diff options
30 files changed, 233 insertions, 111 deletions
diff --git a/SConstruct b/SConstruct index 3734268cae..df1be4bfd7 100644 --- a/SConstruct +++ b/SConstruct @@ -337,17 +337,18 @@ if selected_platform in platform_list: else: # Rest of the world disable_nonessential_warnings = ['-Wno-sign-compare'] shadow_local_warning = [] + all_plus_warnings = ['-Wwrite-strings'] - if 'gcc' in os.path.basename(env["CC"]): + if methods.use_gcc(env): version = methods.get_compiler_version(env) if version != None and version[0] >= '7': shadow_local_warning = ['-Wshadow-local'] if (env["warnings"] == 'extra'): - env.Append(CCFLAGS=['-Wall', '-Wextra'] + shadow_local_warning) + env.Append(CCFLAGS=['-Wall', '-Wextra'] + all_plus_warnings + shadow_local_warning) elif (env["warnings"] == 'all'): - env.Append(CCFLAGS=['-Wall'] + shadow_local_warning + disable_nonessential_warnings) + env.Append(CCFLAGS=['-Wall'] + all_plus_warnings + shadow_local_warning + disable_nonessential_warnings) elif (env["warnings"] == 'moderate'): - env.Append(CCFLAGS=['-Wall', '-Wno-unused'] + shadow_local_warning + disable_nonessential_warnings) + env.Append(CCFLAGS=['-Wall', '-Wno-unused'] + shadow_local_warning + disable_nonessential_warnings) else: # 'no' env.Append(CCFLAGS=['-w']) if (env["werror"]): diff --git a/drivers/gles2/rasterizer_canvas_gles2.cpp b/drivers/gles2/rasterizer_canvas_gles2.cpp index 9a49d5a7e4..320242b773 100644 --- a/drivers/gles2/rasterizer_canvas_gles2.cpp +++ b/drivers/gles2/rasterizer_canvas_gles2.cpp @@ -499,6 +499,23 @@ void RasterizerCanvasGLES2::_canvas_item_render_commands(Item *p_item, Item *cur glDisableVertexAttribArray(VS::ARRAY_COLOR); glVertexAttrib4fv(VS::ARRAY_COLOR, r->modulate.components); + bool can_tile = true; + if (r->texture.is_valid() && r->flags & CANVAS_RECT_TILE && !storage->config.support_npot_repeat_mipmap) { + // workaround for when setting tiling does not work due to hardware limitation + + RasterizerStorageGLES2::Texture *texture = storage->texture_owner.getornull(r->texture); + + if (texture) { + + texture = texture->get_ptr(); + + if (next_power_of_2(texture->alloc_width) != texture->alloc_width && next_power_of_2(texture->alloc_height) != texture->alloc_height) { + state.canvas_shader.set_conditional(CanvasShaderGLES2::USE_FORCE_REPEAT, true); + can_tile = false; + } + } + } + // On some widespread Nvidia cards, the normal draw method can produce some // flickering in draw_rect and especially TileMap rendering (tiles randomly flicker). // See GH-9913. @@ -559,7 +576,7 @@ void RasterizerCanvasGLES2::_canvas_item_render_commands(Item *p_item, Item *cur bool untile = false; - if (r->flags & CANVAS_RECT_TILE && !(texture->flags & VS::TEXTURE_FLAG_REPEAT)) { + if (can_tile && r->flags & CANVAS_RECT_TILE && !(texture->flags & VS::TEXTURE_FLAG_REPEAT)) { glTexParameterf(GL_TEXTURE_2D, GL_TEXTURE_WRAP_S, GL_REPEAT); glTexParameterf(GL_TEXTURE_2D, GL_TEXTURE_WRAP_T, GL_REPEAT); untile = true; @@ -616,7 +633,7 @@ void RasterizerCanvasGLES2::_canvas_item_render_commands(Item *p_item, Item *cur bool untile = false; - if (r->flags & CANVAS_RECT_TILE && !(tex->flags & VS::TEXTURE_FLAG_REPEAT)) { + if (can_tile && r->flags & CANVAS_RECT_TILE && !(tex->flags & VS::TEXTURE_FLAG_REPEAT)) { glTexParameterf(GL_TEXTURE_2D, GL_TEXTURE_WRAP_S, GL_REPEAT); glTexParameterf(GL_TEXTURE_2D, GL_TEXTURE_WRAP_T, GL_REPEAT); untile = true; @@ -664,6 +681,9 @@ void RasterizerCanvasGLES2::_canvas_item_render_commands(Item *p_item, Item *cur glBindBuffer(GL_ARRAY_BUFFER, 0); glBindBuffer(GL_ELEMENT_ARRAY_BUFFER, 0); } + + state.canvas_shader.set_conditional(CanvasShaderGLES2::USE_FORCE_REPEAT, false); + } break; case Item::Command::TYPE_NINEPATCH: { diff --git a/drivers/gles2/rasterizer_storage_gles2.cpp b/drivers/gles2/rasterizer_storage_gles2.cpp index c02eeaa0ee..29e125e8a2 100644 --- a/drivers/gles2/rasterizer_storage_gles2.cpp +++ b/drivers/gles2/rasterizer_storage_gles2.cpp @@ -67,7 +67,7 @@ void RasterizerStorageGLES2::bind_quad_array() const { glEnableVertexAttribArray(VS::ARRAY_TEX_UV); } -Ref<Image> RasterizerStorageGLES2::_get_gl_image_and_format(const Ref<Image> &p_image, Image::Format p_format, uint32_t p_flags, Image::Format &r_real_format, GLenum &r_gl_format, GLenum &r_gl_internal_format, GLenum &r_gl_type, bool &r_compressed) const { +Ref<Image> RasterizerStorageGLES2::_get_gl_image_and_format(const Ref<Image> &p_image, Image::Format p_format, uint32_t p_flags, Image::Format &r_real_format, GLenum &r_gl_format, GLenum &r_gl_internal_format, GLenum &r_gl_type, bool &r_compressed, bool p_will_need_resize) const { r_gl_format = 0; Ref<Image> image = p_image; @@ -195,7 +195,7 @@ Ref<Image> RasterizerStorageGLES2::_get_gl_image_and_format(const Ref<Image> &p_ } break; case Image::FORMAT_DXT1: { - if (config.s3tc_supported) { + if (config.s3tc_supported && !p_will_need_resize) { r_gl_internal_format = _EXT_COMPRESSED_RGBA_S3TC_DXT1_EXT; r_gl_format = GL_RGBA; r_gl_type = GL_UNSIGNED_BYTE; @@ -207,7 +207,7 @@ Ref<Image> RasterizerStorageGLES2::_get_gl_image_and_format(const Ref<Image> &p_ } break; case Image::FORMAT_DXT3: { - if (config.s3tc_supported) { + if (config.s3tc_supported && !p_will_need_resize) { r_gl_internal_format = _EXT_COMPRESSED_RGBA_S3TC_DXT3_EXT; r_gl_format = GL_RGBA; r_gl_type = GL_UNSIGNED_BYTE; @@ -219,7 +219,7 @@ Ref<Image> RasterizerStorageGLES2::_get_gl_image_and_format(const Ref<Image> &p_ } break; case Image::FORMAT_DXT5: { - if (config.s3tc_supported) { + if (config.s3tc_supported && !p_will_need_resize) { r_gl_internal_format = _EXT_COMPRESSED_RGBA_S3TC_DXT5_EXT; r_gl_format = GL_RGBA; r_gl_type = GL_UNSIGNED_BYTE; @@ -269,7 +269,7 @@ Ref<Image> RasterizerStorageGLES2::_get_gl_image_and_format(const Ref<Image> &p_ } break; case Image::FORMAT_ETC: { - if (config.etc1_supported) { + if (config.etc1_supported && !p_will_need_resize) { r_gl_internal_format = _EXT_ETC1_RGB8_OES; r_gl_format = GL_RGBA; r_gl_type = GL_UNSIGNED_BYTE; @@ -315,17 +315,38 @@ Ref<Image> RasterizerStorageGLES2::_get_gl_image_and_format(const Ref<Image> &p_ if (need_decompress) { if (!image.is_null()) { + image = image->duplicate(); + print_line("decompressing..."); image->decompress(); ERR_FAIL_COND_V(image->is_compressed(), image); - image->convert(Image::FORMAT_RGBA8); + switch (image->get_format()) { + case Image::FORMAT_RGB8: { + r_gl_format = GL_RGB; + r_gl_internal_format = GL_RGB; + r_gl_type = GL_UNSIGNED_BYTE; + r_real_format = Image::FORMAT_RGB8; + r_compressed = false; + } break; + case Image::FORMAT_RGBA8: { + r_gl_format = GL_RGBA; + r_gl_internal_format = GL_RGBA; + r_gl_type = GL_UNSIGNED_BYTE; + r_real_format = Image::FORMAT_RGBA8; + r_compressed = false; + } break; + default: { + image->convert(Image::FORMAT_RGBA8); + r_gl_format = GL_RGBA; + r_gl_internal_format = GL_RGBA; + r_gl_type = GL_UNSIGNED_BYTE; + r_real_format = Image::FORMAT_RGBA8; + r_compressed = false; + + } break; + } } - r_gl_format = GL_RGBA; - r_gl_internal_format = GL_RGBA; - r_gl_type = GL_UNSIGNED_BYTE; - r_real_format = Image::FORMAT_RGBA8; - return image; } @@ -395,11 +416,31 @@ void RasterizerStorageGLES2::texture_allocate(RID p_texture, int p_width, int p_ } } - Image::Format real_format; - _get_gl_image_and_format(Ref<Image>(), texture->format, texture->flags, real_format, format, internal_format, type, compressed); - texture->alloc_width = texture->width; texture->alloc_height = texture->height; + texture->resize_to_po2 = false; + if (!config.support_npot_repeat_mipmap) { + int po2_width = next_power_of_2(p_width); + int po2_height = next_power_of_2(p_height); + + bool is_po2 = p_width == po2_width && p_height == po2_height; + + if (!is_po2 && (p_flags & VS::TEXTURE_FLAG_REPEAT || p_flags & VS::TEXTURE_FLAG_MIPMAPS)) { + + if (p_flags & VS::TEXTURE_FLAG_USED_FOR_STREAMING) { + //not supported + ERR_PRINTS("Streaming texture for non power of 2 or has mipmaps on this hardware: " + texture->path + "'. Mipmaps and repeat disabled."); + texture->flags &= ~(VS::TEXTURE_FLAG_REPEAT | VS::TEXTURE_FLAG_MIPMAPS); + } else { + texture->alloc_height = po2_height; + texture->alloc_width = po2_width; + texture->resize_to_po2 = true; + } + } + } + + Image::Format real_format; + _get_gl_image_and_format(Ref<Image>(), texture->format, texture->flags, real_format, format, internal_format, type, compressed, texture->resize_to_po2); texture->gl_format_cache = format; texture->gl_type_cache = type; @@ -414,7 +455,7 @@ void RasterizerStorageGLES2::texture_allocate(RID p_texture, int p_width, int p_ if (p_flags & VS::TEXTURE_FLAG_USED_FOR_STREAMING) { //prealloc if video - glTexImage2D(texture->target, 0, internal_format, p_width, p_height, 0, format, type, NULL); + glTexImage2D(texture->target, 0, internal_format, texture->alloc_width, texture->alloc_height, 0, format, type, NULL); } texture->active = true; @@ -439,7 +480,18 @@ void RasterizerStorageGLES2::texture_set_data(RID p_texture, const Ref<Image> &p } Image::Format real_format; - Ref<Image> img = _get_gl_image_and_format(p_image, p_image->get_format(), texture->flags, real_format, format, internal_format, type, compressed); + Ref<Image> img = _get_gl_image_and_format(p_image, p_image->get_format(), texture->flags, real_format, format, internal_format, type, compressed, texture->resize_to_po2); + + if (texture->resize_to_po2) { + if (p_image->is_compressed()) { + ERR_PRINTS("Texture '" + texture->path + "' was required to be a power of 2 (because it uses either mipmaps or repeat), so it was decompressed. This will hurt performance and memory usage."); + } + + if (img == p_image) { + img = img->duplicate(); + } + img->resize_to_po2(false); + } if (config.shrink_textures_x2 && (p_image->has_mipmaps() || !p_image->is_compressed()) && !(texture->flags & VS::TEXTURE_FLAG_USED_FOR_STREAMING)) { @@ -575,7 +627,7 @@ Ref<Image> RasterizerStorageGLES2::texture_get_data(RID p_texture, int p_layer) GLenum gl_internal_format; GLenum gl_type; bool compressed; - _get_gl_image_and_format(Ref<Image>(), texture->format, texture->flags, real_format, gl_format, gl_internal_format, gl_type, compressed); + _get_gl_image_and_format(Ref<Image>(), texture->format, texture->flags, real_format, gl_format, gl_internal_format, gl_type, compressed, false); PoolVector<uint8_t> data; @@ -620,7 +672,7 @@ Ref<Image> RasterizerStorageGLES2::texture_get_data(RID p_texture, int p_layer) GLenum gl_internal_format; GLenum gl_type; bool compressed; - _get_gl_image_and_format(Ref<Image>(), texture->format, texture->flags, real_format, gl_format, gl_internal_format, gl_type, compressed); + _get_gl_image_and_format(Ref<Image>(), texture->format, texture->flags, real_format, gl_format, gl_internal_format, gl_type, compressed, texture->resize_to_po2); PoolVector<uint8_t> data; @@ -4900,10 +4952,13 @@ void RasterizerStorageGLES2::initialize() { config.float_texture_supported = true; config.s3tc_supported = true; config.etc1_supported = false; + config.support_npot_repeat_mipmap = true; #else config.float_texture_supported = config.extensions.has("GL_ARB_texture_float") || config.extensions.has("GL_OES_texture_float"); config.s3tc_supported = config.extensions.has("GL_EXT_texture_compression_s3tc") || config.extensions.has("WEBGL_compressed_texture_s3tc"); config.etc1_supported = config.extensions.has("GL_OES_compressed_ETC1_RGB8_texture") || config.extensions.has("WEBGL_compressed_texture_etc1"); + config.support_npot_repeat_mipmap = config.extensions.has("GL_OES_texture_npot"); + #endif #ifdef GLES_OVER_GL config.use_rgba_2d_shadows = false; diff --git a/drivers/gles2/rasterizer_storage_gles2.h b/drivers/gles2/rasterizer_storage_gles2.h index a0fa2aacc5..0f1d81f70d 100644 --- a/drivers/gles2/rasterizer_storage_gles2.h +++ b/drivers/gles2/rasterizer_storage_gles2.h @@ -82,6 +82,7 @@ public: bool support_32_bits_indices; bool support_write_depth; bool support_half_float_vertices; + bool support_npot_repeat_mipmap; } config; struct Resources { @@ -240,6 +241,8 @@ public: int mipmaps; + bool resize_to_po2; + bool active; GLenum tex_id; @@ -275,6 +278,7 @@ public: ignore_mipmaps(false), compressed(false), mipmaps(0), + resize_to_po2(false), active(false), tex_id(0), stored_cube_sides(0), @@ -313,7 +317,7 @@ public: mutable RID_Owner<Texture> texture_owner; - Ref<Image> _get_gl_image_and_format(const Ref<Image> &p_image, Image::Format p_format, uint32_t p_flags, Image::Format &r_real_format, GLenum &r_gl_format, GLenum &r_gl_internal_format, GLenum &r_gl_type, bool &r_compressed) const; + Ref<Image> _get_gl_image_and_format(const Ref<Image> &p_image, Image::Format p_format, uint32_t p_flags, Image::Format &r_real_format, GLenum &r_gl_format, GLenum &r_gl_internal_format, GLenum &r_gl_type, bool &r_compressed, bool p_will_need_resize) const; virtual RID texture_create(); virtual void texture_allocate(RID p_texture, int p_width, int p_height, int p_depth_3d, Image::Format p_format, VS::TextureType p_type, uint32_t p_flags = VS::TEXTURE_FLAGS_DEFAULT); diff --git a/drivers/gles2/shader_compiler_gles2.cpp b/drivers/gles2/shader_compiler_gles2.cpp index fdd905458b..d00b03fb8a 100644 --- a/drivers/gles2/shader_compiler_gles2.cpp +++ b/drivers/gles2/shader_compiler_gles2.cpp @@ -770,7 +770,7 @@ ShaderCompilerGLES2::ShaderCompilerGLES2() { /** CANVAS ITEM SHADER **/ actions[VS::SHADER_CANVAS_ITEM].renames["VERTEX"] = "outvec.xy"; - actions[VS::SHADER_CANVAS_ITEM].renames["UV"] = "uv_interp"; + actions[VS::SHADER_CANVAS_ITEM].renames["UV"] = "uv"; actions[VS::SHADER_CANVAS_ITEM].renames["POINT_SIZE"] = "gl_PointSize"; actions[VS::SHADER_CANVAS_ITEM].renames["WORLD_MATRIX"] = "modelview_matrix"; diff --git a/drivers/gles2/shaders/canvas.glsl b/drivers/gles2/shaders/canvas.glsl index e74c7d7166..c292897ad0 100644 --- a/drivers/gles2/shaders/canvas.glsl +++ b/drivers/gles2/shaders/canvas.glsl @@ -107,6 +107,7 @@ vec2 select(vec2 a, vec2 b, bvec2 c) { void main() { vec4 color = color_attrib; + vec2 uv; #ifdef USE_INSTANCING mat4 extra_matrix_instance = extra_matrix * transpose(mat4(instance_xform0, instance_xform1, instance_xform2, vec4(0.0, 0.0, 0.0, 1.0))); @@ -121,9 +122,9 @@ void main() { #ifdef USE_TEXTURE_RECT if (dst_rect.z < 0.0) { // Transpose is encoded as negative dst_rect.z - uv_interp = src_rect.xy + abs(src_rect.zw) * vertex.yx; + uv = src_rect.xy + abs(src_rect.zw) * vertex.yx; } else { - uv_interp = src_rect.xy + abs(src_rect.zw) * vertex; + uv = src_rect.xy + abs(src_rect.zw) * vertex; } vec4 outvec = vec4(0.0, 0.0, 0.0, 1.0); @@ -140,7 +141,7 @@ void main() { #else vec4 outvec = vec4(vertex.xy, 0.0, 1.0); - uv_interp = uv_attrib; + uv = uv_attrib; #endif { @@ -189,6 +190,7 @@ VERTEX_SHADER_CODE #endif + uv_interp = uv; gl_Position = projection_matrix * outvec; #ifdef USE_LIGHTING @@ -345,10 +347,14 @@ void main() { vec4 color = color_interp; vec2 uv = uv_interp; +#ifdef USE_FORCE_REPEAT + //needs to use this to workaround GLES2/WebGL1 forcing tiling that textures that dont support it + uv = mod(uv,vec2(1.0,1.0)); +#endif #if !defined(COLOR_USED) //default behavior, texture by color - color *= texture2D(color_texture, uv_interp); + color *= texture2D(color_texture, uv); #endif #ifdef SCREEN_UV_USED @@ -365,7 +371,7 @@ void main() { #endif if (use_default_normal) { - normal.xy = texture2D(normal_texture, uv_interp).xy * 2.0 - 1.0; + normal.xy = texture2D(normal_texture, uv).xy * 2.0 - 1.0; normal.z = sqrt(1.0 - dot(normal.xy, normal.xy)); normal_used = true; } else { @@ -480,7 +486,7 @@ FRAGMENT_SHADER_CODE #ifdef USE_RGBA_SHADOWS -#define SHADOW_DEPTH(m_tex, m_uv) dot(texture2D((m_tex), (m_uv)), vec4(1.0 / (256.0 * 256.0 * 256.0), 1.0 / (256.0 * 256.0), 1.0 / 256.0, 1)) +#define SHADOW_DEPTH(m_tex, m_uv) dot(texture2D((m_tex), (m_uv)), vec4(1.0 / (256.0 * 256.0 * 256.0), 1.0 / (256.0 * 256.0), 1.0 / 256.0, 1.0)) #else diff --git a/drivers/gles2/shaders/canvas_shadow.glsl b/drivers/gles2/shaders/canvas_shadow.glsl index d39212826e..01b2c59325 100644 --- a/drivers/gles2/shaders/canvas_shadow.glsl +++ b/drivers/gles2/shaders/canvas_shadow.glsl @@ -48,7 +48,7 @@ void main() { #ifdef USE_RGBA_SHADOWS highp vec4 comp = fract(depth * vec4(256.0 * 256.0 * 256.0, 256.0 * 256.0, 256.0, 1.0)); - comp -= comp.xxyz * vec4(0, 1.0 / 256.0, 1.0 / 256.0, 1.0 / 256.0); + comp -= comp.xxyz * vec4(0.0, 1.0 / 256.0, 1.0 / 256.0, 1.0 / 256.0); gl_FragColor = comp; #else diff --git a/drivers/gles2/shaders/scene.glsl b/drivers/gles2/shaders/scene.glsl index faf88cf33d..f90c6a6651 100644 --- a/drivers/gles2/shaders/scene.glsl +++ b/drivers/gles2/shaders/scene.glsl @@ -1598,14 +1598,14 @@ FRAGMENT_SHADER_CODE #ifdef USE_LIGHTMAP_CAPTURE { vec3 cone_dirs[12] = vec3[]( - vec3(0, 0, 1), - vec3(0.866025, 0, 0.5), + vec3(0.0, 0.0, 1.0), + vec3(0.866025, 0.0, 0.5), vec3(0.267617, 0.823639, 0.5), vec3(-0.700629, 0.509037, 0.5), vec3(-0.700629, -0.509037, 0.5), vec3(0.267617, -0.823639, 0.5), - vec3(0, 0, -1), - vec3(0.866025, 0, -0.5), + vec3(0.0, 0.0, -1.0), + vec3(0.866025, 0.0, -0.5), vec3(0.267617, 0.823639, -0.5), vec3(-0.700629, 0.509037, -0.5), vec3(-0.700629, -0.509037, -0.5), diff --git a/drivers/gles3/shaders/canvas.glsl b/drivers/gles3/shaders/canvas.glsl index 88368516c1..6c1806a657 100644 --- a/drivers/gles3/shaders/canvas.glsl +++ b/drivers/gles3/shaders/canvas.glsl @@ -495,7 +495,7 @@ FRAGMENT_SHADER_CODE #endif } #ifdef DEBUG_ENCODED_32 - highp float enc32 = dot(color, highp vec4(1.0 / (256.0 * 256.0 * 256.0), 1.0 / (256.0 * 256.0), 1.0 / 256.0, 1)); + highp float enc32 = dot(color, highp vec4(1.0 / (256.0 * 256.0 * 256.0), 1.0 / (256.0 * 256.0), 1.0 / 256.0, 1.0)); color = vec4(vec3(enc32), 1.0); #endif @@ -589,7 +589,7 @@ FRAGMENT_SHADER_CODE #ifdef USE_RGBA_SHADOWS -#define SHADOW_DEPTH(m_tex, m_uv) dot(texture((m_tex), (m_uv)), vec4(1.0 / (256.0 * 256.0 * 256.0), 1.0 / (256.0 * 256.0), 1.0 / 256.0, 1)) +#define SHADOW_DEPTH(m_tex, m_uv) dot(texture((m_tex), (m_uv)), vec4(1.0 / (256.0 * 256.0 * 256.0), 1.0 / (256.0 * 256.0), 1.0 / 256.0, 1.0)) #else diff --git a/drivers/gles3/shaders/canvas_shadow.glsl b/drivers/gles3/shaders/canvas_shadow.glsl index 68d0713385..13fff7f4d1 100644 --- a/drivers/gles3/shaders/canvas_shadow.glsl +++ b/drivers/gles3/shaders/canvas_shadow.glsl @@ -36,7 +36,7 @@ void main() { #ifdef USE_RGBA_SHADOWS highp vec4 comp = fract(depth * vec4(256.0 * 256.0 * 256.0, 256.0 * 256.0, 256.0, 1.0)); - comp -= comp.xxyz * vec4(0, 1.0 / 256.0, 1.0 / 256.0, 1.0 / 256.0); + comp -= comp.xxyz * vec4(0.0, 1.0 / 256.0, 1.0 / 256.0, 1.0 / 256.0); distance_buf = comp; #else diff --git a/drivers/gles3/shaders/scene.glsl b/drivers/gles3/shaders/scene.glsl index 125faa4eac..3b06b08dec 100644 --- a/drivers/gles3/shaders/scene.glsl +++ b/drivers/gles3/shaders/scene.glsl @@ -400,7 +400,7 @@ void main() { texelFetch(skeleton_texture, tex_ofs, 0), texelFetch(skeleton_texture, tex_ofs + ivec2(0, 1), 0), texelFetch(skeleton_texture, tex_ofs + ivec2(0, 2), 0), - vec4(0.0,0.0,0.0,1.0)) * + vec4(0.0, 0.0, 0.0, 1.0)) * bone_weights.x; tex_ofs = ivec2(bone_indicesi.y % 256, (bone_indicesi.y / 256) * 3); @@ -409,7 +409,7 @@ void main() { texelFetch(skeleton_texture, tex_ofs, 0), texelFetch(skeleton_texture, tex_ofs + ivec2(0, 1), 0), texelFetch(skeleton_texture, tex_ofs + ivec2(0, 2), 0), - vec4(0.0,0.0,0.0,1.0)) * + vec4(0.0, 0.0, 0.0, 1.0)) * bone_weights.y; tex_ofs = ivec2(bone_indicesi.z % 256, (bone_indicesi.z / 256) * 3); @@ -418,7 +418,7 @@ void main() { texelFetch(skeleton_texture, tex_ofs, 0), texelFetch(skeleton_texture, tex_ofs + ivec2(0, 1), 0), texelFetch(skeleton_texture, tex_ofs + ivec2(0, 2), 0), - vec4(0.0,0.0,0.0,1.0)) * + vec4(0.0, 0.0, 0.0, 1.0)) * bone_weights.z; tex_ofs = ivec2(bone_indicesi.w % 256, (bone_indicesi.w / 256) * 3); @@ -427,7 +427,7 @@ void main() { texelFetch(skeleton_texture, tex_ofs, 0), texelFetch(skeleton_texture, tex_ofs + ivec2(0, 1), 0), texelFetch(skeleton_texture, tex_ofs + ivec2(0, 2), 0), - vec4(0.0,0.0,0.0,1.0)) * + vec4(0.0, 0.0, 0.0, 1.0)) * bone_weights.w; world_matrix = transpose(m) * world_matrix; @@ -1521,8 +1521,8 @@ void gi_probe_compute(mediump sampler3D probe, mat4 probe_xform, vec3 bounds, ve #define MAX_CONE_DIRS 6 vec3 cone_dirs[MAX_CONE_DIRS] = vec3[]( - vec3(0, 0, 1), - vec3(0.866025, 0, 0.5), + vec3(0.0, 0.0, 1.0), + vec3(0.866025, 0.0, 0.5), vec3(0.267617, 0.823639, 0.5), vec3(-0.700629, 0.509037, 0.5), vec3(-0.700629, -0.509037, 0.5), @@ -1536,10 +1536,10 @@ void gi_probe_compute(mediump sampler3D probe, mat4 probe_xform, vec3 bounds, ve #define MAX_CONE_DIRS 4 vec3 cone_dirs[MAX_CONE_DIRS] = vec3[]( - vec3(0.707107, 0, 0.707107), - vec3(0, 0.707107, 0.707107), - vec3(-0.707107, 0, 0.707107), - vec3(0, -0.707107, 0.707107)); + vec3(0.707107, 0.0, 0.707107), + vec3(0.0, 0.707107, 0.707107), + vec3(-0.707107, 0.0, 0.707107), + vec3(0.0, -0.707107, 0.707107)); float cone_weights[MAX_CONE_DIRS] = float[](0.25, 0.25, 0.25, 0.25); float cone_angle_tan = 0.98269; @@ -1575,7 +1575,7 @@ void gi_probes_compute(vec3 pos, vec3 normal, float roughness, inout vec3 out_sp vec3 ref_vec = normalize(reflect(normalize(pos), normal)); //find arbitrary tangent and bitangent, then build a matrix - vec3 v0 = abs(normal.z) < 0.999 ? vec3(0, 0, 1) : vec3(0, 1, 0); + vec3 v0 = abs(normal.z) < 0.999 ? vec3(0.0, 0.0, 1.0) : vec3(0.0, 1.0, 0.0); vec3 tangent = normalize(cross(v0, normal)); vec3 bitangent = normalize(cross(tangent, normal)); mat3 normal_mat = mat3(tangent, bitangent, normal); @@ -1963,14 +1963,14 @@ FRAGMENT_SHADER_CODE #ifdef USE_LIGHTMAP_CAPTURE { vec3 cone_dirs[12] = vec3[]( - vec3(0, 0, 1), - vec3(0.866025, 0, 0.5), + vec3(0.0, 0.0, 1.0), + vec3(0.866025, 0.0, 0.5), vec3(0.267617, 0.823639, 0.5), vec3(-0.700629, 0.509037, 0.5), vec3(-0.700629, -0.509037, 0.5), vec3(0.267617, -0.823639, 0.5), - vec3(0, 0, -1), - vec3(0.866025, 0, -0.5), + vec3(0.0, 0.0, -1.0), + vec3(0.866025, 0.0, -0.5), vec3(0.267617, 0.823639, -0.5), vec3(-0.700629, 0.509037, -0.5), vec3(-0.700629, -0.509037, -0.5), diff --git a/drivers/gles3/shaders/screen_space_reflection.glsl b/drivers/gles3/shaders/screen_space_reflection.glsl index 86546319a0..39f1ea6155 100644 --- a/drivers/gles3/shaders/screen_space_reflection.glsl +++ b/drivers/gles3/shaders/screen_space_reflection.glsl @@ -77,7 +77,7 @@ void main() { return; } //ray_dir = normalize(view_dir - normal * dot(normal,view_dir) * 2.0); - //ray_dir = normalize(vec3(1, 1, -1)); + //ray_dir = normalize(vec3(1.0, 1.0, -1.0)); //////////////// diff --git a/editor/import/resource_importer_texture.cpp b/editor/import/resource_importer_texture.cpp index a3bc9d0ec9..112e39cb4a 100644 --- a/editor/import/resource_importer_texture.cpp +++ b/editor/import/resource_importer_texture.cpp @@ -219,7 +219,7 @@ void ResourceImporterTexture::get_import_options(List<ImportOption> *r_options, r_options->push_back(ImportOption(PropertyInfo(Variant::REAL, "svg/scale", PROPERTY_HINT_RANGE, "0.001,100,0.001"), 1.0)); } -void ResourceImporterTexture::_save_stex(const Ref<Image> &p_image, const String &p_to_path, int p_compress_mode, float p_lossy_quality, Image::CompressMode p_vram_compression, bool p_mipmaps, int p_texture_flags, bool p_streamable, bool p_detect_3d, bool p_detect_srgb, bool p_force_rgbe, bool p_detect_normal, bool p_force_normal) { +void ResourceImporterTexture::_save_stex(const Ref<Image> &p_image, const String &p_to_path, int p_compress_mode, float p_lossy_quality, Image::CompressMode p_vram_compression, bool p_mipmaps, int p_texture_flags, bool p_streamable, bool p_detect_3d, bool p_detect_srgb, bool p_force_rgbe, bool p_detect_normal, bool p_force_normal, bool p_force_po2_for_compressed) { FileAccess *f = FileAccess::open(p_to_path, FileAccess::WRITE); f->store_8('G'); @@ -227,8 +227,21 @@ void ResourceImporterTexture::_save_stex(const Ref<Image> &p_image, const String f->store_8('S'); f->store_8('T'); //godot streamable texture - f->store_32(p_image->get_width()); - f->store_32(p_image->get_height()); + bool resize_to_po2 = false; + + if (p_compress_mode == COMPRESS_VIDEO_RAM && p_force_po2_for_compressed && (p_mipmaps || p_texture_flags & Texture::FLAG_REPEAT)) { + resize_to_po2 = true; + f->store_16(next_power_of_2(p_image->get_width())); + f->store_16(p_image->get_width()); + f->store_16(next_power_of_2(p_image->get_height())); + f->store_16(p_image->get_height()); + f->store_16(0); + } else { + f->store_16(p_image->get_width()); + f->store_16(0); + f->store_16(p_image->get_height()); + f->store_16(0); + } f->store_32(p_texture_flags); uint32_t format = 0; @@ -310,6 +323,9 @@ void ResourceImporterTexture::_save_stex(const Ref<Image> &p_image, const String case COMPRESS_VIDEO_RAM: { Ref<Image> image = p_image->duplicate(); + if (resize_to_po2) { + image->resize_to_po2(); + } if (p_mipmaps) { image->generate_mipmaps(p_force_normal); } @@ -478,25 +494,25 @@ Error ResourceImporterTexture::import(const String &p_source_file, const String } if (can_bptc || can_s3tc) { - _save_stex(image, p_save_path + ".s3tc.stex", compress_mode, lossy, can_bptc ? Image::COMPRESS_BPTC : Image::COMPRESS_S3TC, mipmaps, tex_flags, stream, detect_3d, detect_srgb, force_rgbe, detect_normal, force_normal); + _save_stex(image, p_save_path + ".s3tc.stex", compress_mode, lossy, can_bptc ? Image::COMPRESS_BPTC : Image::COMPRESS_S3TC, mipmaps, tex_flags, stream, detect_3d, detect_srgb, force_rgbe, detect_normal, force_normal, false); r_platform_variants->push_back("s3tc"); ok_on_pc = true; } if (ProjectSettings::get_singleton()->get("rendering/vram_compression/import_etc2")) { - _save_stex(image, p_save_path + ".etc2.stex", compress_mode, lossy, Image::COMPRESS_ETC2, mipmaps, tex_flags, stream, detect_3d, detect_srgb, force_rgbe, detect_normal, force_normal); + _save_stex(image, p_save_path + ".etc2.stex", compress_mode, lossy, Image::COMPRESS_ETC2, mipmaps, tex_flags, stream, detect_3d, detect_srgb, force_rgbe, detect_normal, force_normal, false); r_platform_variants->push_back("etc2"); } if (ProjectSettings::get_singleton()->get("rendering/vram_compression/import_etc")) { - _save_stex(image, p_save_path + ".etc.stex", compress_mode, lossy, Image::COMPRESS_ETC, mipmaps, tex_flags, stream, detect_3d, detect_srgb, force_rgbe, detect_normal, force_normal); + _save_stex(image, p_save_path + ".etc.stex", compress_mode, lossy, Image::COMPRESS_ETC, mipmaps, tex_flags, stream, detect_3d, detect_srgb, force_rgbe, detect_normal, force_normal, true); r_platform_variants->push_back("etc"); } if (ProjectSettings::get_singleton()->get("rendering/vram_compression/import_pvrtc")) { - _save_stex(image, p_save_path + ".pvrtc.stex", compress_mode, lossy, Image::COMPRESS_PVRTC4, mipmaps, tex_flags, stream, detect_3d, detect_srgb, force_rgbe, detect_normal, force_normal); + _save_stex(image, p_save_path + ".pvrtc.stex", compress_mode, lossy, Image::COMPRESS_PVRTC4, mipmaps, tex_flags, stream, detect_3d, detect_srgb, force_rgbe, detect_normal, force_normal, true); r_platform_variants->push_back("pvrtc"); } @@ -505,7 +521,7 @@ Error ResourceImporterTexture::import(const String &p_source_file, const String } } else { //import normally - _save_stex(image, p_save_path + ".stex", compress_mode, lossy, Image::COMPRESS_S3TC /*this is ignored */, mipmaps, tex_flags, stream, detect_3d, detect_srgb, force_rgbe, detect_normal, force_normal); + _save_stex(image, p_save_path + ".stex", compress_mode, lossy, Image::COMPRESS_S3TC /*this is ignored */, mipmaps, tex_flags, stream, detect_3d, detect_srgb, force_rgbe, detect_normal, force_normal, false); } return OK; diff --git a/editor/import/resource_importer_texture.h b/editor/import/resource_importer_texture.h index 408af1edcf..49a8b52b04 100644 --- a/editor/import/resource_importer_texture.h +++ b/editor/import/resource_importer_texture.h @@ -83,7 +83,7 @@ public: virtual void get_import_options(List<ImportOption> *r_options, int p_preset = 0) const; virtual bool get_option_visibility(const String &p_option, const Map<StringName, Variant> &p_options) const; - void _save_stex(const Ref<Image> &p_image, const String &p_to_path, int p_compress_mode, float p_lossy_quality, Image::CompressMode p_vram_compression, bool p_mipmaps, int p_texture_flags, bool p_streamable, bool p_detect_3d, bool p_detect_srgb, bool p_force_rgbe, bool p_detect_normal, bool p_force_normal); + void _save_stex(const Ref<Image> &p_image, const String &p_to_path, int p_compress_mode, float p_lossy_quality, Image::CompressMode p_vram_compression, bool p_mipmaps, int p_texture_flags, bool p_streamable, bool p_detect_3d, bool p_detect_srgb, bool p_force_rgbe, bool p_detect_normal, bool p_force_normal, bool p_force_po2_for_compressed); virtual Error import(const String &p_source_file, const String &p_save_path, const Map<StringName, Variant> &p_options, List<String> *r_platform_variants, List<String> *r_gen_files = NULL); diff --git a/methods.py b/methods.py index d8e90a8da5..2d4dc4921f 100644 --- a/methods.py +++ b/methods.py @@ -667,3 +667,6 @@ def get_compiler_version(env): return match.group().split('.') else: return None + +def use_gcc(env): + return 'gcc' in os.path.basename(env["CC"]) diff --git a/platform/javascript/audio_driver_javascript.cpp b/platform/javascript/audio_driver_javascript.cpp index 16fdc267f3..11104007e2 100644 --- a/platform/javascript/audio_driver_javascript.cpp +++ b/platform/javascript/audio_driver_javascript.cpp @@ -209,7 +209,7 @@ Error AudioDriverJavaScript::capture_start() { } function gotMediaInputError(e) { - console.log(e); + out(e); } if (navigator.mediaDevices.getUserMedia) { diff --git a/platform/javascript/detect.py b/platform/javascript/detect.py index c3f3946ee0..3cc79097f8 100644 --- a/platform/javascript/detect.py +++ b/platform/javascript/detect.py @@ -129,10 +129,6 @@ def configure(env): # us since we don't know requirements at compile-time. env.Append(LINKFLAGS=['-s', 'ALLOW_MEMORY_GROWTH=1']) - # Since we use both memory growth and MEMFS preloading, - # this avoids unnecessary copying on start-up. - env.Append(LINKFLAGS=['--no-heap-copy']) - # This setting just makes WebGL 2 APIs available, it does NOT disable WebGL 1. env.Append(LINKFLAGS=['-s', 'USE_WEBGL2=1']) diff --git a/platform/javascript/engine.js b/platform/javascript/engine.js index 91458eb4c3..860d6707ff 100644 --- a/platform/javascript/engine.js +++ b/platform/javascript/engine.js @@ -199,7 +199,8 @@ } LIBS.FS.mkdirTree(dir); } - LIBS.FS.createDataFile('/', file.path, new Uint8Array(file.buffer), true, true, true); + // With memory growth, canOwn should be false. + LIBS.FS.createDataFile(file.path, null, new Uint8Array(file.buffer), true, true, false); }, this); preloadedFiles = null; diff --git a/platform/javascript/http_request.js b/platform/javascript/http_request.js index 7acd32d8bf..66dacfc3d4 100644 --- a/platform/javascript/http_request.js +++ b/platform/javascript/http_request.js @@ -82,7 +82,7 @@ var GodotHTTPRequest = { godot_xhr_send_string: function(xhrId, strPtr) { if (!strPtr) { - console.warn("Failed to send string per XHR: null pointer"); + err("Failed to send string per XHR: null pointer"); return; } GodotHTTPRequest.requests[xhrId].send(UTF8ToString(strPtr)); @@ -90,11 +90,11 @@ var GodotHTTPRequest = { godot_xhr_send_data: function(xhrId, ptr, len) { if (!ptr) { - console.warn("Failed to send data per XHR: null pointer"); + err("Failed to send data per XHR: null pointer"); return; } if (len < 0) { - console.warn("Failed to send data per XHR: buffer length less than 0"); + err("Failed to send data per XHR: buffer length less than 0"); return; } GodotHTTPRequest.requests[xhrId].send(HEAPU8.subarray(ptr, ptr + len)); diff --git a/platform/javascript/javascript_eval.cpp b/platform/javascript/javascript_eval.cpp index bb43e2d46b..dd3eba74e4 100644 --- a/platform/javascript/javascript_eval.cpp +++ b/platform/javascript/javascript_eval.cpp @@ -69,7 +69,7 @@ Variant JavaScript::eval(const String &p_code, bool p_use_global_exec_context) { eval_ret = eval(UTF8ToString(CODE)); } } catch (e) { - console.warn(e); + err(e); eval_ret = null; } @@ -97,7 +97,7 @@ Variant JavaScript::eval(const String &p_code, bool p_use_global_exec_context) { if (array_ptr!==0) { _free(array_ptr) } - console.warn(e); + err(e); // fall through } break; diff --git a/platform/javascript/os_javascript.cpp b/platform/javascript/os_javascript.cpp index e820d07a2a..594c0a46cc 100644 --- a/platform/javascript/os_javascript.cpp +++ b/platform/javascript/os_javascript.cpp @@ -986,8 +986,8 @@ bool OS_JavaScript::main_loop_iterate() { if (sync_wait_time < 0) { /* clang-format off */ EM_ASM( - FS.syncfs(function(err) { - if (err) { console.warn('Failed to save IDB file system: ' + err.message); } + FS.syncfs(function(error) { + if (error) { err('Failed to save IDB file system: ' + error.message); } }); ); /* clang-format on */ diff --git a/platform/windows/os_windows.cpp b/platform/windows/os_windows.cpp index 9f15e7aad7..6e31f5b21d 100644 --- a/platform/windows/os_windows.cpp +++ b/platform/windows/os_windows.cpp @@ -1395,6 +1395,8 @@ Error OS_Windows::initialize(const VideoMode &p_desired, int p_video_driver, int SetThreadPriority(GetCurrentThread(), THREAD_PRIORITY_TIME_CRITICAL); } + update_real_mouse_position(); + return OK; } @@ -1596,6 +1598,19 @@ Point2 OS_Windows::get_mouse_position() const { return Point2(old_x, old_y); } +void OS_Windows::update_real_mouse_position() { + + POINT mouse_pos; + if (GetCursorPos(&mouse_pos) && ScreenToClient(hWnd, &mouse_pos)) { + if (mouse_pos.x > 0 && mouse_pos.y > 0 && mouse_pos.x <= video_mode.width && mouse_pos.y <= video_mode.height) { + old_x = mouse_pos.x; + old_y = mouse_pos.y; + old_invalid = false; + input->set_mouse_position(Point2i(mouse_pos.x, mouse_pos.y)); + } + } +} + int OS_Windows::get_mouse_button_state() const { return last_button_state; @@ -1738,6 +1753,7 @@ void OS_Windows::set_window_position(const Point2 &p_position) { } last_pos = p_position; + update_real_mouse_position(); } Size2 OS_Windows::get_window_size() const { diff --git a/platform/windows/os_windows.h b/platform/windows/os_windows.h index e27dbbe530..6c257016ec 100644 --- a/platform/windows/os_windows.h +++ b/platform/windows/os_windows.h @@ -197,6 +197,7 @@ public: virtual void warp_mouse_position(const Point2 &p_to); virtual Point2 get_mouse_position() const; + void update_real_mouse_position(); virtual int get_mouse_button_state() const; virtual void set_window_title(const String &p_title); diff --git a/platform/x11/detect.py b/platform/x11/detect.py index 16760f9407..1355ae542d 100644 --- a/platform/x11/detect.py +++ b/platform/x11/detect.py @@ -2,7 +2,7 @@ import os import platform import sys from compat import decode_utf8 -from methods import get_compiler_version +from methods import get_compiler_version, use_gcc def is_active(): return True @@ -162,10 +162,11 @@ def configure(env): env.Append(LINKFLAGS=['-pipe']) # Check for gcc version >= 6 before adding -no-pie - version = get_compiler_version(env) - if version != None and version[0] > '6': - env.Append(CCFLAGS=['-fpie']) - env.Append(LINKFLAGS=['-no-pie']) + if use_gcc(env): + version = get_compiler_version(env) + if version != None and version[0] >= '6': + env.Append(CCFLAGS=['-fpie']) + env.Append(LINKFLAGS=['-no-pie']) ## Dependencies diff --git a/scene/resources/material.cpp b/scene/resources/material.cpp index 9a52e9a6bb..78bc9bd15b 100644 --- a/scene/resources/material.cpp +++ b/scene/resources/material.cpp @@ -607,17 +607,17 @@ void SpatialMaterial::_update_shader() { code += "\tMODELVIEW_MATRIX = INV_CAMERA_MATRIX * mat4(CAMERA_MATRIX[0],CAMERA_MATRIX[1],CAMERA_MATRIX[2],WORLD_MATRIX[3]);\n"; if (flags[FLAG_BILLBOARD_KEEP_SCALE]) { - code += "\tMODELVIEW_MATRIX = MODELVIEW_MATRIX * mat4(vec4(length(WORLD_MATRIX[0].xyz),0,0,0),vec4(0,length(WORLD_MATRIX[1].xyz),0,0),vec4(0,0,length(WORLD_MATRIX[2].xyz),0),vec4(0,0,0,1));\n"; + code += "\tMODELVIEW_MATRIX = MODELVIEW_MATRIX * mat4(vec4(length(WORLD_MATRIX[0].xyz), 0.0, 0.0, 0.0),vec4(0.0, length(WORLD_MATRIX[1].xyz), 0.0, 0.0),vec4(0.0, 0.0, length(WORLD_MATRIX[2].xyz), 0.0),vec4(0.0, 0.0, 0.0, 1.0));\n"; } } break; case BILLBOARD_FIXED_Y: { - code += "\tMODELVIEW_MATRIX = INV_CAMERA_MATRIX * mat4(CAMERA_MATRIX[0],WORLD_MATRIX[1],vec4(normalize(cross(CAMERA_MATRIX[0].xyz,WORLD_MATRIX[1].xyz)),0.0),WORLD_MATRIX[3]);\n"; + code += "\tMODELVIEW_MATRIX = INV_CAMERA_MATRIX * mat4(CAMERA_MATRIX[0],WORLD_MATRIX[1],vec4(normalize(cross(CAMERA_MATRIX[0].xyz,WORLD_MATRIX[1].xyz)), 0.0),WORLD_MATRIX[3]);\n"; if (flags[FLAG_BILLBOARD_KEEP_SCALE]) { - code += "\tMODELVIEW_MATRIX = MODELVIEW_MATRIX * mat4(vec4(length(WORLD_MATRIX[0].xyz),0,0,0),vec4(0,1,0,0),vec4(0,0,length(WORLD_MATRIX[2].xyz),0),vec4(0,0,0,1));\n"; + code += "\tMODELVIEW_MATRIX = MODELVIEW_MATRIX * mat4(vec4(length(WORLD_MATRIX[0].xyz), 0.0, 0.0, 0.0),vec4(0.0, 1.0, 0.0, 0.0),vec4(0.0, 0.0, length(WORLD_MATRIX[2].xyz), 0.0), vec4(0.0, 0.0, 0.0, 1.0));\n"; } else { - code += "\tMODELVIEW_MATRIX = MODELVIEW_MATRIX * mat4(vec4(1,0,0,0),vec4(0,1.0/length(WORLD_MATRIX[1].xyz),0,0),vec4(0,0,1,0),vec4(0,0,0,1));\n"; + code += "\tMODELVIEW_MATRIX = MODELVIEW_MATRIX * mat4(vec4(1.0, 0.0, 0.0, 0.0),vec4(0.0, 1.0/length(WORLD_MATRIX[1].xyz), 0.0, 0.0), vec4(0.0, 0.0, 1.0, 0.0),vec4(0.0, 0.0, 0.0 ,1.0));\n"; } } break; case BILLBOARD_PARTICLES: { @@ -625,7 +625,7 @@ void SpatialMaterial::_update_shader() { //make billboard code += "\tmat4 mat_world = mat4(normalize(CAMERA_MATRIX[0])*length(WORLD_MATRIX[0]),normalize(CAMERA_MATRIX[1])*length(WORLD_MATRIX[0]),normalize(CAMERA_MATRIX[2])*length(WORLD_MATRIX[2]),WORLD_MATRIX[3]);\n"; //rotate by rotation - code += "\tmat_world = mat_world * mat4( vec4(cos(INSTANCE_CUSTOM.x),-sin(INSTANCE_CUSTOM.x),0.0,0.0), vec4(sin(INSTANCE_CUSTOM.x),cos(INSTANCE_CUSTOM.x),0.0,0.0),vec4(0.0,0.0,1.0,0.0),vec4(0.0,0.0,0.0,1.0));\n"; + code += "\tmat_world = mat_world * mat4( vec4(cos(INSTANCE_CUSTOM.x),-sin(INSTANCE_CUSTOM.x), 0.0, 0.0), vec4(sin(INSTANCE_CUSTOM.x), cos(INSTANCE_CUSTOM.x), 0.0, 0.0),vec4(0.0, 0.0, 1.0, 0.0),vec4(0.0, 0.0, 0.0, 1.0));\n"; //set modelview code += "\tMODELVIEW_MATRIX = INV_CAMERA_MATRIX * mat_world;\n"; diff --git a/scene/resources/particles_material.cpp b/scene/resources/particles_material.cpp index 59fde7787e..ef67e6ea80 100644 --- a/scene/resources/particles_material.cpp +++ b/scene/resources/particles_material.cpp @@ -311,8 +311,8 @@ void ParticlesMaterial::_update_shader() { //initiate velocity spread in 3D code += " float angle1_rad = rand_from_seed_m1_p1(alt_seed) * spread_rad;\n"; code += " float angle2_rad = rand_from_seed_m1_p1(alt_seed) * spread_rad * (1.0 - flatness);\n"; - code += " vec3 direction_xz = vec3(sin(angle1_rad), 0, cos(angle1_rad));\n"; - code += " vec3 direction_yz = vec3(0, sin(angle2_rad), cos(angle2_rad));\n"; + code += " vec3 direction_xz = vec3(sin(angle1_rad), 0.0, cos(angle1_rad));\n"; + code += " vec3 direction_yz = vec3(0.0, sin(angle2_rad), cos(angle2_rad));\n"; code += " direction_yz.z = direction_yz.z / max(0.0001,sqrt(abs(direction_yz.z))); // better uniform distribution\n"; code += " vec3 direction = vec3(direction_xz.x * direction_yz.z, direction_yz.y, direction_xz.z * direction_yz.z);\n"; code += " direction = normalize(direction);\n"; @@ -347,7 +347,7 @@ void ParticlesMaterial::_update_shader() { code += " VELOCITY.xy = rotm * VELOCITY.xy;\n"; } else { code += " vec3 normal = texelFetch(emission_texture_normal, emission_tex_ofs, 0).xyz;\n"; - code += " vec3 v0 = abs(normal.z) < 0.999 ? vec3(0.0, 0.0, 1.0) : vec3(0, 1.0, 0.0);\n"; + code += " vec3 v0 = abs(normal.z) < 0.999 ? vec3(0.0, 0.0, 1.0) : vec3(0.0, 1.0, 0.0);\n"; code += " vec3 tangent = normalize(cross(v0, normal));\n"; code += " vec3 bitangent = normalize(cross(tangent, normal));\n"; code += " VELOCITY = mat3(tangent, bitangent, normal) * VELOCITY;\n"; diff --git a/scene/resources/texture.cpp b/scene/resources/texture.cpp index e8cc160acf..430db59e44 100644 --- a/scene/resources/texture.cpp +++ b/scene/resources/texture.cpp @@ -481,7 +481,7 @@ Image::Format StreamTexture::get_format() const { return format; } -Error StreamTexture::_load_data(const String &p_path, int &tw, int &th, int &flags, Ref<Image> &image, int p_size_limit) { +Error StreamTexture::_load_data(const String &p_path, int &tw, int &th, int &tw_custom, int &th_custom, int &flags, Ref<Image> &image, int p_size_limit) { alpha_cache.unref(); @@ -497,8 +497,11 @@ Error StreamTexture::_load_data(const String &p_path, int &tw, int &th, int &fla ERR_FAIL_COND_V(header[0] != 'G' || header[1] != 'D' || header[2] != 'S' || header[3] != 'T', ERR_FILE_CORRUPT); } - tw = f->get_32(); - th = f->get_32(); + tw = f->get_16(); + tw_custom = f->get_16(); + th = f->get_16(); + th_custom = f->get_16(); + flags = f->get_32(); //texture flags! uint32_t df = f->get_32(); //data format @@ -705,18 +708,26 @@ Error StreamTexture::_load_data(const String &p_path, int &tw, int &th, int &fla Error StreamTexture::load(const String &p_path) { - int lw, lh, lflags; + int lw, lh, lwc, lhc, lflags; Ref<Image> image; image.instance(); - Error err = _load_data(p_path, lw, lh, lflags, image); + Error err = _load_data(p_path, lw, lh, lwc, lhc, lflags, image); if (err) return err; + if (get_path() == String()) { + //temporarily set path if no path set for resource, helps find errors + VisualServer::get_singleton()->texture_set_path(texture, p_path); + } VS::get_singleton()->texture_allocate(texture, image->get_width(), image->get_height(), 0, image->get_format(), VS::TEXTURE_TYPE_2D, lflags); VS::get_singleton()->texture_set_data(texture, image); + if (lwc || lhc) { + VS::get_singleton()->texture_set_size_override(texture, lwc, lhc, 0); + } else { + } - w = lw; - h = lh; + w = lwc ? lwc : lw; + h = lhc ? lhc : lh; flags = lflags; path_to_file = p_path; format = image->get_format(); @@ -784,6 +795,7 @@ bool StreamTexture::is_pixel_opaque(int p_x, int p_y) const { decom->decompress(); img = decom; } + alpha_cache.instance(); alpha_cache->create_from_image_alpha(img); } diff --git a/scene/resources/texture.h b/scene/resources/texture.h index dc3a3e0c74..21d3782897 100644 --- a/scene/resources/texture.h +++ b/scene/resources/texture.h @@ -187,7 +187,7 @@ public: }; private: - Error _load_data(const String &p_path, int &tw, int &th, int &flags, Ref<Image> &image, int p_size_limit = 0); + Error _load_data(const String &p_path, int &tw, int &th, int &tw_custom, int &th_custom, int &flags, Ref<Image> &image, int p_size_limit = 0); String path_to_file; RID texture; Image::Format format; diff --git a/thirdparty/libvpx/rtcd/vpx_dsp_rtcd_x86.h b/thirdparty/libvpx/rtcd/vpx_dsp_rtcd_x86.h index 82574e096c..c2a68330ac 100644 --- a/thirdparty/libvpx/rtcd/vpx_dsp_rtcd_x86.h +++ b/thirdparty/libvpx/rtcd/vpx_dsp_rtcd_x86.h @@ -22,7 +22,6 @@ extern "C" { void vpx_convolve8_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); void vpx_convolve8_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); void vpx_convolve8_ssse3(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); -void vpx_convolve8_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); RTCD_EXTERN void (*vpx_convolve8)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); void vpx_convolve8_avg_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); @@ -43,13 +42,11 @@ RTCD_EXTERN void (*vpx_convolve8_avg_vert)(const uint8_t *src, ptrdiff_t src_str void vpx_convolve8_horiz_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); void vpx_convolve8_horiz_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); void vpx_convolve8_horiz_ssse3(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); -void vpx_convolve8_horiz_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); RTCD_EXTERN void (*vpx_convolve8_horiz)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); void vpx_convolve8_vert_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); void vpx_convolve8_vert_sse2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); void vpx_convolve8_vert_ssse3(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); -void vpx_convolve8_vert_avx2(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); RTCD_EXTERN void (*vpx_convolve8_vert)(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); void vpx_convolve_avg_c(const uint8_t *src, ptrdiff_t src_stride, uint8_t *dst, ptrdiff_t dst_stride, const int16_t *filter_x, int x_step_q4, const int16_t *filter_y, int y_step_q4, int w, int h); @@ -343,12 +340,10 @@ RTCD_EXTERN void (*vpx_lpf_horizontal_8_dual)(uint8_t *s, int pitch, const uint8 void vpx_lpf_horizontal_edge_16_c(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh); void vpx_lpf_horizontal_edge_16_sse2(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh); -void vpx_lpf_horizontal_edge_16_avx2(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh); RTCD_EXTERN void (*vpx_lpf_horizontal_edge_16)(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh); void vpx_lpf_horizontal_edge_8_c(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh); void vpx_lpf_horizontal_edge_8_sse2(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh); -void vpx_lpf_horizontal_edge_8_avx2(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh); RTCD_EXTERN void (*vpx_lpf_horizontal_edge_8)(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh); void vpx_lpf_vertical_16_c(uint8_t *s, int pitch, const uint8_t *blimit, const uint8_t *limit, const uint8_t *thresh); @@ -440,7 +435,6 @@ static void setup_rtcd_internal(void) vpx_convolve8 = vpx_convolve8_c; if (flags & HAS_SSE2) vpx_convolve8 = vpx_convolve8_sse2; if (flags & HAS_SSSE3) vpx_convolve8 = vpx_convolve8_ssse3; - if (flags & HAS_AVX2) vpx_convolve8 = vpx_convolve8_avx2; vpx_convolve8_avg = vpx_convolve8_avg_c; if (flags & HAS_SSE2) vpx_convolve8_avg = vpx_convolve8_avg_sse2; if (flags & HAS_SSSE3) vpx_convolve8_avg = vpx_convolve8_avg_ssse3; @@ -453,11 +447,9 @@ static void setup_rtcd_internal(void) vpx_convolve8_horiz = vpx_convolve8_horiz_c; if (flags & HAS_SSE2) vpx_convolve8_horiz = vpx_convolve8_horiz_sse2; if (flags & HAS_SSSE3) vpx_convolve8_horiz = vpx_convolve8_horiz_ssse3; - if (flags & HAS_AVX2) vpx_convolve8_horiz = vpx_convolve8_horiz_avx2; vpx_convolve8_vert = vpx_convolve8_vert_c; if (flags & HAS_SSE2) vpx_convolve8_vert = vpx_convolve8_vert_sse2; if (flags & HAS_SSSE3) vpx_convolve8_vert = vpx_convolve8_vert_ssse3; - if (flags & HAS_AVX2) vpx_convolve8_vert = vpx_convolve8_vert_avx2; vpx_convolve_avg = vpx_convolve_avg_c; if (flags & HAS_SSE2) vpx_convolve_avg = vpx_convolve_avg_sse2; vpx_convolve_copy = vpx_convolve_copy_c; @@ -570,10 +562,8 @@ static void setup_rtcd_internal(void) if (flags & HAS_SSE2) vpx_lpf_horizontal_8_dual = vpx_lpf_horizontal_8_dual_sse2; vpx_lpf_horizontal_edge_16 = vpx_lpf_horizontal_edge_16_c; if (flags & HAS_SSE2) vpx_lpf_horizontal_edge_16 = vpx_lpf_horizontal_edge_16_sse2; - if (flags & HAS_AVX2) vpx_lpf_horizontal_edge_16 = vpx_lpf_horizontal_edge_16_avx2; vpx_lpf_horizontal_edge_8 = vpx_lpf_horizontal_edge_8_c; if (flags & HAS_SSE2) vpx_lpf_horizontal_edge_8 = vpx_lpf_horizontal_edge_8_sse2; - if (flags & HAS_AVX2) vpx_lpf_horizontal_edge_8 = vpx_lpf_horizontal_edge_8_avx2; vpx_lpf_vertical_16 = vpx_lpf_vertical_16_c; if (flags & HAS_SSE2) vpx_lpf_vertical_16 = vpx_lpf_vertical_16_sse2; vpx_lpf_vertical_16_dual = vpx_lpf_vertical_16_dual_c; diff --git a/thirdparty/libvpx/vpx_config.h b/thirdparty/libvpx/vpx_config.h index 6caec50c81..e8e91fa6ef 100644 --- a/thirdparty/libvpx/vpx_config.h +++ b/thirdparty/libvpx/vpx_config.h @@ -29,7 +29,7 @@ #define HAVE_MMX 1 #define HAVE_SSE2 1 #define HAVE_SSSE3 1 - #define HAVE_AVX2 1 + #define HAVE_AVX2 0 #elif defined(__x86_64) || defined(__x86_64__) || defined(__amd64) || defined(_M_X64) #define ARCH_X86 0 #define ARCH_X86_64 1 @@ -41,7 +41,7 @@ #define HAVE_MMX 1 #define HAVE_SSE2 1 #define HAVE_SSSE3 1 - #define HAVE_AVX2 1 + #define HAVE_AVX2 0 #elif defined(__arm__) || defined(__TARGET_ARCH_ARM) || defined(_M_ARM) #define ARCH_X86 0 #define ARCH_X86_64 0 |