Age | Commit message (Collapse) | Author | |
---|---|---|---|
2008-03-06 | i965:fix segfault issue when clearing the window which | Xiang, Haihao | |
is created with mode GLUT_SINGLE|GLUT_RGB|GLUT_DEPTH. This issue is introduced by 20b8bff49cba3e8246e29004c5ff38f231d589ff | |||
2008-03-05 | [intel] Add a driconf option to cache freed buffer objects for reuse. | Eric Anholt | |
This is defaulted off as it has potentially large memory costs for a modest performance gain. Ideally we will improve DRM performance to the point where this optimization is not worth the memory cost in any case, or find some middle ground in caching only limited numbers of certain buffers. For now, this provides a modest 4% improvement in openarena on GM965 and 10% in openarena on GM945. | |||
2008-03-04 | r300: replace some hard coded mask by define in stencil area | Christoph Brill | |
2008-03-04 | r300: Fix some issues with masks in stencil buffer area | Christoph Brill | |
2008-03-03 | [dri2] Add tail pointer to reemitDrawableInfo callback. | Kristian Høgsberg | |
When the DRI doesn't parse the event buffer for a while, the X server may overwrite data that the driver didn't get a chance to look at. The reemitDrawableInfo callback requests that the X server reemit all info for the specified drawable. To make use of this, the drive needs to know the new tail pointer so it know where to start reading from. | |||
2008-03-03 | [dri2] Optimize event parsing to skip obsolete events. | Kristian Høgsberg | |
This also fixes the problem where the X server does multiple resizes before the DRI driver gets the events. The obsolete buffer attach events then reference already destroyed buffer objects. | |||
2008-03-03 | [intel] Silence unused variable warning when compiling for i965. | Kristian Høgsberg | |
2008-03-02 | nouveau: compilation fixes | Patrice Mandin | |
2008-03-02 | r300: Corrected a bug with the SUB instruction. | Oliver McFadden | |
2008-03-02 | r300: Corrected a bug with the MAD instruction. | Oliver McFadden | |
The PVS_VECTOR_OPCODE macro should be modified to support macro instructions, too. | |||
2008-03-01 | r300: Added the PVS_SRC_OPERAND documentation from AMD. | Oliver McFadden | |
2008-03-01 | r300: Added the PVS_OP_DST_OPERAND documentation from AMD. | Oliver McFadden | |
2008-03-01 | r300: Added a TODO comment for registers missing from AMD's documentation. | Oliver McFadden | |
2008-03-01 | r300: Moved the vertex program shift/mask defines into the appropriate file. | Oliver McFadden | |
2008-03-01 | r300: Indented the vertex program code with longer lines. | Oliver McFadden | |
2008-03-01 | r300: Moved the PREFER_DP4 define near the position invariant function. | Oliver McFadden | |
2008-03-01 | r300: Added a TODO comment for the MAD opcodes. | Oliver McFadden | |
2008-03-01 | r300: Use the VE_ADD hardware opcode for the SUB opcode. | Oliver McFadden | |
2008-03-01 | r300: Use the VE_MULTIPLY hardware opcode for the MUL opcode. | Oliver McFadden | |
2008-03-01 | r300: Cleaned up the XPD opcode temporary register usage. | Oliver McFadden | |
2008-03-01 | r300: Cleaned up extra white space. | Oliver McFadden | |
2008-03-01 | r300: Prefer to use the VE_ADD for simple MOV style opcodes. | Oliver McFadden | |
The VE_MULTIPLY_ADD has further restrictions on reading temporary memory which may complicate things. See AMD's documentation. | |||
2008-03-01 | r300: Removed the (undocumented) MAD_2 opcode. | Oliver McFadden | |
This opcode is likely a mistake from reverse engineering. MAD_2 isn't included in AMD's documentation, and my testing reviles there is no problem using the documented MAD opcode. | |||
2008-03-01 | r300: Cleaned up the MAD/MAD_2 opcode selection. | Oliver McFadden | |
2008-03-01 | r300: Renamed some misleading macro arguments. | Oliver McFadden | |
2008-03-01 | r300: Cleaned up the vertex program macros. | Oliver McFadden | |
2008-03-01 | r300: Removed duplicate component selection defines. | Oliver McFadden | |
2008-03-01 | r300: Removed duplicate source register class defines. | Oliver McFadden | |
2008-03-01 | r300: Renamed the vertex program source register macro. | Oliver McFadden | |
2008-03-01 | r300: Removed the (obsolete) special source register macros. | Oliver McFadden | |
2008-03-01 | r300: Cleaned up the special vertex program source register macros. | Oliver McFadden | |
2008-03-01 | r300: Added the vertex program swizzle (aka selection) defines. | Oliver McFadden | |
2008-03-01 | r300: Converted to the new src/dest register defines. | Oliver McFadden | |
2008-03-01 | r300: Removed an obsolete comment from the vertex program header file. | Oliver McFadden | |
2008-03-01 | r300: Converted to the new Math Engine defines. | Oliver McFadden | |
2008-03-01 | r300: Added the Math Engine opcode macro. | Oliver McFadden | |
2008-03-01 | r300: Renamed the Vector Engine opcode macro. | Oliver McFadden | |
2008-03-01 | r300: Converted to the new Vector Engine defines. | Oliver McFadden | |
2008-03-01 | r300: Removed the duplicate dest register defines. | Oliver McFadden | |
2008-03-01 | r300: Removed the duplicate "easy" vertex program macros. | Oliver McFadden | |
2008-03-01 | r300: Added the vertex program src/dest register defines. | Oliver McFadden | |
2008-03-01 | r300: Added the Vector Engine and Math Engine defines from AMD's documentation. | Oliver McFadden | |
2008-03-01 | r300: Moved the vertex and fragment program macros into the appropriate files. | Oliver McFadden | |
2008-03-01 | state.depth.range alpha value should be 1, not 0 (bug #14733) | Roland Scheidegger | |
2008-02-29 | Use __DRIextension mechanism providing loader functionality to the driver. | Kristian Høgsberg | |
Instead of passing in a fixed struct, the loader now passes in a list of __DRIextension structs, to advertise the functionality it can provide to the driver. Each extension is individually versioned and can be extended or phased out as the interface develops. | |||
2008-02-29 | Reduce the versioning madness required to create a DRI2 screen. | Kristian Høgsberg | |
Right now the DRI2 screen constructor takes 3 different versions: DRI, DDX and DRM. This is mostly useless, though: DRI: The DRI driver doesn't actually care about the DRI protocol, it only talks to the loader, which in turn speaks DRI protocol. Thus, the DRI protocol version is of not interest to the DRI driver, but it needs to know what functionality the loader provides. At this point that's reflected in the __DRIinterfaceMethods struct and the internal_version integer. DDX: The DDX version number is essentially used to track extensions to the SAREA. With DRI2 the SAREA consists of a number of versioned, self-describing blocks, so the DDX version is no longer interesting. DRM: We have the fd, lets just ask the kernel ourselves. | |||
2008-02-29 | intel: Set the lock flag early to avoid deadlock. | Kristian Høgsberg | |
Another regression from the intel_context.c merge. | |||
2008-02-29 | i965: use _Current pointer instead of Current pointer. | Xiang, Haihao | |
fix double free issue(bug#14710). It also corrects glsl/bitmap demo behavior. | |||
2008-02-28 | [965] Fix conditional sequencing to allow the frame_buffer_texobj case. | Eric Anholt | |
Previously the frame_buffer_texobj case would have been bound as a normal texture, at best resulting in no surface data being associated with it. | |||
2008-02-28 | [intel] Bug #14575: Unmap buffers when deleting/dataing as required. | Eric Anholt | |
Otherwise, we would assertion error when doing the final unreference of the buffer with an outstanding map (catching the memory leak). |