summaryrefslogtreecommitdiffstats
path: root/src/compiler/nir/nir_opt_algebraic.py
Commit message (Expand)AuthorAgeFilesLines
* python: Use the right function for the jobMathieu Bridon2018-08-091-1/+1
* nir: Transform expressions of b2f(a) and b2f(b) to a == bIan Romanick2018-08-041-0/+3
* nir: Transform expressions of b2f(a) and b2f(b) to a ^^ bIan Romanick2018-08-041-0/+3
* nir: Transform expressions of b2f(a) and b2f(b) to !(a && b)Ian Romanick2018-08-041-0/+3
* nir: Transform expressions of b2f(a) and b2f(b) to a && bIan Romanick2018-08-041-0/+3
* nir: Transform expressions of b2f(a) and b2f(b) to !(a || b)Ian Romanick2018-08-041-0/+11
* nir: Transform -fabs(a) >= 0 to a == 0Ian Romanick2018-08-041-0/+9
* nir: Transform expressions of b2f(a) and b2f(b) to a || bIan Romanick2018-08-041-0/+19
* nir: Transform -fabs(a) < 0 to a != 0Ian Romanick2018-08-041-0/+4
* nir: Rearrange bcsel with two bcsel sourcesIan Romanick2018-08-041-0/+4
* nir: Collapse more repeated bcsels on the same argumentIan Romanick2018-08-041-0/+1
* nir: Don't compare i2f or u2i with zeroIan Romanick2018-08-041-0/+13
* nir: Remove f2i(i2f(x)) conversionsIan Romanick2018-08-041-0/+8
* nir: Mark the 0.0 < abs(a) transformation as impreciseIan Romanick2018-08-041-1/+1
* nir: Add a couple of iand/ior optimizationsJason Ekstrand2018-07-241-0/+4
* nir: Add a couple trivial abs optimizationsJason Ekstrand2018-07-231-0/+2
* nir: add a couple of ior opts to nir_opt_algebraicTimothy Arceri2018-07-181-0/+3
* python: Use the print functionMathieu Bridon2018-07-061-5/+7
* python: Stabilize some script outputsMathieu Bridon2018-07-051-1/+2
* nir: Add lowering for find_lsb.Eric Anholt2018-06-061-0/+4
* nir: Add lowering for ifind_msb to ufind_msb.Eric Anholt2018-06-061-0/+4
* nir: Add lowering from ibitfield_extract/ubitfield_extract to shifts.Eric Anholt2018-06-061-0/+16
* nir: Add lowering for bitfieldInsert without using bfi.Eric Anholt2018-06-061-0/+14
* nir: Lower !f2b(x) to x == 0.0Ian Romanick2018-06-011-0/+1
* nir: Add some missing "optimization undo" patternsIan Romanick2018-06-011-1/+19
* nir: optimize iand(ieq(a, 0), ieq(b, 0)) to ieq(ior(a, b), 0)Samuel Pitoiset2018-05-311-0/+2
* nir: add unsigned comparison simplificationsTimothy Arceri2018-05-301-0/+2
* nir: Implement optional b2f->iand loweringAlyssa Rosenzweig2018-05-181-1/+4
* nir: Don't condition 'a-b < 0' -> 'a < b' on is_not_used_by_conditionalIan Romanick2018-03-261-3/+1
* nir: Don't compare b2f or b2i with zeroIan Romanick2018-03-191-0/+5
* nir: Don't i2b a value that is already BooleanIan Romanick2018-03-081-0/+1
* nir: Narrow some dot product operationsIan Romanick2018-03-081-0/+8
* nir: Simplify some comparisons like a+b < aIan Romanick2018-03-061-0/+9
* nir: Use De Morgan's Law on logic compounded comparisonsIan Romanick2018-03-061-0/+9
* nir: Replace fmin(b2f(a), b) with a bcselIan Romanick2018-03-061-0/+9
* nir: Pull b2f out of bcselIan Romanick2018-03-061-0/+1
* nir: Replace an odd comparison involving fmin of -b2fIan Romanick2018-03-061-0/+13
* nir: Mark bcsel-to-fmin (or fmax) transformations as inexactIan Romanick2018-03-061-2/+2
* nir: Recognize some more open-coded fmin / fmaxIan Romanick2018-03-061-0/+2
* nir: add lower_ldexp to nir compiler optionsTimothy Arceri2018-02-281-2/+2
* nir: lower fexp2(fmul(flog2(a), 2)) to fmul(a, a)Samuel Pitoiset2018-02-221-0/+2
* nir: add is_used_once for fmul(fexp2(a), fexp2(b)) to fexp2(fadd(a, b))Samuel Pitoiset2018-02-221-1/+1
* nir: Distribute binary operations with constants into bcselIan Romanick2018-01-301-0/+15
* nir: Rearrange logic op-compounded integer comparesIan Romanick2018-01-301-0/+35
* nir: Rearrange and-compounded float comparesIan Romanick2018-01-301-0/+8
* nir: Separate a weird compare with zero to two compares with zeroIan Romanick2018-01-301-0/+2
* nir: Simplify min and max of b2fIan Romanick2018-01-301-0/+5
* nir: Undo possible damage caused by rearranging or-compounded float comparesIan Romanick2018-01-301-0/+9
* nir: Be more conservative about rearranging or-compounded comparesIan Romanick2018-01-301-4/+9
* nir: See through an fneg to apply existing optimizationsIan Romanick2018-01-301-3/+6