Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | r600g: not fatal if we can't get tiling info from kernel | Dave Airlie | 2010-10-23 | 1 | -1/+1 |
| | |||||
* | winsys/xlib: rename xm->xlib | Brian Paul | 2010-10-21 | 1 | -147/+147 |
| | | | | Move away from the old Mesa-oriented names. | ||||
* | winsys/xlib: fix up allocation/dealloction of XImage | Brian Paul | 2010-10-21 | 1 | -1/+8 |
| | | | | Fixes a crash upon exit when using remote display. | ||||
* | winsys/xlib: use Bool type for shm field | Brian Paul | 2010-10-21 | 1 | -5/+5 |
| | |||||
* | winsys/xlib: formatting fixes | Brian Paul | 2010-10-21 | 1 | -10/+9 |
| | |||||
* | r600g: retrieve tiling info from kernel for shared buffers. | Dave Airlie | 2010-10-18 | 3 | -2/+42 |
| | | | | we need to know if the back is tiled so we can blit from it properly. | ||||
* | r600g: get tiling info from kernel | Dave Airlie | 2010-10-18 | 3 | -0/+68 |
| | |||||
* | wrapper: Add a way to dewrap a pipe screen without destroying it | Jakob Bornecrantz | 2010-10-15 | 2 | -0/+22 |
| | |||||
* | wrapper: Fix spelling | Jakob Bornecrantz | 2010-10-15 | 2 | -2/+2 |
| | |||||
* | r600/drm: fix segfaults in winsys create failure path | Keith Whitwell | 2010-10-15 | 1 | -3/+9 |
| | | | | | | | Would try to destroy radeon->cman, radeon->kman both which were still NULL. Signed-off-by: Dave Airlie <[email protected]> | ||||
* | r600g: Silence uninitialized variable warning. | Vinson Lee | 2010-10-08 | 1 | -1/+2 |
| | |||||
* | r600g: Silence uninitialized variable warning. | Vinson Lee | 2010-10-08 | 1 | -1/+2 |
| | |||||
* | r600g: Silence uninitialized variable warning. | Vinson Lee | 2010-10-08 | 1 | -1/+2 |
| | |||||
* | r600g: Remove unnecessary header. | Vinson Lee | 2010-10-08 | 1 | -1/+0 |
| | |||||
* | r600g: fix dirty state handling | Jerome Glisse | 2010-10-06 | 3 | -7/+7 |
| | | | | | | Avoid having object ending up in dead list of dirty object. Signed-off-by: Jerome Glisse <[email protected]> | ||||
* | r600g: avoid segfault due to unintialized list pointer | Jerome Glisse | 2010-10-06 | 2 | -7/+9 |
| | | | | Signed-off-by: Jerome Glisse <[email protected]> | ||||
* | r600g: userspace fence to avoid kernel call for testing bo busy status | Jerome Glisse | 2010-10-05 | 5 | -47/+99 |
| | | | | Signed-off-by: Jerome Glisse <[email protected]> | ||||
* | r600g: simplify block relocation | Jerome Glisse | 2010-10-05 | 2 | -9/+7 |
| | | | | | | | Since flush rework there could be only one relocation per register in a block. Signed-off-by: Jerome Glisse <[email protected]> | ||||
* | r600g: use dirty list to track dirty blocks | Bas Nieuwenhuizen | 2010-10-05 | 3 | -8/+31 |
| | | | | Got a speed up by tracking the dirty blocks in a seperate list instead of looping through all blocks. This version should work with block that get their dirty state disabled again and I added a dirty check during the flush as some blocks were already dirty. | ||||
* | r600g: improve bo flushing | Jerome Glisse | 2010-10-05 | 3 | -822/+822 |
| | | | | | | | | | Flush read cache before writting register. Track flushing inside of a same cs and avoid reflushing same bo if not necessary. Allmost properly force flush if bo rendered too and then use as a texture in same cs (missing pipeline flush dunno if it's needed or not). Signed-off-by: Jerome Glisse <[email protected]> | ||||
* | r600g: store reloc information in bo structure | Jerome Glisse | 2010-10-05 | 2 | -23/+16 |
| | | | | | | | Allow fast lookup of relocation information & id which was a CPU time consumming operation. Signed-off-by: Jerome Glisse <[email protected]> | ||||
* | r600g: avoid unneeded bo wait | Dave Airlie | 2010-10-05 | 1 | -1/+5 |
| | | | | | | if we know the bo has gone not busy, no need to add another bo wait thanks to Andre (taiu) on irc for pointing this out. | ||||
* | r600g: drop use_mem_constant. | Dave Airlie | 2010-10-05 | 3 | -3/+0 |
| | | | | since we plan on using dx10 constant buffers everywhere. | ||||
* | r600g: drop mman allocator | Dave Airlie | 2010-10-05 | 3 | -8/+1 |
| | | | | we don't use this since constant buffers are now being used on all gpus. | ||||
* | r600g: add bo busy backoff. | Dave Airlie | 2010-10-05 | 2 | -0/+15 |
| | | | | | | When we go to do a lot of bos in one draw like constant bufs we need to avoid bouncing off the busy ioctl, this mitigates by backing off on busy bos for a short amount of times. | ||||
* | r600g: add bo fenced list. | Dave Airlie | 2010-10-05 | 3 | -0/+43 |
| | | | | | this just keeps a list of bos submitted together, and uses them to decide bo busy state for the whole group. | ||||
* | r600g: Fix SCons build. | Vinson Lee | 2010-10-04 | 1 | -1/+1 |
| | |||||
* | r600g: remove dead label & fix indentation | Jerome Glisse | 2010-10-04 | 1 | -11/+9 |
| | | | | Signed-off-by: Jerome Glisse <[email protected]> | ||||
* | r600g: rename radeon_ws_bo to r600_bo | Jerome Glisse | 2010-10-04 | 2 | -1/+1 |
| | | | | Signed-off-by: Jerome Glisse <[email protected]> | ||||
* | r600g: use r600_bo for relocation argument, simplify code | Jerome Glisse | 2010-10-04 | 4 | -19/+29 |
| | | | | Signed-off-by: Jerome Glisse <[email protected]> | ||||
* | r600g: allow r600_bo to be a sub allocation of a big bo | Jerome Glisse | 2010-10-04 | 2 | -4/+4 |
| | | | | | | | Add bo offset everywhere needed if r600_bo is ever a sub bo of a bigger bo. Signed-off-by: Jerome Glisse <[email protected]> | ||||
* | r600g: rename radeon_ws_bo to r600_bo | Jerome Glisse | 2010-10-04 | 4 | -44/+44 |
| | | | | Signed-off-by: Jerome Glisse <[email protected]> | ||||
* | r600g: TODO domain management | Dave Airlie | 2010-10-04 | 1 | -2/+2 |
| | | | | | no wonder it was slow, the code is deliberately forcing stuff into GTT, we used to have domain management but it seems to have disappeared. | ||||
* | r600g: fix wwarning in bo_map function | Dave Airlie | 2010-10-04 | 1 | -0/+1 |
| | |||||
* | r600g: break out of search for reloc bo after finding it. | Dave Airlie | 2010-10-04 | 1 | -0/+1 |
| | | | | this function was taking quite a lot of pointless CPU. | ||||
* | r600g: setup basic loop consts on r600 + evergreen. | Dave Airlie | 2010-10-01 | 2 | -0/+42 |
| | | | | this sets up a single loop constant like r600c does. | ||||
* | r600g: flush SH cache on constant change on evergreen | Dave Airlie | 2010-10-01 | 1 | -2/+2 |
| | |||||
* | r600g: fix evergreen draw-buffers | Dave Airlie | 2010-10-01 | 1 | -1/+1 |
| | | | | just a typo in the register headers. | ||||
* | r600g: add cb flushing for extra buffers + depth buffer on r600/evergreen | Dave Airlie | 2010-10-01 | 2 | -3/+37 |
| | |||||
* | r600g: sync vertex/texture cache on resources on evergreen | Dave Airlie | 2010-10-01 | 1 | -2/+2 |
| | | | | | this gets rid of lots of the instability on evergreen, which isn't surprising since it really broken not to flush caches. | ||||
* | r600g: add reloc for evergreen color attrib | Dave Airlie | 2010-10-01 | 1 | -12/+12 |
| | | | | we'll need this for color tiling on evergreen. | ||||
* | r600g: add winsys support for CTL constants. | Dave Airlie | 2010-10-01 | 3 | -0/+22 |
| | | | | | These need to be emitted, we also need them to do proper vtx start, instead of abusing index offset. | ||||
* | r600g: use Elements macro instead of manual sizeofs | Dave Airlie | 2010-10-01 | 2 | -10/+12 |
| | |||||
* | r600g: keep a mapping around for each bo | John Doe | 2010-09-30 | 2 | -51/+54 |
| | | | | | | Save a lot of call into the kernel and thus improve performances. Signed-off-by: Jerome Glisse <[email protected]> | ||||
* | r600g: don't double count dirty block | John Doe | 2010-09-30 | 3 | -26/+42 |
| | | | | | | | This avoid to overcount the number of dwords we need and thus avoid maximazation of cs buffer use. Signed-off-by: Jerome Glisse <[email protected] | ||||
* | evergreeng: avoid overlapping border color btw VS & PS | Jerome Glisse | 2010-09-30 | 1 | -2/+2 |
| | | | | Signed-off-by: Jerome Glisse <[email protected]> | ||||
* | r600g: use constant buffer instead of register for constant | Jerome Glisse | 2010-09-30 | 2 | -29/+10 |
| | | | | Signed-off-by: Jerome Glisse <[email protected]> | ||||
* | gallium/winsys: remove duplicated include | Nicolas Kaiser | 2010-09-30 | 1 | -1/+0 |
| | | | | | | Remove duplicated include. Signed-off-by: Brian Paul <[email protected]> | ||||
* | r600g: clean up some code from move to new paths. | Dave Airlie | 2010-09-30 | 2 | -9/+2 |
| | | | | mainly remove 2 suffix from function names | ||||
* | r600g: Fix SCons build. | Vinson Lee | 2010-09-29 | 1 | -5/+5 |
| |