aboutsummaryrefslogtreecommitdiff
path: root/src/core/GLES_COMPUTE
AgeCommit message (Collapse)Author
2018-11-02APPBROWSER-364: Rewrite the DepthConcatLayer.cs with the new common codezhenglin
Change-Id: Ib19469bf88aeaf8f4adad8fd6fdd8e203f33327f Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/114601 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Joel Liang <joel.liang@arm.com> Reviewed-by: Pablo Tello <pablo.tello@arm.com>
2018-11-02APPBROWSER-341: Add ResizeNearest operator supportFrank Lei
Change-Id: I1eefcb89b2185514f6eba8f3d0e7b126b20f7d8f Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/113856 Reviewed-by: Joel Liang <joel.liang@arm.com> Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Pablo Tello <pablo.tello@arm.com>
2018-11-02APPBROWSER-358: Fix shader compilation error in matrix additionFrank Lei
Change-Id: Ibe8f9302690759477b64ed2eaa52583809c5b3d0 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/114565 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Stephen Li <stephen.li@arm.com> Reviewed-by: Pablo Tello <pablo.tello@arm.com>
2018-11-02Fix destruction order of singleton objects and resolve hang GLES issue.Ioan-Cristian Szabo
Change-Id: I9df42d4255b371b275a74aa521406e6796ee436b Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/113901 Reviewed-by: Anthony Barbier <anthony.barbier@arm.com> Tested-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02APPBROWSER-304,342: Add exclude padding support for OpenGL ES implementation ↵Xinghang Zhou
and implement MaxPool operators Change-Id: Ie6ba36ff114feec2a21739dba11bbb60b76af443 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/113697 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Stephen Li <stephen.li@arm.com> Reviewed-by: Pablo Tello <pablo.tello@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02APPBROWSER-344: Change data type of the SSBOs to improve the Softmax performanceJoel Liang
Change-Id: I012a6ededcedcc80f1fc11b37ee097ae1f59a737 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/113226 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Stephen Li <stephen.li@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02APPBROWSER-340: Implement NormalizePlanarYUV operatorzhenglin
Change-Id: I943fe27ed83e6c499d8f9919c7083f39f6f25acc Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/113211 Tested-by: Jenkins <bsgcomp@arm.com> Reviewed-by: Stephen Li <stephen.li@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-556: Don't call dlcloseAnthony Barbier
Calling dlclose on a handle meanse the library will get unloaded and therefore its symbols won't be valid anymore. Shared libraries are automatically unloaded when the application exits, so we can just 'leak' the handles. https://stackoverflow.com/questions/26131102/is-it-safe-to-call-dlclose-after-dlsym Change-Id: I8991af0887e5e48108fe787a77e003dd7c872361 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/113264 Reviewed-by: Michalis Spyrou <michalis.spyrou@arm.com> Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com> Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com>
2018-11-02COMPMID-741 - Changelog for major release 17.12Gian Marco
Change-Id: If71190e398217ca6ca44df822554998d047b79db Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/112716 Reviewed-by: Anthony Barbier <anthony.barbier@arm.com> Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com>
2018-11-02APPBROWSER-323: Transpose performance optimizationsteli01
Change-Id: Ib678dee9de43690e4cfb7be1e7ccf7a7ab38233d Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/112085 Reviewed-by: Joel Liang <joel.liang@arm.com> Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02APPBROWSER-289 DirectConvolution1*1 optimization for FP16, and fix a typo in ↵Xinghang Zhou
PoolingLayer Change-Id: I9cd3d872e353a9a404ab1d188d0d48a0965c5916 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/112047 Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com> Reviewed-by: Joel Liang <joel.liang@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02APPBROWSER-313: Performance improvement for softmax layerJoel Liang
Process 8 elements at one time for better performance Change-Id: I90d31e5d0834c5096fdb82f174482ade762b63d2 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/111840 Reviewed-by: Stephen Li <stephen.li@arm.com> Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02APPBROWSER-306, APPBROWSER-298: Code refactoring for fill_border.csJoel Liang
Change-Id: Ibc2ace55e17aa97259220d0411ce10a9c7f547f8 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/112050 Reviewed-by: Stephen Li <stephen.li@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com> Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com>
2018-11-02APPBROWSER-332 Code refactoring for batchnormalization_layer.cszhenglin
Change-Id: Ib695e7551994a10355c823840d3fb6237aef0a65 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/112054 Reviewed-by: Joel Liang <joel.liang@arm.com> Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02APPBROWSER-312 Fully connected performance optimizationFrank Lei
Change-Id: Ie93fd630ebbad7b6ca8812cb5044b3f1908b45fd Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/111830 Reviewed-by: Stephen Li <stephen.li@arm.com> Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02APPBROWSER-290: DC5x5 optimization for FP16ASIAPAC\steli01
Change-Id: I0833437d8353515a29aad69c15c99c7b0fd65156 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/111631 Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com> Reviewed-by: Joel Liang <joel.liang@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02APPBROWSER-314: Performance optimazation for BatchNormalizationLayerzhenglin
Change-Id: Ie3ad9abb64e90720609bb6e67662eaf9dd4f3689 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/111826 Reviewed-by: Joel Liang <joel.liang@arm.com> Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com> (cherry picked from commit 02c1fa663926cc4fcd1995d4d18d7528e0c85d94) Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/111834 Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02APPBROWSER-294: Fix the validation failure issue of DepthConcatLayerzhenglin
Change-Id: I3ee0943eb9177badc17e6419984a50c21152c2c0 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/111462 Reviewed-by: Joel Liang <joel.liang@arm.com> Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02APPBROWSER-324 FP16 Weights optimization in DirectConvolution1x1zhenglin
Change-Id: I94f09caa3d9852b7ecd1d3716ba7853de9dc1c43 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/111636 Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com> Reviewed-by: Joel Liang <joel.liang@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02APPBROWSER-331: Code refactoring for absdiff shader and fix the example ↵Joel Liang
running issue The 8-bits unsigned integer pack/unpack functions have been removed accidentally. Added new pack/unpack functions to the new common shader code. Change-Id: I0ef4507b1758f29686b1dadb76781e4f1f220249 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/111637 Reviewed-by: Stephen Li <stephen.li@arm.com> Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-692 Consistent names for the interfacesGiorgio Arena
Change-Id: I4b1f3f0da9ff5342c7de7083736fe91871d14e5b Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/110351 Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02APPBROWSER-330: Workaround for the free unallocated value issue on AndroidJoel Liang
Added a default destructor in cpp file to workaround the free unallocated value issue on Android Change-Id: I3db84a0504a51b9b90823cbb37aa1cace06870e3 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/111438 Reviewed-by: Frank Lei <frank.lei@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com> Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com>
2018-11-02APPBROWSER-329: Fix the adjusting group size issueszhenglin
Change-Id: I2171b79e85eb316bf6bfb09ab8a54880b7708a4c Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/111435 Reviewed-by: Stephen Li <stephen.li@arm.com> Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com> Reviewed-by: Joel Liang <joel.liang@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02COMPMID-683: Normalization layer API clarification.Georgios Pinitas
Adds a is_scaled parameter in the NormalizationLayerInfo that flags if the alpha parameter should be scaled by the normalization size of not. Unscaled parameter is used by [Krichevksy 2012] which is used in AndroidNN and TensorFlow LRN layer. Change-Id: Iad2aa5e688cf4dcd6cc77a6e28c0663764f34ccb Reviewed-on: http://mpd-gerrit.cambridge.arm.com/96102 Reviewed-by: Diego Lopez Recas <diego.lopezrecas@arm.com> Reviewed-by: Anthony Barbier <anthony.barbier@arm.com> Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
2018-11-02APPBROWSER-298: Workaround Bifrost shader compiler issueJoel Liang
On Bifrost platform, we got this compile error for statement "data = unpack8_half(src_ptr[0])" Error: Compiler log: 2:532: S0001: Function call discards 'readonly' access qualifier. And we need to change it to "data = unpack8_half(uvec4(src_ptr[0]))" Change-Id: I1b4cc66f88a57cf06275a74c796e89ed6043d2e5 Reviewed-on: http://mpd-gerrit.cambridge.arm.com/95773 Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com> Reviewed-by: Stephen Li <stephen.li@arm.com> (cherry picked from commit b2572a8de01cddc7adeacb2eed6a7ba7b8a6cc2d) Reviewed-on: http://mpd-gerrit.cambridge.arm.com/95952 Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
2018-11-02APPBROWSER-298, APPBROWSER-306: Reimplement the common code of compute shaderJoel Liang
The new common code of compute shader is in file helpers_cs.h Rewrite the direct_convolution1x1.cs and softmax_layer.cs to use the new common code. It will also remove the dependence of the token pasting operator (##). We'll remove the "##" support after we rewrite all of the compute shader code. Change-Id: Icd8553ef6b61ad484a8507590ac8ed499bd47061 Reviewed-on: http://mpd-gerrit.cambridge.arm.com/95455 Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com> Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com> Reviewed-by: Frank Lei <frank.lei@arm.com> (cherry picked from commit 0a4f83570d261f839d9866b68979efe8d7a95883) Reviewed-on: http://mpd-gerrit.cambridge.arm.com/95601 Reviewed-by: Jim He <jim.he@arm.com>
2018-11-02COMPMID-671: Add global pooling layer support.Georgios Pinitas
Change-Id: Iead7497cc03e1e7bde440d2965a7bf54cbfa88bf Reviewed-on: http://mpd-gerrit.cambridge.arm.com/95579 Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com> Reviewed-by: Joel Liang <joel.liang@arm.com> Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
2018-11-02COMPMID-631: Merge branches/gles_compute branchAnthony Barbier
Last commit: commit b25c5f68042b0c81bf611d59a1bb8535e1c42497 Author: Xinghang Zhou <xinghang.zhou@arm.com> Date: Wed Oct 25 18:48:10 2017 +0800 Synced validation's tolerances of GCSoftmax from cl side Change-Id: Ibe72054205c1c8721845d679a31af7ed0a7c5cf6 Reviewed-on: http://mpd-gerrit.cambridge.arm.com/93283 Reviewed-by: Anthony Barbier <anthony.barbier@arm.com> Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>