aboutsummaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2018-11-02COMPMID-1086: Optimizing GCGEMM - Add helpers to get target GPU informationMichele Di Giorgio
This patch moves some of the helper functions in CLHelpers.[h,cpp] to GPUTarget.[h,cpp] in order to make them avaialable to both CL and GLES backends. Change-Id: Ib89b05a2084b73eb643b32fac13b8367cc80b6a5 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/128988 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-948: Add validation to CLL2NormalizeLayerJohn Richardson
Change-Id: I452a718a60b81da51cd3e98641fd99c86c4debab Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/129451 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
2018-11-02COMPMID-1050 CL/NEON: Create a function to convert the 2D weights of FC ↵Giorgio Arena
layer from NHWC to NCHW and viceversa Change-Id: If77ffeb92b6eb883e5d2d2c97c2c4d1d23d17c8d Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/129257 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-605: Transition buffer memory managerGeorgios Pinitas
Change-Id: Ide7c6124eb19f13f15f517e62d705646a0cd1ecd Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/130184 Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-959: Remove invalid switch log info on default method.Georgios Pinitas
When Default was set as the convolution method a switch convolution info was incorrectly getting reported. Change-Id: I750b28afec4665799362f77ec3f9b66cf2adb0aa Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/130738 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Isabella Gottardi <isabella.gottardi@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-1052 - Rework validate method in CLGEMMGian Marco Iodice
Change-Id: Iece5bd6478b5fac5164abff30c1e63e8a77291a9 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/130374 Reviewed-by: Anthony Barbier <anthony.barbier@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-959: Fixed Clang-tidy errorAnthony Barbier
Change-Id: Ice1f3b44ae1f503c47b8caf2c3c74f2e976826c7 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/130757 Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-1135 Fix CLGEMMDilatedConvolution mismatchGiorgio Arena
Change-Id: I61364c935c157db352336d126788bfbd18191d2b Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/130748 Reviewed-by: Isabella Gottardi <isabella.gottardi@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-761: Add CL/NEON MeanStdDev benchmark testsJohn Richardson
Change-Id: I6714bd56e4b58e486a41c4f3c00f6f001a7bb782 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/129441 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-810 Add NHWC data format support for NEON convolutionMichalis Spyrou
Change-Id: I2a7b49a12da7f3bc3f04749243b1dc111160de6e Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/129348 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-1112: Enabled multithreading transforms in Winograd.Pablo Tello
Updated RSH code as well. Change-Id: I9452ff5c7f0ff0cd60b8c223cdd71077288eb0c1 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/130177 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-597: Port HOGMultiDetection to new frameworkJohn Richardson
Change-Id: I4b31b7f052a06bea4154d04c9926a0e076e28d73 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/126555 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: John Richardson <john.richardson@arm.com> Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-1109 - Enabling Winograd in the graph when possibleGian Marco Iodice
Change-Id: I524abd28188995ae9c7a43b189b1eb2d7546be93 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/130576 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
2018-11-02COMPMID-959: Add missing header in Workload.hGeorgios Pinitas
Change-Id: I461a113246db73a5ff1e98f5f69cea6d84e2f91f Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/130689 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Pablo Tello <pablo.tello@arm.com>
2018-11-02COMPMID-1131: Fix SIGBUS errorGeorgios Pinitas
Forcing 128-byte alignment on pretransposed buffer. Change-Id: I534c5ebb30ffc9a219e7a3e2d495b4b83b0d1fea Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/130614 Reviewed-by: Pablo Tello <pablo.tello@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-1134: Fixing the Window reconstruction in gemm_nativeAnthony Barbier
Change-Id: Ibab955dbade4805c39cab003362be2bb3c74b166 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/130605 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Pablo Tello <pablo.tello@arm.com>
2018-11-02COMPMID-997: Add support for node's name in GraphAPI.Georgios Pinitas
Change-Id: I0ca02e42807c1ad9afeffb7202a3556feb11442f Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/129701 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
2018-11-02COMPMID-1097: Port mobilenet to NHWCGeorgios Pinitas
Change-Id: I789065bfa0d4ef133388e1904c5caf31e450f80f Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/129495 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-1068 Create validate method to CLDepthWiseConvolutionGiorgio Arena
Change-Id: I3301b66a8a072c6ecd0d7f2dabef350017b55ac4 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/128677 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-1046 - Enabling Winograd VGG19 to use Winograd on NEON and OpenCLGian Marco Iodice
Change-Id: If8a28fc6a3a58473df51c8e7399e6d06d0db10f9 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/127384 Reviewed-by: Anthony Barbier <anthony.barbier@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-959: Assembly kernel marks weights as unused when pretransposeGeorgios Pinitas
Change-Id: I15943c4c97457f4cc9bff3bb1cce8ad9a8d9ed38 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/130474 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-922 - CLGEMM FP16 optimizations - part2Gian Marco Iodice
This patch improves of ~30 % GEMM fp16 when the reshape is required The results have been reported at the following confluence page: https://confluence.arm.com/display/MLENG/GEMM+FP16+performance%3A+ACL+18.05 Change-Id: I8233095a7e9ab06f1f915782a25dd41653b49140 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/128254 Reviewed-by: Anthony Barbier <anthony.barbier@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-1122: Store the cl_context only in one placeAnthony Barbier
Change-Id: I39a8dd34a27f9f891397d58af36bd5f9d152281e Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/130358 Reviewed-by: Pablo Tello <pablo.tello@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-1125: Add support for FP16 in CLDepthwiseConvolutionMichele Di Giorgio
Change-Id: I4838f5a8e4c33ed646cd05e0bb682fca635a29a3 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/130469 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Michalis Spyrou <michalis.spyrou@arm.com>
2018-11-02COMPMID-959: Perform pretranspose if allowed on NEON assemblyGeorgios Pinitas
Change-Id: I281699ce7270aec1317c47b5a13799954cf6c9e8 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/130010 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Pablo Tello <pablo.tello@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-1084 Rework the way validation is performed for NHWC data layoutGiorgio Arena
Change-Id: I00b95f560548da76718298b642c8166f92421097 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/129520 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Michele DiGiorgio <michele.digiorgio@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-1115: Valgrind: asserts triggered in NEPoolingLayerKernelGeorgios Pinitas
Change-Id: I39573d6fe213a273bebeab66724781b1a134a6f8 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/130357 Reviewed-by: Giorgio Arena <giorgio.arena@arm.com> Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-1123 - Enabling Winograd 2x2,3x3 and Winograd 4x4,3x3 without fast_mathGian Marco Iodice
Change-Id: I5ef3db4f84428f7980b97cf8e2c580f78ce021f8 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/130447 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Giorgio Arena <giorgio.arena@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-920: Introduce prepare() stageGeorgios Pinitas
Change-Id: I08ddb7f6e061178e7566518b48e4e18f8f078596 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/129825 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-1110 Add QASYMM8 to supported types for NE/CLActivationLayerMichalis Spyrou
Change-Id: I4c04b6271e923a07ad861cd6663bb48d0a663301 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/130360 Reviewed-by: Michele DiGiorgio <michele.digiorgio@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-761: Add CL/NEON IntegralImage benchmark testsJohn Richardson
Change-Id: I84689a94d4d9bbd4752dbb9fed8ec8dcc5c51eef Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/129433 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Pablo Tello <pablo.tello@arm.com>
2018-11-02COMPMID-1117: TransposeAccessWindow leads to high paddingGeorgios Pinitas
Switches CLGEMMMatrixMultiplyKernel and CLGEMMTranspose1xWKernel to use AccessWindowStatic Change-Id: I21533d4218215d5b8f84b23c603062678eccb1ed Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/130244 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
2018-11-02COMPMID-1116 - Increase the tolerance for CLWinogradConvolutionLayerGian Marco Iodice
COMPMID-1114 - Fix arm_compute_nightly_validation SIGKILL Change-Id: Ie8aba45d13b989ca7aa5208cb4c757f153d6f5ab Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/130214 Reviewed-by: Anthony Barbier <anthony.barbier@arm.com> Tested-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-1107: Add support for ChannelShuffle in CLMichele Di Giorgio
Change-Id: I56d2a02b316f0c69ff1fd7220e732f775414fe69 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/129709 Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Giorgio Arena <giorgio.arena@arm.com> Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
2018-11-02COMPMID-808 Add NHWC data format support for NEON direct convolutionGiorgio Arena
Change-Id: I5d4cc3d5b0d25f3fe4ed998c0f15b1b8e260a43a Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/125697 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-1089 CLPoolingLayer fix padding and add output shape computation to ↵Giorgio Arena
ShapeCalculator Change-Id: Ide83424e9fe6b8102ed9e3c355c099c3912e7e61 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/129635 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Michele DiGiorgio <michele.digiorgio@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-1105: Fix mismatches conv layer when using NativeGemm with multiple ↵Pablo Tello
threads Change-Id: Id5ba16a7e3382070fda936c63d174df53596da04 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/129964 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Giorgio Arena <giorgio.arena@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-761: Add CL/NEON Median3x3 benchmark testsJohn Richardson
Change-Id: I6a4720ed90072d47ddb1908a503b3ab18c442fed Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/129434 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-926 Add depth multiplier support to NEON/CL/GLES depthwise convolutionGiorgio Arena
Change-Id: I03f32c62350e5ea43e77bb15fc5a832d83719e3b Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/126657 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Michele DiGiorgio <michele.digiorgio@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
2018-11-02COMPMID-805 Add NHWC data format support for CL poolingMichalis Spyrou
Change-Id: I3d91fde78b971aba3f6349f633cd9b1c50e5cacf Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/124712 Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com> Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
2018-11-02COMPMID-1096 - Add fast_math flag to CLConvolutionLayerGian Marco Iodice
COMPMID-1103 - CLWinogradConvolutionLayer mismatches Change-Id: Iceaa9482a1790ec39d2720c220261aaea8043978 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/129398 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Giorgio Arena <giorgio.arena@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
2018-11-02COMPMID-761: Add CL/NEON Phase benchmark testsJohn Richardson
Change-Id: I5986d05b576d4d79d4f38e6805b5d0c8f5c3d2b4 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/129447 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
2018-11-02COMPMID-959: Fail when graph is already registeredGeorgios Pinitas
Change-Id: I1cc82bd8535dead60490b8112d7b8507cee45ff6 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/129399 Reviewed-by: Michele DiGiorgio <michele.digiorgio@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-1085 : runtime_error thrown by QASYMM8 CLGEMMConvolutionLayer::validateVidhya Sudhan Loganathan
shape and quantization info were corrected. Error from validate() is forwarded. Validate() tests outside the context of configure()are added. Change-Id: I13f1a02eccda6b595089c4875b21853ca372f2f2 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/129323 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
2018-11-02COMPMID-1010: Remove RSH profiler headerGeorgios Pinitas
Change-Id: I2967ec94c3bead0b92ff1d1581ff6afea21c7f04 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/129405 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Michalis Spyrou <michalis.spyrou@arm.com>
2018-11-02COMPMID-806 Add NHWC data format support format for NEON poolingMichalis Spyrou
Change-Id: I7ab174c72f3d56134fcec259a137739061fd12e9 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/123065 Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-761: Add CL/NEON HOGDescriptor benchmark testsJohn Richardson
Change-Id: I19bb658b0f06b0733fa62a6745c6163e0648a8b0 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/128240 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com> Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
2018-11-02COMPMID-1092: Fix CL GEMMConvolutionLayer FP32 mismatchesGeorgios Pinitas
Change-Id: I76fe0708b50c71b0b4dcda59f6378608ce7a30c9 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/129337 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Michalis Spyrou <michalis.spyrou@arm.com>
2018-11-02COMPMID-1087: Fix CLHOGDetector failure.John Richardson
Change-Id: Ied0fbbd50f1e4dcf930798bc3b735cb70329017d Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/129313 Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-1074: Rename WinograLayer.cpp to WinogradConvolutionLayer.cppGeorgios Pinitas
Change-Id: Iccac7cd6cb458469568d0cd6fb36b262353f4188 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/129261 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Pablo Tello <pablo.tello@arm.com>