Commits · d0e8d251eb5516ac3c27cb607295278587088e86 · Roger Ferrer / llvm-epi · GitLab

Jul 30, 2016
- AMDGPU: Fix shouldConvertConstantLoadToIntImm behavior · 749035b7
  Matt Arsenault authored Jul 30, 2016
  749035b7
Jul 29, 2016
- AMDGPU/SI: Don't handle a loop if there is no loop at all for a terminator BB. · 26fb9d26
  Changpeng Fang authored Jul 28, 2016
  26fb9d26
Jul 28, 2016
- AMDGPU : Add intrinsics for compare with the full wavefront result · 07e03712
  Wei Ding authored Jul 28, 2016
  07e03712
- AMDGPU: add execfix flag to SI_ELSE · 3b572002
  Nicolai Haehnle authored Jul 28, 2016
  3b572002
Jul 27, 2016
- AMDGPU: Use rcp for fdiv 1, x with fpmath metadata · e3862cdc
  Matt Arsenault authored Jul 26, 2016
  e3862cdc
- AMDGPU: Add more tests for LDS size with occupancy · 918e81c3
  Matt Arsenault authored Jul 26, 2016
  918e81c3
Jul 26, 2016
- MIRParser: Use dot instead of colon to mark subregisters · 333e468d
  Matthias Braun authored Jul 26, 2016
  333e468d
- GlobalISel: omit braces on MachineInstr types when there's only one. · 26e40bdb
  Tim Northover authored Jul 26, 2016
  26e40bdb
- AMDGPU: Add missing tests for xnack option for HSA · 07f65718
  Matt Arsenault authored Jul 26, 2016
  07f65718
- AMDGPU: Add fp legacy instruction intrinsics · 32fc527c
  Matt Arsenault authored Jul 26, 2016
  32fc527c
Jul 25, 2016
- AMDGPU: Remove read_workdim intrinsic · b64c8925
  Jan Vesely authored Jul 25, 2016
  b64c8925
- AMDGPU: Fix missing verify-machineinstrs in control flow test · e047b259
  Matt Arsenault authored Jul 25, 2016
  e047b259
Jul 23, 2016
- Revert "[AMDGPU] Emit read-only data to .rodata for hsa" · b8253c88
  Tom Stellard authored Jul 22, 2016
  b8253c88
- GlobalISel: allow multiple types on MachineInstrs. · 98a56eb7
  Tim Northover authored Jul 22, 2016
  98a56eb7
Jul 22, 2016
- Invariant start/end intrinsics overloaded for address space · 0be4a0e6
  Anna Thomas authored Jul 22, 2016
  0be4a0e6
- AMDGPU: Remove redundant test · e2fe67b9
  Matt Arsenault authored Jul 22, 2016
  e2fe67b9
- AMDGPU: Fix groupstaticsize for large LDS · 3c07c813
  Matt Arsenault authored Jul 22, 2016
  3c07c813
- AMDGPU: Add HSA dispatch id intrinsic · 8d718dcf
  Matt Arsenault authored Jul 22, 2016
  8d718dcf
- AMDGPU: Fix i1 fp_to_int · 7fb961f3
  Matt Arsenault authored Jul 22, 2016
  7fb961f3
Jul 21, 2016
- Revert "Invariant start/end intrinsics overloaded for address space" · c858faa2
  Anna Thomas authored Jul 21, 2016
  c858faa2
- Invariant start/end intrinsics overloaded for address space · 29b24dfe
  Anna Thomas authored Jul 21, 2016
  29b24dfe
- [AMDGPU] Emit read-only data to .rodata for hsa · 3c0d8d22
  Konstantin Zhuravlyov authored Jul 21, 2016
  3c0d8d22
- AMDGPU: Fix phis from blocks split due to register indexing · f0ba86a4
  Matt Arsenault authored Jul 21, 2016
  f0ba86a4
Jul 20, 2016
- GlobalISel: implement low-level type with just size & vector lanes. · 62ae568b
  Tim Northover authored Jul 20, 2016
  62ae568b
- AMDGPU: Add missing test coverage for control flow breaks · f14db7a9
  Matt Arsenault authored Jul 20, 2016
  f14db7a9
- AMDGPU: Fix bug causing crash due to invalid opencl version metadata. · 4b1d9f7f
  Yaxun Liu authored Jul 20, 2016
  4b1d9f7f
- Revert "RegScavenging: Add scavengeRegisterBackwards()" · 5b9722d6
  Matthias Braun authored Jul 20, 2016
  5b9722d6
- AMDGPU: Change fdiv lowering based on !fpmath metadata · a1fe17c9
  Matt Arsenault authored Jul 19, 2016
  a1fe17c9
- RegScavenging: Add scavengeRegisterBackwards() · 84fd4bee
  Matthias Braun authored Jul 19, 2016
  84fd4bee
Jul 19, 2016
- AMDGPU: Expand register indexing pseudos in custom inserter · cb540bc0
  Matt Arsenault authored Jul 19, 2016
  cb540bc0
- AMDGPU: Fix test name and broken CHECK-LABEL · 50b76399
  Matt Arsenault authored Jul 18, 2016
  50b76399
Jul 18, 2016
- AMDGPU: Add intrinsic for s_flbit_i32/v_ffbh_i32 · c96e1def
  Matt Arsenault authored Jul 18, 2016
  c96e1def
- AMDGPU/R600: Replace barrier intrinsics · 4c519d35
  Matt Arsenault authored Jul 18, 2016
  4c519d35
- AMDGPU: Remove dead check in AMDGPUPromoteAlloca · efb24540
  Matt Arsenault authored Jul 18, 2016
  efb24540
- AMDGPU: Disable AMDGPUPromoteAlloca pass for shader calling conventions. · bef1ceb8
  Nicolai Haehnle authored Jul 18, 2016
  bef1ceb8
Jul 16, 2016
- Re-commit [AMDGPU] Add metadata for runtime · a711cc79
  Yaxun Liu authored Jul 16, 2016
  a711cc79
- AMDGPU: Fix verifier error from partially undef copy · 73d2f895
  Matt Arsenault authored Jul 15, 2016
  73d2f895
Jul 15, 2016
- AMDGPU: Remove brev intrinsic · a65e6b83
  Matt Arsenault authored Jul 15, 2016
  a65e6b83
- AMDGPU: Fix TargetPrefix for remaining r600 intrinsics · 82e5e1e5
  Matt Arsenault authored Jul 15, 2016
  82e5e1e5
- AMDGPU: Remove AMDGPU.ldexp · 11d3e21f
  Matt Arsenault authored Jul 15, 2016
  11d3e21f