| Commit message (Expand) | Author | Age | Files | Lines |
* | nv50,nvc0: optimize coherent buffer checking at draw time | Samuel Pitoiset | 2016-01-03 | 6 | -68/+82 |
* | nv50,nvc0: make sure there's pushbuf space and that we ref the bo early | Ilia Mirkin | 2016-01-01 | 4 | -6/+5 |
* | nvc0: Set winding order regardless of domain. | Kenneth Graunke | 2015-12-30 | 1 | -2/+4 |
* | nvc0: add ARB_shader_draw_parameters support | Ilia Mirkin | 2015-12-30 | 13 | -15/+73 |
* | gallium: add PIPE_CAP_DRAW_PARAMETERS | Ilia Mirkin | 2015-12-30 | 3 | -0/+3 |
* | nv50/ir: attempt to do more constant folding on mad -> add conversion | Ilia Mirkin | 2015-12-30 | 1 | -11/+10 |
* | nv50/ir: float(s32 & 0xff) = float(u8), not s8 | Ilia Mirkin | 2015-12-29 | 1 | -0/+3 |
* | nvc0: don't forget to reset VTX_TMP bufctx slot after blit completion | Ilia Mirkin | 2015-12-27 | 1 | -0/+2 |
* | nv50,nvc0: add a note when converting vertex elements using CPU | Ilia Mirkin | 2015-12-27 | 2 | -0/+6 |
* | nvc0: remove use of deprecated sw class identifier | Ben Skeggs | 2015-12-22 | 1 | -3/+5 |
* | nv50: fix g98+ vdec class allocation | Ben Skeggs | 2015-12-22 | 1 | -6/+51 |
* | nouveau: remove use of deprecated nouveau_device_wrap() | Ben Skeggs | 2015-12-22 | 2 | -0/+5 |
* | nouveau: fix screen creation failure paths | Ben Skeggs | 2015-12-22 | 4 | -19/+23 |
* | nouveau: return nouveau_screen from hw-specific creation functions | Ben Skeggs | 2015-12-22 | 4 | -9/+9 |
* | nouveau: remove use of deprecated nouveau_device::drm_version | Ben Skeggs | 2015-12-22 | 7 | -12/+15 |
* | nouveau: remove use of deprecated nouveau_device::fd | Ben Skeggs | 2015-12-22 | 2 | -0/+2 |
* | nv50: free memory allocated by the prog which reads MP perf counters | Samuel Pitoiset | 2015-12-16 | 1 | -0/+5 |
* | nv50,nvc0: free memory allocated by performance metrics | Samuel Pitoiset | 2015-12-16 | 6 | -4/+22 |
* | nvc0: free memory allocated by the prog which reads MP perf counters | Samuel Pitoiset | 2015-12-16 | 1 | -0/+1 |
* | nvc0: fix metric-achieved_occupancy calculation on Kepler | Samuel Pitoiset | 2015-12-16 | 1 | -1/+4 |
* | nvc0: remove old comment related to metric calculations | Samuel Pitoiset | 2015-12-15 | 1 | -11/+0 |
* | nvc0: check return value of nvc0_program_validate() | Samuel Pitoiset | 2015-12-14 | 1 | -4/+1 |
* | nv50: check return value of nouveau_object_new() | Samuel Pitoiset | 2015-12-14 | 1 | -4/+4 |
* | nv50,nvc0: make use of unreachable() when invalid texture target happens | Samuel Pitoiset | 2015-12-14 | 2 | -5/+2 |
* | gk104/ir: simplify and fool-proof texbar algorithm | Ilia Mirkin | 2015-12-12 | 2 | -83/+56 |
* | nv50/ir: combine sequences of conversions | Ilia Mirkin | 2015-12-12 | 1 | -0/+43 |
* | nv50/ir: manually optimize multiplication expansion logic | Ilia Mirkin | 2015-12-12 | 1 | -6/+25 |
* | nv50/ir: fix imul emission in the presence of an immediate | Ilia Mirkin | 2015-12-12 | 1 | -4/+7 |
* | nv50/ir: teach post-ra immediate folding into mad about integers | Ilia Mirkin | 2015-12-12 | 1 | -3/+31 |
* | nv50/ir: add short imad support | Ilia Mirkin | 2015-12-12 | 3 | -22/+40 |
* | nv50/ir: can't have predication and immediates | Ilia Mirkin | 2015-12-12 | 1 | -0/+3 |
* | nv50/ir: fix texture grad for cubemaps | Ilia Mirkin | 2015-12-12 | 4 | -7/+6 |
* | nv50/ir: fix assumption that prog->maxGPR is in 32-bit reg units | Ilia Mirkin | 2015-12-12 | 2 | -4/+21 |
* | nv50,nvc0: fix use-after-free when vertex buffers are unbound | Patrick Rudolph | 2015-12-09 | 2 | -7/+6 |
* | nv50/ir: fix cutoff for using r63 vs r127 when replacing zero | Ilia Mirkin | 2015-12-08 | 1 | -1/+2 |
* | nv50/ir: prefer to color mad def and src2 with the same color | Ilia Mirkin | 2015-12-08 | 1 | -0/+14 |
* | nv50/ir: reduce degree limit on ops that can't encode large reg dests | Ilia Mirkin | 2015-12-08 | 1 | -3/+34 |
* | nv50/ir: only unspill once ahead of a group of instructions | Ilia Mirkin | 2015-12-08 | 1 | -5/+20 |
* | nv50/ir: check if the target supports the new offset before inlining | Ilia Mirkin | 2015-12-08 | 4 | -3/+25 |
* | nvc0/ir: fix up mul+add -> mad algebraic opt, enable for integers | Ilia Mirkin | 2015-12-07 | 3 | -12/+11 |
* | gk110/ir: fix imad sat/hi flag emission for immediate args | Ilia Mirkin | 2015-12-07 | 1 | -8/+3 |
* | gk104/ir: sampler doesn't matter for txf | Ilia Mirkin | 2015-12-07 | 1 | -1/+1 |
* | gk110/ir: fix imul hi emission with limm arg | Ilia Mirkin | 2015-12-07 | 1 | -2/+2 |
* | gallium/drivers: Sanitize NULL checks into canonical form | Edward O'Callaghan | 2015-12-06 | 2 | -3/+3 |
* | gallium/drivers: Trivial code-style cleanup | Edward O'Callaghan | 2015-12-06 | 2 | -2/+2 |
* | gallium/drivers/nouveau: Make use of ARRAY_SIZE macro | Edward O'Callaghan | 2015-12-06 | 14 | -22/+20 |
* | nv50/ir: fold shl + mul with immediates | Ilia Mirkin | 2015-12-05 | 1 | -0/+16 |
* | nv50/ir: propagate indirect loads into instructions | Ilia Mirkin | 2015-12-05 | 1 | -0/+52 |
* | nv50/ir: flip shl(add, imm) into add(shl, imm) | Ilia Mirkin | 2015-12-05 | 1 | -4/+34 |
* | nvc0: expose a group of performance metrics for SM30 (Kepler) | Samuel Pitoiset | 2015-12-05 | 2 | -2/+8 |