aboutsummaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2018-11-02COMPMID-1365: Add support for NHWC in CLDepthConcatenateLayerGeorgios Pinitas
Change-Id: I3ed55bdb95d888aff0b0b76fb841bf1669659308 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139963 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-1385: Fixed QASYMM8 mismatchesPablo Tello
Added support for different quantization info in the operands and the output tensors in CLArithmeticAddition. Change-Id: I7704baccc3b609fcc514f947f1b5b5331745ed9e Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139947 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com> Reviewed-by: Giorgio Arena <giorgio.arena@arm.com>
2018-11-02COMPMID-1246 - Fix TensorShape collapse_fromIsabella Gottardi
Change-Id: I076f5e43c76446515d0be7bb1a5d4540d0894760 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/140168 Reviewed-by: Anthony Barbier <anthony.barbier@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-1188: Fixes LSTM IO dimension requirements.Georgios Pinitas
Change-Id: Iee92ccce6422368c19173174e6f58e7aada12233 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/140143 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-1409: Disable BN + Act fusion in the graph if activation is not ↵Georgios Pinitas
supported Change-Id: Ifba641d498638ac48bbc9624338a2168a56fe5c6 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/140075 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02Revert "COMPMID-1246: Disable -pie and -static-libstdc++ when compiling with ↵Anthony Barbier
clang" This reverts commit 15e027e8b4fab55d8bf09e3bf3f589ecf9ef2fe8. Change-Id: I8423b4476b933ca026cd152b8160ad0684f6cb0a Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/140056 Reviewed-by: Pablo Tello <pablo.tello@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-1400: Add command line option to specify the tuner's config fileAnthony Barbier
Change-Id: Ib597e0dff4c8c01f7e6bd46d03824beef4bc1e9a Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139923 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
2018-11-02COMPMID-1246 - Enabled NHWC on graph_mobilenet by defaultGian Marco Iodice
Change-Id: I195d4a0e6fee4cd4ca42403256a9fe639b83fa3d Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/138925 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Giorgio Arena <giorgio.arena@arm.com>
2018-11-02COMPMID-807: NHWC support in CLDirectConvolution.Pablo Tello
Change-Id: I8738aca2cc0104e4c4d7c9605762ab59fce10a33 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/137333 Reviewed-by: Giorgio Arena <giorgio.arena@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-1246: Fixed nightly build: static libraries were linked in the wrong ↵Anthony Barbier
order on Android Change-Id: I5346909c85180eeacd05cd94ee26cb974100c2b9 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139904 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
2018-11-02COMPMID-1271 Avoid memory leak in list of gemm methodsAnthony Barbier
Change-Id: I80764d09bf5fb87b3a98bc0e1803d25c6c682c1f Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139859 Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-1246: Don't link against the Graph library for non-graph examplesAnthony Barbier
Change-Id: Iaba620c72e812e66ab7e0c425b0d8a1fdd305d1e Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139862 Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-1188: Report error on graphs that do not support NHWCGeorgios Pinitas
Change-Id: I14c6ded780339aa75555ea5f62247c509e64d0b0 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139797 Reviewed-by: Anthony Barbier <anthony.barbier@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-1188: Substream copy constructor primitives.Georgios Pinitas
Change-Id: I50b0d9d2f6bfbcd7bfacf31090b94f551fa00bbe Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139876 Reviewed-by: Anthony Barbier <anthony.barbier@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-1288 Optimizing CLGEMMLowp using 8 bit dot product instructionGiorgio Arena
Change-Id: I536174b9381660a94578d6aa1892a6289a820391 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139109 Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-1373 Enable tests for CLWinogradConvolutionLayer for NHWCGiorgio Arena
Change-Id: I2c6a744f174cfb6c78a9923b737f06537debaa0d Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139758 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-1389: Build the validation framework as a standalone static libraryAnthony Barbier
Change-Id: Id0e89cfa3bf5249db3abe110865a67e1777f83a8 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139575 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Pablo Tello <pablo.tello@arm.com>
2018-11-02COMPMID-1399: In ITensor::copy_from, line_size contains twice num_channelsAnthony Barbier
element_size = data_type_size * num_channels line_size = element_size * dimension[0] * num_channels Change-Id: I72dd2ed7ee1f461f86232955c11e002a706cc89a Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139833 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Pablo Tello <pablo.tello@arm.com>
2018-11-02COMPMID-1398: Removed --threads from CommonOptionsAnthony Barbier
Otherwise --threads appear twice for benchmark_graph_* tests Change-Id: I4f5454f6a2fda2558386290583a0ab84cd4666bb Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139827 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Pablo Tello <pablo.tello@arm.com>
2018-11-02COMPMID-1246: Disable -pie and -static-libstdc++ when compiling with clangAnthony Barbier
Remove the following warning: clang50++: warning: argument unused during compilation: '-pie' [-Wunused-command-line-argument] Change-Id: Ifdd53d5cc48542e8bd6b9728c5fdc519ab2c9f06 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139619 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Pablo Tello <pablo.tello@arm.com>
2018-11-02COMPMID-1381: Fix nightly build failure on armv7Anthony Barbier
Change-Id: Ic8e238a03361c04cc0b9daada81d63e5e423d2b8 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139825 Reviewed-by: Pablo Tello <pablo.tello@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-1397: Add option to enable CLTuner in arm_compute_benchmarkAnthony Barbier
Change-Id: Iee062e674e7a20b4786d738d4535e4c89fbbf42e Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139800 Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-761: Add CL/NEON ColorConvert benchmark testsSanghoon Lee
Change-Id: I25b42c5b74567bb50444da393d723b7f5cedafad Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/138694 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-1390: OCLGrind and benchmark tests fail for QASYMM8Georgios Pinitas
COMPMID-1392: OCLGrind failures in im2col1x1_stridex1_dchw COMPMID-1395: OCLGrind failures in output_stage_quantized Change-Id: I35504bd1f701316df122be52d458c71bbd7e7909 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139722 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-1381: Cleaned up the AssemblyHelper interfaceAnthony Barbier
Introduced a new IFunction for when we'll fork the arm_gemm functions Increased encapsulation and abstraction of which method is used Change-Id: I5fd8b14b5c77e7f8ecb09029b5e2eccd10dbdcf4 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139108 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Reviewed-by: Pablo Tello <pablo.tello@arm.com>
2018-11-02COMPMID-1226 Extend CLMeanStdDev to support FP32 / FP16Michalis Spyrou
- Extend support for FP16 in CLReduction. - For F16/F32 MeanStdDev we perform one reduction operation for mean and one for stddev and we calculate the final result in the host CPU. Change-Id: Iad2099f26c0ba7969737d22f00c6c275634d875c Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/135870 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
2018-11-02COMPMID-1388: Change default CLTuner to the one for the detected GPUGeorgios Pinitas
Sets a default tuner for the detected target if no tuner is specified in default_init() Change-Id: I27f1b9bbc0df91c1940315c6cc9042720cd1d3fe Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139630 Reviewed-by: Anthony Barbier <anthony.barbier@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-1396: make validate_example and benchmark_example check if setup was ↵Anthony Barbier
successful Change-Id: I70dbd9f9258174fc80928906c33abecf804284c3 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139716 Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-1167: Validation for NEDepthwiseConvolutionLayerAbe Mbise
Change-Id: I9689e1a0627dc015dd2ce98417e4c97bb55581bb Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/131327 Reviewed-by: Anthony Barbier <anthony.barbier@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-1188 : Rename TNOX to Mali-G76Georgios Pinitas
Change-Id: I136f7aa4bca268abd4fbe4f6ce4bcc2708ec3671 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139689 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-1310: Create graph validation executables.Georgios Pinitas
Change-Id: I9e0b57b1b83fe5a95777cdaeddba6ecef650bafc Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/138697 Reviewed-by: Anthony Barbier <anthony.barbier@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-1271: New system for GEMM heuristicsDavid Mansell
This patch implements a system for separating the "validity" from "preferred" aspect of the current heuristics in gemm_*.cpp. Now, each gemm_*.cpp defines a list of candidate implementations, each of which supplies an is_valid() function (to check for validity), an is_preferred() function (the "heuristic" part), and an instantiate() function which actually produces the GemmCommon object pointer. The actual gemm() function is now templated and uses this list to select an implementation. This patch also implements a mechanism to identify the preferred implementation, and override it via the GemmConfig structure. Change-Id: Id49ab7af8bf2e3e9fd951a9698883ade234d40e1 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139120 Reviewed-by: Anthony Barbier <anthony.barbier@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-1384: graph_mobilenet fails for NHWC on OpenCLGeorgios Pinitas
Makes GEMM3D account top padding when jumping among planes. Change-Id: Ia7c16cfa5498de106774ce42cbc4716e9f43195b Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139612 Reviewed-by: Giorgio Arena <giorgio.arena@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-1383: OCLGrind failure in CLDepthwiseConvolution 3x3 stride 1 NHWCGeorgios Pinitas
Seems OCLGrind to operate wrongly on some intrinsics when there is a mixture of vectors and scalars passed to it. Change-Id: I9e3782e739603ec59bacc3c77d91a70b1899fe3e Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139474 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-1246: Fixed typo in MidgardTuner.h filenameAnthony Barbier
Change-Id: Iade03ee67939f15a455723346a4ee0a890a8278e Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139539 Reviewed-by: Pablo Tello <pablo.tello@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-1387: Removed unnecessary extended_atomics extension for ↵Anthony Barbier
clMeanStdDevKernel Change-Id: I47e55d61a9d9464ef63b0da890e06f8a0b434796 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139527 Reviewed-by: Pablo Tello <pablo.tello@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-1379: Created WindowIterator and TensorAccessorAnthony Barbier
- WindowIterator: used to iterate over arbitrary positions of a window. (More flexible than execute_window_loop which only can iterate over entire dimensions) - TensorAccessor: RSH's code uses pointers to specialised types and strides in element sizes, this helps interfacing with their code. Change-Id: I8ded8758d345668804873409f949b8cec694d289 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139082 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
2018-11-02COMPMID-1124 : Fixes in CLLSTM layerGeorgios Pinitas
Change-Id: Ifc8e12c296d3ef2bf8e0f0bf1b87b7fd47a1fad7 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139248 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Ruomei Yan <ruomei.yan@arm.com> Reviewed-by: Michalis Spyrou <michalis.spyrou@arm.com>
2018-11-02COMPMID-1246 - Adding QASYMM8 on CL Convolution Layer benchmarkGian Marco Iodice
Change-Id: I3d6eb2e4f7c39fa6ddadfe231ba304a325f6a2d2 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139423 Reviewed-by: Pablo Tello <pablo.tello@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-1340 - Implementing Winograd Convolution Layer 1x5/5x1 on OpenCL NHWCGian Marco Iodice
Change-Id: Id5e0795238f77c049df9c109dafc5ef878c1897d Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139234 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Giorgio Arena <giorgio.arena@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02MLCE-19: Guidance in subtensor useGeorgios Pinitas
Change-Id: I1884408981b5a7b28db1cee7827422b52b72bb41 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139344 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Matteo Martincigh <matteo.martincigh@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-1378: Graph: bias in FullyConnectedLayer should always have data ↵Michele Di Giorgio
type S32 Change-Id: Ib4a82b11e810664c663dfc66281ff6bf44d2aa6b Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139029 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-1349: Add support for QASYMM8 LOGISTIC activation in CLActivationLayerMichele Di Giorgio
Change-Id: Ibabce61cf5427de80078a6468023bed05f5e7c2c Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139006 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-1380: Pre-work for SVE support.David Mansell
This patch makes the needed infrastructure changes to allow SVE kernels to be added later on. Change-Id: Ide5bccac2f47278e93fff3d648231aee2d5f8c2e Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/139070 Reviewed-by: Anthony Barbier <anthony.barbier@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-566: Implement reference and CL/NEON validation for ColorConvert ↵Sanghoon Lee
(part 2) Change-Id: I4371c4b6403f55ea7d7baf39df91a45f45b21d5a Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/137975 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-1337 Implementing Winograd Convolution Layer 1x3 and 3x1 kernels on ↵Giorgio Arena
OpenCL NHWC Change-Id: Ia07e0dfcbcd07366c4bcb956e298369fb12a0369 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/138759 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
2018-11-02COMPMID-1339 - Implementing Winograd Convolution Layer 1x5 and 5x1 kernels ↵Gian Marco Iodice
on OpenCL NCHW Change-Id: Ia293cd89651146a0e27e5f7c74ca9c924807e83c Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/138707 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
2018-11-02COMPMID-1246 Change CLLSTM in order to match android testsMichalis Spyrou
Allow cell to input weights to be nullptr if CIFG and peephole are both enabled. Change-Id: I6df705d69551f0fddeedd41b2044278d4575469c Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/137902 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
2018-11-02COMPMID-1354 Add support for NHWC in CLBatchNormalization's documentationGiorgio Arena
Change-Id: I562daf5dc6084ba597e930ae1e4b79e62f7395a7 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/138902 Reviewed-by: Michele DiGiorgio <michele.digiorgio@arm.com> Tested-by: Jenkins <bsgcomp@arm.com>
2018-11-02COMPMID-1246 Fix CLDeconvolution arguments when calculatingMichalis Spyrou
the output dimension in configure Change-Id: I0e5044e5596c065647d0e913fe9122bf6fd9e205 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/138861 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>