summaryrefslogtreecommitdiffstats
path: root/src/compiler
Commit message (Expand)AuthorAgeFilesLines
* spirv/radv: add AMD_gcn_shader capability, remove current extensionsAlejandro Piñeiro2018-03-153-7/+2
* spirv: update arguments for vtn_nir_alu_op_for_spirv_opcode()Samuel Iglesias Gonsálvez2018-03-153-15/+15
* spirv: fix the translation of SPIR-V conversion opcodes to NIRSamuel Iglesias Gonsálvez2018-03-151-3/+37
* glsl: Use hash table cloning in copy propagationThomas Helland2018-03-142-31/+15
* compiler: int8/uint8 supportKarol Herbst2018-03-1411-0/+113
* spirv: Handle doubles when multiplying a mat by a scalarNeil Roberts2018-03-141-3/+3
* nir: lower_load_const_to_scalar fix for 8/16b typesRob Clark2018-03-131-4/+15
* nir/subgroups: Add lowering for vote_ieq/vote_feq to a ballotJason Ekstrand2018-03-132-0/+49
* spirv: Silence compiler warning about undefined srcs[0]Eric Anholt2018-03-131-0/+1
* 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
* spirv: fix autotools buildsTimothy Arceri2018-03-081-0/+1
* spirv: handle AMD_gcn_shader extended instructionsDaniel Schürmann2018-03-076-0/+70
* nir: add AMD_gcn_shader extended instructionsDaniel Schürmann2018-03-072-0/+30
* spirv: import AMD extensions header from glslangDaniel Schürmann2018-03-071-0/+110
* spirv: Add support for subgroup arithmeticJason Ekstrand2018-03-073-8/+94
* nir: Add a helper for getting binop identitiesJason Ekstrand2018-03-072-0/+70
* nir: Add subgroup arithmetic reduction intrinsicsJason Ekstrand2018-03-074-0/+34
* spirv: Add subgroup quad supportJason Ekstrand2018-03-073-2/+28
* nir: Add quad operations and loweringJason Ekstrand2018-03-073-0/+44
* spirv: Add subgroup shuffle supportJason Ekstrand2018-03-073-1/+29
* nir: Add subgroup shuffle intrinsics and loweringJason Ekstrand2018-03-073-3/+69
* nir/lower_subgroups: Add scalarizing for vote_eqJason Ekstrand2018-03-071-0/+29
* spirv: Add subgroup vote supportJason Ekstrand2018-03-073-1/+51
* nir: Generalize nir_intrinsic_vote_eqJason Ekstrand2018-03-074-5/+9
* spirv: Add subgroup ballot supportJason Ekstrand2018-03-074-8/+161
* spirv: Add initial subgroup supportJason Ekstrand2018-03-077-0/+145
* nir: Add new SPIR-V ballot intrinsics and loweringJason Ekstrand2018-03-072-0/+22
* compiler: Add two new system values for subgroupsJason Ekstrand2018-03-074-0/+19
* nir: Add new SPIR-V ballot ALU intrinsics and loweringJason Ekstrand2018-03-072-0/+76
* spirv: Handle the new OpModuleProcessed instructionJason Ekstrand2018-03-071-0/+1
* nir/spirv: Add support for device groupsJason Ekstrand2018-03-077-0/+20
* spirv: Update the SPIR-V headers and json to 1.3.1Jason Ekstrand2018-03-072-41/+683
* spirv: Rework barriersJason Ekstrand2018-03-071-18/+114
* spirv: Add a vtn_constant_value helperJason Ekstrand2018-03-071-0/+6
* glsl: Fix memory leak with known glsl_type instancesSimon Hausmann2018-03-072-87/+47
* spirv: Add SpvCapabilityShaderViewportIndexLayerEXTCaio Marcelo de Oliveira Filho2018-03-073-0/+13
* Revert "nir: bump loop unroll limit to 96."Timothy Arceri2018-03-071-3/+1
* 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: Silence unused parameter warnings in generated nir_constant_expressions ...Ian Romanick2018-03-021-1/+2
* nir/search: Include 8 and 16-bit support in construct_valueJose Maria Casanova Crespo2018-03-011-0/+15
* nir/search: Support 8 and 16-bit constants in match_valueJason Ekstrand2018-03-011-0/+20
* nir/serialize: handle var->name being NULLAlejandro Piñeiro2018-03-011-1/+2
* spirv/i965/anv: Relax push constant offset assertions being 32-bit alignedJose Maria Casanova Crespo2018-02-281-2/+0