// OpenGL helpers // // Copyright (c) 2002-2005 Jan Wassenberg // // This program is free software; you can redistribute it and/or // modify it under the terms of the GNU General Public License as // published by the Free Software Foundation; either version 2 of the // License, or (at your option) any later version. // // This program is distributed in the hope that it will be useful, but // WITHOUT ANY WARRANTY; without even the implied warranty of // MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU // General Public License for more details. // // Contact info: // Jan.Wassenberg@stud.uni-karlsruhe.de // http://www.stud.uni-karlsruhe.de/~urkt/ #include "precompiled.h" #include #include #include #include "lib.h" #include "sdl.h" #include "ogl.h" #include "detect.h" #include "debug.h" #if MSC_VERSION #pragma comment(lib, "opengl32.lib") #pragma comment(lib, "glu32.lib") // glu32 is required - it is apparently pulled in by opengl32, // even though depends.exe marks it as demand-loaded. #endif //---------------------------------------------------------------------------- // extensions //---------------------------------------------------------------------------- // define extension function pointers extern "C" { #define FUNC(ret, name, params) ret (CALL_CONV *name) params; #define FUNC2(ret, nameARB, nameCore, version, params) ret (CALL_CONV *nameARB) params; #include "glext_funcs.h" #undef FUNC2 #undef FUNC } static const char* exts = NULL; static bool have_14, have_13, have_12; // return a C string of unspecified length containing a space-separated // list of all extensions the OpenGL implementation advertises. // (useful for crash logs). const char* oglExtList() { debug_assert(exts && "call oglInit before using this function"); return exts; } // paranoia: newer drivers may forget to advertise an extension // indicating support for something that has been folded into the core. // we therefore check for all extensions known to be offered by the // GL implementation present on the user's system; oglHaveExtension will // take this into account. // the app can therefore just ask for extensions and not worry about this. static bool isImplementedInCore(const char* ext) { #define MATCH(known_ext)\ if(!strcmp(ext, #known_ext))\ return true; if(have_14) { MATCH(GL_SGIS_generate_mipmap); MATCH(GL_NV_blend_square); MATCH(GL_ARB_depth_texture); MATCH(GL_ARB_shadow); MATCH(GL_EXT_fog_coord); MATCH(GL_EXT_multi_draw_arrays); MATCH(GL_ARB_point_parameters); MATCH(GL_EXT_secondary_color); MATCH(GL_EXT_blend_func_separate); MATCH(GL_EXT_stencil_wrap); MATCH(GL_ARB_texture_env_crossbar); MATCH(GL_EXT_texture_lod_bias); MATCH(GL_ARB_texture_mirrored_repeat); MATCH(GL_ARB_window_pos); } if(have_13) { MATCH(GL_ARB_texture_compression); MATCH(GL_ARB_texture_cube_map); MATCH(GL_ARB_multisample); MATCH(GL_ARB_multitexture); MATCH(GL_ARB_transpose_matrix); MATCH(GL_ARB_texture_env_add); MATCH(GL_ARB_texture_env_combine); MATCH(GL_ARB_texture_env_dot3); MATCH(GL_ARB_texture_border_clamp); } if(have_12) { MATCH(GL_EXT_texture3d); MATCH(GL_EXT_bgra); MATCH(GL_EXT_packed_pixels); MATCH(GL_EXT_rescale_normal); MATCH(GL_EXT_separate_specular_color); MATCH(GL_SGIS_texture_edge_clamp); MATCH(GL_SGIS_texture_lod); MATCH(GL_EXT_draw_range_elements); } #undef MATCH return false; } // check if the extension is supported by the OpenGL implementation. // takes subsequently added core support for some extensions into account. bool oglHaveExtension(const char* ext) { debug_assert(exts && "call oglInit before using this function"); if(isImplementedInCore(ext)) return true; const char *p = exts, *end; // make sure ext is valid & doesn't contain spaces if(!ext || ext == '\0' || strchr(ext, ' ')) return false; for(;;) { p = strstr(p, ext); if(!p) return false; // string not found - extension not supported end = p + strlen(ext); // end of current substring // make sure the substring found is an entire extension string, // i.e. it starts and ends with ' ' if((p == exts || p[-1] == ' ') && // valid start AND (*end == ' ' || *end == '\0')) // valid end return true; p = end; } } // check if the OpenGL implementation is at least at . // (format: "%d.%d" major minor) bool oglHaveVersion(const char* desired_version) { int desired_major, desired_minor; if(sscanf(desired_version, "%d.%d", &desired_major, &desired_minor) != 2) { debug_warn("oglHaveVersion: invalid version string"); return false; } int major, minor; const char* version = (const char*)glGetString(GL_VERSION); if(!version || sscanf(version, "%d.%d", &major, &minor) != 2) { debug_warn("oglHaveVersion: GL_VERSION invalid"); return false; } return (major > desired_major) || (major == desired_major && minor >= desired_minor); } // check if all given extension strings (passed as const char* parameters, // terminated by a 0 pointer) are supported by the OpenGL implementation, // as determined by oglHaveExtension. // returns 0 if all are present; otherwise, the first extension in the // list that's not supported (useful for reporting errors). // // note: dummy parameter is necessary to access parameter va_list. // // // rationale: this interface is more convenient than individual // oglHaveExtension calls and allows reporting which extension is missing. // // one disadvantage is that there is no way to indicate that either one // of 2 extensions would be acceptable, e.g. (ARB|EXT)_texture_env_dot3. // this is isn't so bad, since they wouldn't be named differently // if there weren't non-trivial changes between them. for that reason, // we refrain from equivalence checks (which would boil down to // string-matching known extensions to their equivalents). const char* oglHaveExtensions(int dummy, ...) { const char* ext; va_list args; va_start(args, dummy); for(;;) { ext = va_arg(args, const char*); // end of list reached; all were present => return 0. if(!ext) break; // not found => return name of missing extension. if(!oglHaveExtension(ext)) break; } va_end(args); return ext; } static void importExtensionFunctions() { #define FUNC(ret, name, params) *(void**)&name = SDL_GL_GetProcAddress(#name); #define FUNC2(ret, nameARB, nameCore, version, params) \ nameARB = NULL; \ if(oglHaveVersion(version)) \ *(void**)&nameARB = SDL_GL_GetProcAddress(#nameCore); \ if(!nameARB) /* use the ARB name if the driver lied about what version it supports */ \ *(void**)&nameARB = SDL_GL_GetProcAddress(#nameARB); #include "glext_funcs.h" #undef FUNC2 #undef FUNC // It should be safe to load the ARB function pointers even if the // extension isn't advertised, since we won't actually use them without // checking for the extension. } //---------------------------------------------------------------------------- static void dump_gl_error(GLenum err) { debug_printf("GL error: "); #define E(e) case e: debug_printf("%s\n", #e); break; switch (err) { E(GL_INVALID_ENUM) E(GL_INVALID_VALUE) E(GL_INVALID_OPERATION) E(GL_STACK_OVERFLOW) E(GL_STACK_UNDERFLOW) E(GL_OUT_OF_MEMORY) default:; } #undef E } #ifndef oglCheck // don't include this function if it has been defined (in ogl.h) as a no-op void oglCheck() { // glGetError may return multiple errors, so we poll it in a loop. // the debug_warn should only happen once (if this is set), though. bool error_enountered = false; for(;;) { GLenum err = glGetError(); if(err == GL_NO_ERROR) break; error_enountered = true; dump_gl_error(err); } if(error_enountered) debug_warn("oglCheck reports error(s)"); } #endif // ignore and reset the specified error (as returned by glGetError). // any other errors that have occurred are reported as oglCheck would. // // this is useful for suppressing annoying error messages, e.g. // "invalid enum" for GL_CLAMP_TO_EDGE even though we've already // warned the user that their OpenGL implementation is too old. void oglSquelchError(GLenum err_to_ignore) { // glGetError may return multiple errors, so we poll it in a loop. // the debug_warn should only happen once (if this is set), though. bool error_enountered = false; for(;;) { GLenum err = glGetError(); if(err == GL_NO_ERROR) break; if(err == err_to_ignore) continue; error_enountered = true; dump_gl_error(err); } if(error_enountered) debug_warn("oglSquelchError reports other error(s)"); } //---------------------------------------------------------------------------- // feature and limit detect //---------------------------------------------------------------------------- int ogl_max_tex_size = -1; // [pixels] int ogl_max_tex_units = -1; // limit on GL_TEXTUREn int ogl_max_VAR_elements = -1; // GF2: 64K; GF3: 1M int ogl_tex_compression_supported = -1; // S3TC / DXT{1,3,5} // gfx_card and gfx_drv_ver are unchanged on failure. int ogl_get_gfx_info() { const char* vendor = (const char*)glGetString(GL_VENDOR); const char* renderer = (const char*)glGetString(GL_RENDERER); const char* version = (const char*)glGetString(GL_VERSION); // can fail if OpenGL not yet initialized, // or if called between glBegin and glEnd. if(!vendor || !renderer || !version) return -1; strcpy_s(gfx_card, sizeof(gfx_card), vendor); // reduce string to "ATI" or "NVIDIA" if(!strcmp(gfx_card, "ATI Technologies Inc.")) gfx_card[3] = 0; if(!strcmp(gfx_card, "NVIDIA Corporation")) gfx_card[6] = 0; strcat_s(gfx_card, sizeof(gfx_card), renderer); // don't bother cutting off the crap at the end. // too risky, and too many different strings. snprintf(gfx_drv_ver, sizeof(gfx_drv_ver), "OpenGL %s", version); // add "OpenGL" to differentiate this from the real driver version // (returned by platform-specific detect routines). return 0; } static void CALL_CONV emulate_glCompressedTexImage2D(GLenum, GLint, GLenum, GLsizei, GLsizei, GLint, GLsizei, const GLvoid*); static void detectFeatures() { glGetIntegerv(GL_MAX_TEXTURE_SIZE, &ogl_max_tex_size); glGetIntegerv(GL_MAX_TEXTURE_UNITS, &ogl_max_tex_units); // make sure value remains -1 if not supported if(oglHaveExtension("GL_NV_vertex_array_range")) glGetIntegerv(GL_MAX_VERTEX_ARRAY_RANGE_ELEMENT_NV, &ogl_max_VAR_elements); ogl_tex_compression_supported = oglHaveExtensions(0, "GL_ARB_texture_compression", "GL_EXT_texture_compression_s3tc", 0) == 0; // TODO: GL_S3_s3tc? It uses different enumerants (GL_RGB_S3TC etc), so the // texture loading code would need to be changed; and it is not clear whether // it supports the full range of DXT1/3/5. (There seems to be no specification; // and many header files don't have GL_RGBA_DXT5_S3TC, suggesting that the // drivers don't all support that.) if(!ogl_tex_compression_supported) { // If there's no hardware support for compressed textures, do the // decompression in software (but first let the user know it's probably not // going to be very fast). DISPLAY_ERROR(L"Performance warning: your graphics card does not support compressed textures. The game will try to continue anyway, but may be slower than expected. Please try updating your graphics drivers; if that doesn't help, please try upgrading your hardware."); // TODO: i18n glCompressedTexImage2DARB = emulate_glCompressedTexImage2D; // Leave ogl_tex_compression_supported == 0, so that it indicates the presence // of hardware-supported texture compression. } } // call after each video mode change, since thereafter extension functions // may have changed [address]. void oglInit() { // cache extension list and versions for oglHave*. // note: this is less about performance (since the above are not // time-critical) than centralizing the 'OpenGL is ready' check. exts = (const char*)glGetString(GL_EXTENSIONS); if(!exts) { debug_warn("oglInit called before OpenGL is ready for use"); } have_12 = oglHaveVersion("1.2"); have_13 = oglHaveVersion("1.3"); have_14 = oglHaveVersion("1.4"); importExtensionFunctions(); detectFeatures(); } static void CALL_CONV emulate_glCompressedTexImage2D( GLenum target, GLint level, GLenum internalformat, GLsizei width, GLsizei height, GLint border, GLsizei imageSize, const GLvoid* data) { // Software emulation of compressed-texture support, for really old // cards/drivers that can't do it (but which do support everything else // we strictly require). They probably don't have enough VRAM for all the // textures, and are slow anyway, so it's not going to be a pleasant way // of playing; but at least it's better than nothing. GLenum base_fmt = (internalformat == GL_COMPRESSED_RGB_S3TC_DXT1_EXT)? GL_RGB : GL_RGBA; // TODO: handle small (<4x4) images correctly GLsizei blocks_w = (GLsizei)(round_up(width, 4) / 4); GLsizei blocks_h = (GLsizei)(round_up(height, 4) / 4); GLsizei blocks = blocks_w * blocks_h; GLsizei size = blocks * 16 * (base_fmt == GL_RGB ? 3 : 4); GLsizei pitch = size / (blocks_h*4); UNUSED2(pitch); void* rgb_data = malloc(size); bool dxt1 = (internalformat == GL_COMPRESSED_RGB_S3TC_DXT1_EXT || internalformat == GL_COMPRESSED_RGBA_S3TC_DXT1_EXT); bool dxt3 = (internalformat == GL_COMPRESSED_RGBA_S3TC_DXT3_EXT); bool dxt5 = (internalformat == GL_COMPRESSED_RGBA_S3TC_DXT5_EXT); debug_assert(dxt1 || dxt3 || dxt5); debug_assert(imageSize == blocks * (dxt1? 8 : 16)); // This code is inefficient, but I don't care: for(GLsizei block_y = 0; block_y < blocks_h; ++block_y) for(GLsizei block_x = 0; block_x < blocks_w; ++block_x) { int c0_a = 255, c1_a = 255, c2_a = 255, c3_a = 255; u8* alpha = NULL; u8 dxt5alpha[8]; if(!dxt1) { alpha = (u8*)data; data = (char*)data + 8; if(dxt5) { dxt5alpha[0] = alpha[0]; dxt5alpha[1] = alpha[1]; if(alpha[0] > alpha[1]) { dxt5alpha[2] = (6*alpha[0] + 1*alpha[1] + 3)/7; dxt5alpha[3] = (5*alpha[0] + 2*alpha[1] + 3)/7; dxt5alpha[4] = (4*alpha[0] + 3*alpha[1] + 3)/7; dxt5alpha[5] = (3*alpha[0] + 4*alpha[1] + 3)/7; dxt5alpha[6] = (2*alpha[0] + 5*alpha[1] + 3)/7; dxt5alpha[7] = (1*alpha[0] + 6*alpha[1] + 3)/7; } else { dxt5alpha[2] = (4*alpha[0] + 1*alpha[1] + 2)/5; dxt5alpha[3] = (3*alpha[0] + 2*alpha[1] + 2)/5; dxt5alpha[4] = (2*alpha[0] + 3*alpha[1] + 2)/5; dxt5alpha[5] = (1*alpha[0] + 4*alpha[1] + 2)/5; dxt5alpha[6] = 0; dxt5alpha[7] = 255; } } } u16 c0 = *(u16*)( (char*)data + 0 ); u16 c1 = *(u16*)( (char*)data + 2 ); u32 bits = *(u32*)( (char*)data + 4 ); data = (char*)data + 8; // Unpack 565, and copy high bits to low bits int c0_r = ((c0>>8)&0xF8) | ((c0>>13)&7); int c1_r = ((c1>>8)&0xF8) | ((c1>>13)&7); int c0_g = ((c0>>3)&0xFC) | ((c0>>9 )&3); int c1_g = ((c1>>3)&0xFC) | ((c1>>9 )&3); int c0_b = ((c0<<3)&0xF8) | ((c0>>2 )&7); int c1_b = ((c1<<3)&0xF8) | ((c1>>2 )&7); int c2_r, c2_g, c2_b; int c3_r, c3_g, c3_b; if(!dxt1 || c0 > c1) { c2_r = (c0_r*2+c1_r+1)/3; c2_g = (c0_g*2+c1_g+1)/3; c2_b = (c0_b*2+c1_b+1)/3; c3_r = (c0_r+2*c1_r+1)/3; c3_g = (c0_g+2*c1_g+1)/3; c3_b = (c0_b+2*c1_b+1)/3; } else { c2_r = (c0_r+c1_r)/2; c2_g = (c0_g+c1_g)/2; c2_b = (c0_b+c1_b)/2; c3_r = c3_g = c3_b = c3_a = 0; } if(base_fmt == GL_RGB) { int i = 0; for(int y = 0; y < 4; ++y) { u8* out = (u8*)rgb_data + ((block_y*4+y)*blocks_w*4 + block_x*4) * 3; for(int x = 0; x < 4; ++x, ++i) { switch((bits >> (2*i)) & 3) { case 0: *out++ = c0_r; *out++ = c0_g; *out++ = c0_b; break; case 1: *out++ = c1_r; *out++ = c1_g; *out++ = c1_b; break; case 2: *out++ = c2_r; *out++ = c2_g; *out++ = c2_b; break; case 3: *out++ = c3_r; *out++ = c3_g; *out++ = c3_b; break; } } } } else { int i = 0; for(int y = 0; y < 4; ++y) { u8* out = (u8*)rgb_data + ((block_y*4+y)*blocks_w*4 + block_x*4) * 4; for(int x = 0; x < 4; ++x, ++i) { int a = 0; // squelch bogus uninitialized warning switch((bits >> (2*i)) & 3) { case 0: *out++ = c0_r; *out++ = c0_g; *out++ = c0_b; a = c0_a; break; case 1: *out++ = c1_r; *out++ = c1_g; *out++ = c1_b; a = c1_a; break; case 2: *out++ = c2_r; *out++ = c2_g; *out++ = c2_b; a = c2_a; break; case 3: *out++ = c3_r; *out++ = c3_g; *out++ = c3_b; a = c3_a; break; } if(dxt3) { a = (int)((*(u64*)alpha >> (4*i)) & 0xF); a |= a<<4; // copy low bits to high bits } else if(dxt5) { a = dxt5alpha[(*(u64*)(alpha+2) >> (3*i)) & 0x7]; } *out++ = a; } } } } const GLint int_fmt = (base_fmt == GL_RGB)? GL_RGB8 : GL_RGBA8; glTexImage2D(target, level, int_fmt, width, height, border, base_fmt, GL_UNSIGNED_BYTE, rgb_data); free(rgb_data); }