struct vrend_context *ctx);
static void vrender_get_glsl_version(int *glsl_version);
static void vrend_destroy_resource_object(void *obj_ptr);
static void vrend_renderer_detach_res_ctx_p(struct vrend_context *ctx, int res_handle);
static void vrend_destroy_program(struct vrend_linked_shader_program *ent);
static void vrend_apply_sampler_state(struct vrend_context *ctx,
struct vrend_resource *res,
uint32_t shader_type,
int id, int sampler_id, uint32_t srgb_decode);
void vrend_update_stencil_state(struct vrend_context *ctx);
static struct vrend_format_table tex_conv_table[VIRGL_FORMAT_MAX];
static inline bool vrend_format_can_sample(enum virgl_formats format)
{
return tex_conv_table[format].bindings & VREND_BIND_SAMPLER;
}
static inline bool vrend_format_can_render(enum virgl_formats format)
{
return tex_conv_table[format].bindings & VREND_BIND_RENDER;
}
static inline bool vrend_format_is_ds(enum virgl_formats format)
{
return tex_conv_table[format].bindings & VREND_BIND_DEPTHSTENCIL;
}
bool vrend_is_ds_format(enum virgl_formats format)
{
return vrend_format_is_ds(format);
}
bool vrend_format_is_emulated_alpha(enum virgl_formats format)
{
if (!vrend_state.use_core_profile)
return false;
return (format == VIRGL_FORMAT_A8_UNORM ||
format == VIRGL_FORMAT_A16_UNORM);
}
static inline const char *pipe_shader_to_prefix(int shader_type)
{
switch (shader_type) {
case PIPE_SHADER_VERTEX: return "vs";
case PIPE_SHADER_FRAGMENT: return "fs";
case PIPE_SHADER_GEOMETRY: return "gs";
default:
return NULL;
};
}
static const char *vrend_ctx_error_strings[] = { "None", "Unknown", "Illegal shader", "Illegal handle", "Illegal resource", "Illegal surface", "Illegal vertex format", "Illegal command buffer" };
static void __report_context_error(const char *fname, struct vrend_context *ctx, enum virgl_ctx_errors error, uint32_t value)
{
ctx->in_error = true;
ctx->last_error = error;
fprintf(stderr,"%s: context error reported %d \"%s\" %s %d\n", fname, ctx->ctx_id, ctx->debug_name, vrend_ctx_error_strings[error], value);
}
#define report_context_error(ctx, error, value) __report_context_error(__func__, ctx, error, value)
void vrend_report_buffer_error(struct vrend_context *ctx, int cmd)
{
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_CMD_BUFFER, cmd);
}
#define CORE_PROFILE_WARN_NONE 0
#define CORE_PROFILE_WARN_STIPPLE 1
#define CORE_PROFILE_WARN_POLYGON_MODE 2
#define CORE_PROFILE_WARN_TWO_SIDE 3
#define CORE_PROFILE_WARN_CLAMP 4
#define CORE_PROFILE_WARN_SHADE_MODEL 5
static const char *vrend_core_profile_warn_strings[] = { "None", "Stipple", "Polygon Mode", "Two Side", "Clamping", "Shade Model" };
static void __report_core_warn(const char *fname, struct vrend_context *ctx, enum virgl_ctx_errors error, uint32_t value)
{
fprintf(stderr,"%s: core profile violation reported %d \"%s\" %s %d\n", fname, ctx->ctx_id, ctx->debug_name, vrend_core_profile_warn_strings[error], value);
}
#define report_core_warn(ctx, error, value) __report_core_warn(__func__, ctx, error, value)
#define GLES_WARN_NONE 0
#define GLES_WARN_STIPPLE 1
#define GLES_WARN_POLYGON_MODE 2
#define GLES_WARN_DEPTH_RANGE 3
#define GLES_WARN_POINT_SIZE 4
#define GLES_WARN_LOD_BIAS 5
#define GLES_WARN_SRGB_FB 6
#define GLES_WARN_TEXTURE_RECT 7
#define GLES_WARN_OFFSET_LINE 8
#define GLES_WARN_OFFSET_POINT 9
#define GLES_WARN_DEPTH_CLIP 10
#define GLES_WARN_FLATSHADE_FIRST 11
#define GLES_WARN_LINE_SMOOTH 12
#define GLES_WARN_POLY_SMOOTH 13
#define GLES_WARN_DEPTH_CLEAR 14
#define GLES_WARN_LOGIC_OP 15
#define GLES_WARN_TIMESTAMP 16
static const char *vrend_gles_warn_strings[] = {
"None", "Stipple", "Polygon Mode", "Depth Range", "Point Size", "Lod Bias",
"SRGB Framebuffer", "Texture Rect", "Offset Line", "Offset Point",
"Depth Clip", "Flatshade First", "Line Smooth", "Poly Smooth",
"Depth Clear", "LogicOp", "GL_TIMESTAMP"
};
static void __report_gles_warn(const char *fname, struct vrend_context *ctx, enum virgl_ctx_errors error, uint32_t value)
{
int id = ctx ? ctx->ctx_id : -1;
const char *name = ctx ? ctx->debug_name : "NO_CONTEXT";
fprintf(stderr,"%s: gles violation reported %d \"%s\" %s %d\n", fname, id, ctx->debug_name, vrend_gles_warn_strings[error], value);
}
#define report_gles_warn(ctx, error, value) __report_gles_warn(__func__, ctx, error, value)
static void __report_gles_missing_func(const char *fname, struct vrend_context *ctx, const char *missf)
{
int id = ctx ? ctx->ctx_id : -1;
const char *name = ctx ? ctx->debug_name : "NO_CONTEXT";
fprintf(stderr,"%s: gles violation reported %d \"%s\" %s is missing\n", fname, id, ctx->debug_name, missf);
}
#define report_gles_missing_func(ctx, missf) __report_gles_missing_func(__func__, ctx, missf)
static inline bool should_invert_viewport(struct vrend_context *ctx)
{
/* if we have a negative viewport then gallium wanted to invert it,
however since we are rendering to GL FBOs we need to invert it
again unless we are rendering upside down already
- confused?
so if gallium asks for a negative viewport */
return !(ctx->sub->viewport_is_negative ^ ctx->sub->inverted_fbo_content);
}
static void vrend_destroy_surface(struct vrend_surface *surf)
{
vrend_resource_reference(&surf->texture, NULL);
free(surf);
}
static inline void
vrend_surface_reference(struct vrend_surface **ptr, struct vrend_surface *surf)
{
struct vrend_surface *old_surf = *ptr;
if (pipe_reference(&(*ptr)->reference, &surf->reference))
vrend_destroy_surface(old_surf);
*ptr = surf;
}
static void vrend_destroy_sampler_view(struct vrend_sampler_view *samp)
{
vrend_resource_reference(&samp->texture, NULL);
free(samp);
}
static inline void
vrend_sampler_view_reference(struct vrend_sampler_view **ptr, struct vrend_sampler_view *view)
{
struct vrend_sampler_view *old_view = *ptr;
if (pipe_reference(&(*ptr)->reference, &view->reference))
vrend_destroy_sampler_view(old_view);
*ptr = view;
}
static void vrend_destroy_so_target(struct vrend_so_target *target)
{
vrend_resource_reference(&target->buffer, NULL);
free(target);
}
static inline void
vrend_so_target_reference(struct vrend_so_target **ptr, struct vrend_so_target *target)
{
struct vrend_so_target *old_target = *ptr;
if (pipe_reference(&(*ptr)->reference, &target->reference))
vrend_destroy_so_target(old_target);
*ptr = target;
}
static void vrend_shader_destroy(struct vrend_shader *shader)
{
struct vrend_linked_shader_program *ent, *tmp;
LIST_FOR_EACH_ENTRY_SAFE(ent, tmp, &shader->programs, sl[shader->sel->type]) {
vrend_destroy_program(ent);
}
glDeleteShader(shader->id);
free(shader->glsl_prog);
free(shader);
}
static void vrend_destroy_shader_selector(struct vrend_shader_selector *sel)
{
struct vrend_shader *p = sel->current, *c;
int i;
while (p) {
c = p->next_variant;
vrend_shader_destroy(p);
p = c;
}
if (sel->sinfo.so_names)
for (i = 0; i < sel->sinfo.so_info.num_outputs; i++)
free(sel->sinfo.so_names[i]);
free(sel->tmp_buf);
free(sel->sinfo.so_names);
free(sel->sinfo.interpinfo);
free(sel->tokens);
free(sel);
}
static bool vrend_compile_shader(struct vrend_context *ctx,
struct vrend_shader *shader)
{
GLint param;
glShaderSource(shader->id, 1, (const char **)&shader->glsl_prog, NULL);
glCompileShader(shader->id);
glGetShaderiv(shader->id, GL_COMPILE_STATUS, &param);
if (param == GL_FALSE) {
char infolog[65536];
int len;
glGetShaderInfoLog(shader->id, 65536, &len, infolog);
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_SHADER, 0);
fprintf(stderr,"shader failed to compile\n%s\n", infolog);
fprintf(stderr,"GLSL:\n%s\n", shader->glsl_prog);
return false;
}
return true;
}
static inline void
vrend_shader_state_reference(struct vrend_shader_selector **ptr, struct vrend_shader_selector *shader)
{
struct vrend_shader_selector *old_shader = *ptr;
if (pipe_reference(&(*ptr)->reference, &shader->reference))
vrend_destroy_shader_selector(old_shader);
*ptr = shader;
}
void
vrend_insert_format(struct vrend_format_table *entry, uint32_t bindings)
{
tex_conv_table[entry->format] = *entry;
tex_conv_table[entry->format].bindings = bindings;
}
void
vrend_insert_format_swizzle(int override_format, struct vrend_format_table *entry, uint32_t bindings, uint8_t swizzle[4])
{
int i;
tex_conv_table[override_format] = *entry;
tex_conv_table[override_format].bindings = bindings;
tex_conv_table[override_format].flags = VREND_BIND_NEED_SWIZZLE;
for (i = 0; i < 4; i++)
tex_conv_table[override_format].swizzle[i] = swizzle[i];
}
static bool vrend_is_timer_query(GLenum gltype)
{
return gltype == GL_TIMESTAMP ||
gltype == GL_TIME_ELAPSED;
}
static void vrend_use_program(struct vrend_context *ctx, GLuint program_id)
{
if (ctx->sub->program_id != program_id) {
glUseProgram(program_id);
ctx->sub->program_id = program_id;
}
}
static void vrend_init_pstipple_texture(struct vrend_context *ctx)
{
glGenTextures(1, &ctx->pstipple_tex_id);
glBindTexture(GL_TEXTURE_2D, ctx->pstipple_tex_id);
glTexImage2D(GL_TEXTURE_2D, 0, GL_R8, 32, 32, 0, GL_RED, GL_UNSIGNED_BYTE, NULL);
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_S, GL_REPEAT);
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_T, GL_REPEAT);
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_NEAREST);
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, GL_NEAREST);
ctx->pstip_inited = true;
}
static void vrend_blend_enable(struct vrend_context *ctx, bool blend_enable)
{
if (ctx->sub->blend_enabled != blend_enable) {
ctx->sub->blend_enabled = blend_enable;
if (blend_enable)
glEnable(GL_BLEND);
else
glDisable(GL_BLEND);
}
}
static void vrend_depth_test_enable(struct vrend_context *ctx, bool depth_test_enable)
{
if (ctx->sub->depth_test_enabled != depth_test_enable) {
ctx->sub->depth_test_enabled = depth_test_enable;
if (depth_test_enable)
glEnable(GL_DEPTH_TEST);
else
glDisable(GL_DEPTH_TEST);
}
}
static void vrend_alpha_test_enable(struct vrend_context *ctx, bool alpha_test_enable)
{
if (vrend_state.use_core_profile) {
/* handled in shaders */
return;
}
if (ctx->sub->alpha_test_enabled != alpha_test_enable) {
ctx->sub->alpha_test_enabled = alpha_test_enable;
if (alpha_test_enable)
glEnable(GL_ALPHA_TEST);
else
glDisable(GL_ALPHA_TEST);
}
}
static void vrend_stencil_test_enable(struct vrend_context *ctx, bool stencil_test_enable)
{
if (ctx->sub->stencil_test_enabled != stencil_test_enable) {
ctx->sub->stencil_test_enabled = stencil_test_enable;
if (stencil_test_enable)
glEnable(GL_STENCIL_TEST);
else
glDisable(GL_STENCIL_TEST);
}
}
static void dump_stream_out(struct pipe_stream_output_info *so)
{
int i;
if (!so)
return;
printf("streamout: %d\n", so->num_outputs);
printf("strides: ");
for (i = 0; i < 4; i++)
printf("%d ", so->stride[i]);
printf("\n");
printf("outputs:\n");
for (i = 0; i < so->num_outputs; i++) {
printf("\t%d: reg: %d sc: %d, nc: %d ob: %d do: %d\n",
i,
so->output[i].register_index,
so->output[i].start_component,
so->output[i].num_components,
so->output[i].output_buffer,
so->output[i].dst_offset);
}
}
static char *get_skip_str(int *skip_val)
{
char *start_skip = NULL;
if (*skip_val < 0) {
*skip_val = 0;
return NULL;
}
if (*skip_val == 1) {
start_skip = strdup("gl_SkipComponents1");
*skip_val -= 1;
} else if (*skip_val == 2) {
start_skip = strdup("gl_SkipComponents2");
*skip_val -= 2;
} else if (*skip_val == 3) {
start_skip = strdup("gl_SkipComponents3");
*skip_val -= 3;
} else if (*skip_val >= 4) {
start_skip = strdup("gl_SkipComponents4");
*skip_val -= 4;
}
return start_skip;
}
static void set_stream_out_varyings(int prog_id, struct vrend_shader_info *sinfo)
{
struct pipe_stream_output_info *so = &sinfo->so_info;
char *varyings[PIPE_MAX_SHADER_OUTPUTS*2];
int i, j;
int n_outputs = 0;
int last_buffer = 0;
char *start_skip;
int buf_offset = 0;
int skip;
if (!so->num_outputs)
return;
if (vrend_dump_shaders)
dump_stream_out(so);
for (i = 0; i < so->num_outputs; i++) {
if (last_buffer != so->output[i].output_buffer) {
skip = so->stride[last_buffer] - buf_offset;
while (skip) {
start_skip = get_skip_str(&skip);
if (start_skip)
varyings[n_outputs++] = start_skip;
}
for (j = last_buffer; j < so->output[i].output_buffer; j++)
varyings[n_outputs++] = strdup("gl_NextBuffer");
last_buffer = so->output[i].output_buffer;
buf_offset = 0;
}
skip = so->output[i].dst_offset - buf_offset;
while (skip) {
start_skip = get_skip_str(&skip);
if (start_skip)
varyings[n_outputs++] = start_skip;
}
buf_offset = so->output[i].dst_offset;
buf_offset += so->output[i].num_components;
if (sinfo->so_names[i])
varyings[n_outputs++] = strdup(sinfo->so_names[i]);
}
skip = so->stride[last_buffer] - buf_offset;
while (skip) {
start_skip = get_skip_str(&skip);
if (start_skip)
varyings[n_outputs++] = start_skip;
}
glTransformFeedbackVaryings(prog_id, n_outputs,
(const GLchar **)varyings, GL_INTERLEAVED_ATTRIBS_EXT);
for (i = 0; i < n_outputs; i++)
if (varyings[i])
free(varyings[i]);
}
static struct vrend_linked_shader_program *add_shader_program(struct vrend_context *ctx,
struct vrend_shader *vs,
struct vrend_shader *fs,
struct vrend_shader *gs)
{
struct vrend_linked_shader_program *sprog = CALLOC_STRUCT(vrend_linked_shader_program);
char name[32];
int i;
GLuint prog_id;
GLint lret;
int id;
int last_shader;
if (!sprog)
return NULL;
/* need to rewrite VS code to add interpolation params */
if ((gs && gs->compiled_fs_id != fs->id) ||
(!gs && vs->compiled_fs_id != fs->id)) {
bool ret;
if (gs)
vrend_patch_vertex_shader_interpolants(&ctx->shader_cfg, gs->glsl_prog,
&gs->sel->sinfo,
&fs->sel->sinfo, "gso", fs->key.flatshade);
else
vrend_patch_vertex_shader_interpolants(&ctx->shader_cfg, vs->glsl_prog,
&vs->sel->sinfo,
&fs->sel->sinfo, "vso", fs->key.flatshade);
ret = vrend_compile_shader(ctx, gs ? gs : vs);
if (ret == false) {
glDeleteShader(gs ? gs->id : vs->id);
free(sprog);
return NULL;
}
if (gs)
gs->compiled_fs_id = fs->id;
else
vs->compiled_fs_id = fs->id;
}
prog_id = glCreateProgram();
glAttachShader(prog_id, vs->id);
if (gs) {
if (gs->id > 0)
glAttachShader(prog_id, gs->id);
set_stream_out_varyings(prog_id, &gs->sel->sinfo);
}
else
set_stream_out_varyings(prog_id, &vs->sel->sinfo);
glAttachShader(prog_id, fs->id);
if (fs->sel->sinfo.num_outputs > 1) {
if (util_blend_state_is_dual(&ctx->sub->blend_state, 0)) {
glBindFragDataLocationIndexed(prog_id, 0, 0, "fsout_c0");
glBindFragDataLocationIndexed(prog_id, 0, 1, "fsout_c1");
sprog->dual_src_linked = true;
} else {
glBindFragDataLocationIndexed(prog_id, 0, 0, "fsout_c0");
glBindFragDataLocationIndexed(prog_id, 1, 0, "fsout_c1");
sprog->dual_src_linked = false;
}
} else
sprog->dual_src_linked = false;
if (vrend_state.have_vertex_attrib_binding) {
uint32_t mask = vs->sel->sinfo.attrib_input_mask;
while (mask) {
i = u_bit_scan(&mask);
snprintf(name, 32, "in_%d", i);
glBindAttribLocation(prog_id, i, name);
}
}
glLinkProgram(prog_id);
glGetProgramiv(prog_id, GL_LINK_STATUS, &lret);
if (lret == GL_FALSE) {
char infolog[65536];
int len;
glGetProgramInfoLog(prog_id, 65536, &len, infolog);
fprintf(stderr,"got error linking\n%s\n", infolog);
/* dump shaders */
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_SHADER, 0);
fprintf(stderr,"vert shader: %d GLSL\n%s\n", vs->id, vs->glsl_prog);
if (gs)
fprintf(stderr,"geom shader: %d GLSL\n%s\n", gs->id, gs->glsl_prog);
fprintf(stderr,"frag shader: %d GLSL\n%s\n", fs->id, fs->glsl_prog);
glDeleteProgram(prog_id);
free(sprog);
return NULL;
}
sprog->ss[PIPE_SHADER_VERTEX] = vs;
sprog->ss[PIPE_SHADER_FRAGMENT] = fs;
sprog->ss[PIPE_SHADER_GEOMETRY] = gs;
list_add(&sprog->sl[PIPE_SHADER_VERTEX], &vs->programs);
list_add(&sprog->sl[PIPE_SHADER_FRAGMENT], &fs->programs);
if (gs)
list_add(&sprog->sl[PIPE_SHADER_GEOMETRY], &gs->programs);
last_shader = gs ? PIPE_SHADER_GEOMETRY : PIPE_SHADER_FRAGMENT;
sprog->id = prog_id;
list_addtail(&sprog->head, &ctx->sub->programs);
if (fs->key.pstipple_tex)
sprog->fs_stipple_loc = glGetUniformLocation(prog_id, "pstipple_sampler");
else
sprog->fs_stipple_loc = -1;
sprog->vs_ws_adjust_loc = glGetUniformLocation(prog_id, "winsys_adjust_y");
for (id = PIPE_SHADER_VERTEX; id <= last_shader; id++) {
if (sprog->ss[id]->sel->sinfo.samplers_used_mask) {
uint32_t mask = sprog->ss[id]->sel->sinfo.samplers_used_mask;
int nsamp = util_bitcount(sprog->ss[id]->sel->sinfo.samplers_used_mask);
int index;
sprog->shadow_samp_mask[id] = sprog->ss[id]->sel->sinfo.shadow_samp_mask;
if (sprog->ss[id]->sel->sinfo.shadow_samp_mask) {
sprog->shadow_samp_mask_locs[id] = calloc(nsamp, sizeof(uint32_t));
sprog->shadow_samp_add_locs[id] = calloc(nsamp, sizeof(uint32_t));
} else {
sprog->shadow_samp_mask_locs[id] = sprog->shadow_samp_add_locs[id] = NULL;
}
sprog->samp_locs[id] = calloc(nsamp, sizeof(uint32_t));
if (sprog->samp_locs[id]) {
const char *prefix = pipe_shader_to_prefix(id);
index = 0;
while(mask) {
i = u_bit_scan(&mask);
snprintf(name, 32, "%ssamp%d", prefix, i);
sprog->samp_locs[id][index] = glGetUniformLocation(prog_id, name);
if (sprog->ss[id]->sel->sinfo.shadow_samp_mask & (1 << i)) {
snprintf(name, 32, "%sshadmask%d", prefix, i);
sprog->shadow_samp_mask_locs[id][index] = glGetUniformLocation(prog_id, name);
snprintf(name, 32, "%sshadadd%d", prefix, i);
sprog->shadow_samp_add_locs[id][index] = glGetUniformLocation(prog_id, name);
}
index++;
}
}
} else {
sprog->samp_locs[id] = NULL;
sprog->shadow_samp_mask_locs[id] = NULL;
sprog->shadow_samp_add_locs[id] = NULL;
sprog->shadow_samp_mask[id] = 0;
}
sprog->samplers_used_mask[id] = sprog->ss[id]->sel->sinfo.samplers_used_mask;
}
for (id = PIPE_SHADER_VERTEX; id <= last_shader; id++) {
if (sprog->ss[id]->sel->sinfo.num_consts) {
sprog->const_locs[id] = calloc(sprog->ss[id]->sel->sinfo.num_consts, sizeof(uint32_t));
if (sprog->const_locs[id]) {
const char *prefix = pipe_shader_to_prefix(id);
for (i = 0; i < sprog->ss[id]->sel->sinfo.num_consts; i++) {
snprintf(name, 32, "%sconst0[%d]", prefix, i);
sprog->const_locs[id][i] = glGetUniformLocation(prog_id, name);
}
}
} else
sprog->const_locs[id] = NULL;
}
if (!vrend_state.have_vertex_attrib_binding) {
if (vs->sel->sinfo.num_inputs) {
sprog->attrib_locs = calloc(vs->sel->sinfo.num_inputs, sizeof(uint32_t));
if (sprog->attrib_locs) {
for (i = 0; i < vs->sel->sinfo.num_inputs; i++) {
snprintf(name, 32, "in_%d", i);
sprog->attrib_locs[i] = glGetAttribLocation(prog_id, name);
}
}
} else
sprog->attrib_locs = NULL;
}
for (id = PIPE_SHADER_VERTEX; id <= last_shader; id++) {
if (sprog->ss[id]->sel->sinfo.num_ubos) {
const char *prefix = pipe_shader_to_prefix(id);
sprog->ubo_locs[id] = calloc(sprog->ss[id]->sel->sinfo.num_ubos, sizeof(uint32_t));
for (i = 0; i < sprog->ss[id]->sel->sinfo.num_ubos; i++) {
snprintf(name, 32, "%subo%d", prefix, i + 1);
sprog->ubo_locs[id][i] = glGetUniformBlockIndex(prog_id, name);
}
} else
sprog->ubo_locs[id] = NULL;
}
if (vs->sel->sinfo.num_ucp) {
for (i = 0; i < vs->sel->sinfo.num_ucp; i++) {
snprintf(name, 32, "clipp[%d]", i);
sprog->clip_locs[i] = glGetUniformLocation(prog_id, name);
}
}
return sprog;
}
static struct vrend_linked_shader_program *lookup_shader_program(struct vrend_context *ctx,
GLuint vs_id, GLuint fs_id, GLuint gs_id, bool dual_src)
{
struct vrend_linked_shader_program *ent;
LIST_FOR_EACH_ENTRY(ent, &ctx->sub->programs, head) {
if (ent->dual_src_linked != dual_src)
continue;
if (ent->ss[PIPE_SHADER_VERTEX]->id == vs_id && ent->ss[PIPE_SHADER_FRAGMENT]->id == fs_id) {
if (!ent->ss[PIPE_SHADER_GEOMETRY] && gs_id == 0)
return ent;
if (ent->ss[PIPE_SHADER_GEOMETRY] && ent->ss[PIPE_SHADER_GEOMETRY]->id == gs_id)
return ent;
}
}
return NULL;
}
static void vrend_destroy_program(struct vrend_linked_shader_program *ent)
{
int i;
glDeleteProgram(ent->id);
list_del(&ent->head);
for (i = PIPE_SHADER_VERTEX; i <= PIPE_SHADER_GEOMETRY; i++) {
if (ent->ss[i])
list_del(&ent->sl[i]);
free(ent->shadow_samp_mask_locs[i]);
free(ent->shadow_samp_add_locs[i]);
free(ent->samp_locs[i]);
free(ent->const_locs[i]);
free(ent->ubo_locs[i]);
}
free(ent->attrib_locs);
free(ent);
}
static void vrend_free_programs(struct vrend_sub_context *sub)
{
struct vrend_linked_shader_program *ent, *tmp;
if (LIST_IS_EMPTY(&sub->programs))
return;
LIST_FOR_EACH_ENTRY_SAFE(ent, tmp, &sub->programs, head) {
vrend_destroy_program(ent);
}
}
static void vrend_destroy_streamout_object(struct vrend_streamout_object *obj)
{
int i;
list_del(&obj->head);
for (i = 0; i < obj->num_targets; i++)
vrend_so_target_reference(&obj->so_targets[i], NULL);
if (vrend_state.have_tf2)
glDeleteTransformFeedbacks(1, &obj->id);
FREE(obj);
}
int vrend_create_surface(struct vrend_context *ctx,
uint32_t handle,
uint32_t res_handle, uint32_t format,
uint32_t val0, uint32_t val1)
{
struct vrend_surface *surf;
struct vrend_resource *res;
uint32_t ret_handle;
if (format >= PIPE_FORMAT_COUNT) {
return EINVAL;
}
res = vrend_renderer_ctx_res_lookup(ctx, res_handle);
if (!res) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_RESOURCE, res_handle);
return EINVAL;
}
surf = CALLOC_STRUCT(vrend_surface);
if (!surf)
return ENOMEM;
surf->res_handle = res_handle;
surf->format = format;
surf->val0 = val0;
surf->val1 = val1;
pipe_reference_init(&surf->reference, 1);
vrend_resource_reference(&surf->texture, res);
ret_handle = vrend_renderer_object_insert(ctx, surf, sizeof(*surf), handle, VIRGL_OBJECT_SURFACE);
if (ret_handle == 0) {
FREE(surf);
return ENOMEM;
}
return 0;
}
static void vrend_destroy_surface_object(void *obj_ptr)
{
struct vrend_surface *surface = obj_ptr;
vrend_surface_reference(&surface, NULL);
}
static void vrend_destroy_sampler_view_object(void *obj_ptr)
{
struct vrend_sampler_view *samp = obj_ptr;
vrend_sampler_view_reference(&samp, NULL);
}
static void vrend_destroy_so_target_object(void *obj_ptr)
{
struct vrend_so_target *target = obj_ptr;
struct vrend_sub_context *sub_ctx = target->sub_ctx;
struct vrend_streamout_object *obj, *tmp;
bool found;
int i;
LIST_FOR_EACH_ENTRY_SAFE(obj, tmp, &sub_ctx->streamout_list, head) {
found = false;
for (i = 0; i < obj->num_targets; i++) {
if (obj->so_targets[i] == target) {
found = true;
break;
}
}
if (found) {
if (obj == sub_ctx->current_so)
sub_ctx->current_so = NULL;
if (obj->xfb_state == XFB_STATE_PAUSED) {
if (vrend_state.have_tf2)
glBindTransformFeedback(GL_TRANSFORM_FEEDBACK, obj->id);
glEndTransformFeedback();
if (sub_ctx->current_so && vrend_state.have_tf2)
glBindTransformFeedback(GL_TRANSFORM_FEEDBACK, sub_ctx->current_so->id);
}
vrend_destroy_streamout_object(obj);
}
}
vrend_so_target_reference(&target, NULL);
}
static void vrend_destroy_vertex_elements_object(void *obj_ptr)
{
struct vrend_vertex_element_array *v = obj_ptr;
if (vrend_state.have_vertex_attrib_binding) {
glDeleteVertexArrays(1, &v->id);
}
FREE(v);
}
static void vrend_destroy_sampler_state_object(void *obj_ptr)
{
struct vrend_sampler_state *state = obj_ptr;
glDeleteSamplers(1, &state->id);
FREE(state);
}
static GLuint convert_wrap(int wrap)
{
switch(wrap){
case PIPE_TEX_WRAP_REPEAT: return GL_REPEAT;
case PIPE_TEX_WRAP_CLAMP: if (vrend_state.use_core_profile == false) return GL_CLAMP; else return GL_CLAMP_TO_EDGE;
case PIPE_TEX_WRAP_CLAMP_TO_EDGE: return GL_CLAMP_TO_EDGE;
case PIPE_TEX_WRAP_CLAMP_TO_BORDER: return GL_CLAMP_TO_BORDER;
case PIPE_TEX_WRAP_MIRROR_REPEAT: return GL_MIRRORED_REPEAT;
case PIPE_TEX_WRAP_MIRROR_CLAMP: return GL_MIRROR_CLAMP_EXT;
case PIPE_TEX_WRAP_MIRROR_CLAMP_TO_EDGE: return GL_MIRROR_CLAMP_TO_EDGE_EXT;
case PIPE_TEX_WRAP_MIRROR_CLAMP_TO_BORDER: return GL_MIRROR_CLAMP_TO_BORDER_EXT;
default:
assert(0);
return -1;
}
}
static inline GLenum convert_mag_filter(unsigned int filter)
{
if (filter == PIPE_TEX_FILTER_NEAREST)
return GL_NEAREST;
return GL_LINEAR;
}
static inline GLenum convert_min_filter(unsigned int filter, unsigned int mip_filter)
{
if (mip_filter == PIPE_TEX_MIPFILTER_NONE)
return convert_mag_filter(filter);
else if (mip_filter == PIPE_TEX_MIPFILTER_LINEAR) {
if (filter == PIPE_TEX_FILTER_NEAREST)
return GL_NEAREST_MIPMAP_LINEAR;
else
return GL_LINEAR_MIPMAP_LINEAR;
} else if (mip_filter == PIPE_TEX_MIPFILTER_NEAREST) {
if (filter == PIPE_TEX_FILTER_NEAREST)
return GL_NEAREST_MIPMAP_NEAREST;
else
return GL_LINEAR_MIPMAP_NEAREST;
}
assert(0);
return 0;
}
int vrend_create_sampler_state(struct vrend_context *ctx,
uint32_t handle,
struct pipe_sampler_state *templ)
{
struct vrend_sampler_state *state = CALLOC_STRUCT(vrend_sampler_state);
int ret_handle;
if (!state)
return ENOMEM;
state->base = *templ;
if (vrend_state.have_samplers) {
glGenSamplers(1, &state->id);
glSamplerParameteri(state->id, GL_TEXTURE_WRAP_S, convert_wrap(templ->wrap_s));
glSamplerParameteri(state->id, GL_TEXTURE_WRAP_T, convert_wrap(templ->wrap_t));
glSamplerParameteri(state->id, GL_TEXTURE_WRAP_R, convert_wrap(templ->wrap_r));
glSamplerParameterf(state->id, GL_TEXTURE_MIN_FILTER, convert_min_filter(templ->min_img_filter, templ->min_mip_filter));
glSamplerParameterf(state->id, GL_TEXTURE_MAG_FILTER, convert_mag_filter(templ->mag_img_filter));
glSamplerParameterf(state->id, GL_TEXTURE_MIN_LOD, templ->min_lod);
glSamplerParameterf(state->id, GL_TEXTURE_MAX_LOD, templ->max_lod);
glSamplerParameteri(state->id, GL_TEXTURE_COMPARE_MODE, templ->compare_mode ? GL_COMPARE_R_TO_TEXTURE : GL_NONE);
glSamplerParameteri(state->id, GL_TEXTURE_COMPARE_FUNC, GL_NEVER + templ->compare_func);
glSamplerParameteri(state->id, GL_TEXTURE_CUBE_MAP_SEAMLESS, templ->seamless_cube_map);
if (vrend_state.use_gles) {
if (templ->lod_bias != 0.0f) {
report_gles_warn(ctx, GLES_WARN_LOD_BIAS, 0);
}
} else {
glSamplerParameterf(state->id, GL_TEXTURE_LOD_BIAS, templ->lod_bias);
}
glSamplerParameterIuiv(state->id, GL_TEXTURE_BORDER_COLOR, templ->border_color.ui);
}
ret_handle = vrend_renderer_object_insert(ctx, state, sizeof(struct vrend_sampler_state), handle,
VIRGL_OBJECT_SAMPLER_STATE);
if (!ret_handle) {
if (vrend_state.have_samplers)
glDeleteSamplers(1, &state->id);
FREE(state);
return ENOMEM;
}
return 0;
}
static inline GLenum to_gl_swizzle(int swizzle)
{
switch (swizzle) {
case PIPE_SWIZZLE_RED: return GL_RED;
case PIPE_SWIZZLE_GREEN: return GL_GREEN;
case PIPE_SWIZZLE_BLUE: return GL_BLUE;
case PIPE_SWIZZLE_ALPHA: return GL_ALPHA;
case PIPE_SWIZZLE_ZERO: return GL_ZERO;
case PIPE_SWIZZLE_ONE: return GL_ONE;
default:
assert(0);
return 0;
}
}
int vrend_create_sampler_view(struct vrend_context *ctx,
uint32_t handle,
uint32_t res_handle, uint32_t format,
uint32_t val0, uint32_t val1, uint32_t swizzle_packed)
{
struct vrend_sampler_view *view;
struct vrend_resource *res;
int ret_handle;
uint8_t swizzle[4];
res = vrend_renderer_ctx_res_lookup(ctx, res_handle);
if (!res) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_RESOURCE, res_handle);
return EINVAL;
}
view = CALLOC_STRUCT(vrend_sampler_view);
if (!view)
return ENOMEM;
pipe_reference_init(&view->reference, 1);
view->format = format;
view->val0 = val0;
view->val1 = val1;
view->cur_base = -1;
view->cur_max = 10000;
swizzle[0] = swizzle_packed & 0x7;
swizzle[1] = (swizzle_packed >> 3) & 0x7;
swizzle[2] = (swizzle_packed >> 6) & 0x7;
swizzle[3] = (swizzle_packed >> 9) & 0x7;
vrend_resource_reference(&view->texture, res);
view->srgb_decode = GL_DECODE_EXT;
if (view->format != view->texture->base.format) {
if (util_format_is_srgb(view->texture->base.format) &&
!util_format_is_srgb(view->format))
view->srgb_decode = GL_SKIP_DECODE_EXT;
}
if (!(util_format_has_alpha(format) || util_format_is_depth_or_stencil(format))) {
if (swizzle[0] == PIPE_SWIZZLE_ALPHA)
swizzle[0] = PIPE_SWIZZLE_ONE;
if (swizzle[1] == PIPE_SWIZZLE_ALPHA)
swizzle[1] = PIPE_SWIZZLE_ONE;
if (swizzle[2] == PIPE_SWIZZLE_ALPHA)
swizzle[2] = PIPE_SWIZZLE_ONE;
if (swizzle[3] == PIPE_SWIZZLE_ALPHA)
swizzle[3] = PIPE_SWIZZLE_ONE;
}
if (tex_conv_table[format].flags & VREND_BIND_NEED_SWIZZLE) {
if (swizzle[0] <= PIPE_SWIZZLE_ALPHA)
swizzle[0] = tex_conv_table[format].swizzle[swizzle[0]];
if (swizzle[1] <= PIPE_SWIZZLE_ALPHA)
swizzle[1] = tex_conv_table[format].swizzle[swizzle[1]];
if (swizzle[2] <= PIPE_SWIZZLE_ALPHA)
swizzle[2] = tex_conv_table[format].swizzle[swizzle[2]];
if (swizzle[3] <= PIPE_SWIZZLE_ALPHA)
swizzle[3] = tex_conv_table[format].swizzle[swizzle[3]];
}
view->gl_swizzle_r = to_gl_swizzle(swizzle[0]);
view->gl_swizzle_g = to_gl_swizzle(swizzle[1]);
view->gl_swizzle_b = to_gl_swizzle(swizzle[2]);
view->gl_swizzle_a = to_gl_swizzle(swizzle[3]);
ret_handle = vrend_renderer_object_insert(ctx, view, sizeof(*view), handle, VIRGL_OBJECT_SAMPLER_VIEW);
if (ret_handle == 0) {
FREE(view);
return ENOMEM;
}
return 0;
}
void vrend_fb_bind_texture(struct vrend_resource *res,
int idx,
uint32_t level, uint32_t layer)
{
const struct util_format_description *desc = util_format_description(res->base.format);
GLenum attachment = GL_COLOR_ATTACHMENT0_EXT + idx;
if (vrend_format_is_ds(res->base.format)) { {
if (util_format_has_stencil(desc)) {
if (util_format_has_depth(desc))
attachment = GL_DEPTH_STENCIL_ATTACHMENT;
else
attachment = GL_STENCIL_ATTACHMENT;
} else
attachment = GL_DEPTH_ATTACHMENT;
}
}
switch (res->target) {
case GL_TEXTURE_1D_ARRAY:
case GL_TEXTURE_2D_ARRAY:
case GL_TEXTURE_2D_MULTISAMPLE_ARRAY:
case GL_TEXTURE_CUBE_MAP_ARRAY:
if (layer == 0xffffffff)
glFramebufferTexture(GL_FRAMEBUFFER_EXT, attachment,
res->id, level);
else
glFramebufferTextureLayer(GL_FRAMEBUFFER_EXT, attachment,
res->id, level, layer);
break;
case GL_TEXTURE_3D:
if (layer == 0xffffffff)
glFramebufferTexture(GL_FRAMEBUFFER_EXT, attachment,
res->id, level);
else
glFramebufferTexture3DEXT(GL_FRAMEBUFFER_EXT, attachment,
res->target, res->id, level, layer);
break;
case GL_TEXTURE_CUBE_MAP:
if (layer == 0xffffffff)
glFramebufferTexture(GL_FRAMEBUFFER_EXT, attachment,
res->id, level);
else
glFramebufferTexture2DEXT(GL_FRAMEBUFFER_EXT, attachment,
GL_TEXTURE_CUBE_MAP_POSITIVE_X + layer, res->id, level);
break;
case GL_TEXTURE_1D:
glFramebufferTexture1DEXT(GL_FRAMEBUFFER_EXT, attachment,
res->target, res->id, level);
break;
case GL_TEXTURE_2D:
default:
glFramebufferTexture2DEXT(GL_FRAMEBUFFER_EXT, attachment,
res->target, res->id, level);
break;
}
if (attachment == GL_DEPTH_ATTACHMENT) {
switch (res->target) {
case GL_TEXTURE_1D:
glFramebufferTexture1DEXT(GL_FRAMEBUFFER_EXT, GL_STENCIL_ATTACHMENT,
GL_TEXTURE_1D, 0, 0);
break;
case GL_TEXTURE_2D:
default:
glFramebufferTexture2DEXT(GL_FRAMEBUFFER_EXT, GL_STENCIL_ATTACHMENT,
GL_TEXTURE_2D, 0, 0);
break;
}
}
}
static void vrend_hw_set_zsurf_texture(struct vrend_context *ctx)
{
struct vrend_resource *tex;
int first_layer, last_layer;
if (!ctx->sub->zsurf) {
glFramebufferTexture2DEXT(GL_FRAMEBUFFER_EXT, GL_DEPTH_STENCIL_ATTACHMENT,
GL_TEXTURE_2D, 0, 0);
return;
}
tex = ctx->sub->zsurf->texture;
if (!tex)
return;
first_layer = ctx->sub->zsurf->val1 & 0xffff;
last_layer = (ctx->sub->zsurf->val1 >> 16) & 0xffff;
vrend_fb_bind_texture(tex, 0, ctx->sub->zsurf->val0,
first_layer != last_layer ? 0xffffffff : first_layer);
}
static void vrend_hw_set_color_surface(struct vrend_context *ctx, int index)
{
struct vrend_resource *tex;
if (!ctx->sub->surf[index]) {
GLenum attachment = GL_COLOR_ATTACHMENT0 + index;
glFramebufferTexture2DEXT(GL_FRAMEBUFFER_EXT, attachment,
GL_TEXTURE_2D, 0, 0);
} else {
int first_layer = ctx->sub->surf[index]->val1 & 0xffff;
int last_layer = (ctx->sub->surf[index]->val1 >> 16) & 0xffff;
tex = ctx->sub->surf[index]->texture;
vrend_fb_bind_texture(tex, index, ctx->sub->surf[index]->val0,
first_layer != last_layer ? 0xffffffff : first_layer);
}
}
static void vrend_hw_emit_framebuffer_state(struct vrend_context *ctx)
{
static const GLenum buffers[8] = {
GL_COLOR_ATTACHMENT0_EXT,
GL_COLOR_ATTACHMENT1_EXT,
GL_COLOR_ATTACHMENT2_EXT,
GL_COLOR_ATTACHMENT3_EXT,
GL_COLOR_ATTACHMENT4_EXT,
GL_COLOR_ATTACHMENT5_EXT,
GL_COLOR_ATTACHMENT6_EXT,
GL_COLOR_ATTACHMENT7_EXT,
};
glBindFramebufferEXT(GL_FRAMEBUFFER_EXT, ctx->sub->fb_id);
if (ctx->sub->nr_cbufs == 0) {
glReadBuffer(GL_NONE);
if (!vrend_state.use_gles) {
glDisable(GL_FRAMEBUFFER_SRGB_EXT);
}
} else {
struct vrend_surface *surf = NULL;
bool use_srgb = false;
int i;
for (i = 0; i < ctx->sub->nr_cbufs; i++) {
if (ctx->sub->surf[i]) {
surf = ctx->sub->surf[i];
if (util_format_is_srgb(surf->format)) {
use_srgb = true;
}
}
}
if (use_srgb) {
if (!vrend_state.use_gles) {
glEnable(GL_FRAMEBUFFER_SRGB_EXT);
} else {
report_gles_warn(ctx, GLES_WARN_SRGB_FB, 0);
}
} else {
if (!vrend_state.use_gles) {
glDisable(GL_FRAMEBUFFER_SRGB_EXT);
}
}
}
glDrawBuffers(ctx->sub->nr_cbufs, buffers);
}
void vrend_set_framebuffer_state(struct vrend_context *ctx,
uint32_t nr_cbufs, uint32_t surf_handle[PIPE_MAX_COLOR_BUFS],
uint32_t zsurf_handle)
{
struct vrend_surface *surf, *zsurf;
int i;
int old_num;
GLenum status;
GLint new_height = -1;
bool new_ibf = false;
glBindFramebufferEXT(GL_FRAMEBUFFER_EXT, ctx->sub->fb_id);
if (zsurf_handle) {
zsurf = vrend_object_lookup(ctx->sub->object_hash, zsurf_handle, VIRGL_OBJECT_SURFACE);
if (!zsurf) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_SURFACE, zsurf_handle);
return;
}
} else
zsurf = NULL;
if (ctx->sub->zsurf != zsurf) {
vrend_surface_reference(&ctx->sub->zsurf, zsurf);
vrend_hw_set_zsurf_texture(ctx);
}
old_num = ctx->sub->nr_cbufs;
ctx->sub->nr_cbufs = nr_cbufs;
ctx->sub->old_nr_cbufs = old_num;
for (i = 0; i < nr_cbufs; i++) {
if (surf_handle[i] != 0) {
surf = vrend_object_lookup(ctx->sub->object_hash, surf_handle[i], VIRGL_OBJECT_SURFACE);
if (!surf) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_SURFACE, surf_handle[i]);
return;
}
} else
surf = NULL;
if (ctx->sub->surf[i] != surf) {
vrend_surface_reference(&ctx->sub->surf[i], surf);
vrend_hw_set_color_surface(ctx, i);
}
}
if (old_num > ctx->sub->nr_cbufs) {
for (i = ctx->sub->nr_cbufs; i < old_num; i++) {
vrend_surface_reference(&ctx->sub->surf[i], NULL);
vrend_hw_set_color_surface(ctx, i);
}
}
/* find a buffer to set fb_height from */
if (ctx->sub->nr_cbufs == 0 && !ctx->sub->zsurf) {
new_height = 0;
new_ibf = false;
} else if (ctx->sub->nr_cbufs == 0) {
new_height = u_minify(ctx->sub->zsurf->texture->base.height0, ctx->sub->zsurf->val0);
new_ibf = ctx->sub->zsurf->texture->y_0_top ? true : false;
}
else {
surf = NULL;
for (i = 0; i < ctx->sub->nr_cbufs; i++) {
if (ctx->sub->surf[i]) {
surf = ctx->sub->surf[i];
break;
}
}
if (surf == NULL) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_SURFACE, i);
return;
}
new_height = u_minify(surf->texture->base.height0, surf->val0);
new_ibf = surf->texture->y_0_top ? true : false;
}
if (new_height != -1) {
if (ctx->sub->fb_height != new_height || ctx->sub->inverted_fbo_content != new_ibf) {
ctx->sub->fb_height = new_height;
ctx->sub->inverted_fbo_content = new_ibf;
ctx->sub->scissor_state_dirty = (1 << 0);
ctx->sub->viewport_state_dirty = (1 << 0);
}
}
vrend_hw_emit_framebuffer_state(ctx);
if (ctx->sub->nr_cbufs > 0 || ctx->sub->zsurf) {
status = glCheckFramebufferStatus(GL_FRAMEBUFFER);
if (status != GL_FRAMEBUFFER_COMPLETE)
fprintf(stderr,"failed to complete framebuffer 0x%x %s\n", status, ctx->debug_name);
}
ctx->sub->shader_dirty = true;
}
/*
* if the viewport Y scale factor is > 0 then we are rendering to
* an FBO already so don't need to invert rendering?
*/
void vrend_set_viewport_states(struct vrend_context *ctx,
uint32_t start_slot,
uint32_t num_viewports,
const struct pipe_viewport_state *state)
{
/* convert back to glViewport */
GLint x, y;
GLsizei width, height;
GLclampd near_val, far_val;
bool viewport_is_negative = (state[0].scale[1] < 0) ? true : false;
int i, idx;
if (num_viewports > PIPE_MAX_VIEWPORTS ||
start_slot > (PIPE_MAX_VIEWPORTS - num_viewports)) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_CMD_BUFFER, num_viewports);
return;
}
for (i = 0; i < num_viewports; i++) {
GLfloat abs_s1 = fabsf(state[i].scale[1]);
idx = start_slot + i;
width = state[i].scale[0] * 2.0f;
height = abs_s1 * 2.0f;
x = state[i].translate[0] - state[i].scale[0];
y = state[i].translate[1] - state[i].scale[1];
near_val = state[i].translate[2] - state[i].scale[2];
far_val = near_val + (state[i].scale[2] * 2.0);
if (ctx->sub->vps[idx].cur_x != x ||
ctx->sub->vps[idx].cur_y != y ||
ctx->sub->vps[idx].width != width ||
ctx->sub->vps[idx].height != height) {
ctx->sub->viewport_state_dirty |= (1 << idx);
ctx->sub->vps[idx].cur_x = x;
ctx->sub->vps[idx].cur_y = y;
ctx->sub->vps[idx].width = width;
ctx->sub->vps[idx].height = height;
}
if (idx == 0) {
if (ctx->sub->viewport_is_negative != viewport_is_negative)
ctx->sub->viewport_is_negative = viewport_is_negative;
ctx->sub->depth_scale = fabsf(far_val - near_val);
ctx->sub->depth_transform = near_val;
}
if (ctx->sub->vps[idx].near_val != near_val ||
ctx->sub->vps[idx].far_val != far_val) {
ctx->sub->vps[idx].near_val = near_val;
ctx->sub->vps[idx].far_val = far_val;
if (vrend_state.use_gles) {
if (near_val < 0.0f || far_val < 0.0f ||
near_val > 1.0f || far_val > 1.0f || idx) {
report_gles_warn(ctx, GLES_WARN_DEPTH_RANGE, 0);
}
/* Best effort despite the warning, gles will clamp. */
glDepthRangef(ctx->sub->vps[idx].near_val, ctx->sub->vps[idx].far_val);
} else if (idx)
glDepthRangeIndexed(idx, ctx->sub->vps[idx].near_val, ctx->sub->vps[idx].far_val);
else
glDepthRange(ctx->sub->vps[idx].near_val, ctx->sub->vps[idx].far_val);
}
}
}
int vrend_create_vertex_elements_state(struct vrend_context *ctx,
uint32_t handle,
unsigned num_elements,
const struct pipe_vertex_element *elements)
{
struct vrend_vertex_element_array *v;
const struct util_format_description *desc;
GLenum type;
int i;
uint32_t ret_handle;
if (num_elements > PIPE_MAX_ATTRIBS)
return EINVAL;
v = CALLOC_STRUCT(vrend_vertex_element_array);
if (!v)
return ENOMEM;
v->count = num_elements;
for (i = 0; i < num_elements; i++) {
memcpy(&v->elements[i].base, &elements[i], sizeof(struct pipe_vertex_element));
desc = util_format_description(elements[i].src_format);
if (!desc) {
FREE(v);
return EINVAL;
}
type = GL_FALSE;
if (desc->channel[0].type == UTIL_FORMAT_TYPE_FLOAT) {
if (desc->channel[0].size == 32)
type = GL_FLOAT;
else if (desc->channel[0].size == 64)
type = GL_DOUBLE;
else if (desc->channel[0].size == 16)
type = GL_HALF_FLOAT;
} else if (desc->channel[0].type == UTIL_FORMAT_TYPE_UNSIGNED &&
desc->channel[0].size == 8)
type = GL_UNSIGNED_BYTE;
else if (desc->channel[0].type == UTIL_FORMAT_TYPE_SIGNED &&
desc->channel[0].size == 8)
type = GL_BYTE;
else if (desc->channel[0].type == UTIL_FORMAT_TYPE_UNSIGNED &&
desc->channel[0].size == 16)
type = GL_UNSIGNED_SHORT;
else if (desc->channel[0].type == UTIL_FORMAT_TYPE_SIGNED &&
desc->channel[0].size == 16)
type = GL_SHORT;
else if (desc->channel[0].type == UTIL_FORMAT_TYPE_UNSIGNED &&
desc->channel[0].size == 32)
type = GL_UNSIGNED_INT;
else if (desc->channel[0].type == UTIL_FORMAT_TYPE_SIGNED &&
desc->channel[0].size == 32)
type = GL_INT;
else if (elements[i].src_format == PIPE_FORMAT_R10G10B10A2_SSCALED ||
elements[i].src_format == PIPE_FORMAT_R10G10B10A2_SNORM ||
elements[i].src_format == PIPE_FORMAT_B10G10R10A2_SNORM)
type = GL_INT_2_10_10_10_REV;
else if (elements[i].src_format == PIPE_FORMAT_R10G10B10A2_USCALED ||
elements[i].src_format == PIPE_FORMAT_R10G10B10A2_UNORM ||
elements[i].src_format == PIPE_FORMAT_B10G10R10A2_UNORM)
type = GL_UNSIGNED_INT_2_10_10_10_REV;
else if (elements[i].src_format == PIPE_FORMAT_R11G11B10_FLOAT)
type = GL_UNSIGNED_INT_10F_11F_11F_REV;
if (type == GL_FALSE) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_VERTEX_FORMAT, elements[i].src_format);
FREE(v);
return EINVAL;
}
v->elements[i].type = type;
if (desc->channel[0].normalized)
v->elements[i].norm = GL_TRUE;
if (desc->nr_channels == 4 && desc->swizzle[0] == UTIL_FORMAT_SWIZZLE_Z)
v->elements[i].nr_chan = GL_BGRA;
else if (elements[i].src_format == PIPE_FORMAT_R11G11B10_FLOAT)
v->elements[i].nr_chan = 3;
else
v->elements[i].nr_chan = desc->nr_channels;
}
if (vrend_state.have_vertex_attrib_binding) {
glGenVertexArrays(1, &v->id);
glBindVertexArray(v->id);
for (i = 0; i < num_elements; i++) {
struct vrend_vertex_element *ve = &v->elements[i];
if (util_format_is_pure_integer(ve->base.src_format))
glVertexAttribIFormat(i, ve->nr_chan, ve->type, ve->base.src_offset);
else
glVertexAttribFormat(i, ve->nr_chan, ve->type, ve->norm, ve->base.src_offset);
glVertexAttribBinding(i, ve->base.vertex_buffer_index);
glVertexBindingDivisor(i, ve->base.instance_divisor);
glEnableVertexAttribArray(i);
}
}
ret_handle = vrend_renderer_object_insert(ctx, v, sizeof(struct vrend_vertex_element), handle,
VIRGL_OBJECT_VERTEX_ELEMENTS);
if (!ret_handle) {
FREE(v);
return ENOMEM;
}
return 0;
}
void vrend_bind_vertex_elements_state(struct vrend_context *ctx,
uint32_t handle)
{
struct vrend_vertex_element_array *v;
if (!handle) {
ctx->sub->ve = NULL;
return;
}
v = vrend_object_lookup(ctx->sub->object_hash, handle, VIRGL_OBJECT_VERTEX_ELEMENTS);
if (!v) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_HANDLE, handle);
return;
}
if (ctx->sub->ve != v)
ctx->sub->vbo_dirty = true;
ctx->sub->ve = v;
}
void vrend_set_constants(struct vrend_context *ctx,
uint32_t shader,
uint32_t index,
uint32_t num_constant,
float *data)
{
struct vrend_constants *consts;
int i;
consts = &ctx->sub->consts[shader];
ctx->sub->const_dirty[shader] = true;
consts->consts = realloc(consts->consts, num_constant * sizeof(float));
if (!consts->consts)
return;
consts->num_consts = num_constant;
for (i = 0; i < num_constant; i++)
consts->consts[i] = ((unsigned int *)data)[i];
}
void vrend_set_uniform_buffer(struct vrend_context *ctx,
uint32_t shader,
uint32_t index,
uint32_t offset,
uint32_t length,
uint32_t res_handle)
{
struct vrend_resource *res;
if (res_handle) {
res = vrend_renderer_ctx_res_lookup(ctx, res_handle);
if (!res) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_RESOURCE, res_handle);
return;
}
vrend_resource_reference((struct vrend_resource **)&ctx->sub->cbs[shader][index].buffer, res);
ctx->sub->cbs[shader][index].buffer_offset = offset;
ctx->sub->cbs[shader][index].buffer_size = length;
ctx->sub->const_bufs_used_mask[shader] |= (1 << index);
} else {
vrend_resource_reference((struct vrend_resource **)&ctx->sub->cbs[shader][index].buffer, NULL);
ctx->sub->cbs[shader][index].buffer_offset = 0;
ctx->sub->cbs[shader][index].buffer_size = 0;
ctx->sub->const_bufs_used_mask[shader] &= ~(1 << index);
}
}
void vrend_set_index_buffer(struct vrend_context *ctx,
uint32_t res_handle,
uint32_t index_size,
uint32_t offset)
{
struct vrend_resource *res;
ctx->sub->ib.index_size = index_size;
ctx->sub->ib.offset = offset;
if (res_handle) {
if (ctx->sub->index_buffer_res_id != res_handle) {
res = vrend_renderer_ctx_res_lookup(ctx, res_handle);
if (!res) {
vrend_resource_reference((struct vrend_resource **)&ctx->sub->ib.buffer, NULL);
ctx->sub->index_buffer_res_id = 0;
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_RESOURCE, res_handle);
return;
}
vrend_resource_reference((struct vrend_resource **)&ctx->sub->ib.buffer, res);
ctx->sub->index_buffer_res_id = res_handle;
}
} else {
vrend_resource_reference((struct vrend_resource **)&ctx->sub->ib.buffer, NULL);
ctx->sub->index_buffer_res_id = 0;
}
}
void vrend_set_single_vbo(struct vrend_context *ctx,
int index,
uint32_t stride,
uint32_t buffer_offset,
uint32_t res_handle)
{
struct vrend_resource *res;
if (ctx->sub->vbo[index].stride != stride ||
ctx->sub->vbo[index].buffer_offset != buffer_offset ||
ctx->sub->vbo_res_ids[index] != res_handle)
ctx->sub->vbo_dirty = true;
ctx->sub->vbo[index].stride = stride;
ctx->sub->vbo[index].buffer_offset = buffer_offset;
if (res_handle == 0) {
vrend_resource_reference((struct vrend_resource **)&ctx->sub->vbo[index].buffer, NULL);
ctx->sub->vbo_res_ids[index] = 0;
} else if (ctx->sub->vbo_res_ids[index] != res_handle) {
res = vrend_renderer_ctx_res_lookup(ctx, res_handle);
if (!res) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_RESOURCE, res_handle);
ctx->sub->vbo_res_ids[index] = 0;
return;
}
vrend_resource_reference((struct vrend_resource **)&ctx->sub->vbo[index].buffer, res);
ctx->sub->vbo_res_ids[index] = res_handle;
}
}
void vrend_set_num_vbo(struct vrend_context *ctx,
int num_vbo)
{
int old_num = ctx->sub->num_vbos;
int i;
ctx->sub->num_vbos = num_vbo;
ctx->sub->old_num_vbos = old_num;
if (old_num != num_vbo)
ctx->sub->vbo_dirty = true;
for (i = num_vbo; i < old_num; i++) {
vrend_resource_reference((struct vrend_resource **)&ctx->sub->vbo[i].buffer, NULL);
ctx->sub->vbo_res_ids[i] = 0;
}
}
void vrend_set_single_sampler_view(struct vrend_context *ctx,
uint32_t shader_type,
int index,
uint32_t handle)
{
struct vrend_sampler_view *view = NULL;
struct vrend_texture *tex;
if (handle) {
view = vrend_object_lookup(ctx->sub->object_hash, handle, VIRGL_OBJECT_SAMPLER_VIEW);
if (!view) {
ctx->sub->views[shader_type].views[index] = NULL;
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_HANDLE, handle);
return;
}
if (ctx->sub->views[shader_type].views[index] == view) {
return;
}
/* we should have a reference to this texture taken at create time */
tex = (struct vrend_texture *)view->texture;
if (!tex) {
return;
}
if (view->texture->target != GL_TEXTURE_BUFFER) {
glBindTexture(view->texture->target, view->texture->id);
if (util_format_is_depth_or_stencil(view->format)) {
if (vrend_state.use_core_profile == false) {
/* setting depth texture mode is deprecated in core profile */
if (view->depth_texture_mode != GL_RED) {
glTexParameteri(view->texture->target, GL_DEPTH_TEXTURE_MODE, GL_RED);
view->depth_texture_mode = GL_RED;
}
}
if (vrend_state.have_stencil_texturing) {
const struct util_format_description *desc = util_format_description(view->format);
if (!util_format_has_depth(desc)) {
glTexParameteri(view->texture->target, GL_DEPTH_STENCIL_TEXTURE_MODE, GL_STENCIL_INDEX);
} else {
glTexParameteri(view->texture->target, GL_DEPTH_STENCIL_TEXTURE_MODE, GL_DEPTH_COMPONENT);
}
}
}
if (view->cur_base != (view->val1 & 0xff)) {
view->cur_base = view->val1 & 0xff;
glTexParameteri(view->texture->target, GL_TEXTURE_BASE_LEVEL, view->cur_base);
}
if (view->cur_max != ((view->val1 >> 8) & 0xff)) {
view->cur_max = (view->val1 >> 8) & 0xff;
glTexParameteri(view->texture->target, GL_TEXTURE_MAX_LEVEL, view->cur_max);
}
if (tex->cur_swizzle_r != view->gl_swizzle_r) {
glTexParameteri(view->texture->target, GL_TEXTURE_SWIZZLE_R, view->gl_swizzle_r);
tex->cur_swizzle_r = view->gl_swizzle_r;
}
if (tex->cur_swizzle_g != view->gl_swizzle_g) {
glTexParameteri(view->texture->target, GL_TEXTURE_SWIZZLE_G, view->gl_swizzle_g);
tex->cur_swizzle_g = view->gl_swizzle_g;
}
if (tex->cur_swizzle_b != view->gl_swizzle_b) {
glTexParameteri(view->texture->target, GL_TEXTURE_SWIZZLE_B, view->gl_swizzle_b);
tex->cur_swizzle_b = view->gl_swizzle_b;
}
if (tex->cur_swizzle_a != view->gl_swizzle_a) {
glTexParameteri(view->texture->target, GL_TEXTURE_SWIZZLE_A, view->gl_swizzle_a);
tex->cur_swizzle_a = view->gl_swizzle_a;
}
if (tex->srgb_decode != view->srgb_decode && util_format_is_srgb(tex->base.base.format)) {
if (vrend_state.have_samplers)
ctx->sub->sampler_state_dirty = true;
else {
glTexParameteri(view->texture->target, GL_TEXTURE_SRGB_DECODE_EXT,
view->srgb_decode);
tex->srgb_decode = view->srgb_decode;
}
}
} else {
GLenum internalformat;
glBindTexture(GL_TEXTURE_BUFFER, view->texture->tbo_tex_id);
internalformat = tex_conv_table[view->format].internalformat;
glTexBuffer(GL_TEXTURE_BUFFER, internalformat, view->texture->id);
}
}
vrend_sampler_view_reference(&ctx->sub->views[shader_type].views[index], view);
}
void vrend_set_num_sampler_views(struct vrend_context *ctx,
uint32_t shader_type,
uint32_t start_slot,
int num_sampler_views)
{
if (start_slot + num_sampler_views < ctx->sub->views[shader_type].num_views) {
int i;
for (i = start_slot + num_sampler_views; i < ctx->sub->views[shader_type].num_views; i++)
vrend_sampler_view_reference(&ctx->sub->views[shader_type].views[i], NULL);
}
ctx->sub->views[shader_type].num_views = start_slot + num_sampler_views;
}
static void vrend_destroy_shader_object(void *obj_ptr)
{
struct vrend_shader_selector *state = obj_ptr;
vrend_shader_state_reference(&state, NULL);
}
static inline void vrend_fill_shader_key(struct vrend_context *ctx,
unsigned type,
struct vrend_shader_key *key)
{
if (vrend_state.use_core_profile == true) {
int i;
bool add_alpha_test = true;
key->cbufs_are_a8_bitmask = 0;
for (i = 0; i < ctx->sub->nr_cbufs; i++) {
if (!ctx->sub->surf[i])
continue;
if (vrend_format_is_emulated_alpha(ctx->sub->surf[i]->format))
key->cbufs_are_a8_bitmask |= (1 << i);
if (util_format_is_pure_integer(ctx->sub->surf[i]->format))
add_alpha_test = false;
}
if (add_alpha_test) {
key->add_alpha_test = ctx->sub->dsa_state.alpha.enabled;
key->alpha_test = ctx->sub->dsa_state.alpha.func;
key->alpha_ref_val = ctx->sub->dsa_state.alpha.ref_value;
}
key->pstipple_tex = ctx->sub->rs_state.poly_stipple_enable;
key->color_two_side = ctx->sub->rs_state.light_twoside;
key->clip_plane_enable = ctx->sub->rs_state.clip_plane_enable;
key->flatshade = ctx->sub->rs_state.flatshade ? true : false;
} else {
key->add_alpha_test = 0;
key->pstipple_tex = 0;
}
key->invert_fs_origin = !ctx->sub->inverted_fbo_content;
key->coord_replace = ctx->sub->rs_state.point_quad_rasterization ? ctx->sub->rs_state.sprite_coord_enable : 0;
if (ctx->sub->shaders[PIPE_SHADER_GEOMETRY])
key->gs_present = true;
if ((type == PIPE_SHADER_GEOMETRY || type == PIPE_SHADER_FRAGMENT) && ctx->sub->shaders[PIPE_SHADER_VERTEX]) {
key->prev_stage_pervertex_out = ctx->sub->shaders[PIPE_SHADER_VERTEX]->sinfo.has_pervertex_out;
key->prev_stage_num_clip_out = ctx->sub->shaders[PIPE_SHADER_VERTEX]->sinfo.num_clip_out;
key->prev_stage_num_cull_out = ctx->sub->shaders[PIPE_SHADER_VERTEX]->sinfo.num_cull_out;
}
}
static inline int conv_shader_type(int type)
{
switch (type) {
case PIPE_SHADER_VERTEX: return GL_VERTEX_SHADER;
case PIPE_SHADER_FRAGMENT: return GL_FRAGMENT_SHADER;
case PIPE_SHADER_GEOMETRY: return GL_GEOMETRY_SHADER;
default:
return 0;
};
}
static int vrend_shader_create(struct vrend_context *ctx,
struct vrend_shader *shader,
struct vrend_shader_key key)
{
if (!shader->sel->tokens) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_SHADER, 0);
return -1;
}
shader->id = glCreateShader(conv_shader_type(shader->sel->type));
shader->compiled_fs_id = 0;
shader->glsl_prog = vrend_convert_shader(&ctx->shader_cfg, shader->sel->tokens, &key, &shader->sel->sinfo);
if (!shader->glsl_prog) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_SHADER, 0);
glDeleteShader(shader->id);
return -1;
}
shader->key = key;
if (1) {//shader->sel->type == PIPE_SHADER_FRAGMENT || shader->sel->type == PIPE_SHADER_GEOMETRY) {
bool ret;
ret = vrend_compile_shader(ctx, shader);
if (ret == false) {
glDeleteShader(shader->id);
free(shader->glsl_prog);
return -1;
}
}
return 0;
}
static int vrend_shader_select(struct vrend_context *ctx,
struct vrend_shader_selector *sel,
bool *dirty)
{
struct vrend_shader_key key;
struct vrend_shader *shader = NULL;
int r;
memset(&key, 0, sizeof(key));
vrend_fill_shader_key(ctx, sel->type, &key);
if (sel->current && !memcmp(&sel->current->key, &key, sizeof(key)))
return 0;
if (sel->num_shaders > 1) {
struct vrend_shader *p = sel->current, *c = p->next_variant;
while (c && memcmp(&c->key, &key, sizeof(key)) != 0) {
p = c;
c = c->next_variant;
}
if (c) {
p->next_variant = c->next_variant;
shader = c;
}
}
if (!shader) {
shader = CALLOC_STRUCT(vrend_shader);
shader->sel = sel;
list_inithead(&shader->programs);
r = vrend_shader_create(ctx, shader, key);
if (r) {
sel->current = NULL;
FREE(shader);
return r;
}
sel->num_shaders++;
}
if (dirty)
*dirty = true;
shader->next_variant = sel->current;
sel->current = shader;
return 0;
}
static void *vrend_create_shader_state(struct vrend_context *ctx,
const struct pipe_stream_output_info *so_info,
unsigned pipe_shader_type)
{
struct vrend_shader_selector *sel = CALLOC_STRUCT(vrend_shader_selector);
if (!sel)
return NULL;
sel->type = pipe_shader_type;
sel->sinfo.so_info = *so_info;
pipe_reference_init(&sel->reference, 1);
return sel;
}
static int vrend_finish_shader(struct vrend_context *ctx,
struct vrend_shader_selector *sel,
const struct tgsi_token *tokens)
{
int r;
sel->tokens = tgsi_dup_tokens(tokens);
r = vrend_shader_select(ctx, sel, NULL);
if (r) {
return EINVAL;
}
return 0;
}
int vrend_create_shader(struct vrend_context *ctx,
uint32_t handle,
const struct pipe_stream_output_info *so_info,
const char *shd_text, uint32_t offlen, uint32_t num_tokens,
uint32_t type, uint32_t pkt_length)
{
struct vrend_shader_selector *sel = NULL;
int ret_handle;
bool new_shader = true, long_shader = false;
bool finished = false;
int ret;
if (type > PIPE_SHADER_GEOMETRY)
return EINVAL;
if (offlen & VIRGL_OBJ_SHADER_OFFSET_CONT)
new_shader = false;
else if (((offlen + 3) / 4) > pkt_length)
long_shader = true;
/* if we have an in progress one - don't allow a new shader
of that type or a different handle. */
if (ctx->sub->long_shader_in_progress_handle[type]) {
if (new_shader == true)
return EINVAL;
if (handle != ctx->sub->long_shader_in_progress_handle[type])
return EINVAL;
}
if (new_shader) {
sel = vrend_create_shader_state(ctx, so_info, type);
if (sel == NULL)
return ENOMEM;
if (long_shader) {
sel->buf_len = ((offlen + 3) / 4) * 4; /* round up buffer size */
sel->tmp_buf = malloc(sel->buf_len);
if (!sel->tmp_buf) {
ret = ENOMEM;
goto error;
}
memcpy(sel->tmp_buf, shd_text, pkt_length * 4);
sel->buf_offset = pkt_length * 4;
ctx->sub->long_shader_in_progress_handle[type] = handle;
} else
finished = true;
} else {
sel = vrend_object_lookup(ctx->sub->object_hash, handle, VIRGL_OBJECT_SHADER);
if (!sel) {
fprintf(stderr, "got continuation without original shader %d\n", handle);
ret = EINVAL;
goto error;
}
offlen &= ~VIRGL_OBJ_SHADER_OFFSET_CONT;
if (offlen != sel->buf_offset) {
fprintf(stderr, "Got mismatched shader continuation %d vs %d\n",
offlen, sel->buf_offset);
ret = EINVAL;
goto error;
}
/*make sure no overflow */
if (pkt_length * 4 < pkt_length ||
pkt_length * 4 + sel->buf_offset < pkt_length * 4 ||
pkt_length * 4 + sel->buf_offset < sel->buf_offset) {
ret = EINVAL;
goto error;
}
if ((pkt_length * 4 + sel->buf_offset) > sel->buf_len) {
fprintf(stderr, "Got too large shader continuation %d vs %d\n",
pkt_length * 4 + sel->buf_offset, sel->buf_len);
ret = EINVAL;
goto error;
}
memcpy(sel->tmp_buf + sel->buf_offset, shd_text, pkt_length * 4);
sel->buf_offset += pkt_length * 4;
if (sel->buf_offset >= sel->buf_len) {
finished = true;
shd_text = sel->tmp_buf;
}
}
if (finished) {
struct tgsi_token *tokens;
tokens = calloc(num_tokens + 10, sizeof(struct tgsi_token));
if (!tokens) {
ret = ENOMEM;
goto error;
}
if (vrend_dump_shaders)
fprintf(stderr,"shader\n%s\n", shd_text);
if (!tgsi_text_translate((const char *)shd_text, tokens, num_tokens + 10)) {
free(tokens);
ret = EINVAL;
goto error;
}
if (vrend_finish_shader(ctx, sel, tokens)) {
free(tokens);
ret = EINVAL;
goto error;
} else {
free(sel->tmp_buf);
sel->tmp_buf = NULL;
}
free(tokens);
ctx->sub->long_shader_in_progress_handle[type] = 0;
}
if (new_shader) {
ret_handle = vrend_renderer_object_insert(ctx, sel, sizeof(*sel), handle, VIRGL_OBJECT_SHADER);
if (ret_handle == 0) {
ret = ENOMEM;
goto error;
}
}
return 0;
error:
if (new_shader)
vrend_destroy_shader_selector(sel);
else
vrend_renderer_object_destroy(ctx, handle);
return ret;
}
void vrend_bind_shader(struct vrend_context *ctx,
uint32_t handle, uint32_t type)
{
struct vrend_shader_selector *sel;
if (type > PIPE_SHADER_GEOMETRY)
return;
if (handle == 0) {
ctx->sub->shader_dirty = true;
vrend_shader_state_reference(&ctx->sub->shaders[type], NULL);
return;
}
sel = vrend_object_lookup(ctx->sub->object_hash, handle, VIRGL_OBJECT_SHADER);
if (!sel)
return;
if (sel->type != type)
return;
if (ctx->sub->shaders[sel->type] != sel) {
ctx->sub->shader_dirty = true;
ctx->sub->prog_ids[sel->type] = 0;
}
vrend_shader_state_reference(&ctx->sub->shaders[sel->type], sel);
}
void vrend_clear(struct vrend_context *ctx,
unsigned buffers,
const union pipe_color_union *color,
double depth, unsigned stencil)
{
GLbitfield bits = 0;
if (ctx->in_error)
return;
if (ctx->ctx_switch_pending)
vrend_finish_context_switch(ctx);
glBindFramebufferEXT(GL_FRAMEBUFFER_EXT, ctx->sub->fb_id);
vrend_update_frontface_state(ctx);
if (ctx->sub->stencil_state_dirty)
vrend_update_stencil_state(ctx);
if (ctx->sub->scissor_state_dirty)
vrend_update_scissor_state(ctx);
if (ctx->sub->viewport_state_dirty)
vrend_update_viewport_state(ctx);
vrend_use_program(ctx, 0);
if (buffers & PIPE_CLEAR_COLOR) {
if (ctx->sub->nr_cbufs && ctx->sub->surf[0] && vrend_format_is_emulated_alpha(ctx->sub->surf[0]->format)) {
glClearColor(color->f[3], 0.0, 0.0, 0.0);
} else {
glClearColor(color->f[0], color->f[1], color->f[2], color->f[3]);
}
/* This function implements Gallium's full clear callback (st->pipe->clear) on the host. This
callback requires no color component be masked. We must unmask all components before
calling glClear* and restore the previous colormask afterwards, as Gallium expects. */
if (ctx->sub->hw_blend_state.independent_blend_enable) {
/* ARB_draw_buffers_blend is required for this */
int i;
for (i = 0; i < PIPE_MAX_COLOR_BUFS; i++)
glColorMaskIndexedEXT(i, GL_TRUE, GL_TRUE, GL_TRUE, GL_TRUE);
} else
glColorMask(GL_TRUE, GL_TRUE, GL_TRUE, GL_TRUE);
}
if (buffers & PIPE_CLEAR_DEPTH) {
/* gallium clears don't respect depth mask */
glDepthMask(GL_TRUE);
if (vrend_state.use_gles) {
if (0.0f < depth && depth > 1.0f) {
// Only warn, it is clamped by the function.
report_gles_warn(ctx, GLES_WARN_DEPTH_CLEAR, 0);
}
glClearDepthf(depth);
} else {
glClearDepth(depth);
}
}
if (buffers & PIPE_CLEAR_STENCIL) {
glStencilMask(~0u);
glClearStencil(stencil);
}
if (buffers & PIPE_CLEAR_COLOR) {
uint32_t mask = 0;
int i;
for (i = 0; i < ctx->sub->nr_cbufs; i++) {
if (ctx->sub->surf[i])
mask |= (1 << i);
}
if (mask != (buffers >> 2)) {
mask = buffers >> 2;
while (mask) {
i = u_bit_scan(&mask);
if (i < PIPE_MAX_COLOR_BUFS && ctx->sub->surf[i] && util_format_is_pure_uint(ctx->sub->surf[i] && ctx->sub->surf[i]->format))
glClearBufferuiv(GL_COLOR,
i, (GLuint *)color);
else if (i < PIPE_MAX_COLOR_BUFS && ctx->sub->surf[i] && util_format_is_pure_sint(ctx->sub->surf[i] && ctx->sub->surf[i]->format))
glClearBufferiv(GL_COLOR,
i, (GLint *)color);
else
glClearBufferfv(GL_COLOR,
i, (GLfloat *)color);
}
}
else
bits |= GL_COLOR_BUFFER_BIT;
}
if (buffers & PIPE_CLEAR_DEPTH)
bits |= GL_DEPTH_BUFFER_BIT;
if (buffers & PIPE_CLEAR_STENCIL)
bits |= GL_STENCIL_BUFFER_BIT;
if (bits)
glClear(bits);
if (buffers & PIPE_CLEAR_DEPTH)
if (!ctx->sub->dsa_state.depth.writemask)
glDepthMask(GL_FALSE);
/* Restore previous stencil buffer write masks for both front and back faces */
if (buffers & PIPE_CLEAR_STENCIL) {
glStencilMaskSeparate(GL_FRONT, ctx->sub->dsa_state.stencil[0].writemask);
glStencilMaskSeparate(GL_BACK, ctx->sub->dsa_state.stencil[1].writemask);
}
/* Restore previous colormask */
if (buffers & PIPE_CLEAR_COLOR) {
if (ctx->sub->hw_blend_state.independent_blend_enable) {
/* ARB_draw_buffers_blend is required for this */
int i;
for (i = 0; i < PIPE_MAX_COLOR_BUFS; i++) {
struct pipe_blend_state *blend = &ctx->sub->hw_blend_state;
glColorMaskIndexedEXT(i, blend->rt[i].colormask & PIPE_MASK_R ? GL_TRUE : GL_FALSE,
blend->rt[i].colormask & PIPE_MASK_G ? GL_TRUE : GL_FALSE,
blend->rt[i].colormask & PIPE_MASK_B ? GL_TRUE : GL_FALSE,
blend->rt[i].colormask & PIPE_MASK_A ? GL_TRUE : GL_FALSE);
}
} else {
glColorMask(ctx->sub->hw_blend_state.rt[0].colormask & PIPE_MASK_R ? GL_TRUE : GL_FALSE,
ctx->sub->hw_blend_state.rt[0].colormask & PIPE_MASK_G ? GL_TRUE : GL_FALSE,
ctx->sub->hw_blend_state.rt[0].colormask & PIPE_MASK_B ? GL_TRUE : GL_FALSE,
ctx->sub->hw_blend_state.rt[0].colormask & PIPE_MASK_A ? GL_TRUE : GL_FALSE);
}
}
}
static void vrend_update_scissor_state(struct vrend_context *ctx)
{
struct pipe_scissor_state *ss;
struct pipe_rasterizer_state *state = &ctx->sub->rs_state;
GLint y;
GLuint idx;
unsigned mask = ctx->sub->scissor_state_dirty;
if (state->scissor)
glEnable(GL_SCISSOR_TEST);
else
glDisable(GL_SCISSOR_TEST);
while (mask) {
idx = u_bit_scan(&mask);
if (idx >= PIPE_MAX_VIEWPORTS) {
vrend_report_buffer_error(ctx, 0);
break;
}
ss = &ctx->sub->ss[idx];
if (ctx->sub->viewport_is_negative)
y = ss->miny;
else
y = ss->miny;
if (idx > 0)
glScissorIndexed(idx, ss->minx, y, ss->maxx - ss->minx, ss->maxy - ss->miny);
else
glScissor(ss->minx, y, ss->maxx - ss->minx, ss->maxy - ss->miny);
}
ctx->sub->scissor_state_dirty = 0;
}
static void vrend_update_viewport_state(struct vrend_context *ctx)
{
GLint cy;
unsigned mask = ctx->sub->viewport_state_dirty;
int idx;
while (mask) {
idx = u_bit_scan(&mask);
if (ctx->sub->viewport_is_negative)
cy = ctx->sub->vps[idx].cur_y - ctx->sub->vps[idx].height;
else
cy = ctx->sub->vps[idx].cur_y;
if (idx > 0)
glViewportIndexedf(idx, ctx->sub->vps[idx].cur_x, cy, ctx->sub->vps[idx].width, ctx->sub->vps[idx].height);
else
glViewport(ctx->sub->vps[idx].cur_x, cy, ctx->sub->vps[idx].width, ctx->sub->vps[idx].height);
}
ctx->sub->viewport_state_dirty = 0;
}
static GLenum get_gs_xfb_mode(GLenum mode)
{
switch (mode) {
case GL_POINTS:
return GL_POINTS;
case GL_LINE_STRIP:
return GL_LINES;
case GL_TRIANGLE_STRIP:
return GL_TRIANGLES;
default:
fprintf(stderr, "illegal gs transform feedback mode %d\n", mode);
return GL_POINTS;
}
}
static GLenum get_xfb_mode(GLenum mode)
{
switch (mode) {
case GL_POINTS:
return GL_POINTS;
case GL_TRIANGLES:
case GL_TRIANGLE_STRIP:
case GL_TRIANGLE_FAN:
case GL_QUADS:
case GL_QUAD_STRIP:
case GL_POLYGON:
return GL_TRIANGLES;
case GL_LINES:
case GL_LINE_LOOP:
case GL_LINE_STRIP:
return GL_LINES;
default:
fprintf(stderr, "failed to translate TFB %d\n", mode);
return GL_POINTS;
}
}
static void vrend_draw_bind_vertex_legacy(struct vrend_context *ctx,
struct vrend_vertex_element_array *va)
{
uint32_t num_enable;
uint32_t enable_bitmask;
uint32_t disable_bitmask;
int i;
num_enable = va->count;
enable_bitmask = 0;
disable_bitmask = ~((1ull << num_enable) - 1);
for (i = 0; i < va->count; i++) {
struct vrend_vertex_element *ve = &va->elements[i];
int vbo_index = ve->base.vertex_buffer_index;
struct vrend_resource *res;
GLint loc;
if (i >= ctx->sub->prog->ss[PIPE_SHADER_VERTEX]->sel->sinfo.num_inputs) {
/* XYZZY: debug this? */
num_enable = ctx->sub->prog->ss[PIPE_SHADER_VERTEX]->sel->sinfo.num_inputs;
break;
}
res = (struct vrend_resource *)ctx->sub->vbo[vbo_index].buffer;
if (!res) {
fprintf(stderr,"cannot find vbo buf %d %d %d\n", i, va->count, ctx->sub->prog->ss[PIPE_SHADER_VERTEX]->sel->sinfo.num_inputs);
continue;
}
if (vrend_state.use_explicit_locations || vrend_state.have_vertex_attrib_binding) {
loc = i;
} else {
if (ctx->sub->prog->attrib_locs) {
loc = ctx->sub->prog->attrib_locs[i];
} else loc = -1;
if (loc == -1) {
fprintf(stderr,"%s: cannot find loc %d %d %d\n", ctx->debug_name, i, va->count, ctx->sub->prog->ss[PIPE_SHADER_VERTEX]->sel->sinfo.num_inputs);
num_enable--;
if (i == 0) {
fprintf(stderr,"%s: shader probably didn't compile - skipping rendering\n", ctx->debug_name);
return;
}
continue;
}
}
if (ve->type == GL_FALSE) {
fprintf(stderr,"failed to translate vertex type - skipping render\n");
return;
}
glBindBuffer(GL_ARRAY_BUFFER, res->id);
if (ctx->sub->vbo[vbo_index].stride == 0) {
void *data;
/* for 0 stride we are kinda screwed */
data = glMapBufferRange(GL_ARRAY_BUFFER, ctx->sub->vbo[vbo_index].buffer_offset, ve->nr_chan * sizeof(GLfloat), GL_MAP_READ_BIT);
switch (ve->nr_chan) {
case 1:
glVertexAttrib1fv(loc, data);
break;
case 2:
glVertexAttrib2fv(loc, data);
break;
case 3:
glVertexAttrib3fv(loc, data);
break;
case 4:
default:
glVertexAttrib4fv(loc, data);
break;
}
glUnmapBuffer(GL_ARRAY_BUFFER);
disable_bitmask |= (1 << loc);
} else {
enable_bitmask |= (1 << loc);
if (util_format_is_pure_integer(ve->base.src_format)) {
glVertexAttribIPointer(loc, ve->nr_chan, ve->type, ctx->sub->vbo[vbo_index].stride, (void *)(unsigned long)(ve->base.src_offset + ctx->sub->vbo[vbo_index].buffer_offset));
} else {
glVertexAttribPointer(loc, ve->nr_chan, ve->type, ve->norm, ctx->sub->vbo[vbo_index].stride, (void *)(unsigned long)(ve->base.src_offset + ctx->sub->vbo[vbo_index].buffer_offset));
}
glVertexAttribDivisorARB(loc, ve->base.instance_divisor);
}
}
if (ctx->sub->enabled_attribs_bitmask != enable_bitmask) {
uint32_t mask = ctx->sub->enabled_attribs_bitmask & disable_bitmask;
while (mask) {
i = u_bit_scan(&mask);
glDisableVertexAttribArray(i);
}
ctx->sub->enabled_attribs_bitmask &= ~disable_bitmask;
mask = ctx->sub->enabled_attribs_bitmask ^ enable_bitmask;
while (mask) {
i = u_bit_scan(&mask);
glEnableVertexAttribArray(i);
}
ctx->sub->enabled_attribs_bitmask = enable_bitmask;
}
}
static void vrend_draw_bind_vertex_binding(struct vrend_context *ctx,
struct vrend_vertex_element_array *va)
{
int i;
glBindVertexArray(va->id);
if (ctx->sub->vbo_dirty) {
for (i = 0; i < ctx->sub->num_vbos; i++) {
struct vrend_resource *res = (struct vrend_resource *)ctx->sub->vbo[i].buffer;
if (!res)
glBindVertexBuffer(i, 0, 0, 0);
else
glBindVertexBuffer(i,
res->id,
ctx->sub->vbo[i].buffer_offset,
ctx->sub->vbo[i].stride);
}
for (i = ctx->sub->num_vbos; i < ctx->sub->old_num_vbos; i++) {
glBindVertexBuffer(i, 0, 0, 0);
}
ctx->sub->vbo_dirty = false;
}
}
static void vrend_draw_bind_samplers(struct vrend_context *ctx)
{
int sampler_id;
int i;
int shader_type;
sampler_id = 0;
for (shader_type = PIPE_SHADER_VERTEX; shader_type <= ctx->sub->last_shader_idx; shader_type++) {
int index = 0;
for (i = 0; i < ctx->sub->views[shader_type].num_views; i++) {
struct vrend_resource *texture = NULL;
if (ctx->sub->views[shader_type].views[i]) {
texture = ctx->sub->views[shader_type].views[i]->texture;
}
if (!(ctx->sub->prog->samplers_used_mask[shader_type] & (1 << i)))
continue;
if (ctx->sub->prog->samp_locs[shader_type])
glUniform1i(ctx->sub->prog->samp_locs[shader_type][index], sampler_id);
if (ctx->sub->prog->shadow_samp_mask[shader_type] & (1 << i)) {
struct vrend_sampler_view *tview = ctx->sub->views[shader_type].views[i];
glUniform4f(ctx->sub->prog->shadow_samp_mask_locs[shader_type][index],
(tview->gl_swizzle_r == GL_ZERO || tview->gl_swizzle_r == GL_ONE) ? 0.0 : 1.0,
(tview->gl_swizzle_g == GL_ZERO || tview->gl_swizzle_g == GL_ONE) ? 0.0 : 1.0,
(tview->gl_swizzle_b == GL_ZERO || tview->gl_swizzle_b == GL_ONE) ? 0.0 : 1.0,
(tview->gl_swizzle_a == GL_ZERO || tview->gl_swizzle_a == GL_ONE) ? 0.0 : 1.0);
glUniform4f(ctx->sub->prog->shadow_samp_add_locs[shader_type][index],
tview->gl_swizzle_r == GL_ONE ? 1.0 : 0.0,
tview->gl_swizzle_g == GL_ONE ? 1.0 : 0.0,
tview->gl_swizzle_b == GL_ONE ? 1.0 : 0.0,
tview->gl_swizzle_a == GL_ONE ? 1.0 : 0.0);
}
glActiveTexture(GL_TEXTURE0 + sampler_id);
if (texture) {
int id;
if (texture->target == GL_TEXTURE_BUFFER)
id = texture->tbo_tex_id;
else
id = texture->id;
glBindTexture(texture->target, id);
if (ctx->sub->views[shader_type].old_ids[i] != id || ctx->sub->sampler_state_dirty) {
vrend_apply_sampler_state(ctx, texture, shader_type, i, sampler_id, ctx->sub->views[shader_type].views[i]->srgb_decode);
ctx->sub->views[shader_type].old_ids[i] = id;
}
if (ctx->sub->rs_state.point_quad_rasterization) {
if (vrend_state.use_core_profile == false) {
if (ctx->sub->rs_state.sprite_coord_enable & (1 << i))
glTexEnvi(GL_POINT_SPRITE_ARB, GL_COORD_REPLACE_ARB, GL_TRUE);
else
glTexEnvi(GL_POINT_SPRITE_ARB, GL_COORD_REPLACE_ARB, GL_FALSE);
}
}
sampler_id++;
}
index++;
}
}
if (vrend_state.use_core_profile && ctx->sub->prog->fs_stipple_loc != -1) {
glActiveTexture(GL_TEXTURE0 + sampler_id);
glBindTexture(GL_TEXTURE_2D, ctx->pstipple_tex_id);
glUniform1i(ctx->sub->prog->fs_stipple_loc, sampler_id);
}
ctx->sub->sampler_state_dirty = false;
}
static void vrend_draw_bind_ubo(struct vrend_context *ctx)
{
int i;
int ubo_id;
int shader_type;
ubo_id = 0;
for (shader_type = PIPE_SHADER_VERTEX; shader_type <= ctx->sub->last_shader_idx; shader_type++) {
uint32_t mask;
int shader_ubo_idx = 0;
struct pipe_constant_buffer *cb;
struct vrend_resource *res;
if (!ctx->sub->const_bufs_used_mask[shader_type])
continue;
if (!ctx->sub->prog->ubo_locs[shader_type])
continue;
mask = ctx->sub->const_bufs_used_mask[shader_type];
while (mask) {
i = u_bit_scan(&mask);
cb = &ctx->sub->cbs[shader_type][i];
res = (struct vrend_resource *)cb->buffer;
glBindBufferRange(GL_UNIFORM_BUFFER, ubo_id, res->id,
cb->buffer_offset, cb->buffer_size);
glUniformBlockBinding(ctx->sub->prog->id, ctx->sub->prog->ubo_locs[shader_type][shader_ubo_idx], ubo_id);
shader_ubo_idx++;
ubo_id++;
}
}
}
void vrend_draw_vbo(struct vrend_context *ctx,
const struct pipe_draw_info *info,
uint32_t cso, uint32_t indirect_handle,
uint32_t indirect_draw_count_handle)
{
int i;
bool new_program = false;
uint32_t shader_type;
struct vrend_resource *indirect_res = NULL;
if (ctx->in_error)
return;
if (indirect_handle) {
indirect_res = vrend_renderer_ctx_res_lookup(ctx, indirect_handle);
if (!indirect_res) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_RESOURCE, indirect_handle);
return;
}
}
/* this must be zero until we support the feature */
if (indirect_draw_count_handle) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_RESOURCE, indirect_handle);
return;
}
if (ctx->ctx_switch_pending)
vrend_finish_context_switch(ctx);
vrend_update_frontface_state(ctx);
if (ctx->sub->stencil_state_dirty)
vrend_update_stencil_state(ctx);
if (ctx->sub->scissor_state_dirty)
vrend_update_scissor_state(ctx);
if (ctx->sub->viewport_state_dirty)
vrend_update_viewport_state(ctx);
vrend_patch_blend_state(ctx);
if (ctx->sub->shader_dirty) {
struct vrend_linked_shader_program *prog;
bool fs_dirty, vs_dirty, gs_dirty;
bool dual_src = util_blend_state_is_dual(&ctx->sub->blend_state, 0);
bool same_prog;
if (!ctx->sub->shaders[PIPE_SHADER_VERTEX] || !ctx->sub->shaders[PIPE_SHADER_FRAGMENT]) {
fprintf(stderr,"dropping rendering due to missing shaders: %s\n", ctx->debug_name);
return;
}
vrend_shader_select(ctx, ctx->sub->shaders[PIPE_SHADER_FRAGMENT], &fs_dirty);
vrend_shader_select(ctx, ctx->sub->shaders[PIPE_SHADER_VERTEX], &vs_dirty);
if (ctx->sub->shaders[PIPE_SHADER_GEOMETRY])
vrend_shader_select(ctx, ctx->sub->shaders[PIPE_SHADER_GEOMETRY], &gs_dirty);
if (!ctx->sub->shaders[PIPE_SHADER_VERTEX]->current || !ctx->sub->shaders[PIPE_SHADER_FRAGMENT]->current || (ctx->sub->shaders[PIPE_SHADER_GEOMETRY] && !ctx->sub->shaders[PIPE_SHADER_GEOMETRY]->current)) {
fprintf(stderr, "failure to compile shader variants: %s\n", ctx->debug_name);
return;
}
same_prog = true;
if (ctx->sub->shaders[PIPE_SHADER_VERTEX]->current->id != ctx->sub->prog_ids[PIPE_SHADER_VERTEX])
same_prog = false;
if (ctx->sub->shaders[PIPE_SHADER_FRAGMENT]->current->id != ctx->sub->prog_ids[PIPE_SHADER_FRAGMENT])
same_prog = false;
if (ctx->sub->shaders[PIPE_SHADER_GEOMETRY] && ctx->sub->shaders[PIPE_SHADER_GEOMETRY]->current->id != ctx->sub->prog_ids[PIPE_SHADER_GEOMETRY])
same_prog = false;
if (!same_prog) {
prog = lookup_shader_program(ctx, ctx->sub->shaders[PIPE_SHADER_VERTEX]->current->id, ctx->sub->shaders[PIPE_SHADER_FRAGMENT]->current->id, ctx->sub->shaders[PIPE_SHADER_GEOMETRY] ? ctx->sub->shaders[PIPE_SHADER_GEOMETRY]->current->id : 0, dual_src);
if (!prog) {
prog = add_shader_program(ctx,
ctx->sub->shaders[PIPE_SHADER_VERTEX]->current,
ctx->sub->shaders[PIPE_SHADER_FRAGMENT]->current, ctx->sub->shaders[PIPE_SHADER_GEOMETRY] ? ctx->sub->shaders[PIPE_SHADER_GEOMETRY]->current : NULL);
if (!prog)
return;
}
ctx->sub->last_shader_idx = ctx->sub->shaders[PIPE_SHADER_GEOMETRY] ? PIPE_SHADER_GEOMETRY : PIPE_SHADER_FRAGMENT;
} else
prog = ctx->sub->prog;
if (ctx->sub->prog != prog) {
new_program = true;
ctx->sub->prog_ids[PIPE_SHADER_VERTEX] = ctx->sub->shaders[PIPE_SHADER_VERTEX]->current->id;
ctx->sub->prog_ids[PIPE_SHADER_FRAGMENT] = ctx->sub->shaders[PIPE_SHADER_FRAGMENT]->current->id;
if (ctx->sub->shaders[PIPE_SHADER_GEOMETRY])
ctx->sub->prog_ids[PIPE_SHADER_GEOMETRY] = ctx->sub->shaders[PIPE_SHADER_GEOMETRY]->current->id;
ctx->sub->prog = prog;
}
}
if (!ctx->sub->prog) {
fprintf(stderr,"dropping rendering due to missing shaders: %s\n", ctx->debug_name);
return;
}
glBindFramebufferEXT(GL_FRAMEBUFFER_EXT, ctx->sub->fb_id);
vrend_use_program(ctx, ctx->sub->prog->id);
for (shader_type = PIPE_SHADER_VERTEX; shader_type <= ctx->sub->last_shader_idx; shader_type++) {
if (ctx->sub->consts[shader_type].consts &&
ctx->sub->prog->const_locs[shader_type] &&
(ctx->sub->const_dirty[shader_type] || new_program)) {
for (i = 0; i < ctx->sub->shaders[shader_type]->sinfo.num_consts; i++) {
if (ctx->sub->prog->const_locs[shader_type][i] != -1)
glUniform4uiv(ctx->sub->prog->const_locs[shader_type][i], 1, &ctx->sub->consts[shader_type].consts[i * 4]);
}
ctx->sub->const_dirty[shader_type] = false;
}
}
vrend_draw_bind_samplers(ctx);
vrend_draw_bind_ubo(ctx);
if (!ctx->sub->ve) {
fprintf(stderr,"illegal VE setup - skipping renderering\n");
return;
}
glUniform1f(ctx->sub->prog->vs_ws_adjust_loc, ctx->sub->viewport_is_negative ? -1.0 : 1.0);
if (ctx->sub->rs_state.clip_plane_enable) {
for (i = 0 ; i < 8; i++) {
glUniform4fv(ctx->sub->prog->clip_locs[i], 1, (const GLfloat *)&ctx->sub->ucp_state.ucp[i]);
}
}
if (vrend_state.have_vertex_attrib_binding)
vrend_draw_bind_vertex_binding(ctx, ctx->sub->ve);
else
vrend_draw_bind_vertex_legacy(ctx, ctx->sub->ve);
for (i = 0 ; i < ctx->sub->prog->ss[PIPE_SHADER_VERTEX]->sel->sinfo.num_inputs; i++) {
struct vrend_vertex_element_array *va = ctx->sub->ve;
struct vrend_vertex_element *ve = &va->elements[i];
int vbo_index = ve->base.vertex_buffer_index;
if (!ctx->sub->vbo[vbo_index].buffer) {
fprintf(stderr, "VBO missing vertex buffer\n");
return;
}
}
if (info->indexed) {
struct vrend_resource *res = (struct vrend_resource *)ctx->sub->ib.buffer;
glBindBuffer(GL_ELEMENT_ARRAY_BUFFER, res->id);
} else
glBindBuffer(GL_ELEMENT_ARRAY_BUFFER, 0);
if (ctx->sub->current_so) {
if (ctx->sub->current_so->xfb_state == XFB_STATE_STARTED_NEED_BEGIN) {
if (ctx->sub->shaders[PIPE_SHADER_GEOMETRY])
glBeginTransformFeedback(get_gs_xfb_mode(ctx->sub->shaders[PIPE_SHADER_GEOMETRY]->sinfo.gs_out_prim));
else
glBeginTransformFeedback(get_xfb_mode(info->mode));
ctx->sub->current_so->xfb_state = XFB_STATE_STARTED;
} else if (ctx->sub->current_so->xfb_state == XFB_STATE_PAUSED) {
glResumeTransformFeedback();
ctx->sub->current_so->xfb_state = XFB_STATE_STARTED;
}
}
if (info->primitive_restart) {
if (vrend_state.have_nv_prim_restart) {
glEnableClientState(GL_PRIMITIVE_RESTART_NV);
glPrimitiveRestartIndexNV(info->restart_index);
} else {
glEnable(GL_PRIMITIVE_RESTART);
glPrimitiveRestartIndex(info->restart_index);
}
}
if (indirect_res)
glBindBuffer(GL_DRAW_INDIRECT_BUFFER, indirect_res->id);
else
glBindBuffer(GL_DRAW_INDIRECT_BUFFER, 0);
/* set the vertex state up now on a delay */
if (!info->indexed) {
GLenum mode = info->mode;
int count = cso ? cso : info->count;
int start = cso ? 0 : info->start;
if (indirect_handle)
glDrawArraysIndirect(mode, (GLvoid const *)(unsigned long)info->indirect.offset);
else if (info->instance_count <= 1)
glDrawArrays(mode, start, count);
else if (info->start_instance)
glDrawArraysInstancedBaseInstance(mode, start, count, info->instance_count, info->start_instance);
else
glDrawArraysInstancedARB(mode, start, count, info->instance_count);
} else {
GLenum elsz;
GLenum mode = info->mode;
switch (ctx->sub->ib.index_size) {
case 1:
elsz = GL_UNSIGNED_BYTE;
break;
case 2:
elsz = GL_UNSIGNED_SHORT;
break;
case 4:
default:
elsz = GL_UNSIGNED_INT;
break;
}
if (indirect_handle)
glDrawElementsIndirect(mode, elsz, (GLvoid const *)(unsigned long)info->indirect.offset);
else if (info->index_bias) {
if (info->instance_count > 1)
glDrawElementsInstancedBaseVertex(mode, info->count, elsz, (void *)(unsigned long)ctx->sub->ib.offset, info->instance_count, info->index_bias);
else if (info->min_index != 0 || info->max_index != -1)
glDrawRangeElementsBaseVertex(mode, info->min_index, info->max_index, info->count, elsz, (void *)(unsigned long)ctx->sub->ib.offset, info->index_bias);
else
glDrawElementsBaseVertex(mode, info->count, elsz, (void *)(unsigned long)ctx->sub->ib.offset, info->index_bias);
} else if (info->instance_count > 1) {
glDrawElementsInstancedARB(mode, info->count, elsz, (void *)(unsigned long)ctx->sub->ib.offset, info->instance_count);
} else if (info->min_index != 0 || info->max_index != -1)
glDrawRangeElements(mode, info->min_index, info->max_index, info->count, elsz, (void *)(unsigned long)ctx->sub->ib.offset);
else
glDrawElements(mode, info->count, elsz, (void *)(unsigned long)ctx->sub->ib.offset);
}
if (info->primitive_restart) {
if (vrend_state.have_nv_prim_restart)
glDisableClientState(GL_PRIMITIVE_RESTART_NV);
else if (vrend_state.have_gl_prim_restart)
glDisable(GL_PRIMITIVE_RESTART);
}
if (ctx->sub->current_so && vrend_state.have_tf2) {
if (ctx->sub->current_so->xfb_state == XFB_STATE_STARTED) {
glPauseTransformFeedback();
ctx->sub->current_so->xfb_state = XFB_STATE_PAUSED;
}
}
}
static GLenum translate_blend_func(uint32_t pipe_blend)
{
switch(pipe_blend){
case PIPE_BLEND_ADD: return GL_FUNC_ADD;
case PIPE_BLEND_SUBTRACT: return GL_FUNC_SUBTRACT;
case PIPE_BLEND_REVERSE_SUBTRACT: return GL_FUNC_REVERSE_SUBTRACT;
case PIPE_BLEND_MIN: return GL_MIN;
case PIPE_BLEND_MAX: return GL_MAX;
default:
assert("invalid blend token()" == NULL);
return 0;
}
}
static GLenum translate_blend_factor(uint32_t pipe_factor)
{
switch (pipe_factor) {
case PIPE_BLENDFACTOR_ONE: return GL_ONE;
case PIPE_BLENDFACTOR_SRC_COLOR: return GL_SRC_COLOR;
case PIPE_BLENDFACTOR_SRC_ALPHA: return GL_SRC_ALPHA;
case PIPE_BLENDFACTOR_DST_COLOR: return GL_DST_COLOR;
case PIPE_BLENDFACTOR_DST_ALPHA: return GL_DST_ALPHA;
case PIPE_BLENDFACTOR_CONST_COLOR: return GL_CONSTANT_COLOR;
case PIPE_BLENDFACTOR_CONST_ALPHA: return GL_CONSTANT_ALPHA;
case PIPE_BLENDFACTOR_SRC1_COLOR: return GL_SRC1_COLOR;
case PIPE_BLENDFACTOR_SRC1_ALPHA: return GL_SRC1_ALPHA;
case PIPE_BLENDFACTOR_SRC_ALPHA_SATURATE: return GL_SRC_ALPHA_SATURATE;
case PIPE_BLENDFACTOR_ZERO: return GL_ZERO;
case PIPE_BLENDFACTOR_INV_SRC_COLOR: return GL_ONE_MINUS_SRC_COLOR;
case PIPE_BLENDFACTOR_INV_SRC_ALPHA: return GL_ONE_MINUS_SRC_ALPHA;
case PIPE_BLENDFACTOR_INV_DST_COLOR: return GL_ONE_MINUS_DST_COLOR;
case PIPE_BLENDFACTOR_INV_DST_ALPHA: return GL_ONE_MINUS_DST_ALPHA;
case PIPE_BLENDFACTOR_INV_CONST_COLOR: return GL_ONE_MINUS_CONSTANT_COLOR;
case PIPE_BLENDFACTOR_INV_CONST_ALPHA: return GL_ONE_MINUS_CONSTANT_ALPHA;
case PIPE_BLENDFACTOR_INV_SRC1_COLOR: return GL_ONE_MINUS_SRC1_COLOR;
case PIPE_BLENDFACTOR_INV_SRC1_ALPHA: return GL_ONE_MINUS_SRC1_ALPHA;
default:
assert("invalid blend token()" == NULL);
return 0;
}
}
static GLenum
translate_logicop(GLuint pipe_logicop)
{
switch (pipe_logicop) {
#define CASE(x) case PIPE_LOGICOP_##x: return GL_##x
CASE(CLEAR);
CASE(NOR);
CASE(AND_INVERTED);
CASE(COPY_INVERTED);
CASE(AND_REVERSE);
CASE(INVERT);
CASE(XOR);
CASE(NAND);
CASE(AND);
CASE(EQUIV);
CASE(NOOP);
CASE(OR_INVERTED);
CASE(COPY);
CASE(OR_REVERSE);
CASE(OR);
CASE(SET);
default:
assert("invalid logicop token()" == NULL);
return 0;
}
#undef CASE
}
static GLenum
translate_stencil_op(GLuint op)
{
switch (op) {
#define CASE(x) case PIPE_STENCIL_OP_##x: return GL_##x
CASE(KEEP);
CASE(ZERO);
CASE(REPLACE);
CASE(INCR);
CASE(DECR);
CASE(INCR_WRAP);
CASE(DECR_WRAP);
CASE(INVERT);
default:
assert("invalid stencilop token()" == NULL);
return 0;
}
#undef CASE
}
static inline bool is_dst_blend(int blend_factor)
{
return (blend_factor == PIPE_BLENDFACTOR_DST_ALPHA ||
blend_factor == PIPE_BLENDFACTOR_INV_DST_ALPHA);
}
static inline int conv_a8_blend(int blend_factor)
{
if (blend_factor == PIPE_BLENDFACTOR_DST_ALPHA)
return PIPE_BLENDFACTOR_DST_COLOR;
if (blend_factor == PIPE_BLENDFACTOR_INV_DST_ALPHA)
return PIPE_BLENDFACTOR_INV_DST_COLOR;
return blend_factor;
}
static inline int conv_dst_blend(int blend_factor)
{
if (blend_factor == PIPE_BLENDFACTOR_DST_ALPHA)
return PIPE_BLENDFACTOR_ONE;
if (blend_factor == PIPE_BLENDFACTOR_INV_DST_ALPHA)
return PIPE_BLENDFACTOR_ZERO;
return blend_factor;
}
static void vrend_hw_emit_blend(struct vrend_context *ctx, struct pipe_blend_state *state)
{
if (state->logicop_enable != ctx->sub->hw_blend_state.logicop_enable) {
ctx->sub->hw_blend_state.logicop_enable = state->logicop_enable;
if (vrend_state.use_gles) {
if (state->logicop_enable) {
report_gles_warn(ctx, GLES_WARN_LOGIC_OP, 0);
}
} else if (state->logicop_enable) {
glEnable(GL_COLOR_LOGIC_OP);
glLogicOp(translate_logicop(state->logicop_func));
} else {
glDisable(GL_COLOR_LOGIC_OP);
}
}
if (state->independent_blend_enable) {
/* ARB_draw_buffers_blend is required for this */
int i;
for (i = 0; i < PIPE_MAX_COLOR_BUFS; i++) {
if (state->rt[i].blend_enable) {
glBlendFuncSeparateiARB(i, translate_blend_factor(state->rt[i].rgb_src_factor),
translate_blend_factor(state->rt[i].rgb_dst_factor),
translate_blend_factor(state->rt[i].alpha_src_factor),
translate_blend_factor(state->rt[i].alpha_dst_factor));
glBlendEquationSeparateiARB(i, translate_blend_func(state->rt[i].rgb_func),
translate_blend_func(state->rt[i].alpha_func));
glEnableIndexedEXT(GL_BLEND, i);
} else
glDisableIndexedEXT(GL_BLEND, i);
if (state->rt[i].colormask != ctx->sub->hw_blend_state.rt[i].colormask) {
ctx->sub->hw_blend_state.rt[i].colormask = state->rt[i].colormask;
glColorMaskIndexedEXT(i, state->rt[i].colormask & PIPE_MASK_R ? GL_TRUE : GL_FALSE,
state->rt[i].colormask & PIPE_MASK_G ? GL_TRUE : GL_FALSE,
state->rt[i].colormask & PIPE_MASK_B ? GL_TRUE : GL_FALSE,
state->rt[i].colormask & PIPE_MASK_A ? GL_TRUE : GL_FALSE);
}
}
} else {
if (state->rt[0].blend_enable) {
glBlendFuncSeparate(translate_blend_factor(state->rt[0].rgb_src_factor),
translate_blend_factor(state->rt[0].rgb_dst_factor),
translate_blend_factor(state->rt[0].alpha_src_factor),
translate_blend_factor(state->rt[0].alpha_dst_factor));
glBlendEquationSeparate(translate_blend_func(state->rt[0].rgb_func),
translate_blend_func(state->rt[0].alpha_func));
vrend_blend_enable(ctx, true);
}
else
vrend_blend_enable(ctx, false);
if (state->rt[0].colormask != ctx->sub->hw_blend_state.rt[0].colormask) {
int i;
for (i = 0; i < PIPE_MAX_COLOR_BUFS; i++)
ctx->sub->hw_blend_state.rt[i].colormask = state->rt[i].colormask;
glColorMask(state->rt[0].colormask & PIPE_MASK_R ? GL_TRUE : GL_FALSE,
state->rt[0].colormask & PIPE_MASK_G ? GL_TRUE : GL_FALSE,
state->rt[0].colormask & PIPE_MASK_B ? GL_TRUE : GL_FALSE,
state->rt[0].colormask & PIPE_MASK_A ? GL_TRUE : GL_FALSE);
}
}
if (vrend_state.have_multisample) {
if (state->alpha_to_coverage)
glEnable(GL_SAMPLE_ALPHA_TO_COVERAGE);
else
glDisable(GL_SAMPLE_ALPHA_TO_COVERAGE);
if (state->alpha_to_one)
glEnable(GL_SAMPLE_ALPHA_TO_ONE);
else
glDisable(GL_SAMPLE_ALPHA_TO_ONE);
}
if (state->dither)
glEnable(GL_DITHER);
else
glDisable(GL_DITHER);
}
/* there are a few reasons we might need to patch the blend state.
a) patching blend factors for dst with no alpha
b) patching colormask/blendcolor/blendfactors for A8/A16 format
emulation using GL_R8/GL_R16.
*/
static void vrend_patch_blend_state(struct vrend_context *ctx)
{
struct pipe_blend_state new_state = ctx->sub->blend_state;
struct pipe_blend_state *state = &ctx->sub->blend_state;
bool dest_alpha_only = false, dest_has_no_alpha = false;
struct pipe_blend_color blend_color = ctx->sub->blend_color;
int i;
if (ctx->sub->nr_cbufs == 0)
return;
for (i = 0; i < ctx->sub->nr_cbufs; i++) {
if (!ctx->sub->surf[i])
continue;
if (vrend_format_is_emulated_alpha(ctx->sub->surf[i]->format)) {
dest_alpha_only = true;
}
if (!util_format_has_alpha(ctx->sub->surf[i]->format)) {
dest_has_no_alpha = true;
}
}
if (dest_alpha_only) {
for (i = 0; i < (state->independent_blend_enable ? PIPE_MAX_COLOR_BUFS : 1); i++) {
if (state->rt[i].blend_enable) {
new_state.rt[i].rgb_src_factor = conv_a8_blend(state->rt[i].alpha_src_factor);
new_state.rt[i].rgb_dst_factor = conv_a8_blend(state->rt[i].alpha_dst_factor);
new_state.rt[i].alpha_src_factor = PIPE_BLENDFACTOR_ZERO;
new_state.rt[i].alpha_dst_factor = PIPE_BLENDFACTOR_ZERO;
}
new_state.rt[i].colormask = 0;
if (state->rt[i].colormask & PIPE_MASK_A)
new_state.rt[i].colormask |= PIPE_MASK_R;
}
blend_color.color[0] = blend_color.color[3];
blend_color.color[1] = 0.0f;
blend_color.color[2] = 0.0f;
blend_color.color[3] = 0.0f;
} else if (dest_has_no_alpha) {
for (i = 0; i < (state->independent_blend_enable ? PIPE_MAX_COLOR_BUFS : 1); i++) {
if (!(is_dst_blend(state->rt[i].rgb_src_factor) ||
is_dst_blend(state->rt[i].rgb_dst_factor) ||
is_dst_blend(state->rt[i].alpha_src_factor) ||
is_dst_blend(state->rt[i].alpha_dst_factor)))
continue;
new_state.rt[i].rgb_src_factor = conv_dst_blend(state->rt[i].rgb_src_factor);
new_state.rt[i].rgb_dst_factor = conv_dst_blend(state->rt[i].rgb_dst_factor);
new_state.rt[i].alpha_src_factor = conv_dst_blend(state->rt[i].alpha_src_factor);
new_state.rt[i].alpha_dst_factor = conv_dst_blend(state->rt[i].alpha_dst_factor);
}
}
vrend_hw_emit_blend(ctx, &new_state);
glBlendColor(blend_color.color[0],
blend_color.color[1],
blend_color.color[2],
blend_color.color[3]);
}
void vrend_object_bind_blend(struct vrend_context *ctx,
uint32_t handle)
{
struct pipe_blend_state *state;
if (handle == 0) {
memset(&ctx->sub->blend_state, 0, sizeof(ctx->sub->blend_state));
vrend_blend_enable(ctx, false);
return;
}
state = vrend_object_lookup(ctx->sub->object_hash, handle, VIRGL_OBJECT_BLEND);
if (!state) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_HANDLE, handle);
return;
}
ctx->sub->blend_state = *state;
vrend_hw_emit_blend(ctx, &ctx->sub->blend_state);
}
static void vrend_hw_emit_dsa(struct vrend_context *ctx)
{
struct pipe_depth_stencil_alpha_state *state = &ctx->sub->dsa_state;
if (state->depth.enabled) {
vrend_depth_test_enable(ctx, true);
glDepthFunc(GL_NEVER + state->depth.func);
if (state->depth.writemask)
glDepthMask(GL_TRUE);
else
glDepthMask(GL_FALSE);
} else
vrend_depth_test_enable(ctx, false);
if (state->alpha.enabled) {
vrend_alpha_test_enable(ctx, true);
if (!vrend_state.use_core_profile)
glAlphaFunc(GL_NEVER + state->alpha.func, state->alpha.ref_value);
} else
vrend_alpha_test_enable(ctx, false);
}
void vrend_object_bind_dsa(struct vrend_context *ctx,
uint32_t handle)
{
struct pipe_depth_stencil_alpha_state *state;
if (handle == 0) {
memset(&ctx->sub->dsa_state, 0, sizeof(ctx->sub->dsa_state));
ctx->sub->dsa = NULL;
ctx->sub->stencil_state_dirty = true;
ctx->sub->shader_dirty = true;
vrend_hw_emit_dsa(ctx);
return;
}
state = vrend_object_lookup(ctx->sub->object_hash, handle, VIRGL_OBJECT_DSA);
if (!state) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_HANDLE, handle);
return;
}
if (ctx->sub->dsa != state) {
ctx->sub->stencil_state_dirty = true;
ctx->sub->shader_dirty = true;
}
ctx->sub->dsa_state = *state;
ctx->sub->dsa = state;
vrend_hw_emit_dsa(ctx);
}
static void vrend_update_frontface_state(struct vrend_context *ctx)
{
struct pipe_rasterizer_state *state = &ctx->sub->rs_state;
int front_ccw = state->front_ccw;
front_ccw ^= (ctx->sub->inverted_fbo_content ? 0 : 1);
if (front_ccw)
glFrontFace(GL_CCW);
else
glFrontFace(GL_CW);
}
void vrend_update_stencil_state(struct vrend_context *ctx)
{
struct pipe_depth_stencil_alpha_state *state = ctx->sub->dsa;
int i;
if (!state)
return;
if (!state->stencil[1].enabled) {
if (state->stencil[0].enabled) {
vrend_stencil_test_enable(ctx, true);
glStencilOp(translate_stencil_op(state->stencil[0].fail_op),
translate_stencil_op(state->stencil[0].zfail_op),
translate_stencil_op(state->stencil[0].zpass_op));
glStencilFunc(GL_NEVER + state->stencil[0].func,
ctx->sub->stencil_refs[0],
state->stencil[0].valuemask);
glStencilMask(state->stencil[0].writemask);
} else
vrend_stencil_test_enable(ctx, false);
} else {
vrend_stencil_test_enable(ctx, true);
for (i = 0; i < 2; i++) {
GLenum face = (i == 1) ? GL_BACK : GL_FRONT;
glStencilOpSeparate(face,
translate_stencil_op(state->stencil[i].fail_op),
translate_stencil_op(state->stencil[i].zfail_op),
translate_stencil_op(state->stencil[i].zpass_op));
glStencilFuncSeparate(face, GL_NEVER + state->stencil[i].func,
ctx->sub->stencil_refs[i],
state->stencil[i].valuemask);
glStencilMaskSeparate(face, state->stencil[i].writemask);
}
}
ctx->sub->stencil_state_dirty = false;
}
static inline GLenum translate_fill(uint32_t mode)
{
switch (mode) {
case PIPE_POLYGON_MODE_POINT:
return GL_POINT;
case PIPE_POLYGON_MODE_LINE:
return GL_LINE;
case PIPE_POLYGON_MODE_FILL:
return GL_FILL;
default:
assert(0);
return 0;
}
}
static void vrend_hw_emit_rs(struct vrend_context *ctx)
{
struct pipe_rasterizer_state *state = &ctx->sub->rs_state;
int i;
if (vrend_state.use_gles) {
if (!state->depth_clip) {
report_gles_warn(ctx, GLES_WARN_DEPTH_CLIP, 0);
}
} else if (state->depth_clip) {
glDisable(GL_DEPTH_CLAMP);
} else {
glEnable(GL_DEPTH_CLAMP);
}
if (vrend_state.use_gles) {
/* guest send invalid glPointSize parameter */
if (!state->point_size_per_vertex &&
state->point_size != 1.0f &&
state->point_size != 0.0f) {
report_gles_warn(ctx, GLES_WARN_POINT_SIZE, 0);
}
} else if (state->point_size_per_vertex) {
glEnable(GL_PROGRAM_POINT_SIZE);
} else {
glDisable(GL_PROGRAM_POINT_SIZE);
if (state->point_size) {
glPointSize(state->point_size);
}
}
/* line_width < 0 is invalid, the guest sometimes forgot to set it. */
glLineWidth(state->line_width <= 0 ? 1.0f : state->line_width);
if (state->rasterizer_discard != ctx->sub->hw_rs_state.rasterizer_discard) {
ctx->sub->hw_rs_state.rasterizer_discard = state->rasterizer_discard;
if (state->rasterizer_discard)
glEnable(GL_RASTERIZER_DISCARD);
else
glDisable(GL_RASTERIZER_DISCARD);
}
if (vrend_state.use_gles == true) {
if (translate_fill(state->fill_front) != GL_FILL) {
report_gles_warn(ctx, GLES_WARN_POLYGON_MODE, 0);
}
if (translate_fill(state->fill_back) != GL_FILL) {
report_gles_warn(ctx, GLES_WARN_POLYGON_MODE, 0);
}
} else if (vrend_state.use_core_profile == false) {
glPolygonMode(GL_FRONT, translate_fill(state->fill_front));
glPolygonMode(GL_BACK, translate_fill(state->fill_back));
} else if (state->fill_front == state->fill_back) {
glPolygonMode(GL_FRONT_AND_BACK, translate_fill(state->fill_front));
} else
report_core_warn(ctx, CORE_PROFILE_WARN_POLYGON_MODE, 0);
if (state->offset_tri) {
glEnable(GL_POLYGON_OFFSET_FILL);
} else {
glDisable(GL_POLYGON_OFFSET_FILL);
}
if (vrend_state.use_gles) {
if (state->offset_line) {
report_gles_warn(ctx, GLES_WARN_OFFSET_LINE, 0);
}
} else if (state->offset_line) {
glEnable(GL_POLYGON_OFFSET_LINE);
} else {
glDisable(GL_POLYGON_OFFSET_LINE);
}
if (vrend_state.use_gles) {
if (state->offset_point) {
report_gles_warn(ctx, GLES_WARN_OFFSET_POINT, 0);
}
} else if (state->offset_point) {
glEnable(GL_POLYGON_OFFSET_POINT);
} else {
glDisable(GL_POLYGON_OFFSET_POINT);
}
if (state->flatshade != ctx->sub->hw_rs_state.flatshade) {
ctx->sub->hw_rs_state.flatshade = state->flatshade;
if (vrend_state.use_core_profile == false) {
if (state->flatshade) {
glShadeModel(GL_FLAT);
} else {
glShadeModel(GL_SMOOTH);
}
}
}
if (state->flatshade_first != ctx->sub->hw_rs_state.flatshade_first) {
ctx->sub->hw_rs_state.flatshade_first = state->flatshade_first;
if (vrend_state.use_gles) {
if (state->flatshade_first) {
report_gles_warn(ctx, GLES_WARN_FLATSHADE_FIRST, 0);
}
} else if (state->flatshade_first) {
glProvokingVertexEXT(GL_FIRST_VERTEX_CONVENTION_EXT);
} else {
glProvokingVertexEXT(GL_LAST_VERTEX_CONVENTION_EXT);
}
}
glPolygonOffset(state->offset_scale, state->offset_units);
if (vrend_state.use_core_profile == false) {
if (state->poly_stipple_enable)
glEnable(GL_POLYGON_STIPPLE);
else
glDisable(GL_POLYGON_STIPPLE);
} else if (state->poly_stipple_enable) {
if (!ctx->pstip_inited)
vrend_init_pstipple_texture(ctx);
}
if (state->point_quad_rasterization) {
if (vrend_state.use_core_profile == false &&
vrend_state.use_gles == false) {
glEnable(GL_POINT_SPRITE);
}
if (vrend_state.use_gles == false) {
glPointParameteri(GL_POINT_SPRITE_COORD_ORIGIN, state->sprite_coord_mode ? GL_UPPER_LEFT : GL_LOWER_LEFT);
}
} else {
if (vrend_state.use_core_profile == false &&
vrend_state.use_gles == false) {
glDisable(GL_POINT_SPRITE);
}
}
if (state->cull_face != PIPE_FACE_NONE) {
switch (state->cull_face) {
case PIPE_FACE_FRONT:
glCullFace(GL_FRONT);
break;
case PIPE_FACE_BACK:
glCullFace(GL_BACK);
break;
case PIPE_FACE_FRONT_AND_BACK:
glCullFace(GL_FRONT_AND_BACK);
break;
default:
fprintf(stderr, "unhandled cull-face: %x\n", state->cull_face);
}
glEnable(GL_CULL_FACE);
} else
glDisable(GL_CULL_FACE);
/* two sided lighting handled in shader for core profile */
if (vrend_state.use_core_profile == false) {
if (state->light_twoside)
glEnable(GL_VERTEX_PROGRAM_TWO_SIDE);
else
glDisable(GL_VERTEX_PROGRAM_TWO_SIDE);
}
if (state->clip_plane_enable != ctx->sub->hw_rs_state.clip_plane_enable) {
ctx->sub->hw_rs_state.clip_plane_enable = state->clip_plane_enable;
for (i = 0; i < 8; i++) {
if (state->clip_plane_enable & (1 << i))
glEnable(GL_CLIP_PLANE0 + i);
else
glDisable(GL_CLIP_PLANE0 + i);
}
}
if (vrend_state.use_core_profile == false) {
glLineStipple(state->line_stipple_factor, state->line_stipple_pattern);
if (state->line_stipple_enable)
glEnable(GL_LINE_STIPPLE);
else
glDisable(GL_LINE_STIPPLE);
} else if (state->line_stipple_enable) {
if (vrend_state.use_gles)
report_core_warn(ctx, GLES_WARN_STIPPLE, 0);
else
report_core_warn(ctx, CORE_PROFILE_WARN_STIPPLE, 0);
}
if (vrend_state.use_gles) {
if (state->line_smooth) {
report_gles_warn(ctx, GLES_WARN_LINE_SMOOTH, 0);
}
} else if (state->line_smooth) {
glEnable(GL_LINE_SMOOTH);
} else {
glDisable(GL_LINE_SMOOTH);
}
if (vrend_state.use_gles) {
if (state->poly_smooth) {
report_gles_warn(ctx, GLES_WARN_POLY_SMOOTH, 0);
}
} else if (state->poly_smooth) {
glEnable(GL_POLYGON_SMOOTH);
} else {
glDisable(GL_POLYGON_SMOOTH);
}
if (vrend_state.use_core_profile == false) {
if (state->clamp_vertex_color)
glClampColor(GL_CLAMP_VERTEX_COLOR_ARB, GL_TRUE);
else
glClampColor(GL_CLAMP_VERTEX_COLOR_ARB, GL_FALSE);
if (state->clamp_fragment_color)
glClampColor(GL_CLAMP_FRAGMENT_COLOR_ARB, GL_TRUE);
else
glClampColor(GL_CLAMP_FRAGMENT_COLOR_ARB, GL_FALSE);
} else {
if (state->clamp_vertex_color || state->clamp_fragment_color)
report_core_warn(ctx, CORE_PROFILE_WARN_CLAMP, 0);
}
if (vrend_state.have_multisample) {
if (state->multisample) {
glEnable(GL_MULTISAMPLE);
glEnable(GL_SAMPLE_MASK);
} else {
glDisable(GL_MULTISAMPLE);
glDisable(GL_SAMPLE_MASK);
}
if (vrend_state.have_sample_shading) {
if (state->force_persample_interp)
glEnable(GL_SAMPLE_SHADING);
else
glDisable(GL_SAMPLE_SHADING);
}
}
}
void vrend_object_bind_rasterizer(struct vrend_context *ctx,
uint32_t handle)
{
struct pipe_rasterizer_state *state;
if (handle == 0) {
memset(&ctx->sub->rs_state, 0, sizeof(ctx->sub->rs_state));
return;
}
state = vrend_object_lookup(ctx->sub->object_hash, handle, VIRGL_OBJECT_RASTERIZER);
if (!state) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_HANDLE, handle);
return;
}
ctx->sub->rs_state = *state;
ctx->sub->scissor_state_dirty = (1 << 0);
ctx->sub->shader_dirty = true;
vrend_hw_emit_rs(ctx);
}
void vrend_bind_sampler_states(struct vrend_context *ctx,
uint32_t shader_type,
uint32_t start_slot,
uint32_t num_states,
uint32_t *handles)
{
int i;
struct vrend_sampler_state *state;
if (shader_type >= PIPE_SHADER_TYPES) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_CMD_BUFFER, shader_type);
return;
}
if (num_states > PIPE_MAX_SAMPLERS ||
start_slot > (PIPE_MAX_SAMPLERS - num_states)) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_CMD_BUFFER, num_states);
return;
}
ctx->sub->num_sampler_states[shader_type] = num_states;
for (i = 0; i < num_states; i++) {
if (handles[i] == 0)
state = NULL;
else
state = vrend_object_lookup(ctx->sub->object_hash, handles[i], VIRGL_OBJECT_SAMPLER_STATE);
ctx->sub->sampler_state[shader_type][i + start_slot] = state;
}
ctx->sub->sampler_state_dirty = true;
}
static void vrend_apply_sampler_state(struct vrend_context *ctx,
struct vrend_resource *res,
uint32_t shader_type,
int id,
int sampler_id,
uint32_t srgb_decode)
{
struct vrend_texture *tex = (struct vrend_texture *)res;
struct vrend_sampler_state *vstate = ctx->sub->sampler_state[shader_type][id];
struct pipe_sampler_state *state = &vstate->base;
bool set_all = false;
GLenum target = tex->base.target;
if (!state) {
fprintf(stderr, "cannot find sampler state for %d %d\n", shader_type, id);
return;
}
if (res->base.nr_samples > 1) {
tex->state = *state;
return;
}
if (target == GL_TEXTURE_BUFFER) {
tex->state = *state;
return;
}
if (vrend_state.have_samplers) {
glBindSampler(sampler_id, vstate->id);
glSamplerParameteri(vstate->id, GL_TEXTURE_SRGB_DECODE_EXT,
srgb_decode);
return;
}
if (tex->state.max_lod == -1)
set_all = true;
if (tex->state.wrap_s != state->wrap_s || set_all)
glTexParameteri(target, GL_TEXTURE_WRAP_S, convert_wrap(state->wrap_s));
if (tex->state.wrap_t != state->wrap_t || set_all)
glTexParameteri(target, GL_TEXTURE_WRAP_T, convert_wrap(state->wrap_t));
if (tex->state.wrap_r != state->wrap_r || set_all)
glTexParameteri(target, GL_TEXTURE_WRAP_R, convert_wrap(state->wrap_r));
if (tex->state.min_img_filter != state->min_img_filter ||
tex->state.min_mip_filter != state->min_mip_filter || set_all)
glTexParameterf(target, GL_TEXTURE_MIN_FILTER, convert_min_filter(state->min_img_filter, state->min_mip_filter));
if (tex->state.mag_img_filter != state->mag_img_filter || set_all)
glTexParameterf(target, GL_TEXTURE_MAG_FILTER, convert_mag_filter(state->mag_img_filter));
if (res->target != GL_TEXTURE_RECTANGLE) {
if (tex->state.min_lod != state->min_lod || set_all)
glTexParameterf(target, GL_TEXTURE_MIN_LOD, state->min_lod);
if (tex->state.max_lod != state->max_lod || set_all)
glTexParameterf(target, GL_TEXTURE_MAX_LOD, state->max_lod);
if (tex->state.lod_bias != state->lod_bias || set_all) {
if (vrend_state.use_gles) {
if (state->lod_bias) {
report_gles_warn(ctx, GLES_WARN_LOD_BIAS, 0);
}
} else {
glTexParameterf(target, GL_TEXTURE_LOD_BIAS, state->lod_bias);
}
}
}
if (tex->state.compare_mode != state->compare_mode || set_all)
glTexParameteri(target, GL_TEXTURE_COMPARE_MODE, state->compare_mode ? GL_COMPARE_R_TO_TEXTURE : GL_NONE);
if (tex->state.compare_func != state->compare_func || set_all)
glTexParameteri(target, GL_TEXTURE_COMPARE_FUNC, GL_NEVER + state->compare_func);
/*
* Oh this is a fun one. On GLES 2.0 all cubemap MUST NOT be seamless.
* But on GLES 3.0 all cubemaps MUST be seamless. Either way there is no
* way to toggle between the behaviour when running on GLES. And adding
* warnings will spew the logs quite bad. Ignore and hope for the best.
*/
if (!vrend_state.use_gles) {
if (state->seamless_cube_map) {
glEnable(GL_TEXTURE_CUBE_MAP_SEAMLESS);
} else {
glDisable(GL_TEXTURE_CUBE_MAP_SEAMLESS);
}
}
if (memcmp(&tex->state.border_color, &state->border_color, 16) || set_all)
glTexParameterIuiv(target, GL_TEXTURE_BORDER_COLOR, state->border_color.ui);
tex->state = *state;
}
static GLenum tgsitargettogltarget(const enum pipe_texture_target target, int nr_samples)
{
switch(target) {
case PIPE_TEXTURE_1D:
return GL_TEXTURE_1D;
case PIPE_TEXTURE_2D:
return (nr_samples > 1) ? GL_TEXTURE_2D_MULTISAMPLE : GL_TEXTURE_2D;
case PIPE_TEXTURE_3D:
return GL_TEXTURE_3D;
case PIPE_TEXTURE_RECT:
return GL_TEXTURE_RECTANGLE_NV;
case PIPE_TEXTURE_CUBE:
return GL_TEXTURE_CUBE_MAP;
case PIPE_TEXTURE_1D_ARRAY:
return GL_TEXTURE_1D_ARRAY;
case PIPE_TEXTURE_2D_ARRAY:
return (nr_samples > 1) ? GL_TEXTURE_2D_MULTISAMPLE_ARRAY : GL_TEXTURE_2D_ARRAY;
case PIPE_TEXTURE_CUBE_ARRAY:
return GL_TEXTURE_CUBE_MAP_ARRAY;
case PIPE_BUFFER:
default:
return PIPE_BUFFER;
}
return PIPE_BUFFER;
}
renderer: use a thread to block for fences. Instead of polling the fences regularly, have a thread that blocks for a single fence using a separate shared context, then uses eventfd to wake up the main thread when something happens. Inside the guest, glmark2 typicially runs twice as fast with the thread sync. Although in general, the performances seems to be about +30%. The benefits is mostly for CPU-bounds tasks (when main the thread hits 100%) A naive perf stat of the vtest renderer with glmark2 "build" test with a fixed number of frames (500) results in the following stats data: (do not value timing related informations, since the renderer is ran and stopped manually) without thread: 3032.282265 task-clock (msec) # 0.420 CPUs utilized 4,277 context-switches # 0.001 M/sec 102 cpu-migrations # 0.034 K/sec 9,020 page-faults # 0.003 M/sec 7,884,098,254 cycles # 2.600 GHz 4,440,126,451 stalled-cycles-frontend # 56.32% frontend cycles idle <not supported> stalled-cycles-backend 11,024,091,578 instructions # 1.40 insns per cycle # 0.40 stalled # cycles per insn 1,091,831,588 branches # 360.069 M/sec 5,426,846 branch-misses # 0.50% of all branches with thread: 3403.592921 task-clock (msec) # 0.452 CPUs utilized 7,145 context-switches # 0.002 M/sec 410 cpu-migrations # 0.120 K/sec 6,191 page-faults # 0.002 M/sec 7,475,038,064 cycles # 2.196 GHz 4,487,043,071 stalled-cycles-frontend # 60.03% frontend cycles idle <not supported> stalled-cycles-backend 9,925,205,494 instructions # 1.33 insns per cycle # 0.45 stalled # cycles per insn 834,375,503 branches # 245.146 M/sec 4,919,995 branch-misses # 0.59% of all branches Signed-off-by: Marc-André Lureau <marcandre.lureau@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
9 years ago
static void vrend_free_sync_thread(void)
{
if (!vrend_state.sync_thread)
return;
pipe_mutex_lock(vrend_state.fence_mutex);
vrend_state.stop_sync_thread = true;
pipe_mutex_unlock(vrend_state.fence_mutex);
pipe_condvar_signal(vrend_state.fence_cond);
pipe_thread_wait(vrend_state.sync_thread);
vrend_state.sync_thread = 0;
pipe_condvar_destroy(vrend_state.fence_cond);
pipe_mutex_destroy(vrend_state.fence_mutex);
renderer: use a thread to block for fences. Instead of polling the fences regularly, have a thread that blocks for a single fence using a separate shared context, then uses eventfd to wake up the main thread when something happens. Inside the guest, glmark2 typicially runs twice as fast with the thread sync. Although in general, the performances seems to be about +30%. The benefits is mostly for CPU-bounds tasks (when main the thread hits 100%) A naive perf stat of the vtest renderer with glmark2 "build" test with a fixed number of frames (500) results in the following stats data: (do not value timing related informations, since the renderer is ran and stopped manually) without thread: 3032.282265 task-clock (msec) # 0.420 CPUs utilized 4,277 context-switches # 0.001 M/sec 102 cpu-migrations # 0.034 K/sec 9,020 page-faults # 0.003 M/sec 7,884,098,254 cycles # 2.600 GHz 4,440,126,451 stalled-cycles-frontend # 56.32% frontend cycles idle <not supported> stalled-cycles-backend 11,024,091,578 instructions # 1.40 insns per cycle # 0.40 stalled # cycles per insn 1,091,831,588 branches # 360.069 M/sec 5,426,846 branch-misses # 0.50% of all branches with thread: 3403.592921 task-clock (msec) # 0.452 CPUs utilized 7,145 context-switches # 0.002 M/sec 410 cpu-migrations # 0.120 K/sec 6,191 page-faults # 0.002 M/sec 7,475,038,064 cycles # 2.196 GHz 4,487,043,071 stalled-cycles-frontend # 60.03% frontend cycles idle <not supported> stalled-cycles-backend 9,925,205,494 instructions # 1.33 insns per cycle # 0.45 stalled # cycles per insn 834,375,503 branches # 245.146 M/sec 4,919,995 branch-misses # 0.59% of all branches Signed-off-by: Marc-André Lureau <marcandre.lureau@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
9 years ago
}
#ifdef HAVE_EVENTFD
renderer: use a thread to block for fences. Instead of polling the fences regularly, have a thread that blocks for a single fence using a separate shared context, then uses eventfd to wake up the main thread when something happens. Inside the guest, glmark2 typicially runs twice as fast with the thread sync. Although in general, the performances seems to be about +30%. The benefits is mostly for CPU-bounds tasks (when main the thread hits 100%) A naive perf stat of the vtest renderer with glmark2 "build" test with a fixed number of frames (500) results in the following stats data: (do not value timing related informations, since the renderer is ran and stopped manually) without thread: 3032.282265 task-clock (msec) # 0.420 CPUs utilized 4,277 context-switches # 0.001 M/sec 102 cpu-migrations # 0.034 K/sec 9,020 page-faults # 0.003 M/sec 7,884,098,254 cycles # 2.600 GHz 4,440,126,451 stalled-cycles-frontend # 56.32% frontend cycles idle <not supported> stalled-cycles-backend 11,024,091,578 instructions # 1.40 insns per cycle # 0.40 stalled # cycles per insn 1,091,831,588 branches # 360.069 M/sec 5,426,846 branch-misses # 0.50% of all branches with thread: 3403.592921 task-clock (msec) # 0.452 CPUs utilized 7,145 context-switches # 0.002 M/sec 410 cpu-migrations # 0.120 K/sec 6,191 page-faults # 0.002 M/sec 7,475,038,064 cycles # 2.196 GHz 4,487,043,071 stalled-cycles-frontend # 60.03% frontend cycles idle <not supported> stalled-cycles-backend 9,925,205,494 instructions # 1.33 insns per cycle # 0.45 stalled # cycles per insn 834,375,503 branches # 245.146 M/sec 4,919,995 branch-misses # 0.59% of all branches Signed-off-by: Marc-André Lureau <marcandre.lureau@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
9 years ago
static ssize_t
write_full(int fd, const void *ptr, size_t count)
renderer: use a thread to block for fences. Instead of polling the fences regularly, have a thread that blocks for a single fence using a separate shared context, then uses eventfd to wake up the main thread when something happens. Inside the guest, glmark2 typicially runs twice as fast with the thread sync. Although in general, the performances seems to be about +30%. The benefits is mostly for CPU-bounds tasks (when main the thread hits 100%) A naive perf stat of the vtest renderer with glmark2 "build" test with a fixed number of frames (500) results in the following stats data: (do not value timing related informations, since the renderer is ran and stopped manually) without thread: 3032.282265 task-clock (msec) # 0.420 CPUs utilized 4,277 context-switches # 0.001 M/sec 102 cpu-migrations # 0.034 K/sec 9,020 page-faults # 0.003 M/sec 7,884,098,254 cycles # 2.600 GHz 4,440,126,451 stalled-cycles-frontend # 56.32% frontend cycles idle <not supported> stalled-cycles-backend 11,024,091,578 instructions # 1.40 insns per cycle # 0.40 stalled # cycles per insn 1,091,831,588 branches # 360.069 M/sec 5,426,846 branch-misses # 0.50% of all branches with thread: 3403.592921 task-clock (msec) # 0.452 CPUs utilized 7,145 context-switches # 0.002 M/sec 410 cpu-migrations # 0.120 K/sec 6,191 page-faults # 0.002 M/sec 7,475,038,064 cycles # 2.196 GHz 4,487,043,071 stalled-cycles-frontend # 60.03% frontend cycles idle <not supported> stalled-cycles-backend 9,925,205,494 instructions # 1.33 insns per cycle # 0.45 stalled # cycles per insn 834,375,503 branches # 245.146 M/sec 4,919,995 branch-misses # 0.59% of all branches Signed-off-by: Marc-André Lureau <marcandre.lureau@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
9 years ago
{
const char *buf = ptr;
ssize_t ret = 0;
ssize_t total = 0;
renderer: use a thread to block for fences. Instead of polling the fences regularly, have a thread that blocks for a single fence using a separate shared context, then uses eventfd to wake up the main thread when something happens. Inside the guest, glmark2 typicially runs twice as fast with the thread sync. Although in general, the performances seems to be about +30%. The benefits is mostly for CPU-bounds tasks (when main the thread hits 100%) A naive perf stat of the vtest renderer with glmark2 "build" test with a fixed number of frames (500) results in the following stats data: (do not value timing related informations, since the renderer is ran and stopped manually) without thread: 3032.282265 task-clock (msec) # 0.420 CPUs utilized 4,277 context-switches # 0.001 M/sec 102 cpu-migrations # 0.034 K/sec 9,020 page-faults # 0.003 M/sec 7,884,098,254 cycles # 2.600 GHz 4,440,126,451 stalled-cycles-frontend # 56.32% frontend cycles idle <not supported> stalled-cycles-backend 11,024,091,578 instructions # 1.40 insns per cycle # 0.40 stalled # cycles per insn 1,091,831,588 branches # 360.069 M/sec 5,426,846 branch-misses # 0.50% of all branches with thread: 3403.592921 task-clock (msec) # 0.452 CPUs utilized 7,145 context-switches # 0.002 M/sec 410 cpu-migrations # 0.120 K/sec 6,191 page-faults # 0.002 M/sec 7,475,038,064 cycles # 2.196 GHz 4,487,043,071 stalled-cycles-frontend # 60.03% frontend cycles idle <not supported> stalled-cycles-backend 9,925,205,494 instructions # 1.33 insns per cycle # 0.45 stalled # cycles per insn 834,375,503 branches # 245.146 M/sec 4,919,995 branch-misses # 0.59% of all branches Signed-off-by: Marc-André Lureau <marcandre.lureau@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
9 years ago
while (count) {
ret = write(fd, buf, count);
if (ret < 0) {
if (errno == EINTR)
continue;
break;
}
count -= ret;
buf += ret;
total += ret;
}
return total;
renderer: use a thread to block for fences. Instead of polling the fences regularly, have a thread that blocks for a single fence using a separate shared context, then uses eventfd to wake up the main thread when something happens. Inside the guest, glmark2 typicially runs twice as fast with the thread sync. Although in general, the performances seems to be about +30%. The benefits is mostly for CPU-bounds tasks (when main the thread hits 100%) A naive perf stat of the vtest renderer with glmark2 "build" test with a fixed number of frames (500) results in the following stats data: (do not value timing related informations, since the renderer is ran and stopped manually) without thread: 3032.282265 task-clock (msec) # 0.420 CPUs utilized 4,277 context-switches # 0.001 M/sec 102 cpu-migrations # 0.034 K/sec 9,020 page-faults # 0.003 M/sec 7,884,098,254 cycles # 2.600 GHz 4,440,126,451 stalled-cycles-frontend # 56.32% frontend cycles idle <not supported> stalled-cycles-backend 11,024,091,578 instructions # 1.40 insns per cycle # 0.40 stalled # cycles per insn 1,091,831,588 branches # 360.069 M/sec 5,426,846 branch-misses # 0.50% of all branches with thread: 3403.592921 task-clock (msec) # 0.452 CPUs utilized 7,145 context-switches # 0.002 M/sec 410 cpu-migrations # 0.120 K/sec 6,191 page-faults # 0.002 M/sec 7,475,038,064 cycles # 2.196 GHz 4,487,043,071 stalled-cycles-frontend # 60.03% frontend cycles idle <not supported> stalled-cycles-backend 9,925,205,494 instructions # 1.33 insns per cycle # 0.45 stalled # cycles per insn 834,375,503 branches # 245.146 M/sec 4,919,995 branch-misses # 0.59% of all branches Signed-off-by: Marc-André Lureau <marcandre.lureau@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
9 years ago
}
static void wait_sync(struct vrend_fence *fence)
{
GLenum glret;
ssize_t n;
uint64_t value = 1;
do {
glret = glClientWaitSync(fence->syncobj, 0, 1000000000);
switch (glret) {
case GL_WAIT_FAILED:
fprintf(stderr, "wait sync failed: illegal fence object %p\n", fence->syncobj);
break;
case GL_ALREADY_SIGNALED:
case GL_CONDITION_SATISFIED:
break;
default:
break;
}
} while (glret == GL_TIMEOUT_EXPIRED);
pipe_mutex_lock(vrend_state.fence_mutex);
list_addtail(&fence->fences, &vrend_state.fence_list);
pipe_mutex_unlock(vrend_state.fence_mutex);
n = write_full(vrend_state.eventfd, &value, sizeof(value));
if (n != sizeof(value)) {
perror("failed to write to eventfd\n");
}
}
static int thread_sync(void *arg)
{
virgl_gl_context gl_context = vrend_state.sync_context;
struct vrend_fence *fence, *stor;
pipe_mutex_lock(vrend_state.fence_mutex);
vrend_clicbs->make_current(0, gl_context);
while (!vrend_state.stop_sync_thread) {
if (LIST_IS_EMPTY(&vrend_state.fence_wait_list) &&
pipe_condvar_wait(vrend_state.fence_cond, vrend_state.fence_mutex) != 0) {
fprintf(stderr, "error while waiting on condition\n");
break;
}
LIST_FOR_EACH_ENTRY_SAFE(fence, stor, &vrend_state.fence_wait_list, fences) {
if (vrend_state.stop_sync_thread)
break;
list_del(&fence->fences);
pipe_mutex_unlock(vrend_state.fence_mutex);
wait_sync(fence);
pipe_mutex_lock(vrend_state.fence_mutex);
}
}
vrend_clicbs->make_current(0, 0);
vrend_clicbs->destroy_gl_context(vrend_state.sync_context);
pipe_mutex_unlock(vrend_state.fence_mutex);
return 0;
}
static void vrend_renderer_use_threaded_sync(void)
{
struct virgl_gl_ctx_param ctx_params;
if (getenv("VIRGL_DISABLE_MT"))
return;
ctx_params.shared = true;
ctx_params.major_ver = vrend_state.gl_major_ver;
ctx_params.minor_ver = vrend_state.gl_minor_ver;
vrend_state.stop_sync_thread = false;
vrend_state.sync_context = vrend_clicbs->create_gl_context(0, &ctx_params);
if (vrend_state.sync_context == NULL) {
fprintf(stderr, "failed to create sync opengl context\n");
return;
}
vrend_state.eventfd = eventfd(0, EFD_CLOEXEC | EFD_NONBLOCK);
if (vrend_state.eventfd == -1) {
fprintf(stderr, "Failed to create eventfd\n");
vrend_clicbs->destroy_gl_context(vrend_state.sync_context);
return;
}
pipe_condvar_init(vrend_state.fence_cond);
pipe_mutex_init(vrend_state.fence_mutex);
vrend_state.sync_thread = pipe_thread_create(thread_sync, NULL);
if (!vrend_state.sync_thread) {
close(vrend_state.eventfd);
vrend_state.eventfd = -1;
vrend_clicbs->destroy_gl_context(vrend_state.sync_context);
pipe_condvar_destroy(vrend_state.fence_cond);
pipe_mutex_destroy(vrend_state.fence_mutex);
renderer: use a thread to block for fences. Instead of polling the fences regularly, have a thread that blocks for a single fence using a separate shared context, then uses eventfd to wake up the main thread when something happens. Inside the guest, glmark2 typicially runs twice as fast with the thread sync. Although in general, the performances seems to be about +30%. The benefits is mostly for CPU-bounds tasks (when main the thread hits 100%) A naive perf stat of the vtest renderer with glmark2 "build" test with a fixed number of frames (500) results in the following stats data: (do not value timing related informations, since the renderer is ran and stopped manually) without thread: 3032.282265 task-clock (msec) # 0.420 CPUs utilized 4,277 context-switches # 0.001 M/sec 102 cpu-migrations # 0.034 K/sec 9,020 page-faults # 0.003 M/sec 7,884,098,254 cycles # 2.600 GHz 4,440,126,451 stalled-cycles-frontend # 56.32% frontend cycles idle <not supported> stalled-cycles-backend 11,024,091,578 instructions # 1.40 insns per cycle # 0.40 stalled # cycles per insn 1,091,831,588 branches # 360.069 M/sec 5,426,846 branch-misses # 0.50% of all branches with thread: 3403.592921 task-clock (msec) # 0.452 CPUs utilized 7,145 context-switches # 0.002 M/sec 410 cpu-migrations # 0.120 K/sec 6,191 page-faults # 0.002 M/sec 7,475,038,064 cycles # 2.196 GHz 4,487,043,071 stalled-cycles-frontend # 60.03% frontend cycles idle <not supported> stalled-cycles-backend 9,925,205,494 instructions # 1.33 insns per cycle # 0.45 stalled # cycles per insn 834,375,503 branches # 245.146 M/sec 4,919,995 branch-misses # 0.59% of all branches Signed-off-by: Marc-André Lureau <marcandre.lureau@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
9 years ago
}
}
#else
static void vrend_renderer_use_threaded_sync(void)
{
}
#endif
static void vrend_debug_cb(GLenum source, GLenum type, GLuint id,
GLenum severity, GLsizei length,
const GLchar* message, const void* userParam)
{
if (type != GL_DEBUG_TYPE_ERROR) {
return;
}
fprintf(stderr, "ERROR: %s\n", message);
}
renderer: use a thread to block for fences. Instead of polling the fences regularly, have a thread that blocks for a single fence using a separate shared context, then uses eventfd to wake up the main thread when something happens. Inside the guest, glmark2 typicially runs twice as fast with the thread sync. Although in general, the performances seems to be about +30%. The benefits is mostly for CPU-bounds tasks (when main the thread hits 100%) A naive perf stat of the vtest renderer with glmark2 "build" test with a fixed number of frames (500) results in the following stats data: (do not value timing related informations, since the renderer is ran and stopped manually) without thread: 3032.282265 task-clock (msec) # 0.420 CPUs utilized 4,277 context-switches # 0.001 M/sec 102 cpu-migrations # 0.034 K/sec 9,020 page-faults # 0.003 M/sec 7,884,098,254 cycles # 2.600 GHz 4,440,126,451 stalled-cycles-frontend # 56.32% frontend cycles idle <not supported> stalled-cycles-backend 11,024,091,578 instructions # 1.40 insns per cycle # 0.40 stalled # cycles per insn 1,091,831,588 branches # 360.069 M/sec 5,426,846 branch-misses # 0.50% of all branches with thread: 3403.592921 task-clock (msec) # 0.452 CPUs utilized 7,145 context-switches # 0.002 M/sec 410 cpu-migrations # 0.120 K/sec 6,191 page-faults # 0.002 M/sec 7,475,038,064 cycles # 2.196 GHz 4,487,043,071 stalled-cycles-frontend # 60.03% frontend cycles idle <not supported> stalled-cycles-backend 9,925,205,494 instructions # 1.33 insns per cycle # 0.45 stalled # cycles per insn 834,375,503 branches # 245.146 M/sec 4,919,995 branch-misses # 0.59% of all branches Signed-off-by: Marc-André Lureau <marcandre.lureau@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
9 years ago
int vrend_renderer_init(struct vrend_if_cbs *cbs, uint32_t flags)
{
bool gles;
int gl_ver;
virgl_gl_context gl_context;
struct virgl_gl_ctx_param ctx_params;
if (!vrend_state.inited) {
vrend_state.inited = true;
vrend_object_init_resource_table();
vrend_clicbs = cbs;
}
ctx_params.shared = false;
for (uint32_t i = 0; i < ARRAY_SIZE(gl_versions); i++) {
ctx_params.major_ver = gl_versions[i].major;
ctx_params.minor_ver = gl_versions[i].minor;
gl_context = vrend_clicbs->create_gl_context(0, &ctx_params);
if (gl_context)
break;
}
vrend_clicbs->make_current(0, gl_context);
gl_ver = epoxy_gl_version();
/* enable error output as early as possible */
if (vrend_use_debug_cb && epoxy_has_gl_extension("GL_KHR_debug")) {
glDebugMessageCallback(vrend_debug_cb, NULL);
glEnable(GL_DEBUG_OUTPUT);
glDisable(GL_DEBUG_OUTPUT_SYNCHRONOUS);
vrend_state.have_debug_cb = true;
}
/* make sure you have the latest version of libepoxy */
gles = epoxy_is_desktop_gl() == 0;
vrend_state.gl_major_ver = gl_ver / 10;
vrend_state.gl_minor_ver = gl_ver % 10;
if (gles) {
fprintf(stderr, "gl_version %d - es profile enabled\n", gl_ver);
vrend_state.use_gles = true;
/* for now, makes the rest of the code use the most GLES 3.x like path */
vrend_state.use_core_profile = 1;
} else if (gl_ver > 30 && !epoxy_has_gl_extension("GL_ARB_compatibility")) {
fprintf(stderr, "gl_version %d - core profile enabled\n", gl_ver);
vrend_state.use_core_profile = 1;
} else {
fprintf(stderr, "gl_version %d - compat profile\n", gl_ver);
}
if (epoxy_has_gl_extension("GL_ARB_robustness")) {
vrend_state.have_arb_robustness = true;
} else if (gles && epoxy_has_gl_extension("GL_KHR_robustness")) {
vrend_state.have_gles_khr_robustness = true;
} else {
fprintf(stderr,"WARNING: running without ARB/KHR robustness in place may crash\n");
}
/* used for buffers that we want to texture from */
if (epoxy_has_gl_extension("GL_ARB_texture_buffer_object")) {
vrend_state.have_arb_or_gles_ext_texture_buffer = true;
}
if (gles && epoxy_has_gl_extension("GL_EXT_texture_buffer")) {
vrend_state.have_arb_or_gles_ext_texture_buffer = true;
}
if (epoxy_has_gl_extension("GL_MESA_pack_invert"))
vrend_state.have_mesa_invert = true;
if (gl_ver >= 43 || epoxy_has_gl_extension("GL_ARB_vertex_attrib_binding"))
vrend_state.have_vertex_attrib_binding = true;
if (gl_ver >= 33 || epoxy_has_gl_extension("GL_ARB_sampler_objects"))
vrend_state.have_samplers = true;
if (gl_ver >= 33 || epoxy_has_gl_extension("GL_ARB_shader_bit_encoding"))
vrend_state.have_bit_encoding = true;
if (gl_ver >= 31)
vrend_state.have_gl_prim_restart = true;
else if (epoxy_has_gl_extension("GL_NV_primitive_restart"))
vrend_state.have_nv_prim_restart = true;
if (gl_ver >= 40 || epoxy_has_gl_extension("GL_ARB_transform_feedback2"))
vrend_state.have_tf2 = true;
if (epoxy_has_gl_extension("GL_ARB_stencil_texturing"))
vrend_state.have_stencil_texturing = true;
if (epoxy_has_gl_extension("GL_EXT_framebuffer_multisample") && epoxy_has_gl_extension("GL_ARB_texture_multisample")) {
vrend_state.have_multisample = true;
if (epoxy_has_gl_extension("GL_EXT_framebuffer_multisample_blit_scaled"))
vrend_state.have_ms_scaled_blit = true;
}
if (gl_ver >= 40 || epoxy_has_gl_extension("GL_ARB_sample_shading"))
vrend_state.have_sample_shading = true;
/* callbacks for when we are cleaning up the object table */
vrend_resource_set_destroy_callback(vrend_destroy_resource_object);
vrend_object_set_destroy_callback(VIRGL_OBJECT_QUERY, vrend_destroy_query_object);
vrend_object_set_destroy_callback(VIRGL_OBJECT_SURFACE, vrend_destroy_surface_object);
vrend_object_set_destroy_callback(VIRGL_OBJECT_SHADER, vrend_destroy_shader_object);
vrend_object_set_destroy_callback(VIRGL_OBJECT_SAMPLER_VIEW, vrend_destroy_sampler_view_object);
vrend_object_set_destroy_callback(VIRGL_OBJECT_STREAMOUT_TARGET, vrend_destroy_so_target_object);
vrend_object_set_destroy_callback(VIRGL_OBJECT_SAMPLER_STATE, vrend_destroy_sampler_state_object);
vrend_object_set_destroy_callback(VIRGL_OBJECT_VERTEX_ELEMENTS, vrend_destroy_vertex_elements_object);
/* disable for format testing, spews a lot of errors */
if (vrend_state.have_debug_cb) {
glDisable(GL_DEBUG_OUTPUT);
}
if (vrend_state.use_gles) {
vrend_build_format_list_gles();
} else {
vrend_build_format_list();
}
/* disable for format testing */
if (vrend_state.have_debug_cb) {
glDisable(GL_DEBUG_OUTPUT);
}
vrend_clicbs->destroy_gl_context(gl_context);
list_inithead(&vrend_state.fence_list);
renderer: use a thread to block for fences. Instead of polling the fences regularly, have a thread that blocks for a single fence using a separate shared context, then uses eventfd to wake up the main thread when something happens. Inside the guest, glmark2 typicially runs twice as fast with the thread sync. Although in general, the performances seems to be about +30%. The benefits is mostly for CPU-bounds tasks (when main the thread hits 100%) A naive perf stat of the vtest renderer with glmark2 "build" test with a fixed number of frames (500) results in the following stats data: (do not value timing related informations, since the renderer is ran and stopped manually) without thread: 3032.282265 task-clock (msec) # 0.420 CPUs utilized 4,277 context-switches # 0.001 M/sec 102 cpu-migrations # 0.034 K/sec 9,020 page-faults # 0.003 M/sec 7,884,098,254 cycles # 2.600 GHz 4,440,126,451 stalled-cycles-frontend # 56.32% frontend cycles idle <not supported> stalled-cycles-backend 11,024,091,578 instructions # 1.40 insns per cycle # 0.40 stalled # cycles per insn 1,091,831,588 branches # 360.069 M/sec 5,426,846 branch-misses # 0.50% of all branches with thread: 3403.592921 task-clock (msec) # 0.452 CPUs utilized 7,145 context-switches # 0.002 M/sec 410 cpu-migrations # 0.120 K/sec 6,191 page-faults # 0.002 M/sec 7,475,038,064 cycles # 2.196 GHz 4,487,043,071 stalled-cycles-frontend # 60.03% frontend cycles idle <not supported> stalled-cycles-backend 9,925,205,494 instructions # 1.33 insns per cycle # 0.45 stalled # cycles per insn 834,375,503 branches # 245.146 M/sec 4,919,995 branch-misses # 0.59% of all branches Signed-off-by: Marc-André Lureau <marcandre.lureau@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
9 years ago
list_inithead(&vrend_state.fence_wait_list);
list_inithead(&vrend_state.waiting_query_list);
list_inithead(&vrend_state.active_ctx_list);
/* create 0 context */
vrend_renderer_context_create_internal(0, 0, NULL);
renderer: use a thread to block for fences. Instead of polling the fences regularly, have a thread that blocks for a single fence using a separate shared context, then uses eventfd to wake up the main thread when something happens. Inside the guest, glmark2 typicially runs twice as fast with the thread sync. Although in general, the performances seems to be about +30%. The benefits is mostly for CPU-bounds tasks (when main the thread hits 100%) A naive perf stat of the vtest renderer with glmark2 "build" test with a fixed number of frames (500) results in the following stats data: (do not value timing related informations, since the renderer is ran and stopped manually) without thread: 3032.282265 task-clock (msec) # 0.420 CPUs utilized 4,277 context-switches # 0.001 M/sec 102 cpu-migrations # 0.034 K/sec 9,020 page-faults # 0.003 M/sec 7,884,098,254 cycles # 2.600 GHz 4,440,126,451 stalled-cycles-frontend # 56.32% frontend cycles idle <not supported> stalled-cycles-backend 11,024,091,578 instructions # 1.40 insns per cycle # 0.40 stalled # cycles per insn 1,091,831,588 branches # 360.069 M/sec 5,426,846 branch-misses # 0.50% of all branches with thread: 3403.592921 task-clock (msec) # 0.452 CPUs utilized 7,145 context-switches # 0.002 M/sec 410 cpu-migrations # 0.120 K/sec 6,191 page-faults # 0.002 M/sec 7,475,038,064 cycles # 2.196 GHz 4,487,043,071 stalled-cycles-frontend # 60.03% frontend cycles idle <not supported> stalled-cycles-backend 9,925,205,494 instructions # 1.33 insns per cycle # 0.45 stalled # cycles per insn 834,375,503 branches # 245.146 M/sec 4,919,995 branch-misses # 0.59% of all branches Signed-off-by: Marc-André Lureau <marcandre.lureau@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
9 years ago
vrend_state.eventfd = -1;
if (flags & VREND_USE_THREAD_SYNC) {
vrend_renderer_use_threaded_sync();
}
return 0;
}
void
vrend_renderer_fini(void)
{
if (!vrend_state.inited)
return;
renderer: use a thread to block for fences. Instead of polling the fences regularly, have a thread that blocks for a single fence using a separate shared context, then uses eventfd to wake up the main thread when something happens. Inside the guest, glmark2 typicially runs twice as fast with the thread sync. Although in general, the performances seems to be about +30%. The benefits is mostly for CPU-bounds tasks (when main the thread hits 100%) A naive perf stat of the vtest renderer with glmark2 "build" test with a fixed number of frames (500) results in the following stats data: (do not value timing related informations, since the renderer is ran and stopped manually) without thread: 3032.282265 task-clock (msec) # 0.420 CPUs utilized 4,277 context-switches # 0.001 M/sec 102 cpu-migrations # 0.034 K/sec 9,020 page-faults # 0.003 M/sec 7,884,098,254 cycles # 2.600 GHz 4,440,126,451 stalled-cycles-frontend # 56.32% frontend cycles idle <not supported> stalled-cycles-backend 11,024,091,578 instructions # 1.40 insns per cycle # 0.40 stalled # cycles per insn 1,091,831,588 branches # 360.069 M/sec 5,426,846 branch-misses # 0.50% of all branches with thread: 3403.592921 task-clock (msec) # 0.452 CPUs utilized 7,145 context-switches # 0.002 M/sec 410 cpu-migrations # 0.120 K/sec 6,191 page-faults # 0.002 M/sec 7,475,038,064 cycles # 2.196 GHz 4,487,043,071 stalled-cycles-frontend # 60.03% frontend cycles idle <not supported> stalled-cycles-backend 9,925,205,494 instructions # 1.33 insns per cycle # 0.45 stalled # cycles per insn 834,375,503 branches # 245.146 M/sec 4,919,995 branch-misses # 0.59% of all branches Signed-off-by: Marc-André Lureau <marcandre.lureau@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
9 years ago
vrend_free_sync_thread();
if (vrend_state.eventfd != -1) {
close(vrend_state.eventfd);
vrend_state.eventfd = -1;
}
renderer: use a thread to block for fences. Instead of polling the fences regularly, have a thread that blocks for a single fence using a separate shared context, then uses eventfd to wake up the main thread when something happens. Inside the guest, glmark2 typicially runs twice as fast with the thread sync. Although in general, the performances seems to be about +30%. The benefits is mostly for CPU-bounds tasks (when main the thread hits 100%) A naive perf stat of the vtest renderer with glmark2 "build" test with a fixed number of frames (500) results in the following stats data: (do not value timing related informations, since the renderer is ran and stopped manually) without thread: 3032.282265 task-clock (msec) # 0.420 CPUs utilized 4,277 context-switches # 0.001 M/sec 102 cpu-migrations # 0.034 K/sec 9,020 page-faults # 0.003 M/sec 7,884,098,254 cycles # 2.600 GHz 4,440,126,451 stalled-cycles-frontend # 56.32% frontend cycles idle <not supported> stalled-cycles-backend 11,024,091,578 instructions # 1.40 insns per cycle # 0.40 stalled # cycles per insn 1,091,831,588 branches # 360.069 M/sec 5,426,846 branch-misses # 0.50% of all branches with thread: 3403.592921 task-clock (msec) # 0.452 CPUs utilized 7,145 context-switches # 0.002 M/sec 410 cpu-migrations # 0.120 K/sec 6,191 page-faults # 0.002 M/sec 7,475,038,064 cycles # 2.196 GHz 4,487,043,071 stalled-cycles-frontend # 60.03% frontend cycles idle <not supported> stalled-cycles-backend 9,925,205,494 instructions # 1.33 insns per cycle # 0.45 stalled # cycles per insn 834,375,503 branches # 245.146 M/sec 4,919,995 branch-misses # 0.59% of all branches Signed-off-by: Marc-André Lureau <marcandre.lureau@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
9 years ago
vrend_decode_reset(false);
vrend_object_fini_resource_table();
vrend_decode_reset(true);
vrend_state.current_ctx = NULL;
vrend_state.current_hw_ctx = NULL;
vrend_state.inited = false;
}
static void vrend_destroy_sub_context(struct vrend_sub_context *sub)
{
int i, j;
struct vrend_streamout_object *obj, *tmp;
if (sub->fb_id)
glDeleteFramebuffers(1, &sub->fb_id);
if (sub->blit_fb_ids[0])
glDeleteFramebuffers(2, sub->blit_fb_ids);
glBindBuffer(GL_ELEMENT_ARRAY_BUFFER, 0);
if (!vrend_state.have_vertex_attrib_binding) {
while (sub->enabled_attribs_bitmask) {
i = u_bit_scan(&sub->enabled_attribs_bitmask);
glDisableVertexAttribArray(i);
}
glDeleteVertexArrays(1, &sub->vaoid);
}
glBindVertexArray(0);
if (sub->current_so)
glBindTransformFeedback(GL_TRANSFORM_FEEDBACK, 0);
LIST_FOR_EACH_ENTRY_SAFE(obj, tmp, &sub->streamout_list, head) {
vrend_destroy_streamout_object(obj);
}
vrend_shader_state_reference(&sub->shaders[PIPE_SHADER_VERTEX], NULL);
vrend_shader_state_reference(&sub->shaders[PIPE_SHADER_FRAGMENT], NULL);
vrend_shader_state_reference(&sub->shaders[PIPE_SHADER_GEOMETRY], NULL);
vrend_free_programs(sub);
for (i = 0; i < PIPE_SHADER_TYPES; i++) {
free(sub->consts[i].consts);
sub->consts[i].consts = NULL;
for (j = 0; j < PIPE_MAX_SHADER_SAMPLER_VIEWS; j++) {
vrend_sampler_view_reference(&sub->views[i].views[j], NULL);
}
}
if (sub->zsurf)
vrend_surface_reference(&sub->zsurf, NULL);
for (i = 0; i < sub->nr_cbufs; i++) {
if (!sub->surf[i])
continue;
vrend_surface_reference(&sub->surf[i], NULL);
}
vrend_resource_reference((struct vrend_resource **)&sub->ib.buffer, NULL);
vrend_object_fini_ctx_table(sub->object_hash);
vrend_clicbs->destroy_gl_context(sub->gl_context);
list_del(&sub->head);
FREE(sub);
}
bool vrend_destroy_context(struct vrend_context *ctx)
{
bool switch_0 = (ctx == vrend_state.current_ctx);
struct vrend_context *cur = vrend_state.current_ctx;
struct vrend_sub_context *sub, *tmp;
if (switch_0) {
vrend_state.current_ctx = NULL;
vrend_state.current_hw_ctx = NULL;
}
if (vrend_state.use_core_profile) {
if (ctx->pstip_inited)
glDeleteTextures(1, &ctx->pstipple_tex_id);
ctx->pstip_inited = false;
}
/* reset references on framebuffers */
vrend_set_framebuffer_state(ctx, 0, NULL, 0);
vrend_set_num_sampler_views(ctx, PIPE_SHADER_VERTEX, 0, 0);
vrend_set_num_sampler_views(ctx, PIPE_SHADER_FRAGMENT, 0, 0);
vrend_set_num_sampler_views(ctx, PIPE_SHADER_GEOMETRY, 0, 0);
vrend_set_streamout_targets(ctx, 0, 0, NULL);
vrend_set_num_vbo(ctx, 0);
vrend_set_index_buffer(ctx, 0, 0, 0);
vrend_renderer_force_ctx_0();
LIST_FOR_EACH_ENTRY_SAFE(sub, tmp, &ctx->sub_ctxs, head)
vrend_destroy_sub_context(sub);
vrend_object_fini_ctx_table(ctx->res_hash);
list_del(&ctx->ctx_entry);
FREE(ctx);
if (!switch_0 && cur)
vrend_hw_switch_context(cur, true);
return switch_0;
}
struct vrend_context *vrend_create_context(int id, uint32_t nlen, const char *debug_name)
{
struct vrend_context *grctx = CALLOC_STRUCT(vrend_context);
if (!grctx)
return NULL;
if (nlen && debug_name) {
strncpy(grctx->debug_name, debug_name, 64);
}
grctx->ctx_id = id;
list_inithead(&grctx->sub_ctxs);
list_inithead(&grctx->active_nontimer_query_list);
grctx->res_hash = vrend_object_init_ctx_table();
grctx->shader_cfg.use_gles = vrend_state.use_gles;
grctx->shader_cfg.use_core_profile = vrend_state.use_core_profile;
grctx->shader_cfg.use_explicit_locations = vrend_state.use_explicit_locations;
vrend_renderer_create_sub_ctx(grctx, 0);
vrend_renderer_set_sub_ctx(grctx, 0);
vrender_get_glsl_version(&grctx->shader_cfg.glsl_version);
list_addtail(&grctx->ctx_entry, &vrend_state.active_ctx_list);
return grctx;
}
int vrend_renderer_resource_attach_iov(int res_handle, struct iovec *iov,
int num_iovs)
{
struct vrend_resource *res;
res = vrend_resource_lookup(res_handle, 0);
if (!res)
return EINVAL;
if (res->iov)
return 0;
/* work out size and max resource size */
res->iov = iov;
res->num_iovs = num_iovs;
return 0;
}
void vrend_renderer_resource_detach_iov(int res_handle,
struct iovec **iov_p,
int *num_iovs_p)
{
struct vrend_resource *res;
res = vrend_resource_lookup(res_handle, 0);
if (!res) {
return;
}
if (iov_p)
*iov_p = res->iov;
if (num_iovs_p)
*num_iovs_p = res->num_iovs;
res->iov = NULL;
res->num_iovs = 0;
}
static int check_resource_valid(struct vrend_renderer_resource_create_args *args)
{
/* do not accept handle 0 */
if (args->handle == 0)
return -1;
/* limit the target */
if (args->target >= PIPE_MAX_TEXTURE_TYPES)
return -1;
if (args->format >= VIRGL_FORMAT_MAX)
return -1;
/* only texture 2d and 2d array can have multiple samples */
if (args->nr_samples > 1) {
if (args->target != PIPE_TEXTURE_2D && args->target != PIPE_TEXTURE_2D_ARRAY)
return -1;
/* multisample can't have miplevels */
if (args->last_level > 0)
return -1;
}
if (args->last_level > 0) {
/* buffer and rect textures can't have mipmaps */
if (args->target == PIPE_BUFFER || args->target == PIPE_TEXTURE_RECT)
return -1;
if (args->last_level > (floor(log2(MAX2(args->width, args->height))) + 1))
return -1;
}
if (args->flags != 0 && args->flags != VIRGL_RESOURCE_Y_0_TOP)
return -1;
if (args->flags & VIRGL_RESOURCE_Y_0_TOP)
if (args->target != PIPE_TEXTURE_2D && args->target != PIPE_TEXTURE_RECT)
return -1;
/* array size for array textures only */
if (args->target == PIPE_TEXTURE_CUBE) {
if (args->array_size != 6)
return -1;
} else if (args->target == PIPE_TEXTURE_CUBE_ARRAY) {
if (args->array_size % 6)
return -1;
} else if (args->array_size > 1) {
if (args->target != PIPE_TEXTURE_2D_ARRAY &&
args->target != PIPE_TEXTURE_1D_ARRAY)
return -1;
}
if (args->bind == 0 ||
args->bind == VREND_RES_BIND_CUSTOM ||
args->bind == VREND_RES_BIND_INDEX_BUFFER ||
args->bind == VREND_RES_BIND_STREAM_OUTPUT ||
args->bind == VREND_RES_BIND_VERTEX_BUFFER ||
args->bind == VREND_RES_BIND_CONSTANT_BUFFER) {
if (args->target != PIPE_BUFFER)
return -1;
if (args->height != 1 || args->depth != 1)
return -1;
} else {
if (!((args->bind & VREND_RES_BIND_SAMPLER_VIEW) ||
(args->bind & VREND_RES_BIND_DEPTH_STENCIL) ||
(args->bind & VREND_RES_BIND_RENDER_TARGET) ||
(args->bind & VREND_RES_BIND_CURSOR)))
return -1;
if (args->target == PIPE_TEXTURE_2D ||
args->target == PIPE_TEXTURE_RECT ||
args->target == PIPE_TEXTURE_CUBE ||
args->target == PIPE_TEXTURE_2D_ARRAY ||
args->target == PIPE_TEXTURE_CUBE_ARRAY) {
if (args->depth != 1)
return -1;
}
if (args->target == PIPE_TEXTURE_1D ||
args->target == PIPE_TEXTURE_1D_ARRAY) {
if (args->height != 1 || args->depth != 1)
return -1;
}
}
return 0;
}
int vrend_renderer_resource_create(struct vrend_renderer_resource_create_args *args, struct iovec *iov, uint32_t num_iovs)
{
struct vrend_resource *gr;
int level;
int ret;
ret = check_resource_valid(args);
if (ret)
return EINVAL;
gr = (struct vrend_resource *)CALLOC_STRUCT(vrend_texture);
if (!gr)
return ENOMEM;
gr->handle = args->handle;
gr->iov = iov;
gr->num_iovs = num_iovs;
gr->base.width0 = args->width;
gr->base.height0 = args->height;
gr->base.depth0 = args->depth;
gr->base.format = args->format;
gr->base.target = args->target;
gr->base.last_level = args->last_level;
gr->base.nr_samples = args->nr_samples;
gr->base.array_size = args->array_size;
if (args->flags & VIRGL_RESOURCE_Y_0_TOP)
gr->y_0_top = true;
pipe_reference_init(&gr->base.reference, 1);
if (args->bind == VREND_RES_BIND_CUSTOM) {
/* custom should only be for buffers */
gr->ptr = malloc(args->width);
if (!gr->ptr) {
FREE(gr);
return ENOMEM;
}
} else if (args->bind == VREND_RES_BIND_INDEX_BUFFER) {
gr->target = GL_ELEMENT_ARRAY_BUFFER_ARB;
glGenBuffersARB(1, &gr->id);
glBindBufferARB(gr->target, gr->id);
glBufferData(gr->target, args->width, NULL, GL_STREAM_DRAW);
} else if (args->bind == VREND_RES_BIND_STREAM_OUTPUT) {
gr->target = GL_TRANSFORM_FEEDBACK_BUFFER;
glGenBuffersARB(1, &gr->id);
glBindBuffer(gr->target, gr->id);
glBufferData(gr->target, args->width, NULL, GL_STREAM_DRAW);
} else if (args->bind == VREND_RES_BIND_VERTEX_BUFFER) {
gr->target = GL_ARRAY_BUFFER_ARB;
glGenBuffersARB(1, &gr->id);
glBindBufferARB(gr->target, gr->id);
glBufferData(gr->target, args->width, NULL, GL_STREAM_DRAW);
} else if (args->bind == VREND_RES_BIND_CONSTANT_BUFFER) {
gr->target = GL_UNIFORM_BUFFER;
glGenBuffersARB(1, &gr->id);
glBindBufferARB(gr->target, gr->id);
glBufferData(gr->target, args->width, NULL, GL_STREAM_DRAW);
} else if (args->target == PIPE_BUFFER && args->bind == 0) {
gr->target = GL_ARRAY_BUFFER_ARB;
glGenBuffersARB(1, &gr->id);
glBindBufferARB(gr->target, gr->id);
glBufferData(gr->target, args->width, NULL, GL_STREAM_DRAW);
} else if (args->target == PIPE_BUFFER && (args->bind & VREND_RES_BIND_SAMPLER_VIEW)) {
GLenum internalformat;
/*
* On Desktop we use GL_ARB_texture_buffer_object on GLES we use
* GL_EXT_texture_buffer (it is in the ANDRIOD extension pack).
*/
#if GL_TEXTURE_BUFFER != GL_TEXTURE_BUFFER_EXT
#error "GL_TEXTURE_BUFFER enums differ, they shouldn't."
#endif
/* need to check GL version here */
if (vrend_state.have_arb_or_gles_ext_texture_buffer) {
gr->target = GL_TEXTURE_BUFFER;
glGenBuffersARB(1, &gr->id);
glBindBufferARB(gr->target, gr->id);
glGenTextures(1, &gr->tbo_tex_id);
glBufferData(gr->target, args->width, NULL, GL_STREAM_DRAW);
glBindTexture(gr->target, gr->tbo_tex_id);
internalformat = tex_conv_table[args->format].internalformat;
glTexBuffer(gr->target, internalformat, gr->id);
} else {
gr->target = GL_PIXEL_PACK_BUFFER_ARB;
glGenBuffersARB(1, &gr->id);
glBindBufferARB(gr->target, gr->id);
glBufferData(gr->target, args->width, NULL, GL_STREAM_DRAW);
}
} else {
struct vrend_texture *gt = (struct vrend_texture *)gr;
GLenum internalformat, glformat, gltype;
gr->target = tgsitargettogltarget(args->target, args->nr_samples);
/* ugly workaround for texture rectangle missing on GLES */
if (vrend_state.use_gles && gr->target == GL_TEXTURE_RECTANGLE_NV) {
/* for some guests this is the only usage of rect */
if (args->width != 1 || args->height != 1) {
report_gles_warn(NULL, GLES_WARN_TEXTURE_RECT, 0);
}
gr->target = GL_TEXTURE_2D;
}
/* fallback for 1D textures */
if (vrend_state.use_gles && gr->target == GL_TEXTURE_1D) {
gr->target = GL_TEXTURE_2D;
}
/* fallback for 1D array textures */
if (vrend_state.use_gles && gr->target == GL_TEXTURE_1D_ARRAY) {
gr->target = GL_TEXTURE_2D_ARRAY;
}
glGenTextures(1, &gr->id);
glBindTexture(gr->target, gr->id);
internalformat = tex_conv_table[args->format].internalformat;
glformat = tex_conv_table[args->format].glformat;
gltype = tex_conv_table[args->format].gltype;
if (internalformat == 0) {
fprintf(stderr,"unknown format is %d\n", args->format);
FREE(gr);
return EINVAL;
}
if (args->nr_samples > 1) {
if (vrend_state.use_gles) {
report_gles_missing_func(NULL, "glTexImage[2,3]DMultisample");
} else if (gr->target == GL_TEXTURE_2D_MULTISAMPLE) {
glTexImage2DMultisample(gr->target, args->nr_samples,
internalformat, args->width, args->height,
GL_TRUE);
} else {
glTexImage3DMultisample(gr->target, args->nr_samples,
internalformat, args->width, args->height, args->array_size,
GL_TRUE);
}
} else if (gr->target == GL_TEXTURE_CUBE_MAP) {
int i;
for (i = 0; i < 6; i++) {
GLenum ctarget = GL_TEXTURE_CUBE_MAP_POSITIVE_X + i;
for (level = 0; level <= args->last_level; level++) {
unsigned mwidth = u_minify(args->width, level);
unsigned mheight = u_minify(args->height, level);
glTexImage2D(ctarget, level, internalformat, mwidth, mheight, 0, glformat,
gltype, NULL);
}
}
} else if (gr->target == GL_TEXTURE_3D ||
gr->target == GL_TEXTURE_2D_ARRAY ||
gr->target == GL_TEXTURE_CUBE_MAP_ARRAY) {
for (level = 0; level <= args->last_level; level++) {
unsigned depth_param = (gr->target == GL_TEXTURE_2D_ARRAY || gr->target == GL_TEXTURE_CUBE_MAP_ARRAY) ? args->array_size : u_minify(args->depth, level);
unsigned mwidth = u_minify(args->width, level);
unsigned mheight = u_minify(args->height, level);
glTexImage3D(gr->target, level, internalformat, mwidth, mheight, depth_param, 0,
glformat,
gltype, NULL);
}
} else if (gr->target == GL_TEXTURE_1D && vrend_state.use_gles) {
report_gles_missing_func(NULL, "glTexImage1D");
} else if (gr->target == GL_TEXTURE_1D) {
for (level = 0; level <= args->last_level; level++) {
unsigned mwidth = u_minify(args->width, level);
glTexImage1D(gr->target, level, internalformat, mwidth, 0,
glformat,
gltype, NULL);
}
} else {
for (level = 0; level <= args->last_level; level++) {
unsigned mwidth = u_minify(args->width, level);
unsigned mheight = u_minify(args->height, level);
glTexImage2D(gr->target, level, internalformat, mwidth, gr->target == GL_TEXTURE_1D_ARRAY ? args->array_size : mheight, 0, glformat,
gltype, NULL);
}
}
gt->state.max_lod = -1;
gt->cur_swizzle_r = gt->cur_swizzle_g = gt->cur_swizzle_b = gt->cur_swizzle_a = -1;
}
ret = vrend_resource_insert(gr, args->handle);
if (ret == 0) {
vrend_renderer_resource_destroy(gr, true);
return ENOMEM;
}
return 0;
}
void vrend_renderer_resource_destroy(struct vrend_resource *res, bool remove)
{
if (res->readback_fb_id)
glDeleteFramebuffers(1, &res->readback_fb_id);
if (res->ptr)
free(res->ptr);
if (res->id) {
if (res->target == GL_ELEMENT_ARRAY_BUFFER_ARB ||
res->target == GL_ARRAY_BUFFER_ARB ||
res->target == GL_UNIFORM_BUFFER||
res->target == GL_TEXTURE_BUFFER||
res->target == GL_TRANSFORM_FEEDBACK_BUFFER) {
glDeleteBuffers(1, &res->id);
if (res->target == GL_TEXTURE_BUFFER)
glDeleteTextures(1, &res->tbo_tex_id);
} else
glDeleteTextures(1, &res->id);
}
if (res->handle && remove)
vrend_resource_remove(res->handle);
free(res);
}
static void vrend_destroy_resource_object(void *obj_ptr)
{
struct vrend_resource *res = obj_ptr;
if (pipe_reference(&res->base.reference, NULL))
vrend_renderer_resource_destroy(res, false);
}
void vrend_renderer_resource_unref(uint32_t res_handle)
{
struct vrend_resource *res;
struct vrend_context *ctx;
res = vrend_resource_lookup(res_handle, 0);
if (!res)
return;
/* find in all contexts and detach also */
/* remove from any contexts */
LIST_FOR_EACH_ENTRY(ctx, &vrend_state.active_ctx_list, ctx_entry) {
vrend_renderer_detach_res_ctx_p(ctx, res->handle);
}
vrend_resource_remove(res->handle);
}
static int use_sub_data = 0;
struct virgl_sub_upload_data {
GLenum target;
struct pipe_box *box;
};
static void iov_buffer_upload(void *cookie, uint32_t doff, void *src, int len)
{
struct virgl_sub_upload_data *d = cookie;
glBufferSubData(d->target, d->box->x + doff, len, src);
}
static void vrend_scale_depth(void *ptr, int size, float scale_val)
{
GLuint *ival = ptr;
const GLfloat myscale = 1.0f / 0xffffff;
int i;
for (i = 0; i < size / 4; i++) {
GLuint value = ival[i];
GLfloat d = ((float)(value >> 8) * myscale) * scale_val;
d = CLAMP(d, 0.0F, 1.0F);
ival[i] = (int)(d / myscale) << 8;
}
}
static void read_transfer_data(struct pipe_resource *res,
struct iovec *iov,
unsigned int num_iovs,
char *data,
uint32_t src_stride,
struct pipe_box *box,
uint32_t level,
uint64_t offset,
bool invert)
{
int blsize = util_format_get_blocksize(res->format);
uint32_t size = vrend_get_iovec_size(iov, num_iovs);
uint32_t send_size = util_format_get_nblocks(res->format, box->width,
box->height) * blsize * box->depth;
uint32_t bwx = util_format_get_nblocksx(res->format, box->width) * blsize;
uint32_t bh = util_format_get_nblocksy(res->format, box->height);
int d, h;
if ((send_size == size || bh == 1) && !invert && box->depth == 1)
vrend_read_from_iovec(iov, num_iovs, offset, data, send_size);
else {
if (invert) {
for (d = 0; d < box->depth; d++) {
uint32_t myoffset = offset + d * src_stride * u_minify(res->height0, level);
for (h = bh - 1; h >= 0; h--) {
void *ptr = data + (h * bwx) + d * (bh * bwx);
vrend_read_from_iovec(iov, num_iovs, myoffset, ptr, bwx);
myoffset += src_stride;
}
}
} else {
for (d = 0; d < box->depth; d++) {
uint32_t myoffset = offset + d * src_stride * u_minify(res->height0, level);
for (h = 0; h < bh; h++) {
void *ptr = data + (h * bwx) + d * (bh * bwx);
vrend_read_from_iovec(iov, num_iovs, myoffset, ptr, bwx);
myoffset += src_stride;
}
}
}
}
}
static void write_transfer_data(struct pipe_resource *res,
struct iovec *iov,
unsigned num_iovs,
char *data,
uint32_t dst_stride,
struct pipe_box *box,
uint32_t level,
uint64_t offset,
bool invert)
{
int blsize = util_format_get_blocksize(res->format);
uint32_t size = vrend_get_iovec_size(iov, num_iovs);
uint32_t send_size = util_format_get_nblocks(res->format, box->width,
box->height) * blsize * box->depth;
uint32_t bwx = util_format_get_nblocksx(res->format, box->width) * blsize;
uint32_t bh = util_format_get_nblocksy(res->format, box->height);
int d, h;
uint32_t stride = dst_stride ? dst_stride : util_format_get_nblocksx(res->format, u_minify(res->width0, level)) * blsize;
if ((send_size == size || bh == 1) && !invert && box->depth == 1) {
vrend_write_to_iovec(iov, num_iovs, offset, data, send_size);
} else if (invert) {
for (d = 0; d < box->depth; d++) {
uint32_t myoffset = offset + d * stride * u_minify(res->height0, level);
for (h = bh - 1; h >= 0; h--) {
void *ptr = data + (h * bwx) + d * (bh * bwx);
vrend_write_to_iovec(iov, num_iovs, myoffset, ptr, bwx);
myoffset += stride;
}
}
} else {
for (d = 0; d < box->depth; d++) {
uint32_t myoffset = offset + d * stride * u_minify(res->height0, level);
for (h = 0; h < bh; h++) {
void *ptr = data + (h * bwx) + d * (bh * bwx);
vrend_write_to_iovec(iov, num_iovs, myoffset, ptr, bwx);
myoffset += stride;
}
}
}
}
static bool check_transfer_bounds(struct vrend_resource *res,
const struct vrend_transfer_info *info)
{
int lwidth, lheight;
/* check mipmap level is in bounds */
if (info->level > res->base.last_level)
return false;
if (info->box->x < 0 || info->box->y < 0)
return false;
/* these will catch bad y/z/w/d with 1D textures etc */
lwidth = u_minify(res->base.width0, info->level);
if (info->box->width > lwidth)
return false;
if (info->box->x > lwidth)
return false;
if (info->box->width + info->box->x > lwidth)
return false;
lheight = u_minify(res->base.height0, info->level);
if (info->box->height > lheight)
return false;
if (info->box->y > lheight)
return false;
if (info->box->height + info->box->y > lheight)
return false;
if (res->base.target == PIPE_TEXTURE_3D) {
int ldepth = u_minify(res->base.depth0, info->level);
if (info->box->depth > ldepth)
return false;
if (info->box->z > ldepth)
return false;
if (info->box->z + info->box->depth > ldepth)
return false;
} else {
if (info->box->depth > res->base.array_size)
return false;
if (info->box->z > res->base.array_size)
return false;
if (info->box->z + info->box->depth > res->base.array_size)
return false;
}
return true;
}
static bool check_iov_bounds(struct vrend_resource *res,
const struct vrend_transfer_info *info,
struct iovec *iov, int num_iovs)
{
GLuint send_size;
GLuint iovsize = vrend_get_iovec_size(iov, num_iovs);
GLuint valid_stride, valid_layer_stride;
/* validate the send size */
valid_stride = util_format_get_stride(res->base.format, info->box->width);
if (info->stride) {
/* only validate passed in stride for boxes with height */
if (info->box->height > 1) {
if (info->stride < valid_stride)
return false;
valid_stride = info->stride;
}
}
valid_layer_stride = util_format_get_2d_size(res->base.format, valid_stride,
info->box->height);
/* layer stride only makes sense for 3d,cube and arrays */
if (info->layer_stride) {
if ((res->base.target != PIPE_TEXTURE_3D &&
res->base.target != PIPE_TEXTURE_CUBE &&
res->base.target != PIPE_TEXTURE_1D_ARRAY &&
res->base.target != PIPE_TEXTURE_2D_ARRAY &&
res->base.target != PIPE_TEXTURE_CUBE_ARRAY))
return false;
/* only validate passed in layer_stride for boxes with depth */
if (info->box->depth > 1) {
if (info->layer_stride < valid_layer_stride)
return false;
valid_layer_stride = info->layer_stride;
}
}
send_size = valid_layer_stride * info->box->depth;
if (iovsize < info->offset)
return false;
if (iovsize < send_size)
return false;
if (iovsize < info->offset + send_size)
return false;
return true;
}
static int vrend_renderer_transfer_write_iov(struct vrend_context *ctx,
struct vrend_resource *res,
struct iovec *iov, int num_iovs,
const struct vrend_transfer_info *info)
{
void *data;
if (res->target == 0 && res->ptr) {
vrend_read_from_iovec(iov, num_iovs, info->offset, res->ptr + info->box->x, info->box->width);
return 0;
}
if (res->target == GL_TRANSFORM_FEEDBACK_BUFFER ||
res->target == GL_ELEMENT_ARRAY_BUFFER_ARB ||
res->target == GL_ARRAY_BUFFER_ARB ||
res->target == GL_TEXTURE_BUFFER ||
res->target == GL_UNIFORM_BUFFER) {
struct virgl_sub_upload_data d;
d.box = info->box;
d.target = res->target;
glBindBufferARB(res->target, res->id);
if (use_sub_data == 1) {
vrend_read_from_iovec_cb(iov, num_iovs, info->offset, info->box->width, &iov_buffer_upload, &d);
} else {
data = glMapBufferRange(res->target, info->box->x, info->box->width, GL_MAP_INVALIDATE_RANGE_BIT | GL_MAP_UNSYNCHRONIZED_BIT | GL_MAP_WRITE_BIT);
if (data == NULL) {
fprintf(stderr,"map failed for element buffer\n");
vrend_read_from_iovec_cb(iov, num_iovs, info->offset, info->box->width, &iov_buffer_upload, &d);
} else {
vrend_read_from_iovec(iov, num_iovs, info->offset, data, info->box->width);
glUnmapBuffer(res->target);
}
}
} else {
GLenum glformat;
GLenum gltype;
int need_temp = 0;
int elsize = util_format_get_blocksize(res->base.format);
int x = 0, y = 0;
bool compressed;
bool invert = false;
float depth_scale;
GLuint send_size = 0;
uint32_t stride = info->stride;
vrend_use_program(ctx, 0);
if (!stride)
stride = util_format_get_nblocksx(res->base.format, u_minify(res->base.width0, info->level)) * elsize;
compressed = util_format_is_compressed(res->base.format);
if (num_iovs > 1 || compressed) {
need_temp = true;
}
if (vrend_state.use_core_profile == true && (res->y_0_top || (res->base.format == (enum pipe_format)VIRGL_FORMAT_Z24X8_UNORM))) {
need_temp = true;
if (res->y_0_top)
invert = true;
}
if (need_temp) {
send_size = util_format_get_nblocks(res->base.format, info->box->width,
info->box->height) * elsize * info->box->depth;
data = malloc(send_size);
if (!data)
return ENOMEM;
read_transfer_data(&res->base, iov, num_iovs, data, stride,
info->box, info->level, info->offset, invert);
} else {
data = (char*)iov[0].iov_base + info->offset;
}
if (stride && !need_temp) {
glPixelStorei(GL_UNPACK_ROW_LENGTH, stride / elsize);
glPixelStorei(GL_UNPACK_IMAGE_HEIGHT, u_minify(res->base.height0, info->level));
} else
glPixelStorei(GL_UNPACK_ROW_LENGTH, 0);
switch (elsize) {
case 1:
case 3:
glPixelStorei(GL_UNPACK_ALIGNMENT, 1);
break;
case 2:
case 6:
glPixelStorei(GL_UNPACK_ALIGNMENT, 2);
break;
case 4:
default:
glPixelStorei(GL_UNPACK_ALIGNMENT, 4);
break;
case 8:
glPixelStorei(GL_UNPACK_ALIGNMENT, 8);
break;
}
glformat = tex_conv_table[res->base.format].glformat;
gltype = tex_conv_table[res->base.format].gltype;
if ((!vrend_state.use_core_profile) && (res->y_0_top)) {
GLuint buffers;
if (res->readback_fb_id == 0 || res->readback_fb_level != info->level) {
GLuint fb_id;
if (res->readback_fb_id)
glDeleteFramebuffers(1, &res->readback_fb_id);
glGenFramebuffers(1, &fb_id);
glBindFramebufferEXT(GL_FRAMEBUFFER_EXT, fb_id);
vrend_fb_bind_texture(res, 0, info->level, 0);
res->readback_fb_id = fb_id;
res->readback_fb_level = info->level;
} else {
glBindFramebufferEXT(GL_FRAMEBUFFER_EXT, res->readback_fb_id);
}
buffers = GL_COLOR_ATTACHMENT0_EXT;
glDrawBuffers(1, &buffers);
vrend_blend_enable(ctx, false);
vrend_depth_test_enable(ctx, false);
vrend_alpha_test_enable(ctx, false);
vrend_stencil_test_enable(ctx, false);
glPixelZoom(1.0f, res->y_0_top ? -1.0f : 1.0f);
glWindowPos2i(info->box->x, res->y_0_top ? res->base.height0 - info->box->y : info->box->y);
glDrawPixels(info->box->width, info->box->height, glformat, gltype,
data);
} else {
uint32_t comp_size;
glBindTexture(res->target, res->id);
if (compressed) {
glformat = tex_conv_table[res->base.format].internalformat;
comp_size = util_format_get_nblocks(res->base.format, info->box->width,
info->box->height) * util_format_get_blocksize(res->base.format);
}
if (glformat == 0) {
glformat = GL_BGRA;
gltype = GL_UNSIGNED_BYTE;
}
x = info->box->x;
y = invert ? res->base.height0 - info->box->y - info->box->height : info->box->y;
if (res->base.format == (enum pipe_format)VIRGL_FORMAT_Z24X8_UNORM) {
/* we get values from the guest as 24-bit scaled integers
but we give them to the host GL and it interprets them
as 32-bit scaled integers, so we need to scale them here */
depth_scale = 256.0;
if (!vrend_state.use_core_profile)
glPixelTransferf(GL_DEPTH_SCALE, depth_scale);
else
vrend_scale_depth(data, send_size, depth_scale);
}
if (res->target == GL_TEXTURE_CUBE_MAP) {
GLenum ctarget = GL_TEXTURE_CUBE_MAP_POSITIVE_X + info->box->z;
if (compressed) {
glCompressedTexSubImage2D(ctarget, info->level, x, y,
info->box->width, info->box->height,
glformat, comp_size, data);
} else {
glTexSubImage2D(ctarget, info->level, x, y, info->box->width, info->box->height,
glformat, gltype, data);
}
} else if (res->target == GL_TEXTURE_3D || res->target == GL_TEXTURE_2D_ARRAY || res->target == GL_TEXTURE_CUBE_MAP_ARRAY) {
if (compressed) {
glCompressedTexSubImage3D(res->target, info->level, x, y, info->box->z,
info->box->width, info->box->height, info->box->depth,
glformat, comp_size, data);
} else {
glTexSubImage3D(res->target, info->level, x, y, info->box->z,
info->box->width, info->box->height, info->box->depth,
glformat, gltype, data);
}
} else if (res->target == GL_TEXTURE_1D) {
if (vrend_state.use_gles) {
/* Covers both compressed and none compressed. */
report_gles_missing_func(ctx, "gl[Compressed]TexSubImage1D");
} else if (compressed) {
glCompressedTexSubImage1D(res->target, info->level, info->box->x,
info->box->width,
glformat, comp_size, data);
} else {
glTexSubImage1D(res->target, info->level, info->box->x, info->box->width,
glformat, gltype, data);
}
} else {
if (compressed) {
glCompressedTexSubImage2D(res->target, info->level, x, res->target == GL_TEXTURE_1D_ARRAY ? info->box->z : y,
info->box->width, info->box->height,
glformat, comp_size, data);
} else {
glTexSubImage2D(res->target, info->level, x, res->target == GL_TEXTURE_1D_ARRAY ? info->box->z : y,
info->box->width,
res->target == GL_TEXTURE_1D_ARRAY ? info->box->depth : info->box->height,
glformat, gltype, data);
}
}
if (res->base.format == (enum pipe_format)VIRGL_FORMAT_Z24X8_UNORM) {
if (!vrend_state.use_core_profile)
glPixelTransferf(GL_DEPTH_SCALE, 1.0);
}
}
if (stride && !need_temp) {
glPixelStorei(GL_UNPACK_ROW_LENGTH, 0);
glPixelStorei(GL_UNPACK_IMAGE_HEIGHT, 0);
}
glPixelStorei(GL_UNPACK_ALIGNMENT, 4);
if (need_temp)
free(data);
}
return 0;
}
static uint32_t vrend_get_texture_depth(struct vrend_resource *res, uint32_t level)
{
uint32_t depth = 1;
if (res->target == GL_TEXTURE_3D)
depth = u_minify(res->base.depth0, level);
else if (res->target == GL_TEXTURE_1D_ARRAY || res->target == GL_TEXTURE_2D_ARRAY ||
res->target == GL_TEXTURE_CUBE_MAP || res->target == GL_TEXTURE_CUBE_MAP_ARRAY)
depth = res->base.array_size;
return depth;
}
static int vrend_transfer_send_getteximage(struct vrend_context *ctx,
struct vrend_resource *res,
struct iovec *iov, int num_iovs,
const struct vrend_transfer_info *info)
{
GLenum format, type;
uint32_t tex_size;
char *data;
int elsize = util_format_get_blocksize(res->base.format);
int compressed = util_format_is_compressed(res->base.format);
GLenum target;
uint32_t send_offset = 0;
format = tex_conv_table[res->base.format].glformat;
type = tex_conv_table[res->base.format].gltype;
if (compressed)
format = tex_conv_table[res->base.format].internalformat;
tex_size = util_format_get_nblocks(res->base.format, u_minify(res->base.width0, info->level), u_minify(res->base.height0, info->level)) *
util_format_get_blocksize(res->base.format) * vrend_get_texture_depth(res, info->level);
if (info->box->z && res->target != GL_TEXTURE_CUBE_MAP) {
send_offset = util_format_get_nblocks(res->base.format, u_minify(res->base.width0, info->level), u_minify(res->base.height0, info->level)) * util_format_get_blocksize(res->base.format) * info->box->z;
}
data = malloc(tex_size);
if (!data)
return ENOMEM;
switch (elsize) {
case 1:
glPixelStorei(GL_PACK_ALIGNMENT, 1);
break;
case 2:
glPixelStorei(GL_PACK_ALIGNMENT, 2);
break;
case 4:
default:
glPixelStorei(GL_PACK_ALIGNMENT, 4);
break;
case 8:
glPixelStorei(GL_PACK_ALIGNMENT, 8);
break;
}
glBindTexture(res->target, res->id);
if (res->target == GL_TEXTURE_CUBE_MAP) {
target = GL_TEXTURE_CUBE_MAP_POSITIVE_X + info->box->z;
} else
target = res->target;
if (compressed) {
if (vrend_state.have_arb_robustness) {
glGetnCompressedTexImageARB(target, info->level, tex_size, data);
} else if (vrend_state.use_gles) {
report_gles_missing_func(ctx, "glGetCompressedTexImage");
} else {
glGetCompressedTexImage(target, info->level, data);
}
} else {
if (vrend_state.have_arb_robustness) {
glGetnTexImageARB(target, info->level, format, type, tex_size, data);
} else if (vrend_state.use_gles) {
report_gles_missing_func(ctx, "glGetTexImage");
} else {
glGetTexImage(target, info->level, format, type, data);
}
}
glPixelStorei(GL_PACK_ALIGNMENT, 4);
write_transfer_data(&res->base, iov, num_iovs, data + send_offset,
info->stride, info->box, info->level, info->offset,
false);
free(data);
return 0;
}
static int vrend_transfer_send_readpixels(struct vrend_context *ctx,
struct vrend_resource *res,
struct iovec *iov, int num_iovs,
const struct vrend_transfer_info *info)
{
char *myptr = (char*)iov[0].iov_base + info->offset;
int need_temp = 0;
GLuint fb_id;
char *data;
bool actually_invert, separate_invert = false;
GLenum format, type;
GLint y1;
uint32_t send_size = 0;
uint32_t h = u_minify(res->base.height0, info->level);
int elsize = util_format_get_blocksize(res->base.format);
float depth_scale;
vrend_use_program(ctx, 0);
format = tex_conv_table[res->base.format].glformat;
type = tex_conv_table[res->base.format].gltype;
/* if we are asked to invert and reading from a front then don't */
actually_invert = res->y_0_top;
if (actually_invert && !vrend_state.have_mesa_invert)
separate_invert = true;
if (num_iovs > 1 || separate_invert)
need_temp = 1;
if (need_temp) {
send_size = util_format_get_nblocks(res->base.format, info->box->width, info->box->height) * info->box->depth * util_format_get_blocksize(res->base.format);
data = malloc(send_size);
if (!data) {
fprintf(stderr,"malloc failed %d\n", send_size);
return ENOMEM;
}
} else {
send_size = iov[0].iov_len - info->offset;
data = myptr;
}
if (res->readback_fb_id == 0 || res->readback_fb_level != info->level || res->readback_fb_z != info->box->z) {
if (res->readback_fb_id)
glDeleteFramebuffers(1, &res->readback_fb_id);
glGenFramebuffers(1, &fb_id);
glBindFramebufferEXT(GL_FRAMEBUFFER_EXT, fb_id);
vrend_fb_bind_texture(res, 0, info->level, info->box->z);
res->readback_fb_id = fb_id;
res->readback_fb_level = info->level;
res->readback_fb_z = info->box->z;
} else
glBindFramebufferEXT(GL_FRAMEBUFFER_EXT, res->readback_fb_id);
if (actually_invert)
y1 = h - info->box->y - info->box->height;
else
y1 = info->box->y;
if (vrend_state.have_mesa_invert && actually_invert)
glPixelStorei(GL_PACK_INVERT_MESA, 1);
if (!vrend_format_is_ds(res->base.format))
glReadBuffer(GL_COLOR_ATTACHMENT0_EXT);
if (!need_temp && info->stride)
glPixelStorei(GL_PACK_ROW_LENGTH, info->stride / elsize);
switch (elsize) {
case 1:
glPixelStorei(GL_PACK_ALIGNMENT, 1);
break;
case 2:
glPixelStorei(GL_PACK_ALIGNMENT, 2);
break;
case 4:
default:
glPixelStorei(GL_PACK_ALIGNMENT, 4);
break;
case 8:
glPixelStorei(GL_PACK_ALIGNMENT, 8);
break;
}
if (res->base.format == (enum pipe_format)VIRGL_FORMAT_Z24X8_UNORM) {
/* we get values from the guest as 24-bit scaled integers
but we give them to the host GL and it interprets them
as 32-bit scaled integers, so we need to scale them here */
depth_scale = 1.0 / 256.0;
if (!vrend_state.use_core_profile) {
glPixelTransferf(GL_DEPTH_SCALE, depth_scale);
}
}
if (vrend_state.have_arb_robustness)
glReadnPixelsARB(info->box->x, y1, info->box->width, info->box->height, format, type, send_size, data);
else if (vrend_state.have_gles_khr_robustness)
glReadnPixelsKHR(info->box->x, y1, info->box->width, info->box->height, format, type, send_size, data);
else
glReadPixels(info->box->x, y1, info->box->width, info->box->height, format, type, data);
if (res->base.format == (enum pipe_format)VIRGL_FORMAT_Z24X8_UNORM) {
if (!vrend_state.use_core_profile)
glPixelTransferf(GL_DEPTH_SCALE, 1.0);
else
vrend_scale_depth(data, send_size, depth_scale);
}
if (vrend_state.have_mesa_invert && actually_invert)
glPixelStorei(GL_PACK_INVERT_MESA, 0);
if (!need_temp && info->stride)
glPixelStorei(GL_PACK_ROW_LENGTH, 0);
glPixelStorei(GL_PACK_ALIGNMENT, 4);
if (need_temp) {
write_transfer_data(&res->base, iov, num_iovs, data,
info->stride, info->box, info->level, info->offset,
separate_invert);
free(data);
}
return 0;
}
static int vrend_renderer_transfer_send_iov(struct vrend_context *ctx,
struct vrend_resource *res,
struct iovec *iov, int num_iovs,
const struct vrend_transfer_info *info)
{
if (res->target == 0 && res->ptr) {
uint32_t send_size = info->box->width * util_format_get_blocksize(res->base.format);
vrend_write_to_iovec(iov, num_iovs, info->offset, res->ptr + info->box->x, send_size);
return 0;
}
if (res->target == GL_ELEMENT_ARRAY_BUFFER_ARB ||
res->target == GL_ARRAY_BUFFER_ARB ||
res->target == GL_TRANSFORM_FEEDBACK_BUFFER ||
res->target == GL_TEXTURE_BUFFER ||
res->target == GL_UNIFORM_BUFFER) {
uint32_t send_size = info->box->width * util_format_get_blocksize(res->base.format);
void *data;
glBindBufferARB(res->target, res->id);
data = glMapBufferRange(res->target, info->box->x, info->box->width, GL_MAP_READ_BIT);
if (!data)
fprintf(stderr,"unable to open buffer for reading %d\n", res->target);
else
vrend_write_to_iovec(iov, num_iovs, info->offset, data, send_size);
glUnmapBuffer(res->target);
} else {
bool can_readpixels = true;
can_readpixels = vrend_format_can_render(res->base.format) || vrend_format_is_ds(res->base.format);
if (can_readpixels) {
return vrend_transfer_send_readpixels(ctx, res,
iov, num_iovs, info);
}
return vrend_transfer_send_getteximage(ctx, res,
iov, num_iovs, info);
}
return 0;
}
int vrend_renderer_transfer_iov(const struct vrend_transfer_info *info,
int transfer_mode)
{
struct vrend_resource *res;
struct vrend_context *ctx;
struct iovec *iov;
int num_iovs;
if (!info->box)
return EINVAL;
ctx = vrend_lookup_renderer_ctx(info->ctx_id);
if (!ctx)
return EINVAL;
if (info->ctx_id == 0)
res = vrend_resource_lookup(info->handle, 0);
else
res = vrend_renderer_ctx_res_lookup(ctx, info->handle);
if (!res) {
if (info->ctx_id)
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_RESOURCE, info->handle);
return EINVAL;
}
iov = info->iovec;
num_iovs = info->iovec_cnt;
if (res->iov && (!iov || num_iovs == 0)) {
iov = res->iov;
num_iovs = res->num_iovs;
}
if (!iov) {
if (info->ctx_id)
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_RESOURCE, info->handle);
return EINVAL;
}
if (!check_transfer_bounds(res, info))
return EINVAL;
if (!check_iov_bounds(res, info, iov, num_iovs))
return EINVAL;
vrend_hw_switch_context(vrend_lookup_renderer_ctx(0), true);
if (transfer_mode == VREND_TRANSFER_WRITE)
return vrend_renderer_transfer_write_iov(ctx, res, iov, num_iovs,
info);
else
return vrend_renderer_transfer_send_iov(ctx, res, iov, num_iovs,
info);
}
int vrend_transfer_inline_write(struct vrend_context *ctx,
struct vrend_transfer_info *info,
unsigned usage)
{
struct vrend_resource *res;
res = vrend_renderer_ctx_res_lookup(ctx, info->handle);
if (!res) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_RESOURCE, info->handle);
return EINVAL;
}
if (!check_transfer_bounds(res, info)) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_CMD_BUFFER, info->handle);
return EINVAL;
}
if (!check_iov_bounds(res, info, info->iovec, info->iovec_cnt)) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_CMD_BUFFER, info->handle);
return EINVAL;
}
return vrend_renderer_transfer_write_iov(ctx, res, info->iovec, info->iovec_cnt, info);
}
void vrend_set_stencil_ref(struct vrend_context *ctx,
struct pipe_stencil_ref *ref)
{
if (ctx->sub->stencil_refs[0] != ref->ref_value[0] ||
ctx->sub->stencil_refs[1] != ref->ref_value[1]) {
ctx->sub->stencil_refs[0] = ref->ref_value[0];
ctx->sub->stencil_refs[1] = ref->ref_value[1];
ctx->sub->stencil_state_dirty = true;
}
}
void vrend_set_blend_color(struct vrend_context *ctx,
struct pipe_blend_color *color)
{
ctx->sub->blend_color = *color;
glBlendColor(color->color[0], color->color[1], color->color[2],
color->color[3]);
}
void vrend_set_scissor_state(struct vrend_context *ctx,
uint32_t start_slot,
uint32_t num_scissor,
struct pipe_scissor_state *ss)
{
int i, idx;
if (start_slot > PIPE_MAX_VIEWPORTS ||
num_scissor > (PIPE_MAX_VIEWPORTS - start_slot)) {
vrend_report_buffer_error(ctx, 0);
return;
}
for (i = 0; i < num_scissor; i++) {
idx = start_slot + i;
ctx->sub->ss[idx] = ss[i];
ctx->sub->scissor_state_dirty |= (1 << idx);
}
}
void vrend_set_polygon_stipple(struct vrend_context *ctx,
struct pipe_poly_stipple *ps)
{
if (vrend_state.use_core_profile) {
static const unsigned bit31 = 1 << 31;
GLubyte *stip = calloc(1, 1024);
int i, j;
if (!ctx->pstip_inited)
vrend_init_pstipple_texture(ctx);
if (!stip)
return;
for (i = 0; i < 32; i++) {
for (j = 0; j < 32; j++) {
if (ps->stipple[i] & (bit31 >> j))
stip[i * 32 + j] = 0;
else
stip[i * 32 + j] = 255;
}
}
glBindTexture(GL_TEXTURE_2D, ctx->pstipple_tex_id);
glTexSubImage2D(GL_TEXTURE_2D, 0, 0, 0, 32, 32,
GL_RED, GL_UNSIGNED_BYTE, stip);
free(stip);
return;
}
glPolygonStipple((const GLubyte *)ps->stipple);
}
void vrend_set_clip_state(struct vrend_context *ctx, struct pipe_clip_state *ucp)
{
if (vrend_state.use_core_profile) {
ctx->sub->ucp_state = *ucp;
} else {
int i, j;
GLdouble val[4];
for (i = 0; i < 8; i++) {
for (j = 0; j < 4; j++)
val[j] = ucp->ucp[i][j];
glClipPlane(GL_CLIP_PLANE0 + i, val);
}
}
}
void vrend_set_sample_mask(struct vrend_context *ctx, unsigned sample_mask)
{
glSampleMaski(0, sample_mask);
}
static void vrend_hw_emit_streamout_targets(struct vrend_context *ctx, struct vrend_streamout_object *so_obj)
{
int i;
for (i = 0; i < so_obj->num_targets; i++) {
if (so_obj->so_targets[i]->buffer_offset || so_obj->so_targets[i]->buffer_size < so_obj->so_targets[i]->buffer->base.width0)
glBindBufferRange(GL_TRANSFORM_FEEDBACK_BUFFER, i, so_obj->so_targets[i]->buffer->id, so_obj->so_targets[i]->buffer_offset, so_obj->so_targets[i]->buffer_size);
else
glBindBufferBase(GL_TRANSFORM_FEEDBACK_BUFFER, i, so_obj->so_targets[i]->buffer->id);
}
}
void vrend_set_streamout_targets(struct vrend_context *ctx,
uint32_t append_bitmask,
uint32_t num_targets,
uint32_t *handles)
{
struct vrend_so_target *target;
int i;
if (num_targets) {
bool found = false;
struct vrend_streamout_object *obj;
LIST_FOR_EACH_ENTRY(obj, &ctx->sub->streamout_list, head) {
if (obj->num_targets == num_targets) {
if (!memcmp(handles, obj->handles, num_targets * 4)) {
found = true;
break;
}
}
}
if (found) {
ctx->sub->current_so = obj;
glBindTransformFeedback(GL_TRANSFORM_FEEDBACK, obj->id);
return;
}
obj = CALLOC_STRUCT(vrend_streamout_object);
if (vrend_state.have_tf2) {
glGenTransformFeedbacks(1, &obj->id);
glBindTransformFeedback(GL_TRANSFORM_FEEDBACK, obj->id);
}
obj->num_targets = num_targets;
for (i = 0; i < num_targets; i++) {
obj->handles[i] = handles[i];
target = vrend_object_lookup(ctx->sub->object_hash, handles[i], VIRGL_OBJECT_STREAMOUT_TARGET);
if (!target) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_HANDLE, handles[i]);
free(obj);
return;
}
vrend_so_target_reference(&obj->so_targets[i], target);
}
vrend_hw_emit_streamout_targets(ctx, obj);
list_addtail(&obj->head, &ctx->sub->streamout_list);
ctx->sub->current_so = obj;
obj->xfb_state = XFB_STATE_STARTED_NEED_BEGIN;
} else {
if (vrend_state.have_tf2)
glBindTransformFeedback(GL_TRANSFORM_FEEDBACK, 0);
ctx->sub->current_so = NULL;
}
}
static void vrend_resource_buffer_copy(struct vrend_context *ctx,
struct vrend_resource *src_res,
struct vrend_resource *dst_res,
uint32_t dstx, uint32_t srcx,
uint32_t width)
{
glBindBuffer(GL_COPY_READ_BUFFER, src_res->id);
glBindBuffer(GL_COPY_WRITE_BUFFER, dst_res->id);
glCopyBufferSubData(GL_COPY_READ_BUFFER, GL_COPY_WRITE_BUFFER, srcx, dstx, width);
glBindBuffer(GL_COPY_READ_BUFFER, 0);
glBindBuffer(GL_COPY_WRITE_BUFFER, 0);
}
static void vrend_resource_copy_fallback(struct vrend_context *ctx,
struct vrend_resource *src_res,
struct vrend_resource *dst_res,
uint32_t dst_level,
uint32_t dstx, uint32_t dsty,
uint32_t dstz, uint32_t src_level,
const struct pipe_box *src_box)
{
char *tptr;
uint32_t transfer_size;
GLenum glformat, gltype;
int elsize = util_format_get_blocksize(dst_res->base.format);
int compressed = util_format_is_compressed(dst_res->base.format);
int cube_slice = 1;
uint32_t slice_size, slice_offset;
int i;
if (src_res->target == GL_TEXTURE_CUBE_MAP)
cube_slice = 6;
if (src_res->base.format != dst_res->base.format) {
fprintf(stderr, "copy fallback failed due to mismatched formats %d %d\n", src_res->base.format, dst_res->base.format);
return;
}
/* this is ugly need to do a full GetTexImage */
slice_size = util_format_get_nblocks(src_res->base.format, u_minify(src_res->base.width0, src_level), u_minify(src_res->base.height0, src_level)) *
util_format_get_blocksize(src_res->base.format);
transfer_size = slice_size * vrend_get_texture_depth(src_res, src_level);
tptr = malloc(transfer_size);
if (!tptr)
return;
glformat = tex_conv_table[src_res->base.format].glformat;
gltype = tex_conv_table[src_res->base.format].gltype;
if (compressed)
glformat = tex_conv_table[src_res->base.format].internalformat;
switch (elsize) {
case 1:
glPixelStorei(GL_PACK_ALIGNMENT, 1);
break;
case 2:
glPixelStorei(GL_PACK_ALIGNMENT, 2);
break;
case 4:
default:
glPixelStorei(GL_PACK_ALIGNMENT, 4);
break;
case 8:
glPixelStorei(GL_PACK_ALIGNMENT, 8);
break;
}
glBindTexture(src_res->target, src_res->id);
slice_offset = 0;
for (i = 0; i < cube_slice; i++) {
GLenum ctarget = src_res->target == GL_TEXTURE_CUBE_MAP ? GL_TEXTURE_CUBE_MAP_POSITIVE_X + i : src_res->target;
if (compressed) {
if (vrend_state.have_arb_robustness)
glGetnCompressedTexImageARB(ctarget, src_level, transfer_size, tptr + slice_offset);
else if (vrend_state.use_gles)
report_gles_missing_func(ctx, "glGetCompressedTexImage");
else
glGetCompressedTexImage(ctarget, src_level, tptr + slice_offset);
} else {
if (vrend_state.have_arb_robustness)
glGetnTexImageARB(ctarget, src_level, glformat, gltype, transfer_size, tptr + slice_offset);
else if (vrend_state.use_gles)
report_gles_missing_func(ctx, "glGetTexImage");
else
glGetTexImage(ctarget, src_level, glformat, gltype, tptr + slice_offset);
}
slice_offset += slice_size;
}
glPixelStorei(GL_PACK_ALIGNMENT, 4);
switch (elsize) {
case 1:
glPixelStorei(GL_UNPACK_ALIGNMENT, 1);
break;
case 2:
glPixelStorei(GL_UNPACK_ALIGNMENT, 2);
break;
case 4:
default:
glPixelStorei(GL_UNPACK_ALIGNMENT, 4);
break;
case 8:
glPixelStorei(GL_UNPACK_ALIGNMENT, 8);
break;
}
glBindTexture(dst_res->target, dst_res->id);
slice_offset = src_box->z * slice_size;
cube_slice = (src_res->target == GL_TEXTURE_CUBE_MAP) ? src_box->z + src_box->depth : cube_slice;
i = (src_res->target == GL_TEXTURE_CUBE_MAP) ? src_box->z : 0;
for (; i < cube_slice; i++) {
GLenum ctarget = dst_res->target == GL_TEXTURE_CUBE_MAP ? GL_TEXTURE_CUBE_MAP_POSITIVE_X + i : dst_res->target;
if (compressed) {
if (ctarget == GL_TEXTURE_1D) {
glCompressedTexSubImage1D(ctarget, dst_level, dstx,
src_box->width,
glformat, transfer_size, tptr + slice_offset);
} else {
glCompressedTexSubImage2D(ctarget, dst_level, dstx, dsty,
src_box->width, src_box->height,
glformat, transfer_size, tptr + slice_offset);
}
} else {
if (ctarget == GL_TEXTURE_1D) {
glTexSubImage1D(ctarget, dst_level, dstx, src_box->width, glformat, gltype, tptr + slice_offset);
} else if (ctarget == GL_TEXTURE_3D ||
ctarget == GL_TEXTURE_2D_ARRAY ||
ctarget == GL_TEXTURE_CUBE_MAP_ARRAY) {
glTexSubImage3D(ctarget, dst_level, dstx, dsty, dstz, src_box->width, src_box->height, src_box->depth, glformat, gltype, tptr + slice_offset);
} else {
glTexSubImage2D(ctarget, dst_level, dstx, dsty, src_box->width, src_box->height, glformat, gltype, tptr + slice_offset);
}
}
slice_offset += slice_size;
}
glPixelStorei(GL_UNPACK_ALIGNMENT, 4);
free(tptr);
}
void vrend_renderer_resource_copy_region(struct vrend_context *ctx,
uint32_t dst_handle, uint32_t dst_level,
uint32_t dstx, uint32_t dsty, uint32_t dstz,
uint32_t src_handle, uint32_t src_level,
const struct pipe_box *src_box)
{
struct vrend_resource *src_res, *dst_res;
GLbitfield glmask = 0;
GLint sy1, sy2, dy1, dy2;
if (ctx->in_error)
return;
src_res = vrend_renderer_ctx_res_lookup(ctx, src_handle);
dst_res = vrend_renderer_ctx_res_lookup(ctx, dst_handle);
if (!src_res) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_RESOURCE, src_handle);
return;
}
if (!dst_res) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_RESOURCE, dst_handle);
return;
}
if (src_res->base.target == PIPE_BUFFER && dst_res->base.target == PIPE_BUFFER) {
/* do a buffer copy */
vrend_resource_buffer_copy(ctx, src_res, dst_res, dstx,
src_box->x, src_box->width);
return;
}
if (!vrend_format_can_render(src_res->base.format) ||
!vrend_format_can_render(dst_res->base.format)) {
vrend_resource_copy_fallback(ctx, src_res, dst_res, dst_level, dstx,
dsty, dstz, src_level, src_box);
return;
}
glBindFramebuffer(GL_FRAMEBUFFER_EXT, ctx->sub->blit_fb_ids[0]);
/* clean out fb ids */
glFramebufferTexture2DEXT(GL_FRAMEBUFFER_EXT, GL_DEPTH_STENCIL_ATTACHMENT,
GL_TEXTURE_2D, 0, 0);
vrend_fb_bind_texture(src_res, 0, src_level, src_box->z);
glBindFramebuffer(GL_FRAMEBUFFER_EXT, ctx->sub->blit_fb_ids[1]);
glFramebufferTexture2DEXT(GL_FRAMEBUFFER_EXT, GL_DEPTH_STENCIL_ATTACHMENT,
GL_TEXTURE_2D, 0, 0);
vrend_fb_bind_texture(dst_res, 0, dst_level, dstz);
glBindFramebuffer(GL_DRAW_FRAMEBUFFER, ctx->sub->blit_fb_ids[1]);
glBindFramebuffer(GL_READ_FRAMEBUFFER, ctx->sub->blit_fb_ids[0]);
glmask = GL_COLOR_BUFFER_BIT;
glDisable(GL_SCISSOR_TEST);
if (!src_res->y_0_top) {
sy1 = src_box->y;
sy2 = src_box->y + src_box->height;
} else {
sy1 = src_res->base.height0 - src_box->y - src_box->height;
sy2 = src_res->base.height0 - src_box->y;
}
if (!dst_res->y_0_top) {
dy1 = dsty;
dy2 = dsty + src_box->height;
} else {
dy1 = dst_res->base.height0 - dsty - src_box->height;
dy2 = dst_res->base.height0 - dsty;
}
glBlitFramebuffer(src_box->x, sy1,
src_box->x + src_box->width,
sy2,
dstx, dy1,
dstx + src_box->width,
dy2,
glmask, GL_NEAREST);
}
static void vrend_renderer_blit_int(struct vrend_context *ctx,
struct vrend_resource *src_res,
struct vrend_resource *dst_res,
const struct pipe_blit_info *info)
{
GLbitfield glmask = 0;
int src_y1, src_y2, dst_y1, dst_y2;
GLenum filter;
int n_layers = 1, i;
bool use_gl = false;
filter = convert_mag_filter(info->filter);
/* if we can't make FBO's use the fallback path */
if (!vrend_format_can_render(src_res->base.format) &&
!vrend_format_is_ds(src_res->base.format))
use_gl = true;
if (!vrend_format_can_render(dst_res->base.format) &&
!vrend_format_is_ds(dst_res->base.format))
use_gl = true;
if (util_format_is_srgb(src_res->base.format))
use_gl = true;
/* different depth formats */
if (vrend_format_is_ds(src_res->base.format) &&
vrend_format_is_ds(dst_res->base.format)) {
if (src_res->base.format != dst_res->base.format) {
if (!(src_res->base.format == PIPE_FORMAT_S8_UINT_Z24_UNORM &&
(dst_res->base.format == PIPE_FORMAT_Z24X8_UNORM))) {
use_gl = true;
}
}
}
/* glBlitFramebuffer - can support depth stencil with NEAREST
which we use for mipmaps */
if ((info->mask & (PIPE_MASK_Z | PIPE_MASK_S)) && info->filter == PIPE_TEX_FILTER_LINEAR)
use_gl = true;
/* for scaled MS blits we either need extensions or hand roll */
if (info->mask & PIPE_MASK_RGBA &&
src_res->base.nr_samples > 1 &&
src_res->base.nr_samples != dst_res->base.nr_samples &&
(info->src.box.width != info->dst.box.width ||
info->src.box.height != info->dst.box.height)) {
if (vrend_state.have_ms_scaled_blit)
filter = GL_SCALED_RESOLVE_NICEST_EXT;
else
use_gl = true;
}
/* for 3D mipmapped blits - hand roll time */
if (info->src.box.depth != info->dst.box.depth)
use_gl = true;
if (vrend_format_is_emulated_alpha(info->dst.format) ||
vrend_format_is_emulated_alpha(info->src.format))
use_gl = true;
if (use_gl) {
vrend_renderer_blit_gl(ctx, src_res, dst_res, info);
vrend_clicbs->make_current(0, ctx->sub->gl_context);
return;
}
if (info->mask & PIPE_MASK_Z)
glmask |= GL_DEPTH_BUFFER_BIT;
if (info->mask & PIPE_MASK_S)
glmask |= GL_STENCIL_BUFFER_BIT;
if (info->mask & PIPE_MASK_RGBA)
glmask |= GL_COLOR_BUFFER_BIT;
if (!dst_res->y_0_top) {
dst_y1 = info->dst.box.y + info->dst.box.height;
dst_y2 = info->dst.box.y;
} else {
dst_y1 = dst_res->base.height0 - info->dst.box.y - info->dst.box.height;
dst_y2 = dst_res->base.height0 - info->dst.box.y;
}
if (!src_res->y_0_top) {
src_y1 = info->src.box.y + info->src.box.height;
src_y2 = info->src.box.y;
} else {
src_y1 = src_res->base.height0 - info->src.box.y - info->src.box.height;
src_y2 = src_res->base.height0 - info->src.box.y;
}
if (info->scissor_enable) {
glScissor(info->scissor.minx, info->scissor.miny, info->scissor.maxx - info->scissor.minx, info->scissor.maxy - info->scissor.miny);
ctx->sub->scissor_state_dirty = (1 << 0);
glEnable(GL_SCISSOR_TEST);
} else
glDisable(GL_SCISSOR_TEST);
glBindFramebuffer(GL_FRAMEBUFFER_EXT, ctx->sub->blit_fb_ids[0]);
if (info->mask & PIPE_MASK_RGBA)
glFramebufferTexture2DEXT(GL_FRAMEBUFFER_EXT, GL_DEPTH_STENCIL_ATTACHMENT,
GL_TEXTURE_2D, 0, 0);
else
glFramebufferTexture2DEXT(GL_FRAMEBUFFER_EXT, GL_COLOR_ATTACHMENT0,
GL_TEXTURE_2D, 0, 0);
glBindFramebuffer(GL_FRAMEBUFFER_EXT, ctx->sub->blit_fb_ids[1]);
if (info->mask & PIPE_MASK_RGBA)
glFramebufferTexture2DEXT(GL_FRAMEBUFFER_EXT, GL_DEPTH_STENCIL_ATTACHMENT,
GL_TEXTURE_2D, 0, 0);
else if (info->mask & (PIPE_MASK_Z | PIPE_MASK_S))
glFramebufferTexture2DEXT(GL_FRAMEBUFFER_EXT, GL_COLOR_ATTACHMENT0,
GL_TEXTURE_2D, 0, 0);
if (info->src.box.depth == info->dst.box.depth)
n_layers = info->dst.box.depth;
for (i = 0; i < n_layers; i++) {
glBindFramebuffer(GL_FRAMEBUFFER_EXT, ctx->sub->blit_fb_ids[0]);
glFramebufferTexture2DEXT(GL_FRAMEBUFFER_EXT, GL_DEPTH_STENCIL_ATTACHMENT,
GL_TEXTURE_2D, 0, 0);
vrend_fb_bind_texture(src_res, 0, info->src.level, info->src.box.z + i);
glBindFramebuffer(GL_FRAMEBUFFER_EXT, ctx->sub->blit_fb_ids[1]);
vrend_fb_bind_texture(dst_res, 0, info->dst.level, info->dst.box.z + i);
glBindFramebuffer(GL_DRAW_FRAMEBUFFER, ctx->sub->blit_fb_ids[1]);
glBindFramebuffer(GL_READ_FRAMEBUFFER, ctx->sub->blit_fb_ids[0]);
glBlitFramebuffer(info->src.box.x,
src_y1,
info->src.box.x + info->src.box.width,
src_y2,
info->dst.box.x,
dst_y1,
info->dst.box.x + info->dst.box.width,
dst_y2,
glmask, filter);
}
}
void vrend_renderer_blit(struct vrend_context *ctx,
uint32_t dst_handle, uint32_t src_handle,
const struct pipe_blit_info *info)
{
struct vrend_resource *src_res, *dst_res;
src_res = vrend_renderer_ctx_res_lookup(ctx, src_handle);
dst_res = vrend_renderer_ctx_res_lookup(ctx, dst_handle);
if (!src_res) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_RESOURCE, src_handle);
return;
}
if (!dst_res) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_RESOURCE, dst_handle);
return;
}
if (ctx->in_error)
return;
if (info->render_condition_enable == false)
vrend_pause_render_condition(ctx, true);
vrend_renderer_blit_int(ctx, src_res, dst_res, info);
if (info->render_condition_enable == false)
vrend_pause_render_condition(ctx, false);
}
int vrend_renderer_create_fence(int client_fence_id, uint32_t ctx_id)
{
struct vrend_fence *fence;
fence = malloc(sizeof(struct vrend_fence));
if (!fence)
return ENOMEM;
fence->ctx_id = ctx_id;
fence->fence_id = client_fence_id;
fence->syncobj = glFenceSync(GL_SYNC_GPU_COMMANDS_COMPLETE, 0);
renderer: use a thread to block for fences. Instead of polling the fences regularly, have a thread that blocks for a single fence using a separate shared context, then uses eventfd to wake up the main thread when something happens. Inside the guest, glmark2 typicially runs twice as fast with the thread sync. Although in general, the performances seems to be about +30%. The benefits is mostly for CPU-bounds tasks (when main the thread hits 100%) A naive perf stat of the vtest renderer with glmark2 "build" test with a fixed number of frames (500) results in the following stats data: (do not value timing related informations, since the renderer is ran and stopped manually) without thread: 3032.282265 task-clock (msec) # 0.420 CPUs utilized 4,277 context-switches # 0.001 M/sec 102 cpu-migrations # 0.034 K/sec 9,020 page-faults # 0.003 M/sec 7,884,098,254 cycles # 2.600 GHz 4,440,126,451 stalled-cycles-frontend # 56.32% frontend cycles idle <not supported> stalled-cycles-backend 11,024,091,578 instructions # 1.40 insns per cycle # 0.40 stalled # cycles per insn 1,091,831,588 branches # 360.069 M/sec 5,426,846 branch-misses # 0.50% of all branches with thread: 3403.592921 task-clock (msec) # 0.452 CPUs utilized 7,145 context-switches # 0.002 M/sec 410 cpu-migrations # 0.120 K/sec 6,191 page-faults # 0.002 M/sec 7,475,038,064 cycles # 2.196 GHz 4,487,043,071 stalled-cycles-frontend # 60.03% frontend cycles idle <not supported> stalled-cycles-backend 9,925,205,494 instructions # 1.33 insns per cycle # 0.45 stalled # cycles per insn 834,375,503 branches # 245.146 M/sec 4,919,995 branch-misses # 0.59% of all branches Signed-off-by: Marc-André Lureau <marcandre.lureau@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
9 years ago
if (fence->syncobj == NULL)
goto fail;
if (vrend_state.sync_thread) {
pipe_mutex_lock(vrend_state.fence_mutex);
list_addtail(&fence->fences, &vrend_state.fence_wait_list);
pipe_mutex_unlock(vrend_state.fence_mutex);
pipe_condvar_signal(vrend_state.fence_cond);
} else
list_addtail(&fence->fences, &vrend_state.fence_list);
return 0;
renderer: use a thread to block for fences. Instead of polling the fences regularly, have a thread that blocks for a single fence using a separate shared context, then uses eventfd to wake up the main thread when something happens. Inside the guest, glmark2 typicially runs twice as fast with the thread sync. Although in general, the performances seems to be about +30%. The benefits is mostly for CPU-bounds tasks (when main the thread hits 100%) A naive perf stat of the vtest renderer with glmark2 "build" test with a fixed number of frames (500) results in the following stats data: (do not value timing related informations, since the renderer is ran and stopped manually) without thread: 3032.282265 task-clock (msec) # 0.420 CPUs utilized 4,277 context-switches # 0.001 M/sec 102 cpu-migrations # 0.034 K/sec 9,020 page-faults # 0.003 M/sec 7,884,098,254 cycles # 2.600 GHz 4,440,126,451 stalled-cycles-frontend # 56.32% frontend cycles idle <not supported> stalled-cycles-backend 11,024,091,578 instructions # 1.40 insns per cycle # 0.40 stalled # cycles per insn 1,091,831,588 branches # 360.069 M/sec 5,426,846 branch-misses # 0.50% of all branches with thread: 3403.592921 task-clock (msec) # 0.452 CPUs utilized 7,145 context-switches # 0.002 M/sec 410 cpu-migrations # 0.120 K/sec 6,191 page-faults # 0.002 M/sec 7,475,038,064 cycles # 2.196 GHz 4,487,043,071 stalled-cycles-frontend # 60.03% frontend cycles idle <not supported> stalled-cycles-backend 9,925,205,494 instructions # 1.33 insns per cycle # 0.45 stalled # cycles per insn 834,375,503 branches # 245.146 M/sec 4,919,995 branch-misses # 0.59% of all branches Signed-off-by: Marc-André Lureau <marcandre.lureau@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
9 years ago
fail:
fprintf(stderr, "failed to create fence sync object\n");
free(fence);
return ENOMEM;
}
static void free_fence_locked(struct vrend_fence *fence)
{
list_del(&fence->fences);
glDeleteSync(fence->syncobj);
free(fence);
}
static void flush_eventfd(int fd)
{
ssize_t len;
uint64_t value;
do {
len = read(fd, &value, sizeof(value));
} while ((len == -1 && errno == EINTR) || len == sizeof(value));
}
void vrend_renderer_check_fences(void)
{
struct vrend_fence *fence, *stor;
uint32_t latest_id = 0;
GLenum glret;
if (!vrend_state.inited)
return;
renderer: use a thread to block for fences. Instead of polling the fences regularly, have a thread that blocks for a single fence using a separate shared context, then uses eventfd to wake up the main thread when something happens. Inside the guest, glmark2 typicially runs twice as fast with the thread sync. Although in general, the performances seems to be about +30%. The benefits is mostly for CPU-bounds tasks (when main the thread hits 100%) A naive perf stat of the vtest renderer with glmark2 "build" test with a fixed number of frames (500) results in the following stats data: (do not value timing related informations, since the renderer is ran and stopped manually) without thread: 3032.282265 task-clock (msec) # 0.420 CPUs utilized 4,277 context-switches # 0.001 M/sec 102 cpu-migrations # 0.034 K/sec 9,020 page-faults # 0.003 M/sec 7,884,098,254 cycles # 2.600 GHz 4,440,126,451 stalled-cycles-frontend # 56.32% frontend cycles idle <not supported> stalled-cycles-backend 11,024,091,578 instructions # 1.40 insns per cycle # 0.40 stalled # cycles per insn 1,091,831,588 branches # 360.069 M/sec 5,426,846 branch-misses # 0.50% of all branches with thread: 3403.592921 task-clock (msec) # 0.452 CPUs utilized 7,145 context-switches # 0.002 M/sec 410 cpu-migrations # 0.120 K/sec 6,191 page-faults # 0.002 M/sec 7,475,038,064 cycles # 2.196 GHz 4,487,043,071 stalled-cycles-frontend # 60.03% frontend cycles idle <not supported> stalled-cycles-backend 9,925,205,494 instructions # 1.33 insns per cycle # 0.45 stalled # cycles per insn 834,375,503 branches # 245.146 M/sec 4,919,995 branch-misses # 0.59% of all branches Signed-off-by: Marc-André Lureau <marcandre.lureau@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
9 years ago
if (vrend_state.sync_thread) {
flush_eventfd(vrend_state.eventfd);
pipe_mutex_lock(vrend_state.fence_mutex);
LIST_FOR_EACH_ENTRY_SAFE(fence, stor, &vrend_state.fence_list, fences) {
if (fence->fence_id > latest_id)
latest_id = fence->fence_id;
free_fence_locked(fence);
}
renderer: use a thread to block for fences. Instead of polling the fences regularly, have a thread that blocks for a single fence using a separate shared context, then uses eventfd to wake up the main thread when something happens. Inside the guest, glmark2 typicially runs twice as fast with the thread sync. Although in general, the performances seems to be about +30%. The benefits is mostly for CPU-bounds tasks (when main the thread hits 100%) A naive perf stat of the vtest renderer with glmark2 "build" test with a fixed number of frames (500) results in the following stats data: (do not value timing related informations, since the renderer is ran and stopped manually) without thread: 3032.282265 task-clock (msec) # 0.420 CPUs utilized 4,277 context-switches # 0.001 M/sec 102 cpu-migrations # 0.034 K/sec 9,020 page-faults # 0.003 M/sec 7,884,098,254 cycles # 2.600 GHz 4,440,126,451 stalled-cycles-frontend # 56.32% frontend cycles idle <not supported> stalled-cycles-backend 11,024,091,578 instructions # 1.40 insns per cycle # 0.40 stalled # cycles per insn 1,091,831,588 branches # 360.069 M/sec 5,426,846 branch-misses # 0.50% of all branches with thread: 3403.592921 task-clock (msec) # 0.452 CPUs utilized 7,145 context-switches # 0.002 M/sec 410 cpu-migrations # 0.120 K/sec 6,191 page-faults # 0.002 M/sec 7,475,038,064 cycles # 2.196 GHz 4,487,043,071 stalled-cycles-frontend # 60.03% frontend cycles idle <not supported> stalled-cycles-backend 9,925,205,494 instructions # 1.33 insns per cycle # 0.45 stalled # cycles per insn 834,375,503 branches # 245.146 M/sec 4,919,995 branch-misses # 0.59% of all branches Signed-off-by: Marc-André Lureau <marcandre.lureau@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
9 years ago
pipe_mutex_unlock(vrend_state.fence_mutex);
} else {
vrend_renderer_force_ctx_0();
LIST_FOR_EACH_ENTRY_SAFE(fence, stor, &vrend_state.fence_list, fences) {
glret = glClientWaitSync(fence->syncobj, 0, 0);
if (glret == GL_ALREADY_SIGNALED){
latest_id = fence->fence_id;
free_fence_locked(fence);
}
/* don't bother checking any subsequent ones */
else if (glret == GL_TIMEOUT_EXPIRED) {
break;
}
}
}
if (latest_id == 0)
return;
vrend_clicbs->write_fence(latest_id);
}
static bool vrend_get_one_query_result(GLuint query_id, bool use_64, uint64_t *result)
{
GLuint ready;
GLuint passed;
GLuint64 pass64;
glGetQueryObjectuiv(query_id, GL_QUERY_RESULT_AVAILABLE_ARB, &ready);
if (!ready)
return false;
if (use_64) {
glGetQueryObjectui64v(query_id, GL_QUERY_RESULT_ARB, &pass64);
*result = pass64;
} else {
glGetQueryObjectuiv(query_id, GL_QUERY_RESULT_ARB, &passed);
*result = passed;
}
return true;
}
static bool vrend_check_query(struct vrend_query *query)
{
uint64_t result;
struct virgl_host_query_state *state;
bool ret;
ret = vrend_get_one_query_result(query->id, false, &result);
if (ret == false)
return false;
state = (struct virgl_host_query_state *)query->res->ptr;
state->result = result;
state->query_state = VIRGL_QUERY_STATE_DONE;
return true;
}
void vrend_renderer_check_queries(void)
{
struct vrend_query *query, *stor;
if (!vrend_state.inited)
return;
LIST_FOR_EACH_ENTRY_SAFE(query, stor, &vrend_state.waiting_query_list, waiting_queries) {
vrend_hw_switch_context(vrend_lookup_renderer_ctx(query->ctx_id), true);
if (vrend_check_query(query))
list_delinit(&query->waiting_queries);
}
}
bool vrend_hw_switch_context(struct vrend_context *ctx, bool now)
{
if (ctx == vrend_state.current_ctx && ctx->ctx_switch_pending == false)
return true;
if (ctx->ctx_id != 0 && ctx->in_error) {
return false;
}
ctx->ctx_switch_pending = true;
if (now == true) {
vrend_finish_context_switch(ctx);
}
vrend_state.current_ctx = ctx;
return true;
}
static void vrend_finish_context_switch(struct vrend_context *ctx)
{
if (ctx->ctx_switch_pending == false)
return;
ctx->ctx_switch_pending = false;
if (vrend_state.current_hw_ctx == ctx)
return;
vrend_state.current_hw_ctx = ctx;
vrend_clicbs->make_current(0, ctx->sub->gl_context);
}
void
vrend_renderer_object_destroy(struct vrend_context *ctx, uint32_t handle)
{
vrend_object_remove(ctx->sub->object_hash, handle, 0);
}
uint32_t vrend_renderer_object_insert(struct vrend_context *ctx, void *data,
uint32_t size, uint32_t handle, enum virgl_object_type type)
{
return vrend_object_insert(ctx->sub->object_hash, data, size, handle, type);
}
int vrend_create_query(struct vrend_context *ctx, uint32_t handle,
uint32_t query_type, uint32_t query_index,
uint32_t res_handle, uint32_t offset)
{
struct vrend_query *q;
struct vrend_resource *res;
uint32_t ret_handle;
res = vrend_renderer_ctx_res_lookup(ctx, res_handle);
if (!res) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_RESOURCE, res_handle);
return EINVAL;
}
q = CALLOC_STRUCT(vrend_query);
if (!q)
return ENOMEM;
list_inithead(&q->waiting_queries);
q->type = query_type;
q->index = query_index;
q->ctx_id = ctx->ctx_id;
vrend_resource_reference(&q->res, res);
switch (q->type) {
case PIPE_QUERY_OCCLUSION_COUNTER:
q->gltype = GL_SAMPLES_PASSED_ARB;
break;
case PIPE_QUERY_OCCLUSION_PREDICATE:
q->gltype = GL_ANY_SAMPLES_PASSED;
break;
case PIPE_QUERY_TIMESTAMP:
q->gltype = GL_TIMESTAMP;
break;
case PIPE_QUERY_TIME_ELAPSED:
q->gltype = GL_TIME_ELAPSED;
break;
case PIPE_QUERY_PRIMITIVES_GENERATED:
q->gltype = GL_PRIMITIVES_GENERATED;
break;
case PIPE_QUERY_PRIMITIVES_EMITTED:
q->gltype = GL_TRANSFORM_FEEDBACK_PRIMITIVES_WRITTEN;
break;
default:
fprintf(stderr,"unknown query object received %d\n", q->type);
break;
}
glGenQueries(1, &q->id);
ret_handle = vrend_renderer_object_insert(ctx, q, sizeof(struct vrend_query), handle,
VIRGL_OBJECT_QUERY);
if (!ret_handle) {
FREE(q);
return ENOMEM;
}
return 0;
}
static void vrend_destroy_query(struct vrend_query *query)
{
vrend_resource_reference(&query->res, NULL);
list_del(&query->waiting_queries);
glDeleteQueries(1, &query->id);
free(query);
}
static void vrend_destroy_query_object(void *obj_ptr)
{
struct vrend_query *query = obj_ptr;
vrend_destroy_query(query);
}
void vrend_begin_query(struct vrend_context *ctx, uint32_t handle)
{
struct vrend_query *q;
q = vrend_object_lookup(ctx->sub->object_hash, handle, VIRGL_OBJECT_QUERY);
if (!q)
return;
if (q->gltype == GL_TIMESTAMP)
return;
glBeginQuery(q->gltype, q->id);
}
void vrend_end_query(struct vrend_context *ctx, uint32_t handle)
{
struct vrend_query *q;
q = vrend_object_lookup(ctx->sub->object_hash, handle, VIRGL_OBJECT_QUERY);
if (!q)
return;
if (vrend_is_timer_query(q->gltype)) {
if (vrend_state.use_gles && q->gltype == GL_TIMESTAMP) {
report_gles_warn(ctx, GLES_WARN_TIMESTAMP, 0);
} else if (q->gltype == GL_TIMESTAMP) {
glQueryCounter(q->id, q->gltype);
} else {
/* remove from active query list for this context */
glEndQuery(q->gltype);
}
return;
}
glEndQuery(q->gltype);
}
void vrend_get_query_result(struct vrend_context *ctx, uint32_t handle,
uint32_t wait)
{
struct vrend_query *q;
bool ret;
q = vrend_object_lookup(ctx->sub->object_hash, handle, VIRGL_OBJECT_QUERY);
if (!q)
return;
ret = vrend_check_query(q);
if (ret == false)
list_addtail(&q->waiting_queries, &vrend_state.waiting_query_list);
}
static void vrend_pause_render_condition(struct vrend_context *ctx, bool pause)
{
if (pause) {
if (ctx->sub->cond_render_q_id)
glEndConditionalRenderNV();
} else {
if (ctx->sub->cond_render_q_id)
glBeginConditionalRender(ctx->sub->cond_render_q_id,
ctx->sub->cond_render_gl_mode);
}
}
void vrend_render_condition(struct vrend_context *ctx,
uint32_t handle,
bool condition,
uint mode)
{
struct vrend_query *q;
GLenum glmode = 0;
if (handle == 0) {
glEndConditionalRenderNV();
ctx->sub->cond_render_q_id = 0;
ctx->sub->cond_render_gl_mode = 0;
return;
}
q = vrend_object_lookup(ctx->sub->object_hash, handle, VIRGL_OBJECT_QUERY);
if (!q)
return;
switch (mode) {
case PIPE_RENDER_COND_WAIT:
glmode = GL_QUERY_WAIT;
break;
case PIPE_RENDER_COND_NO_WAIT:
glmode = GL_QUERY_NO_WAIT;
break;
case PIPE_RENDER_COND_BY_REGION_WAIT:
glmode = GL_QUERY_BY_REGION_WAIT;
break;
case PIPE_RENDER_COND_BY_REGION_NO_WAIT:
glmode = GL_QUERY_BY_REGION_NO_WAIT;
break;
default:
fprintf(stderr, "unhandled condition %x\n", mode);
}
ctx->sub->cond_render_q_id = q->id;
ctx->sub->cond_render_gl_mode = glmode;
glBeginConditionalRender(q->id, glmode);
}
int vrend_create_so_target(struct vrend_context *ctx,
uint32_t handle,
uint32_t res_handle,
uint32_t buffer_offset,
uint32_t buffer_size)
{
struct vrend_so_target *target;
struct vrend_resource *res;
int ret_handle;
res = vrend_renderer_ctx_res_lookup(ctx, res_handle);
if (!res) {
report_context_error(ctx, VIRGL_ERROR_CTX_ILLEGAL_RESOURCE, res_handle);
return EINVAL;
}
target = CALLOC_STRUCT(vrend_so_target);
if (!target)
return ENOMEM;
pipe_reference_init(&target->reference, 1);
target->res_handle = res_handle;
target->buffer_offset = buffer_offset;
target->buffer_size = buffer_size;
target->sub_ctx = ctx->sub;
vrend_resource_reference(&target->buffer, res);
ret_handle = vrend_renderer_object_insert(ctx, target, sizeof(*target), handle,
VIRGL_OBJECT_STREAMOUT_TARGET);
if (ret_handle == 0) {
FREE(target);
return ENOMEM;
}
return 0;
}
static void vrender_get_glsl_version(int *glsl_version)
{
int major_local, minor_local;
const GLubyte *version_str;
int c;
int version;
version_str = glGetString(GL_SHADING_LANGUAGE_VERSION);
if (vrend_state.use_gles) {
char tmp[20];
c = sscanf((const char *)version_str, "%s %s %s %s %i.%i",
tmp, tmp, tmp, tmp, &major_local, &minor_local);
assert(c == 6);
} else {
c = sscanf((const char *)version_str, "%i.%i",
&major_local, &minor_local);
assert(c == 2);
}
version = (major_local * 100) + minor_local;
if (glsl_version)
*glsl_version = version;
}
/*
* Does all of the common caps setting,
* if it dedects a early out returns true.
*/
bool vrend_renderer_fill_caps_common(uint32_t set, uint32_t version,
union virgl_caps *caps)
{
int i, gl_ver;
GLint max;
if (!caps) {
return true;
}
if (set > 2) {
caps->max_version = 0;
return true;
}
if (set == 1) {
memset(caps, 0, sizeof(struct virgl_caps_v1));
caps->max_version = 1;
} else if (set == 2) {
memset(caps, 0, sizeof(*caps));
caps->max_version = 2;
}
gl_ver = epoxy_gl_version();
/*
* We can't fully support this feature on GLES,
* but it is needed for OpenGL 2.1 so lie.
*/
caps->v1.bset.occlusion_query = 1;
/* Some checks looks at v1.glsl_level so set it here. */
if (vrend_state.use_gles) {
caps->v1.glsl_level = 120;
} else if (vrend_state.use_core_profile) {
if (gl_ver == 31)
caps->v1.glsl_level = 140;
else if (gl_ver == 32)
caps->v1.glsl_level = 150;
else if (gl_ver >= 33)
caps->v1.glsl_level = 330;
} else {
caps->v1.glsl_level = 130;
}
/* Set supported prims here as we now know what shaders we support. */
caps->v1.prim_mask = (1 << PIPE_PRIM_POINTS) | (1 << PIPE_PRIM_LINES) | (1 << PIPE_PRIM_LINE_STRIP) | (1 << PIPE_PRIM_LINE_LOOP) | (1 << PIPE_PRIM_TRIANGLES) | (1 << PIPE_PRIM_TRIANGLE_STRIP) | (1 << PIPE_PRIM_TRIANGLE_FAN);
if (vrend_state.use_gles == false &&
vrend_state.use_core_profile == false) {
caps->v1.prim_mask |= (1 << PIPE_PRIM_QUADS) | (1 << PIPE_PRIM_QUAD_STRIP) | (1 << PIPE_PRIM_POLYGON);
}
if (!vrend_state.use_gles && caps->v1.glsl_level >= 150) {
caps->v1.prim_mask |= (1 << PIPE_PRIM_LINES_ADJACENCY) |
(1 << PIPE_PRIM_LINE_STRIP_ADJACENCY) |
(1 << PIPE_PRIM_TRIANGLES_ADJACENCY) |
(1 << PIPE_PRIM_TRIANGLE_STRIP_ADJACENCY);
}
/* Common limits for all backends. */
glGetIntegerv(GL_MAX_DRAW_BUFFERS, &max);
caps->v1.max_render_targets = max;
glGetIntegerv(GL_MAX_SAMPLES, &max);
caps->v1.max_samples = max;
/* All of the formats are common. */
for (i = 0; i < VIRGL_FORMAT_MAX; i++) {
uint32_t offset = i / 32;
uint32_t index = i % 32;
if (tex_conv_table[i].internalformat != 0) {
if (vrend_format_can_sample(i)) {
caps->v1.sampler.bitmask[offset] |= (1 << index);
if (vrend_format_can_render(i))
caps->v1.render.bitmask[offset] |= (1 << index);
}
}
}
if (!vrend_state.use_gles &&
epoxy_has_gl_extension("GL_ARB_vertex_type_10f_11f_11f_rev")) {
int val = VIRGL_FORMAT_R11G11B10_FLOAT;
uint32_t offset = val / 32;
uint32_t index = val % 32;
caps->v1.vertexbuffer.bitmask[offset] |= (1 << index);
}
/* These are filled in by the init code, so are common. */
if (vrend_state.have_nv_prim_restart ||
vrend_state.have_gl_prim_restart) {
caps->v1.bset.primitive_restart = 1;
}
return false;
}
void vrend_renderer_fill_caps_gles(uint32_t set, uint32_t version,
union virgl_caps *caps)
{
GLint max;
GLfloat range[2];
bool fill_capset2 = false;
int gles_ver = epoxy_gl_version();
if (set == 2) {
fill_capset2 = true;
}
caps->v1.bset.indep_blend_enable = 0;
caps->v1.bset.conditional_render = 0;
caps->v1.bset.poly_stipple = 0;
caps->v1.bset.color_clamping = 0;
caps->v1.bset.instanceid = 0;
caps->v1.bset.primitive_restart = 0;
caps->v1.bset.fragment_coord_conventions = 0;
caps->v1.bset.depth_clip_disable = 0;
caps->v1.bset.seamless_cube_map = 0;
caps->v1.bset.seamless_cube_map_per_texture = 0;
caps->v1.bset.texture_multisample = 0;
caps->v1.bset.mirror_clamp = 0;
caps->v1.bset.indep_blend_func = 0;
caps->v1.bset.cube_map_array = 0;
caps->v1.bset.texture_query_lod = 0;
caps->v1.bset.has_indirect_draw = 0;
caps->v1.bset.has_sample_shading = 0;
caps->v1.bset.start_instance = 0;
caps->v1.bset.shader_stencil_export = 0;
caps->v1.bset.streamout_pause_resume = 0;
caps->v1.max_dual_source_render_targets = 0;
caps->v1.max_dual_source_render_targets = 0;
caps->v1.max_tbo_size = 0;
caps->v1.max_texture_gather_components = 0;
caps->v1.max_viewports = 1;
if (gles_ver >= 30) {
glGetIntegerv(GL_MAX_ARRAY_TEXTURE_LAYERS, &max);
caps->v1.max_texture_array_layers = max;
caps->v1.bset.primitive_restart = 1;
}
if (!fill_capset2) {
return;
}
glGetFloatv(GL_ALIASED_POINT_SIZE_RANGE, range);
caps->v2.min_aliased_point_size = range[0];
caps->v2.max_aliased_point_size = range[1];
/* Not available on GLES */
caps->v2.min_smooth_point_size = 0.0f;
caps->v2.max_smooth_point_size = 0.0f;
glGetFloatv(GL_ALIASED_LINE_WIDTH_RANGE, range);
caps->v2.min_aliased_line_width = range[0];
caps->v2.max_aliased_line_width = range[1];
/* Not available on GLES */
caps->v2.min_smooth_line_width = 0.0f;
caps->v2.max_smooth_line_width = 0.0f;
glGetFloatv(GL_MAX_TEXTURE_LOD_BIAS, &caps->v2.max_texture_lod_bias);
glGetIntegerv(GL_MAX_VERTEX_ATTRIBS, &caps->v2.max_vertex_attribs);
glGetIntegerv(GL_MAX_VERTEX_OUTPUT_COMPONENTS, &max);
caps->v2.max_vertex_outputs = max / 4;
/* Not available on GLES */
caps->v2.max_geom_output_vertices = 0;
caps->v2.max_geom_total_output_components = 0;
caps->v2.max_shader_patch_varyings = 0;
/* Not available on GLES */
caps->v2.min_texture_gather_offset = 0;
caps->v2.max_texture_gather_offset = 0;
glGetIntegerv(GL_MIN_PROGRAM_TEXEL_OFFSET, &caps->v2.min_texel_offset);
glGetIntegerv(GL_MAX_PROGRAM_TEXEL_OFFSET, &caps->v2.max_texel_offset);
glGetIntegerv(GL_UNIFORM_BUFFER_OFFSET_ALIGNMENT, &caps->v2.uniform_buffer_offset_alignment);
/* Not available on GLES */
caps->v2.texture_buffer_offset_alignment = 0;
}
void vrend_renderer_fill_caps(uint32_t set, uint32_t version,
union virgl_caps *caps)
{
int i;
GLint max;
GLfloat range[2];
int gl_ver = epoxy_gl_version();
bool fill_capset2 = false;
/* Returns true if we should early out. */
if (vrend_renderer_fill_caps_common(set, version, caps)) {
return;
}
/* GLES has it's own path */
if (vrend_state.use_gles) {
vrend_renderer_fill_caps_gles(set, version, caps);
return;
}
if (set == 2) {
fill_capset2 = true;
}
if (gl_ver >= 30) {
caps->v1.bset.indep_blend_enable = 1;
caps->v1.bset.conditional_render = 1;
} else {
if (epoxy_has_gl_extension("GL_EXT_draw_buffers2"))
caps->v1.bset.indep_blend_enable = 1;
if (epoxy_has_gl_extension("GL_NV_conditional_render"))
caps->v1.bset.conditional_render = 1;
}
if (vrend_state.use_core_profile) {
caps->v1.bset.poly_stipple = 0;
caps->v1.bset.color_clamping = 0;
} else {
caps->v1.bset.poly_stipple = 1;
caps->v1.bset.color_clamping = 1;
}
if (gl_ver >= 31) {
caps->v1.bset.instanceid = 1;
glGetIntegerv(GL_MAX_VERTEX_UNIFORM_BLOCKS, &max);
vrend_state.max_uniform_blocks = max;
caps->v1.max_uniform_blocks = max + 1;
} else {
if (epoxy_has_gl_extension("GL_ARB_draw_instanced"))
caps->v1.bset.instanceid = 1;
}
if (gl_ver >= 32) {
caps->v1.bset.fragment_coord_conventions = 1;
caps->v1.bset.depth_clip_disable = 1;
caps->v1.bset.seamless_cube_map = 1;
} else {
if (epoxy_has_gl_extension("GL_ARB_fragment_coord_conventions"))
caps->v1.bset.fragment_coord_conventions = 1;
if (epoxy_has_gl_extension("GL_ARB_seamless_cube_map"))
caps->v1.bset.seamless_cube_map = 1;
}
if (epoxy_has_gl_extension("GL_AMD_seamless_cube_map_per_texture")) {
caps->v1.bset.seamless_cube_map_per_texture = 1;
}
if (epoxy_has_gl_extension("GL_ARB_texture_multisample")) {
caps->v1.bset.texture_multisample = 1;
}
if (gl_ver >= 40) {
caps->v1.bset.indep_blend_func = 1;
caps->v1.bset.cube_map_array = 1;
caps->v1.bset.texture_query_lod = 1;
caps->v1.bset.has_indirect_draw = 1;
caps->v1.bset.has_sample_shading = 1;
} else {
if (epoxy_has_gl_extension("GL_ARB_draw_buffers_blend"))
caps->v1.bset.indep_blend_func = 1;
if (epoxy_has_gl_extension("GL_ARB_texture_cube_map_array"))
caps->v1.bset.cube_map_array = 1;
if (epoxy_has_gl_extension("GL_ARB_texture_query_lod"))
caps->v1.bset.texture_query_lod = 1;
if (epoxy_has_gl_extension("GL_ARB_indirect_draw"))
caps->v1.bset.has_indirect_draw = 1;
if (epoxy_has_gl_extension("GL_ARB_sample_shading"))
caps->v1.bset.has_sample_shading = 1;
}
if (gl_ver >= 42) {
caps->v1.bset.start_instance = 1;
} else if (epoxy_has_gl_extension("GL_ARB_base_instance")) {
caps->v1.bset.start_instance = 1;
}
if (epoxy_has_gl_extension("GL_ARB_shader_stencil_export")) {
caps->v1.bset.shader_stencil_export = 1;
}
if (gl_ver >= 45) {
caps->v1.bset.has_cull = 1;
} else {
if (epoxy_has_gl_extension("GL_ARB_cull_distance"))
caps->v1.bset.has_cull = 1;
}
if (epoxy_has_gl_extension("GL_EXT_texture_mirror_clamp")) {
caps->v1.bset.mirror_clamp = true;
}
if (epoxy_has_gl_extension("GL_EXT_texture_array")) {
glGetIntegerv(GL_MAX_ARRAY_TEXTURE_LAYERS, &max);
caps->v1.max_texture_array_layers = max;
}
/* we need tf3 so we can do gallium skip buffers */
if (epoxy_has_gl_extension("GL_ARB_transform_feedback2")) {
caps->v1.bset.streamout_pause_resume = 1;
}
if (epoxy_has_gl_extension("GL_ARB_transform_feedback3")) {
glGetIntegerv(GL_MAX_TRANSFORM_FEEDBACK_BUFFERS, &max);
caps->v1.max_streamout_buffers = max;
} else if (epoxy_has_gl_extension("GL_EXT_transform_feedback")) {
caps->v1.max_streamout_buffers = 4;
}
if (epoxy_has_gl_extension("GL_ARB_blend_func_extended")) {
glGetIntegerv(GL_MAX_DUAL_SOURCE_DRAW_BUFFERS, &max);
caps->v1.max_dual_source_render_targets = max;
}
if (epoxy_has_gl_extension("GL_ARB_texture_buffer_object")) {
glGetIntegerv(GL_MAX_TEXTURE_BUFFER_SIZE, &max);
caps->v1.max_tbo_size = max;
}
if (epoxy_has_gl_extension("GL_ARB_texture_gather")) {
glGetIntegerv(GL_MAX_PROGRAM_TEXTURE_GATHER_COMPONENTS_ARB, &max);
caps->v1.max_texture_gather_components = max;
}
if (epoxy_has_gl_extension("GL_ARB_viewport_array")) {
glGetIntegerv(GL_MAX_VIEWPORTS, &max);
caps->v1.max_viewports = max;
} else {
caps->v1.max_viewports = 1;
}
if (!fill_capset2)
return;
glGetFloatv(GL_ALIASED_POINT_SIZE_RANGE, range);
caps->v2.min_aliased_point_size = range[0];
caps->v2.max_aliased_point_size = range[1];
glGetFloatv(GL_SMOOTH_POINT_SIZE_RANGE, range);
caps->v2.min_smooth_point_size = range[0];
caps->v2.max_smooth_point_size = range[1];
glGetFloatv(GL_ALIASED_LINE_WIDTH_RANGE, range);
caps->v2.min_aliased_line_width = range[0];
caps->v2.max_aliased_line_width = range[1];
glGetFloatv(GL_SMOOTH_LINE_WIDTH_RANGE, range);
caps->v2.min_smooth_line_width = range[0];
caps->v2.max_smooth_line_width = range[1];
glGetFloatv(GL_MAX_TEXTURE_LOD_BIAS, &caps->v2.max_texture_lod_bias);
glGetIntegerv(GL_MAX_VERTEX_ATTRIBS, &caps->v2.max_vertex_attribs);
glGetIntegerv(GL_MAX_VERTEX_OUTPUT_COMPONENTS, &max);
caps->v2.max_vertex_outputs = max / 4;
if (gl_ver >= 32) {
glGetIntegerv(GL_MAX_GEOMETRY_OUTPUT_VERTICES, &caps->v2.max_geom_output_vertices);
glGetIntegerv(GL_MAX_GEOMETRY_TOTAL_OUTPUT_COMPONENTS, &caps->v2.max_geom_total_output_components);
}
caps->v2.max_shader_patch_varyings = 0; // until we do tess.
if (epoxy_has_gl_extension("GL_ARB_texture_gather")) {
glGetIntegerv(GL_MIN_PROGRAM_TEXTURE_GATHER_OFFSET, &caps->v2.min_texture_gather_offset);
glGetIntegerv(GL_MAX_PROGRAM_TEXTURE_GATHER_OFFSET, &caps->v2.max_texture_gather_offset);
}
glGetIntegerv(GL_MIN_PROGRAM_TEXEL_OFFSET, &caps->v2.min_texel_offset);
glGetIntegerv(GL_MAX_PROGRAM_TEXEL_OFFSET, &caps->v2.max_texel_offset);
glGetIntegerv(GL_UNIFORM_BUFFER_OFFSET_ALIGNMENT, &caps->v2.uniform_buffer_offset_alignment);
if (gl_ver >= 43)
glGetIntegerv(GL_TEXTURE_BUFFER_OFFSET_ALIGNMENT, &caps->v2.texture_buffer_offset_alignment);
}
GLint64 vrend_renderer_get_timestamp(void)
{
GLint64 v;
glGetInteger64v(GL_TIMESTAMP, &v);
return v;
}
void *vrend_renderer_get_cursor_contents(uint32_t res_handle, uint32_t *width, uint32_t *height)
{
GLenum format, type;
struct vrend_resource *res;
int blsize;
char *data, *data2;
int size;
int h;
res = vrend_resource_lookup(res_handle, 0);
if (!res)
return NULL;
if (res->base.width0 > 128 || res->base.height0 > 128)
return NULL;
if (res->target != GL_TEXTURE_2D)
return NULL;
if (width)
*width = res->base.width0;
if (height)
*height = res->base.height0;
format = tex_conv_table[res->base.format].glformat;
type = tex_conv_table[res->base.format].gltype;
blsize = util_format_get_blocksize(res->base.format);
size = util_format_get_nblocks(res->base.format, res->base.width0, res->base.height0) * blsize;
data = malloc(size);
data2 = malloc(size);
if (!data || !data2) {
free(data);
free(data2);
return NULL;
}
if (vrend_state.have_arb_robustness) {
glBindTexture(res->target, res->id);
glGetnTexImageARB(res->target, 0, format, type, size, data);
} else if (vrend_state.use_gles) {
GLuint fb_id;
if (res->readback_fb_id == 0 || res->readback_fb_level != 0 || res->readback_fb_z != 0) {
if (res->readback_fb_id)
glDeleteFramebuffers(1, &res->readback_fb_id);
glGenFramebuffers(1, &fb_id);
glBindFramebufferEXT(GL_FRAMEBUFFER_EXT, fb_id);
vrend_fb_bind_texture(res, 0, 0, 0);
res->readback_fb_id = fb_id;
res->readback_fb_level = 0;
res->readback_fb_z = 0;
} else {
glBindFramebufferEXT(GL_FRAMEBUFFER_EXT, res->readback_fb_id);
}
if (vrend_state.have_arb_robustness) {
glReadnPixelsARB(0, 0, *width, *height, format, type, size, data);
} else if (vrend_state.have_gles_khr_robustness) {
glReadnPixelsKHR(0, 0, *width, *height, format, type, size, data);
} else {
glReadPixels(0, 0, *width, *height, format, type, data);
}
} else {
glBindTexture(res->target, res->id);
glGetTexImage(res->target, 0, format, type, data);
}
for (h = 0; h < res->base.height0; h++) {
uint32_t doff = (res->base.height0 - h - 1) * res->base.width0 * blsize;
uint32_t soff = h * res->base.width0 * blsize;
memcpy(data2 + doff, data + soff, res->base.width0 * blsize);
}
free(data);
return data2;
}
void vrend_renderer_force_ctx_0(void)
{
struct vrend_context *ctx0 = vrend_lookup_renderer_ctx(0);
vrend_state.current_ctx = NULL;
vrend_state.current_hw_ctx = NULL;
vrend_hw_switch_context(ctx0, true);
vrend_clicbs->make_current(0, ctx0->sub->gl_context);
}
void vrend_renderer_get_rect(int res_handle, struct iovec *iov, unsigned int num_iovs,
uint32_t offset, int x, int y, int width, int height)
{
struct vrend_resource *res = vrend_resource_lookup(res_handle, 0);
struct vrend_transfer_info transfer_info;
struct pipe_box box;
int elsize;
memset(&transfer_info, 0, sizeof(transfer_info));
elsize = util_format_get_blocksize(res->base.format);
box.x = x;
box.y = y;
box.z = 0;
box.width = width;
box.height = height;
box.depth = 1;
transfer_info.box = &box;
transfer_info.stride = util_format_get_nblocksx(res->base.format, res->base.width0) * elsize;
transfer_info.offset = offset;
transfer_info.handle = res->handle;
transfer_info.iovec = iov;
transfer_info.iovec_cnt = num_iovs;
vrend_renderer_transfer_iov(&transfer_info, VREND_TRANSFER_READ);
}
void vrend_renderer_attach_res_ctx(int ctx_id, int resource_id)
{
struct vrend_context *ctx = vrend_lookup_renderer_ctx(ctx_id);
struct vrend_resource *res;
if (!ctx)
return;
res = vrend_resource_lookup(resource_id, 0);
if (!res)
return;
vrend_object_insert_nofree(ctx->res_hash, res, sizeof(*res), resource_id, 1, false);
}
static void vrend_renderer_detach_res_ctx_p(struct vrend_context *ctx, int res_handle)
{
struct vrend_resource *res;
res = vrend_object_lookup(ctx->res_hash, res_handle, 1);
if (!res)
return;
vrend_object_remove(ctx->res_hash, res_handle, 1);
}
void vrend_renderer_detach_res_ctx(int ctx_id, int res_handle)
{
struct vrend_context *ctx = vrend_lookup_renderer_ctx(ctx_id);
if (!ctx)
return;
vrend_renderer_detach_res_ctx_p(ctx, res_handle);
}
static struct vrend_resource *vrend_renderer_ctx_res_lookup(struct vrend_context *ctx, int res_handle)
{
struct vrend_resource *res = vrend_object_lookup(ctx->res_hash, res_handle, 1);
return res;
}
int vrend_renderer_resource_get_info(int res_handle,
struct vrend_renderer_resource_info *info)
{
struct vrend_resource *res;
int elsize;
if (!info)
return EINVAL;
res = vrend_resource_lookup(res_handle, 0);
if (!res)
return EINVAL;
elsize = util_format_get_blocksize(res->base.format);
info->handle = res_handle;
info->tex_id = res->id;
info->width = res->base.width0;
info->height = res->base.height0;
info->depth = res->base.depth0;
info->format = res->base.format;
info->flags = res->y_0_top ? VIRGL_RESOURCE_Y_0_TOP : 0;
info->stride = util_format_get_nblocksx(res->base.format, u_minify(res->base.width0, 0)) * elsize;
return 0;
}
void vrend_renderer_get_cap_set(uint32_t cap_set, uint32_t *max_ver,
uint32_t *max_size)
{
switch (cap_set) {
case VREND_CAP_SET:
*max_ver = 1;
*max_size = sizeof(struct virgl_caps_v1);
break;
case VREND_CAP_SET2:
/* we should never need to increase this - it should be possible to just grow virgl_caps */
*max_ver = 2;
*max_size = sizeof(struct virgl_caps_v2);
break;
default:
*max_ver = 0;
*max_size = 0;
break;
}
}
void vrend_renderer_create_sub_ctx(struct vrend_context *ctx, int sub_ctx_id)
{
struct vrend_sub_context *sub;
struct virgl_gl_ctx_param ctx_params;
LIST_FOR_EACH_ENTRY(sub, &ctx->sub_ctxs, head) {
if (sub->sub_ctx_id == sub_ctx_id) {
return;
}
}
sub = CALLOC_STRUCT(vrend_sub_context);
if (!sub)
return;
ctx_params.shared = (ctx->ctx_id == 0 && sub_ctx_id == 0) ? false : true;
ctx_params.major_ver = vrend_state.gl_major_ver;
ctx_params.minor_ver = vrend_state.gl_minor_ver;
sub->gl_context = vrend_clicbs->create_gl_context(0, &ctx_params);
vrend_clicbs->make_current(0, sub->gl_context);
/* enable if vrend_renderer_init function has done it as well */
if (vrend_state.have_debug_cb) {
glDebugMessageCallback(vrend_debug_cb, NULL);
glEnable(GL_DEBUG_OUTPUT);
glDisable(GL_DEBUG_OUTPUT_SYNCHRONOUS);
}
sub->sub_ctx_id = sub_ctx_id;
if (!vrend_state.have_vertex_attrib_binding) {
glGenVertexArrays(1, &sub->vaoid);
glBindVertexArray(sub->vaoid);
}
glGenFramebuffers(1, &sub->fb_id);
glGenFramebuffers(2, sub->blit_fb_ids);
list_inithead(&sub->programs);
list_inithead(&sub->streamout_list);
sub->object_hash = vrend_object_init_ctx_table();
ctx->sub = sub;
list_add(&sub->head, &ctx->sub_ctxs);
if (sub_ctx_id == 0)
ctx->sub0 = sub;
}
void vrend_renderer_destroy_sub_ctx(struct vrend_context *ctx, int sub_ctx_id)
{
struct vrend_sub_context *sub, *tofree = NULL;
/* never destroy sub context id 0 */
if (sub_ctx_id == 0)
return;
LIST_FOR_EACH_ENTRY(sub, &ctx->sub_ctxs, head) {
if (sub->sub_ctx_id == sub_ctx_id) {
tofree = sub;
}
}
if (tofree) {
if (ctx->sub == tofree) {
ctx->sub = ctx->sub0;
vrend_clicbs->make_current(0, ctx->sub->gl_context);
}
vrend_destroy_sub_context(tofree);
}
}
void vrend_renderer_set_sub_ctx(struct vrend_context *ctx, int sub_ctx_id)
{
struct vrend_sub_context *sub;
/* find the sub ctx */
if (ctx->sub && ctx->sub->sub_ctx_id == sub_ctx_id)
return;
LIST_FOR_EACH_ENTRY(sub, &ctx->sub_ctxs, head) {
if (sub->sub_ctx_id == sub_ctx_id) {
ctx->sub = sub;
vrend_clicbs->make_current(0, sub->gl_context);
break;
}
}
}
static void vrend_reset_fences(void)
{
struct vrend_fence *fence, *stor;
renderer: use a thread to block for fences. Instead of polling the fences regularly, have a thread that blocks for a single fence using a separate shared context, then uses eventfd to wake up the main thread when something happens. Inside the guest, glmark2 typicially runs twice as fast with the thread sync. Although in general, the performances seems to be about +30%. The benefits is mostly for CPU-bounds tasks (when main the thread hits 100%) A naive perf stat of the vtest renderer with glmark2 "build" test with a fixed number of frames (500) results in the following stats data: (do not value timing related informations, since the renderer is ran and stopped manually) without thread: 3032.282265 task-clock (msec) # 0.420 CPUs utilized 4,277 context-switches # 0.001 M/sec 102 cpu-migrations # 0.034 K/sec 9,020 page-faults # 0.003 M/sec 7,884,098,254 cycles # 2.600 GHz 4,440,126,451 stalled-cycles-frontend # 56.32% frontend cycles idle <not supported> stalled-cycles-backend 11,024,091,578 instructions # 1.40 insns per cycle # 0.40 stalled # cycles per insn 1,091,831,588 branches # 360.069 M/sec 5,426,846 branch-misses # 0.50% of all branches with thread: 3403.592921 task-clock (msec) # 0.452 CPUs utilized 7,145 context-switches # 0.002 M/sec 410 cpu-migrations # 0.120 K/sec 6,191 page-faults # 0.002 M/sec 7,475,038,064 cycles # 2.196 GHz 4,487,043,071 stalled-cycles-frontend # 60.03% frontend cycles idle <not supported> stalled-cycles-backend 9,925,205,494 instructions # 1.33 insns per cycle # 0.45 stalled # cycles per insn 834,375,503 branches # 245.146 M/sec 4,919,995 branch-misses # 0.59% of all branches Signed-off-by: Marc-André Lureau <marcandre.lureau@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
9 years ago
if (vrend_state.sync_thread)
pipe_mutex_lock(vrend_state.fence_mutex);
LIST_FOR_EACH_ENTRY_SAFE(fence, stor, &vrend_state.fence_list, fences) {
renderer: use a thread to block for fences. Instead of polling the fences regularly, have a thread that blocks for a single fence using a separate shared context, then uses eventfd to wake up the main thread when something happens. Inside the guest, glmark2 typicially runs twice as fast with the thread sync. Although in general, the performances seems to be about +30%. The benefits is mostly for CPU-bounds tasks (when main the thread hits 100%) A naive perf stat of the vtest renderer with glmark2 "build" test with a fixed number of frames (500) results in the following stats data: (do not value timing related informations, since the renderer is ran and stopped manually) without thread: 3032.282265 task-clock (msec) # 0.420 CPUs utilized 4,277 context-switches # 0.001 M/sec 102 cpu-migrations # 0.034 K/sec 9,020 page-faults # 0.003 M/sec 7,884,098,254 cycles # 2.600 GHz 4,440,126,451 stalled-cycles-frontend # 56.32% frontend cycles idle <not supported> stalled-cycles-backend 11,024,091,578 instructions # 1.40 insns per cycle # 0.40 stalled # cycles per insn 1,091,831,588 branches # 360.069 M/sec 5,426,846 branch-misses # 0.50% of all branches with thread: 3403.592921 task-clock (msec) # 0.452 CPUs utilized 7,145 context-switches # 0.002 M/sec 410 cpu-migrations # 0.120 K/sec 6,191 page-faults # 0.002 M/sec 7,475,038,064 cycles # 2.196 GHz 4,487,043,071 stalled-cycles-frontend # 60.03% frontend cycles idle <not supported> stalled-cycles-backend 9,925,205,494 instructions # 1.33 insns per cycle # 0.45 stalled # cycles per insn 834,375,503 branches # 245.146 M/sec 4,919,995 branch-misses # 0.59% of all branches Signed-off-by: Marc-André Lureau <marcandre.lureau@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
9 years ago
free_fence_locked(fence);
}
renderer: use a thread to block for fences. Instead of polling the fences regularly, have a thread that blocks for a single fence using a separate shared context, then uses eventfd to wake up the main thread when something happens. Inside the guest, glmark2 typicially runs twice as fast with the thread sync. Although in general, the performances seems to be about +30%. The benefits is mostly for CPU-bounds tasks (when main the thread hits 100%) A naive perf stat of the vtest renderer with glmark2 "build" test with a fixed number of frames (500) results in the following stats data: (do not value timing related informations, since the renderer is ran and stopped manually) without thread: 3032.282265 task-clock (msec) # 0.420 CPUs utilized 4,277 context-switches # 0.001 M/sec 102 cpu-migrations # 0.034 K/sec 9,020 page-faults # 0.003 M/sec 7,884,098,254 cycles # 2.600 GHz 4,440,126,451 stalled-cycles-frontend # 56.32% frontend cycles idle <not supported> stalled-cycles-backend 11,024,091,578 instructions # 1.40 insns per cycle # 0.40 stalled # cycles per insn 1,091,831,588 branches # 360.069 M/sec 5,426,846 branch-misses # 0.50% of all branches with thread: 3403.592921 task-clock (msec) # 0.452 CPUs utilized 7,145 context-switches # 0.002 M/sec 410 cpu-migrations # 0.120 K/sec 6,191 page-faults # 0.002 M/sec 7,475,038,064 cycles # 2.196 GHz 4,487,043,071 stalled-cycles-frontend # 60.03% frontend cycles idle <not supported> stalled-cycles-backend 9,925,205,494 instructions # 1.33 insns per cycle # 0.45 stalled # cycles per insn 834,375,503 branches # 245.146 M/sec 4,919,995 branch-misses # 0.59% of all branches Signed-off-by: Marc-André Lureau <marcandre.lureau@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
9 years ago
if (vrend_state.sync_thread)
pipe_mutex_unlock(vrend_state.fence_mutex);
}
void vrend_renderer_reset(void)
{
renderer: use a thread to block for fences. Instead of polling the fences regularly, have a thread that blocks for a single fence using a separate shared context, then uses eventfd to wake up the main thread when something happens. Inside the guest, glmark2 typicially runs twice as fast with the thread sync. Although in general, the performances seems to be about +30%. The benefits is mostly for CPU-bounds tasks (when main the thread hits 100%) A naive perf stat of the vtest renderer with glmark2 "build" test with a fixed number of frames (500) results in the following stats data: (do not value timing related informations, since the renderer is ran and stopped manually) without thread: 3032.282265 task-clock (msec) # 0.420 CPUs utilized 4,277 context-switches # 0.001 M/sec 102 cpu-migrations # 0.034 K/sec 9,020 page-faults # 0.003 M/sec 7,884,098,254 cycles # 2.600 GHz 4,440,126,451 stalled-cycles-frontend # 56.32% frontend cycles idle <not supported> stalled-cycles-backend 11,024,091,578 instructions # 1.40 insns per cycle # 0.40 stalled # cycles per insn 1,091,831,588 branches # 360.069 M/sec 5,426,846 branch-misses # 0.50% of all branches with thread: 3403.592921 task-clock (msec) # 0.452 CPUs utilized 7,145 context-switches # 0.002 M/sec 410 cpu-migrations # 0.120 K/sec 6,191 page-faults # 0.002 M/sec 7,475,038,064 cycles # 2.196 GHz 4,487,043,071 stalled-cycles-frontend # 60.03% frontend cycles idle <not supported> stalled-cycles-backend 9,925,205,494 instructions # 1.33 insns per cycle # 0.45 stalled # cycles per insn 834,375,503 branches # 245.146 M/sec 4,919,995 branch-misses # 0.59% of all branches Signed-off-by: Marc-André Lureau <marcandre.lureau@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
9 years ago
if (vrend_state.sync_thread) {
vrend_free_sync_thread();
vrend_state.stop_sync_thread = false;
}
vrend_reset_fences();
vrend_decode_reset(false);
vrend_object_fini_resource_table();
vrend_decode_reset(true);
vrend_object_init_resource_table();
vrend_renderer_context_create_internal(0, 0, NULL);
}
renderer: use a thread to block for fences. Instead of polling the fences regularly, have a thread that blocks for a single fence using a separate shared context, then uses eventfd to wake up the main thread when something happens. Inside the guest, glmark2 typicially runs twice as fast with the thread sync. Although in general, the performances seems to be about +30%. The benefits is mostly for CPU-bounds tasks (when main the thread hits 100%) A naive perf stat of the vtest renderer with glmark2 "build" test with a fixed number of frames (500) results in the following stats data: (do not value timing related informations, since the renderer is ran and stopped manually) without thread: 3032.282265 task-clock (msec) # 0.420 CPUs utilized 4,277 context-switches # 0.001 M/sec 102 cpu-migrations # 0.034 K/sec 9,020 page-faults # 0.003 M/sec 7,884,098,254 cycles # 2.600 GHz 4,440,126,451 stalled-cycles-frontend # 56.32% frontend cycles idle <not supported> stalled-cycles-backend 11,024,091,578 instructions # 1.40 insns per cycle # 0.40 stalled # cycles per insn 1,091,831,588 branches # 360.069 M/sec 5,426,846 branch-misses # 0.50% of all branches with thread: 3403.592921 task-clock (msec) # 0.452 CPUs utilized 7,145 context-switches # 0.002 M/sec 410 cpu-migrations # 0.120 K/sec 6,191 page-faults # 0.002 M/sec 7,475,038,064 cycles # 2.196 GHz 4,487,043,071 stalled-cycles-frontend # 60.03% frontend cycles idle <not supported> stalled-cycles-backend 9,925,205,494 instructions # 1.33 insns per cycle # 0.45 stalled # cycles per insn 834,375,503 branches # 245.146 M/sec 4,919,995 branch-misses # 0.59% of all branches Signed-off-by: Marc-André Lureau <marcandre.lureau@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
9 years ago
int vrend_renderer_get_poll_fd(void)
{
if (!vrend_state.inited)
return -1;
return vrend_state.eventfd;
}