| Commit message (Expand) | Author | Age | Files | Lines |
|---|
| * | nir/algebraic: Mark some value range analysis-based optimizations imprecise | Ian Romanick | 2019-09-04 | 1 | -9/+13 |
| * | nir/algrbraic: Don't optimize open-coded bitfield reverse when lowering is en... | Ian Romanick | 2019-09-04 | 1 | -1/+1 |
| * | nir/algebraic: add a few masking-before-unpack optimizations | Rhys Perry | 2019-08-16 | 1 | -1/+9 |
| * | nir/algebraic: Reassociate shift-by-constant of shift-by-constant | Ian Romanick | 2019-08-14 | 1 | -1/+25 |
| * | nir/algebraic: Reassociate add-and-shift to be shift-and-add | Ian Romanick | 2019-08-14 | 1 | -0/+5 |
| * | nir/algebraic: Use value range analysis to eliminate useless unary ops | Ian Romanick | 2019-08-05 | 1 | -2/+8 |
| * | nir/algebraic: Use value range analysis to convert fmin to fsat | Ian Romanick | 2019-08-05 | 1 | -1/+1 |
| * | nir/algebraic: Use value range analysis to eliminate tautological compares | Ian Romanick | 2019-08-05 | 1 | -21/+21 |
| * | nir/algebraic: Use value range analysis to eliminate tautological compares no... | Ian Romanick | 2019-08-05 | 1 | -0/+28 |
| * | nir/algebraic: Simplify some comparisons like a+constant < constant | Ian Romanick | 2019-08-05 | 1 | -0/+8 |
| * | nir/algebraic: Recognize (a < 0 || 0 < b) as min(a, -b) < 0 | Ian Romanick | 2019-08-05 | 1 | -0/+14 |
| * | nir/algebraic: Replace checks that a value is between (or not) [0, 1] | Ian Romanick | 2019-08-05 | 1 | -0/+10 |
| * | nir/algebraic: rename lower_bitshift to lower_bitops | Erico Nunes | 2019-07-31 | 1 | -5/+5 |
| * | nir/algebraic: add new fsum ops and fdot lowering | Erico Nunes | 2019-07-31 | 1 | -0/+6 |
| * | nir/algebraic: add scmp algebraic optimizations | Jonathan Marek | 2019-07-24 | 1 | -0/+16 |
| * | nir/algebraic: add option to lower fall_equalN/fany_nequalN | Jonathan Marek | 2019-07-24 | 1 | -0/+6 |
| * | nir/algebraic: add fdot2 optimizations | Jonathan Marek | 2019-07-24 | 1 | -0/+3 |
| * | nir/algebraic: add option to lower fdph | Jonathan Marek | 2019-07-24 | 1 | -1/+3 |
| * | nir: replace lower_sincos with algebraic opt | Jonathan Marek | 2019-07-24 | 1 | -0/+10 |
| * | nir/algebraic: allow swizzle in nir_algebraic replace expression | Jonathan Marek | 2019-07-24 | 1 | -5/+6 |
| * | nir/algebraic: mark a few comparison simplifications as precise | Rhys Perry | 2019-07-19 | 1 | -2/+2 |
| * | nir/algebraic: optimize contradictory iand operands | Rhys Perry | 2019-07-19 | 1 | -0/+6 |
| * | nir/algebraic: Optimize comparisons and up-casts | Jason Ekstrand | 2019-07-17 | 1 | -0/+67 |
| * | nir/algebraic: Optimize comparing unpacked values | Jason Ekstrand | 2019-07-17 | 1 | -0/+8 |
| * | nir/algebraic: Recognize open-coded flrp(a, b, a) | Ian Romanick | 2019-07-11 | 1 | -0/+24 |
| * | nir/algebraic: Rearrange 1-((1-a) * (1-b)) into flrp-friendly form | Ian Romanick | 2019-07-11 | 1 | -0/+10 |
| * | nir/algebraic: Reassociate fadd into fmul in DPH-like pattern | Ian Romanick | 2019-07-11 | 1 | -0/+12 |
| * | nir/algebraic: Recognize open-coded flrp(-1, 1, a) and flrp(1, -1, a) | Ian Romanick | 2019-07-11 | 1 | -0/+12 |
| * | nir: Add optimization to use ROR/ROL instructions | Sagar Ghuge | 2019-07-01 | 1 | -0/+12 |
| * | nir/algebraic: Fail build when too many commutative expressions are used | Ian Romanick | 2019-06-28 | 1 | -1/+7 |
| * | nir: Fix lowering of bitfield_insert to shifts. | Eric Anholt | 2019-06-28 | 1 | -3/+5 |
| * | nir/algebraic: Add helpers and a rule involving wrapping | Caio Marcelo de Oliveira Filho | 2019-06-26 | 1 | -0/+3 |
| * | nir: remove fnot/fxor/fand/for opcodes | Jonathan Marek | 2019-06-26 | 1 | -2/+0 |
| * | nir: introduce lowering of bitfield_insert to bfm and a new opcode bitfield_s... | Daniel Schürmann | 2019-06-24 | 1 | -0/+6 |
| * | nir/algebraic: Use unsigned comparison when lowering bitfield insert/extract | Daniel Schürmann | 2019-06-24 | 1 | -2/+2 |
| * | nir/algebraic: Remove unnecessary iand of [iu]bfe and bfm sources | Daniel Schürmann | 2019-06-24 | 1 | -0/+8 |
| * | nir: define behavior of nir_op_bfm and nir_op_u/ibfe according to SM5 spec. | Daniel Schürmann | 2019-06-24 | 1 | -11/+4 |
| * | nir/algebraic: add optimization pattern for ('ult', a, ('and', b, a)) and fri... | Daniel Schürmann | 2019-06-24 | 1 | -0/+4 |
| * | nir/opt_algebraic: Fix rules for imadsh_mix16 | Eduardo Lima Mitev | 2019-06-10 | 1 | -2/+2 |
| * | nir_algebraic: Add basic optimizations for umul_low and imadsh_mix16 | Eduardo Lima Mitev | 2019-06-07 | 1 | -0/+9 |
| * | nir: Combine lower_fmod16/32 back into a single lower_fmod. | Kenneth Graunke | 2019-06-05 | 1 | -3/+3 |
| * | nir: Drop lower_fmod64 option. | Kenneth Graunke | 2019-06-05 | 1 | -1/+0 |
| * | nir/algebraic: Simplify max(abs(a), 0.0) -> abs(a) | Alyssa Rosenzweig | 2019-06-04 | 1 | -0/+1 |
| * | nir: add lower_bitshift option | Jonathan Marek | 2019-05-31 | 1 | -3/+5 |
| * | nir/algebraic: Remove problematic "optimization" | Alyssa Rosenzweig | 2019-05-16 | 1 | -3/+0 |
| * | nir/algebraic: Commute 1-fsat(a) to fsat(1-a) for all non-fmul instructions | Ian Romanick | 2019-05-14 | 1 | -5/+8 |
| * | nir/algebraic: Eliminate useless fsat() on operand of comparison w/value in (... | Ian Romanick | 2019-05-14 | 1 | -0/+27 |
| * | nir/algebraic: Strip double negatives from comparison sources | Ian Romanick | 2019-05-14 | 1 | -0/+28 |
| * | intel/compiler: Repeat nir_opt_algebraic_late | Ian Romanick | 2019-05-14 | 1 | -0/+4 |
| * | Revert "nir: add late opt to turn inot/b2f combos back to bcsel" | Ian Romanick | 2019-05-14 | 1 | -3/+0 |
