summaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2003-05-26Reverted forced eye coords hack in run_vertex_stage.embedded-2-branchdok6662-4/+4
Force eye coords in _mesa_update_tnl_spaces. Need to find out why eye coords aren't used where they are needed, e.g. chromium's title screen.
2003-05-26Applied patch from Ville Syrjälä <syrjala@sci.fi>, thanks.dok6662-6/+6
Fix tpitch when firstLevel > baseLevel. There was some inconsistency with dirty_images[0]. In some cases it referred to the actual level and to the hw level in others. I changed it so that it uses the actual level. I also changed the way mgaUploadSubImage() is called. Now it's only called for the levels that are actually used. This got rid of all those ugly "Image[n] = NULL" warnings.
2003-05-21Test program for GL_MESA_ycbcr_texture.dok6663-0/+320
2003-05-21GL_MESA_ycbcr_texture support by Ian Romanick <idr@us.ibm.com>.dok6664-2/+15
2003-05-20Ian fixed his fix (mip map level limit check).dok6661-22/+16
2003-05-20Sync with Ian's version and enhancements of my recent fixes.dok6665-53/+77
2003-05-20Applied patch from Ville Syrjälä <syrjala@sci.fi>, thanks!dok6662-8/+5
- Fix pitch setup for > 1024x1024 textures - G200 should support textures up to 2048x2048 too
2003-05-19Code cleanup.dok6661-10/+3
2003-05-19Reject mip maps smaller than 8x8.dok6663-29/+59
Changed bit order in TEXFILTER mip map setting. Use firstLevel/lastLevel calculation from radeon driver, horing tObj->MinLod and tObj->MaxLod. Set max mip map level from 12 to 11 for G400 as doc says mipmap count can be 0 to 10. Still not sure about that, maybe it's just 10? Where is the 12 from?
2003-05-18Allow zero sized card texture heap (using AGP textures only).dok6661-1/+1
2003-05-18Added missing drmUnmap() calls in mgaDestroyScreen().dok6661-3/+11
2003-05-17In _mesa_set_viewport():dok6661-5/+5
Call ctx->Driver.Viewport() after setting up ctx->Viewport._WindowMap, at least the Matrox driver uses it in mgaViewport(). This fixes the broken 2D graphics in Quake3 I was writing about a few days ago on the mailing list.
2003-05-13Ignore root only flag for now.dok6661-1/+1
2003-05-13Reenabled dstorg check in mga_verify_context(). Possible values aredok6661-3/+3
front offset, back offset and MGA_DSTORG_EXTENDED_CONTEXT. The latter one causes the new drm_mga_extended_context_regs_t info to be used.
2003-05-13Patched MGA drm module is now binary compatible with the original one.dok66611-190/+277
It's now possible to run either XFree86 DRI or DirectFBGL DRI using the same module! Compile drm modules with -D_HAVE_FULL_GL.
2003-05-13Updated _mesa_problem() is required for the new driver.dok6662-6/+13
2003-05-13Ignore...dok6666-0/+7
2003-05-13Re(im)ported Matrox driver from current DRI HEAD branch includingdok66648-2958/+7537
recent texmem merge. Minor problems with textures in quake2 are fixed now.
2003-05-13Fixed mgaInitScreenConfigs() to return both 16 and 32 bit configs.dok6661-43/+37
2003-05-12- Write back pSAREA to driverContext in clients' createScreen.dok6664-1/+12
- Free driverClientMsg allocated by drivers - Fixed mmesa->setup.depth_pitch
2003-05-11Removed one-time depth buffer allocation from the driver.dok66619-603/+415
Removed depth buffer related values from the global (non context/drawable) structures and added them to the drawable. Initialize all visual configurations (16 and 32 bit). Windowing system (dfb, miniglx) has to allocate the depth, too, now depending on the values from the visual config that is to be used. I'm now able to have two OpenGL windows (in one app for now) rendering concurrently (not multi threaded yet, but easy to do) both using their own depth buffer. I'm sorry, but I had to disable the radeon driver until it's updated to the recent changes. I also doubt that MiniGLX is currently working in this branch.
2003-05-09Read back AGP mode to report correct rate to user.dok6661-0/+9
2003-05-09Added glutExtensionSupported().dok6661-0/+42
2003-05-09Fix fog coord generation by updating this file to HEAD version.dok6661-6/+6
2003-05-07Removed accidently added files.dok66611-0/+0
2003-05-07Removed usleep(100) from MGAScreenInit().dok6661-8/+9
Added AGP mode debug message.
2003-05-06Fixed cleanly checked out builds.dok6661-2/+2
2003-05-06Moved module sources from src/kernel/radeon to src/kernel/drm.dok66655-79/+31306
Added patched mga drm module, version 3.2 now. Check for the 3.2 version.
2003-05-06This is the first checkin to the embedded-2-branch.dok66641-769/+649
Architectural changes --------------------- Moved global (and static) front/back buffer offset/pitch/cpp to the drawables for dynamic allocation of multiple buffers. Made driverContext.shared.fbOrigin the amount of video memory that won't be used by the driver for local textures etc. The windowing system will allocate front/back buffers below this origin. Removed front/back buffer offset, pitch, width, height and cpp from mgaScreen and mgaDMAInit. Added them to mga_context_regs_t. Unlock hardware in mgaUnbindContext() and radeonUnbindContext(). Removed nearly all MiniGLX dependencies from DRI Util, it can now be used without allocating any MiniGLXDisplayRec, GLXDrawable, XVisualInfo etc. Fixes to the Matrox driver -------------------------- Proper deinitialization (buffer unmapping etc). Fixed mip mapping by implementing mgaChooseTextureFormat() the supposed way and removing the texture conversion code. Otherwise the texture is marked to be incomplete because the formats don't match. Fixed multi texturing by fixing the texture unit swapping in mgaUpdateTextureState(). Fixed fogging. Fixed mgaClearColor(). Fixed register masking in mgaSetTexFilter(). Fixed mgaDDTexParameter() for texture objects that aren't currently bound to the TMU but to the GL context. Other changes ------------- Disabled src/drv/fb until it's working again. Some signed/unsigned and other warnings fixes (gcc-3.3 -Wall). Major TODOs ----------- Update radeon_dri and fb_dri for the architectural changes. Cleanup/update MiniGLX according to the changes (it already works again on mga). Regain backward compatibility in the mga drm module (already have a complete solution in mind). Cleanup DRI Util code.
2003-05-02Don't call glPolygonModeKeith Whitwell3-20/+1
2003-05-02Use correct condition for subsettingembedded-1-20030502Keith Whitwell1-1/+1
2003-05-02mprotect sarea when focus lostKeith Whitwell1-2/+4
2003-05-02mprotect sarea until focus gainedKeith Whitwell1-2/+7
2003-05-02Remove isosurf targetKeith Whitwell1-1/+0
2003-05-01Move the Doxygen configuration files into the usual places and integrate ā†µJose Fonseca5-15/+1082
with the build system.
2003-05-01Final fixes to the AGPGART and RADEONFB modules' documentation.Jose Fonseca6-55/+59
2003-05-01Use mprotect() to protect the sarea when not in focus.Keith Whitwell7-33/+63
2003-05-01Animate by defaultKeith Whitwell1-2/+2
2003-05-01Final documentation to the AGPGART module.Jose Fonseca8-180/+771
2003-04-30Final documentation for the radeon fb module.Jose Fonseca2-45/+302
2003-04-30Remove debug printfs, Use XCheckMaskEventKeith Whitwell1-2/+1
2003-04-30Remove debug printfsKeith Whitwell3-12/+0
2003-04-29More documentation.Jose Fonseca10-206/+2404
2003-04-29New filesKeith Whitwell2-0/+497
2003-04-29Add -autostart optionKeith Whitwell1-1/+9
2003-04-29Remove bogus assertKeith Whitwell1-2/+1
2003-04-29Minor fixesKeith Whitwell1-6/+4
2003-04-29Shortcircuit ioctl if no focusKeith Whitwell1-0/+3
2003-04-29Protect some swrast-specific code in MakeCurrentKeith Whitwell1-1/+3
2003-04-28Partial documentation of this file.Keith Whitwell1-33/+129