Gitiles
Code Review Sign In
gerrit-public.fairphone.software / platform / external / XNNPACK / e2dc2ec647a2507930bbac3630596a499879b596
  1. e2dc2ec Implement subs for aarch64 assembler by Zhi An Ng · 2 years, 8 months ago
  2. 234d6b4 Implement prfm (only PLDL1KEEP) on aarch64 assembler by Zhi An Ng · 2 years, 8 months ago
  3. 65ccb13 Implement movi for aarch64 assembler by Zhi An Ng · 2 years, 8 months ago
  4. 6e68f54 Implement ld1 for 1, 2, and 3 registers for aarch64 assembler by Zhi An Ng · 2 years, 8 months ago
  5. 5702efb Implement ld2r for aarch64 assembler by Zhi An Ng · 2 years, 8 months ago
  6. 04cdc41 Implement ldr for aarch64 assembler by Zhi An Ng · 2 years, 8 months ago
  7. 0ba29e7 Implement LDP for aarch64 assembler by Zhi An Ng · 2 years, 8 months ago
  8. 70ea0a2 Specialize F32 GEMM A53 JIT microkernel for min/max params by Zhi An Ng · 2 years, 8 months ago
  9. 109a5eb Initial aarch64 assembler structure by Zhi An Ng · 2 years, 8 months ago
  10. 8f920a6 Initialize F16 microkernel pointers on x86 by Marat Dukhan · 2 years, 8 months ago
  11. ffbf7ff Cleanup transpose microkernels in BUILD & CMakeLists by Marat Dukhan · 2 years, 8 months ago
  12. 66eb508 Add missing declarations and unit tests for F16 DWCONV microkernels by Marat Dukhan · 2 years, 8 months ago
  13. 0ec25cf Duplicate test methods in gemm-microkernel-test for JIT codegen, update IGEMM generator signature and test generation script. by Zhi An Ng · 2 years, 8 months ago
  14. e7225eb Specialize F32 GEMM (a53) on kc by Zhi An Ng · 2 years, 8 months ago
  15. 8d07e40 Enable QU8 4x8 NEON MLA Lane microkernel AArch32 assembly language by Frank Barchard · 2 years, 8 months ago
  16. 901845c QU8 4x8 NEON MLA Lane microkernel AArch32 assembly language by Frank Barchard · 2 years, 8 months ago
  17. b26ead1 F16C implementation of F16 GAVGPOOL microkernels by Marat Dukhan · 2 years, 8 months ago
  18. c7c92b0 Generate F16 GAVGPOOL NEONFP16ARITH microkernels from template by Marat Dukhan · 2 years, 8 months ago
  19. 01f6aee Add unreachable check for F32 GEMM a53 generator by Zhi An Ng · 2 years, 8 months ago
  20. e78eb33 Bump shard count for f32_igemm_minmax_test (timing out on coverage runs) by Zhi An Ng · 2 years, 8 months ago
  21. 13599f3 Specialize F32 GEMM (a53) on nc by Zhi An Ng · 2 years, 8 months ago
  22. 1d6b7c9 Support FP32 weights in FP16 NC Fully Connected operator by Marat Dukhan · 2 years, 9 months ago
  23. d2e8d4d Enable QC8 AArch32 4x8 lane GEMM/IGEMM assembly microkernels for ARMv7 NEON by Frank Barchard · 2 years, 9 months ago
  24. 6989ec4 Support FP32 weights in FP16 NHWC Convolution operator by Marat Dukhan · 2 years, 9 months ago
  25. 5e1a303 QC8 GEMM/IGEMM assembly microkernels for ARMv7 NEON by Frank Barchard · 2 years, 9 months ago
  26. 83844ae Change JIT generator signature to accept nc and kc to specialize on those values by Zhi An Ng · 2 years, 9 months ago
  27. b1a869d Merge generate transpose scripts by Alan Kelly · 2 years, 9 months ago
  28. 9dfdfb5 Remove unused transpose function declarations. by Alan Kelly · 2 years, 9 months ago
  29. 667e0f1 Regenerate transpose tests by Alan Kelly · 2 years, 9 months ago
  30. 4b23423 Split test generator for qu8-gavgpool by Frank Barchard · 2 years, 9 months ago
  31. 5da6d38 SSE2 transpose microkernel code generator. by Alan Kelly · 2 years, 9 months ago
  32. d19bde9 Add x64 scalar transpose microkernels by Alan Kelly · 2 years, 9 months ago
  33. cd21b02 Add x8 scalar transpose microkernels by Alan Kelly · 2 years, 9 months ago
  34. 84aae41 Add x16 scalar transpose microkernels by Alan Kelly · 2 years, 9 months ago
  35. 6315472 Remove declarations for scalar transpose microkernels that don't exist by Alan Kelly · 2 years, 9 months ago
  36. d7111a5 Remove F32 GEMM E2E JIT benchmarks (temporarily) as we are changing the JIT generator interface by Zhi An Ng · 2 years, 9 months ago
  37. 58fe65e Change default JIT code buffer size to 16kb by Zhi An Ng · 2 years, 9 months ago
  38. af9ff85 Fix GEMM test templates to use variable n instead of fixed NR and regenerate tests by Zhi An Ng · 2 years, 9 months ago
  39. 2d38e3c Fix more errors in CMakeLists by Marat Dukhan · 2 years, 9 months ago
  40. 1e074d7 Fix CMake build by Marat Dukhan · 2 years, 9 months ago
  41. 8575504 Switch QS8/QU8 GAVGPOOL NEON microkernels to RNDNU requantization by Marat Dukhan · 2 years, 9 months ago
  42. 33a98fa Switch QS8/QU8 VMUL[C] NEON microkernels to RNDNU requantization by Marat Dukhan · 2 years, 9 months ago
  43. d1f53e4 Generate QU8 GAVGPOOL microkernels from QS8 GAVGPOOL templates by Marat Dukhan · 2 years, 9 months ago
  44. d81fa0a Pipeline remaining QS8 AVGPOOL microkernels by Marat Dukhan · 2 years, 9 months ago
  45. 139337c Include vcvtnq_f32 polyfill in QS8 GAVGPOOL NEONV8 microkernels by Marat Dukhan · 2 years, 9 months ago
  46. 9e258d6 Remove multi-accumulator support in QS8 GAVGPOOL microkernels by Marat Dukhan · 2 years, 9 months ago
  47. 7d45d90 Create a new jit-test for jit-related tests that are not architecture specific by Zhi An Ng · 2 years, 9 months ago
  48. 7781786 Enable QU8 3x8 lane for AArch32 by Frank Barchard · 2 years, 9 months ago
  49. d7a4b22 Generate missing QS8 GAVGPOOL microkernels by Marat Dukhan · 2 years, 9 months ago
  50. 6faf955 Reoptimize SSE QS8 GAVGPOOL microkernels by Marat Dukhan · 2 years, 9 months ago
  51. 847ff5e Refactor naming of QS8 GAVGPOOL microkernels by Marat Dukhan · 2 years, 9 months ago
  52. 53f4106 Switch QS8 GAVGPOOL microkernels to use FP32 requantization by Marat Dukhan · 2 years, 9 months ago
  53. e9e9708 Add missing asserts for requantization scale in MUL parameters initialization by Marat Dukhan · 2 years, 9 months ago
  54. 34cb23f Re-generate amalgamated microkernels by Marat Dukhan · 2 years, 9 months ago
  55. 1789a3c Fix CMake builds by Zhi An Ng · 2 years, 9 months ago
  56. 1581248 Specify 8-byte alignment for packed WAsm SIMD parameters by Marat Dukhan · 2 years, 9 months ago
  57. 1a856c1 Change unit tests to depend on jit_test_mode by Zhi An Ng · 2 years, 9 months ago
  58. 90a10b8 Replicate QS8/QU8 MUL WAsm SIMD parameters to 64 bit rather than 128 bit by Marat Dukhan · 2 years, 9 months ago
  59. 3b32963 Fix bug in not changing memory to be executable when we have unused capacity. by Zhi An Ng · 2 years, 9 months ago
  60. 8f2eeee Skip calling __builtin_clear_cache on iOS, iOS uses sys_cache_invalidate by Zhi An Ng · 2 years, 9 months ago
  61. e7242ea Replicate QS8/QU8 ADDSUB WAsm SIMD parameters to 64 bit rather than 128 bit by Marat Dukhan · 2 years, 9 months ago
  62. 48d74c3 Replicate QC8/QS8/QU8 CONV WAsm SIMD parameters to 64 bit rather than 128 bit by Marat Dukhan · 2 years, 9 months ago
  63. b402cbe Bump shard counts for qs8_igemm_minmax_rndnu_test by Zhi An Ng · 2 years, 9 months ago
  64. 44616e1 Bump shard counts for qs8_gemm_minmax_rndnu_test, the test sometimes timeout in coverage runs. by Zhi An Ng · 2 years, 9 months ago
  65. 9441d46 clang-format indents case labels with in switches matching existing XNNPACK style. by Alan Kelly · 2 years, 9 months ago
  66. c27f04b Add missing generated unit tests to BUILD and CMakeLists.txt. by Zhi An Ng · 2 years, 9 months ago
  67. d6e2e1a Remove xnn_qu8_quantize_avgpool and xnn_qs8_quantize_avgpool helpers by Marat Dukhan · 2 years, 9 months ago
  68. 50323b8 Combine requantization with parameter initialization in unit tests by Marat Dukhan · 2 years, 9 months ago
  69. bd7f9a4 F16C implementation of F16 PRELU microkernels by Marat Dukhan · 2 years, 9 months ago
  70. 4c1fd6f Allow generate-gemm-test.py to accept multiple output files, and shard the generated tests across all specified output files. by Zhi An Ng · 2 years, 9 months ago
  71. 4897670 Re-enable up to AVX2 microkernels on Android x86/x86-64 & iOS simulator builds by Marat Dukhan · 2 years, 9 months ago
  72. 603ec5f Remove unused declarations for F16 VRELU microkernels by Marat Dukhan · 2 years, 9 months ago
  73. 085102c Reoptimize pointer updates in PRELU microkernels by Marat Dukhan · 2 years, 9 months ago
  74. 3ab63b0 Rollback "Enable up to AVX2 microkernels on Android x86/x86-64 builds" by XNNPACK Team · 2 years, 9 months ago
  75. d454545 F16C implementation of F16 VBINARY[C] microkernels by Marat Dukhan · 2 years, 9 months ago
  76. 717665f Add JIT microkernels to F32 GEMM E2E benchmarks by Zhi An Ng · 2 years, 9 months ago
  77. 1f1ee2c Enable up to AVX2 microkernels on Android x86/x86-64 builds by Marat Dukhan · 2 years, 9 months ago
  78. a0b45e5 Allow overriding logging settings in Bazel by Marat Dukhan · 2 years, 9 months ago
  79. d90af6f Move gemm-microkernel-tester test code into separate cc file by Zhi An Ng · 2 years, 9 months ago
  80. c7e534f Bump shard_count for slow subtract_nd_test by Zhi An Ng · 2 years, 9 months ago
  81. a30e2df Fix QU8 E2E lane benchmark tile sizes by Frank Barchard · 2 years, 9 months ago
  82. 969e61f Enable 2x16 for QU8 neon lane microkernel in AArch32 by Frank Barchard · 2 years, 9 months ago
  83. 2780863 Scalar transpose microkernel by Alan Kelly · 2 years, 9 months ago
  84. 49979b6 Implement vldr for S registers by Zhi An Ng · 2 years, 9 months ago
  85. a72cde3 Reoptimize pointer updates in VMULCADDC microkernels by Marat Dukhan · 2 years, 9 months ago
  86. e8c1979 Add enable_jit to various targets in BUILD by Zhi An Ng · 2 years, 9 months ago
  87. a248337 Split more of qs8-gemm-minmax-rndnu out into another file, for microkernels with "c4" by Zhi An Ng · 2 years, 9 months ago
  88. 4c738f0 Fix wrong WAsm SIMD parameter initialization in f32-spmm-minmax.yaml by Marat Dukhan · 2 years, 9 months ago
  89. d5a5333 Additional tile sizes for QU8 neon lane microkernel. by Frank Barchard · 2 years, 9 months ago
  90. 751f622 F16C implementation of F16 VHSWISH microkernels by Marat Dukhan · 2 years, 9 months ago
  91. 645af97 FMA3 implementation of F16 DWCONV/VCLAMP/VMULCADDC microkernels by Marat Dukhan · 2 years, 9 months ago
  92. 8459822 Split F32 SCALEMINMAX parameter initialization functions by ISA by Marat Dukhan · 2 years, 9 months ago
  93. f2e2edf Round results to FP16 after multiplication by scale in AVX2 F16 GEMM/IGEMM by Marat Dukhan · 2 years, 9 months ago
  94. ef5560d Use ISA-specific parameter initialization functions in F32 PAVGPOOL tests by Marat Dukhan · 2 years, 9 months ago
  95. 3c949a3 Split QS8/QU8 AVGPOOL parameter initialization functions by ISA by Marat Dukhan · 2 years, 9 months ago
  96. 9f8eac7 Avoid _mm_loadu_si16 and _mm_storeu_si16 unsupported on older compilers by Marat Dukhan · 2 years, 9 months ago
  97. 48c5e98 Fix CMake build on x86-64 by Marat Dukhan · 2 years, 9 months ago
  98. da382d1 Refactor parameter initialization for AVGPOOL/GAVGPOOL/PAVGPOOL microkernels by Marat Dukhan · 2 years, 9 months ago
  99. a7d74b1 Specify parameter initialization function in GAVGPOOL microkernel tests by Marat Dukhan · 2 years, 9 months ago
  100. 4a6dca9 Specify parameter initialization function in [P]AVGPOOL microkernel tests by Marat Dukhan · 2 years, 9 months ago