| <!-- HTML header for doxygen 1.8.15--> |
| <!-- Remember to use version doxygen 1.8.15 +--> |
| <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> |
| <html xmlns="http://www.w3.org/1999/xhtml"> |
| <head> |
| <meta http-equiv="Content-Type" content="text/xhtml;charset=UTF-8"/> |
| <meta http-equiv="X-UA-Compatible" content="IE=9"/> |
| <meta name="generator" content="Doxygen 1.8.15"/> |
| <meta name="robots" content="NOINDEX, NOFOLLOW" /> <!-- Prevent indexing by search engines --> |
| <title>Compute Library: src/runtime/CL/functions/CLGEMMLowpMatrixMultiplyCore.cpp Source File</title> |
| <link href="tabs.css" rel="stylesheet" type="text/css"/> |
| <script type="text/javascript" src="jquery.js"></script> |
| <script type="text/javascript" src="dynsections.js"></script> |
| <link href="navtree.css" rel="stylesheet" type="text/css"/> |
| <script type="text/javascript" src="resize.js"></script> |
| <script type="text/javascript" src="navtreedata.js"></script> |
| <script type="text/javascript" src="navtree.js"></script> |
| <script type="text/javascript"> |
| /* @license magnet:?xt=urn:btih:cf05388f2679ee054f2beb29a391d25f4e673ac3&dn=gpl-2.0.txt GPL-v2 */ |
| $(document).ready(initResizable); |
| /* @license-end */</script> |
| <link href="search/search.css" rel="stylesheet" type="text/css"/> |
| <script type="text/javascript" src="search/searchdata.js"></script> |
| <script type="text/javascript" src="search/search.js"></script> |
| <script type="text/x-mathjax-config"> |
| MathJax.Hub.Config({ |
| extensions: ["tex2jax.js"], |
| jax: ["input/TeX","output/HTML-CSS"], |
| }); |
| </script><script type="text/javascript" async="async" src="http://cdn.mathjax.org/mathjax/latest/MathJax.js"></script> |
| <link href="doxygen.css" rel="stylesheet" type="text/css" /> |
| <link href="stylesheet.css" rel="stylesheet" type="text/css"/> |
| </head> |
| <body> |
| <div id="top"><!-- do not remove this div, it is closed by doxygen! --> |
| <div id="titlearea"> |
| <table cellspacing="0" cellpadding="0"> |
| <tbody> |
| <tr style="height: 56px;"> |
| <img alt="Compute Library" src="https://raw.githubusercontent.com/ARM-software/ComputeLibrary/gh-pages/ACL_logo.png" style="max-width: 100%;margin-top: 15px;margin-left: 10px"/> |
| <td style="padding-left: 0.5em;"> |
| <div id="projectname"> |
|  <span id="projectnumber">20.02.1</span> |
| </div> |
| </td> |
| </tr> |
| </tbody> |
| </table> |
| </div> |
| <!-- end header part --> |
| <!-- Generated by Doxygen 1.8.15 --> |
| <script type="text/javascript"> |
| /* @license magnet:?xt=urn:btih:cf05388f2679ee054f2beb29a391d25f4e673ac3&dn=gpl-2.0.txt GPL-v2 */ |
| var searchBox = new SearchBox("searchBox", "search",false,'Search'); |
| /* @license-end */ |
| </script> |
| <script type="text/javascript" src="menudata.js"></script> |
| <script type="text/javascript" src="menu.js"></script> |
| <script type="text/javascript"> |
| /* @license magnet:?xt=urn:btih:cf05388f2679ee054f2beb29a391d25f4e673ac3&dn=gpl-2.0.txt GPL-v2 */ |
| $(function() { |
| initMenu('',true,false,'search.php','Search'); |
| $(document).ready(function() { init_search(); }); |
| }); |
| /* @license-end */</script> |
| <div id="main-nav"></div> |
| </div><!-- top --> |
| <div id="side-nav" class="ui-resizable side-nav-resizable"> |
| <div id="nav-tree"> |
| <div id="nav-tree-contents"> |
| <div id="nav-sync" class="sync"></div> |
| </div> |
| </div> |
| <div id="splitbar" style="-moz-user-select:none;" |
| class="ui-resizable-handle"> |
| </div> |
| </div> |
| <script type="text/javascript"> |
| /* @license magnet:?xt=urn:btih:cf05388f2679ee054f2beb29a391d25f4e673ac3&dn=gpl-2.0.txt GPL-v2 */ |
| $(document).ready(function(){initNavTree('_c_l_g_e_m_m_lowp_matrix_multiply_core_8cpp_source.xhtml','');}); |
| /* @license-end */ |
| </script> |
| <div id="doc-content"> |
| <!-- window showing the filter options --> |
| <div id="MSearchSelectWindow" |
| onmouseover="return searchBox.OnSearchSelectShow()" |
| onmouseout="return searchBox.OnSearchSelectHide()" |
| onkeydown="return searchBox.OnSearchSelectKey(event)"> |
| </div> |
| |
| <!-- iframe showing the search results (closed by default) --> |
| <div id="MSearchResultsWindow"> |
| <iframe src="javascript:void(0)" frameborder="0" |
| name="MSearchResults" id="MSearchResults"> |
| </iframe> |
| </div> |
| |
| <div class="header"> |
| <div class="headertitle"> |
| <div class="title">CLGEMMLowpMatrixMultiplyCore.cpp</div> </div> |
| </div><!--header--> |
| <div class="contents"> |
| <a href="_c_l_g_e_m_m_lowp_matrix_multiply_core_8cpp.xhtml">Go to the documentation of this file.</a><div class="fragment"><div class="line"><a name="l00001"></a><span class="lineno"> 1</span> <span class="comment">/*</span></div><div class="line"><a name="l00002"></a><span class="lineno"> 2</span> <span class="comment"> * Copyright (c) 2017-2020 ARM Limited.</span></div><div class="line"><a name="l00003"></a><span class="lineno"> 3</span> <span class="comment"> *</span></div><div class="line"><a name="l00004"></a><span class="lineno"> 4</span> <span class="comment"> * SPDX-License-Identifier: MIT</span></div><div class="line"><a name="l00005"></a><span class="lineno"> 5</span> <span class="comment"> *</span></div><div class="line"><a name="l00006"></a><span class="lineno"> 6</span> <span class="comment"> * Permission is hereby granted, free of charge, to any person obtaining a copy</span></div><div class="line"><a name="l00007"></a><span class="lineno"> 7</span> <span class="comment"> * of this software and associated documentation files (the "Software"), to</span></div><div class="line"><a name="l00008"></a><span class="lineno"> 8</span> <span class="comment"> * deal in the Software without restriction, including without limitation the</span></div><div class="line"><a name="l00009"></a><span class="lineno"> 9</span> <span class="comment"> * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or</span></div><div class="line"><a name="l00010"></a><span class="lineno"> 10</span> <span class="comment"> * sell copies of the Software, and to permit persons to whom the Software is</span></div><div class="line"><a name="l00011"></a><span class="lineno"> 11</span> <span class="comment"> * furnished to do so, subject to the following conditions:</span></div><div class="line"><a name="l00012"></a><span class="lineno"> 12</span> <span class="comment"> *</span></div><div class="line"><a name="l00013"></a><span class="lineno"> 13</span> <span class="comment"> * The above copyright notice and this permission notice shall be included in all</span></div><div class="line"><a name="l00014"></a><span class="lineno"> 14</span> <span class="comment"> * copies or substantial portions of the Software.</span></div><div class="line"><a name="l00015"></a><span class="lineno"> 15</span> <span class="comment"> *</span></div><div class="line"><a name="l00016"></a><span class="lineno"> 16</span> <span class="comment"> * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR</span></div><div class="line"><a name="l00017"></a><span class="lineno"> 17</span> <span class="comment"> * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,</span></div><div class="line"><a name="l00018"></a><span class="lineno"> 18</span> <span class="comment"> * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE</span></div><div class="line"><a name="l00019"></a><span class="lineno"> 19</span> <span class="comment"> * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER</span></div><div class="line"><a name="l00020"></a><span class="lineno"> 20</span> <span class="comment"> * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,</span></div><div class="line"><a name="l00021"></a><span class="lineno"> 21</span> <span class="comment"> * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE</span></div><div class="line"><a name="l00022"></a><span class="lineno"> 22</span> <span class="comment"> * SOFTWARE.</span></div><div class="line"><a name="l00023"></a><span class="lineno"> 23</span> <span class="comment"> */</span></div><div class="line"><a name="l00024"></a><span class="lineno"> 24</span> <span class="preprocessor">#include "<a class="code" href="_c_l_g_e_m_m_lowp_matrix_multiply_core_8h.xhtml">arm_compute/runtime/CL/functions/CLGEMMLowpMatrixMultiplyCore.h</a>"</span></div><div class="line"><a name="l00025"></a><span class="lineno"> 25</span> </div><div class="line"><a name="l00026"></a><span class="lineno"> 26</span> <span class="preprocessor">#include "<a class="code" href="_i_c_l_tensor_8h.xhtml">arm_compute/core/CL/ICLTensor.h</a>"</span></div><div class="line"><a name="l00027"></a><span class="lineno"> 27</span> <span class="preprocessor">#include "<a class="code" href="_c_l_g_e_m_m_native_kernel_configuration_8h.xhtml">arm_compute/core/CL/gemm/native/CLGEMMNativeKernelConfiguration.h</a>"</span></div><div class="line"><a name="l00028"></a><span class="lineno"> 28</span> <span class="preprocessor">#include "<a class="code" href="_c_l_g_e_m_m_reshaped_only_r_h_s_kernel_configuration_8h.xhtml">arm_compute/core/CL/gemm/reshaped_only_rhs/CLGEMMReshapedOnlyRHSKernelConfiguration.h</a>"</span></div><div class="line"><a name="l00029"></a><span class="lineno"> 29</span> <span class="preprocessor">#include "<a class="code" href="_error_8h.xhtml">arm_compute/core/Error.h</a>"</span></div><div class="line"><a name="l00030"></a><span class="lineno"> 30</span> <span class="preprocessor">#include "<a class="code" href="arm__compute_2core_2_helpers_8h.xhtml">arm_compute/core/Helpers.h</a>"</span></div><div class="line"><a name="l00031"></a><span class="lineno"> 31</span> <span class="preprocessor">#include "<a class="code" href="_tensor_info_8h.xhtml">arm_compute/core/TensorInfo.h</a>"</span></div><div class="line"><a name="l00032"></a><span class="lineno"> 32</span> <span class="preprocessor">#include "<a class="code" href="arm__compute_2core_2_types_8h.xhtml">arm_compute/core/Types.h</a>"</span></div><div class="line"><a name="l00033"></a><span class="lineno"> 33</span> <span class="preprocessor">#include "<a class="code" href="_validate_8h.xhtml">arm_compute/core/Validate.h</a>"</span></div><div class="line"><a name="l00034"></a><span class="lineno"> 34</span> <span class="preprocessor">#include "<a class="code" href="_shape_calculator_8h.xhtml">arm_compute/core/utils/misc/ShapeCalculator.h</a>"</span></div><div class="line"><a name="l00035"></a><span class="lineno"> 35</span> <span class="preprocessor">#include "<a class="code" href="_asymm_helpers_8h.xhtml">arm_compute/core/utils/quantization/AsymmHelpers.h</a>"</span></div><div class="line"><a name="l00036"></a><span class="lineno"> 36</span> <span class="preprocessor">#include "<a class="code" href="_c_l_scheduler_8h.xhtml">arm_compute/runtime/CL/CLScheduler.h</a>"</span></div><div class="line"><a name="l00037"></a><span class="lineno"> 37</span> </div><div class="line"><a name="l00038"></a><span class="lineno"> 38</span> <span class="keyword">namespace </span><a class="code" href="namespacearm__compute.xhtml">arm_compute</a></div><div class="line"><a name="l00039"></a><span class="lineno"> 39</span> {</div><div class="line"><a name="l00040"></a><span class="lineno"> 40</span> <span class="keyword">using namespace </span><a class="code" href="namespacearm__compute_1_1misc_1_1shape__calculator.xhtml">arm_compute::misc::shape_calculator</a>;</div><div class="line"><a name="l00041"></a><span class="lineno"> 41</span> <span class="keyword">using namespace </span><a class="code" href="namespacearm__compute_1_1cl__gemm.xhtml">arm_compute::cl_gemm</a>;</div><div class="line"><a name="l00042"></a><span class="lineno"> 42</span> </div><div class="line"><a name="l00043"></a><span class="lineno"> 43</span> <span class="keyword">namespace</span></div><div class="line"><a name="l00044"></a><span class="lineno"> 44</span> {</div><div class="line"><a name="l00045"></a><span class="lineno"> 45</span> <span class="keyword">inline</span> <span class="keywordtype">bool</span> is_gemm_reshaped(<span class="keywordtype">bool</span> reshape_b_only_on_first_run, <a class="code" href="namespacearm__compute.xhtml#a735ac6c2a02e320969625308810444f3">GPUTarget</a> gpu_target)</div><div class="line"><a name="l00046"></a><span class="lineno"> 46</span> {</div><div class="line"><a name="l00047"></a><span class="lineno"> 47</span>  <span class="keywordflow">return</span> (<a class="code" href="namespacearm__compute.xhtml#a2355c2bf5d1950088937416baea24fe6">get_arch_from_target</a>(gpu_target) != <a class="code" href="namespacearm__compute.xhtml#a735ac6c2a02e320969625308810444f3afa362f4daec88442a387ff7cda411a38">GPUTarget::MIDGARD</a>) && (reshape_b_only_on_first_run);</div><div class="line"><a name="l00048"></a><span class="lineno"> 48</span> }</div><div class="line"><a name="l00049"></a><span class="lineno"> 49</span> } <span class="comment">// namespace</span></div><div class="line"><a name="l00050"></a><span class="lineno"> 50</span> </div><div class="line"><a name="l00051"></a><span class="lineno"><a class="line" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_core.xhtml#a1c3a177d5a3188691c9946e1a78839d0"> 51</a></span> <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_core.xhtml#a1c3a177d5a3188691c9946e1a78839d0">CLGEMMLowpMatrixMultiplyCore::CLGEMMLowpMatrixMultiplyCore</a>(std::shared_ptr<IMemoryManager> memory_manager)</div><div class="line"><a name="l00052"></a><span class="lineno"> 52</span>  : _memory_group(std::move(memory_manager)),</div><div class="line"><a name="l00053"></a><span class="lineno"> 53</span>  _weights_to_qasymm8(),</div><div class="line"><a name="l00054"></a><span class="lineno"> 54</span>  _mm_midgard_kernel(),</div><div class="line"><a name="l00055"></a><span class="lineno"> 55</span>  _mm_native_kernel(),</div><div class="line"><a name="l00056"></a><span class="lineno"> 56</span>  _mm_reshaped_only_rhs_kernel(),</div><div class="line"><a name="l00057"></a><span class="lineno"> 57</span>  _mtx_b_reshape_kernel(),</div><div class="line"><a name="l00058"></a><span class="lineno"> 58</span>  _mtx_a_reduction_kernel(),</div><div class="line"><a name="l00059"></a><span class="lineno"> 59</span>  _mtx_b_reduction_kernel(),</div><div class="line"><a name="l00060"></a><span class="lineno"> 60</span>  _offset_contribution_kernel(),</div><div class="line"><a name="l00061"></a><span class="lineno"> 61</span>  _offset_contribution_output_stage_kernel(),</div><div class="line"><a name="l00062"></a><span class="lineno"> 62</span>  _qasymm8_weights(),</div><div class="line"><a name="l00063"></a><span class="lineno"> 63</span>  _vector_sum_col(),</div><div class="line"><a name="l00064"></a><span class="lineno"> 64</span>  _vector_sum_row(),</div><div class="line"><a name="l00065"></a><span class="lineno"> 65</span>  _tmp_b(),</div><div class="line"><a name="l00066"></a><span class="lineno"> 66</span>  _mm_result_s32(),</div><div class="line"><a name="l00067"></a><span class="lineno"> 67</span>  _gemm_output_stage_multipliers(),</div><div class="line"><a name="l00068"></a><span class="lineno"> 68</span>  _gemm_output_stage_shifts(),</div><div class="line"><a name="l00069"></a><span class="lineno"> 69</span>  _matrix_a(nullptr),</div><div class="line"><a name="l00070"></a><span class="lineno"> 70</span>  _original_b(nullptr),</div><div class="line"><a name="l00071"></a><span class="lineno"> 71</span>  _output(nullptr),</div><div class="line"><a name="l00072"></a><span class="lineno"> 72</span>  _a_offset(0),</div><div class="line"><a name="l00073"></a><span class="lineno"> 73</span>  _b_offset(0),</div><div class="line"><a name="l00074"></a><span class="lineno"> 74</span>  _is_gemm_reshaped(true),</div><div class="line"><a name="l00075"></a><span class="lineno"> 75</span>  _is_midgard(false),</div><div class="line"><a name="l00076"></a><span class="lineno"> 76</span>  _reshape_b_only_on_first_run(false),</div><div class="line"><a name="l00077"></a><span class="lineno"> 77</span>  _is_prepared(false),</div><div class="line"><a name="l00078"></a><span class="lineno"> 78</span>  _fuse_output_stage(false),</div><div class="line"><a name="l00079"></a><span class="lineno"> 79</span>  _convert_to_qasymm8(false)</div><div class="line"><a name="l00080"></a><span class="lineno"> 80</span> {</div><div class="line"><a name="l00081"></a><span class="lineno"> 81</span> }</div><div class="line"><a name="l00082"></a><span class="lineno"> 82</span> </div><div class="line"><a name="l00083"></a><span class="lineno"><a class="line" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_core.xhtml#a0ae54876c8c3394f2e207f6b884f2b58"> 83</a></span> <span class="keywordtype">void</span> <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_core.xhtml#a0ae54876c8c3394f2e207f6b884f2b58">CLGEMMLowpMatrixMultiplyCore::configure</a>(<span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *a, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#aa76b4a6e74940dabc5b7fc6b2dab3545">b</a>, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *c, <a class="code" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *output, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_g_e_m_m_info.xhtml">GEMMInfo</a> &gemm_info)</div><div class="line"><a name="l00084"></a><span class="lineno"> 84</span> {</div><div class="line"><a name="l00085"></a><span class="lineno"> 85</span>  <a class="code" href="_validate_8h.xhtml#a921b705e9e3e0fe928928447869e62a5">ARM_COMPUTE_ERROR_ON_NULLPTR</a>(a, <a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#aa76b4a6e74940dabc5b7fc6b2dab3545">b</a>, output);</div><div class="line"><a name="l00086"></a><span class="lineno"> 86</span>  <a class="code" href="_error_8h.xhtml#a938dcd406ce611ef5345ad2531cdb948">ARM_COMPUTE_ERROR_THROW_ON</a>(<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_core.xhtml#a8c3cf2d65afb288e39909171ada19566">CLGEMMLowpMatrixMultiplyCore::validate</a>(a-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>(), <a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#aa76b4a6e74940dabc5b7fc6b2dab3545">b</a>->info(), c != <span class="keyword">nullptr</span> ? c-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>() : <span class="keyword">nullptr</span>, output-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>(), gemm_info));</div><div class="line"><a name="l00087"></a><span class="lineno"> 87</span> </div><div class="line"><a name="l00088"></a><span class="lineno"> 88</span>  _is_prepared = <span class="keyword">false</span>;</div><div class="line"><a name="l00089"></a><span class="lineno"> 89</span>  _original_b = <a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#aa76b4a6e74940dabc5b7fc6b2dab3545">b</a>;</div><div class="line"><a name="l00090"></a><span class="lineno"> 90</span>  _reshape_b_only_on_first_run = gemm_info.<a class="code" href="classarm__compute_1_1_g_e_m_m_info.xhtml#a4c8f9fa843de1086c27c86a6b8cf4582">reshape_b_only_on_first_run</a>();</div><div class="line"><a name="l00091"></a><span class="lineno"> 91</span>  _a_offset = a-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>()-><a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml#a3f3e1a3200223e6a304a533b1016e749">quantization_info</a>().<a class="code" href="classarm__compute_1_1_quantization_info.xhtml#a706fc156bcd4c45441bcaad05884b57d">uniform</a>().<a class="code" href="structarm__compute_1_1_uniform_quantization_info.xhtml#a97bd6c077f3c7769f575b82988b9b668">offset</a>;</div><div class="line"><a name="l00092"></a><span class="lineno"> 92</span>  _matrix_a = a;</div><div class="line"><a name="l00093"></a><span class="lineno"> 93</span>  _output = output;</div><div class="line"><a name="l00094"></a><span class="lineno"> 94</span> </div><div class="line"><a name="l00095"></a><span class="lineno"> 95</span>  _convert_to_qasymm8 = <a class="code" href="namespacearm__compute.xhtml#a84437d80241f6a31e1a07c231ee8e3ac">is_data_type_quantized_per_channel</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#aa76b4a6e74940dabc5b7fc6b2dab3545">b</a>->info()->data_type()) && <a class="code" href="namespacearm__compute.xhtml#a922cd85c475c0d725fc030116ea57979">is_data_type_quantized_symmetric</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#aa76b4a6e74940dabc5b7fc6b2dab3545">b</a>->info()->data_type())</div><div class="line"><a name="l00096"></a><span class="lineno"> 96</span>  && <a class="code" href="namespacearm__compute.xhtml#a14f46283f316e7f0fad301d5c1507e9f">is_data_type_quantized_asymmetric</a>(a-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>()-><a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml#a7cfb31af63202568efef5214acfbf3ba">data_type</a>());</div><div class="line"><a name="l00097"></a><span class="lineno"> 97</span>  _b_offset = _convert_to_qasymm8 ? -128 : <a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#aa76b4a6e74940dabc5b7fc6b2dab3545">b</a>->info()->quantization_info().uniform().offset;</div><div class="line"><a name="l00098"></a><span class="lineno"> 98</span> </div><div class="line"><a name="l00099"></a><span class="lineno"> 99</span>  <span class="comment">// Get the GPU target</span></div><div class="line"><a name="l00100"></a><span class="lineno"> 100</span>  <span class="keyword">const</span> <a class="code" href="namespacearm__compute.xhtml#a735ac6c2a02e320969625308810444f3">GPUTarget</a> gpu_target = <a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#a9b58d0eb9a2af8e6d7908695e1557d6c">CLScheduler::get</a>().<a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#a27561688e2fc60176608ef725a4ecb30">target</a>();</div><div class="line"><a name="l00101"></a><span class="lineno"> 101</span> </div><div class="line"><a name="l00102"></a><span class="lineno"> 102</span>  <span class="comment">// Set the target for the kernels</span></div><div class="line"><a name="l00103"></a><span class="lineno"> 103</span>  _mm_midgard_kernel.<a class="code" href="classarm__compute_1_1_i_c_l_kernel.xhtml#ad5ba9d34a3a855bf1dd2e36316ff550a">set_target</a>(gpu_target);</div><div class="line"><a name="l00104"></a><span class="lineno"> 104</span>  _mm_native_kernel.<a class="code" href="classarm__compute_1_1_i_c_l_kernel.xhtml#ad5ba9d34a3a855bf1dd2e36316ff550a">set_target</a>(gpu_target);</div><div class="line"><a name="l00105"></a><span class="lineno"> 105</span>  _mm_reshaped_only_rhs_kernel.<a class="code" href="classarm__compute_1_1_i_c_l_kernel.xhtml#ad5ba9d34a3a855bf1dd2e36316ff550a">set_target</a>(gpu_target);</div><div class="line"><a name="l00106"></a><span class="lineno"> 106</span> </div><div class="line"><a name="l00107"></a><span class="lineno"> 107</span>  <a class="code" href="structarm__compute_1_1_g_e_m_m_r_h_s_matrix_info.xhtml">GEMMRHSMatrixInfo</a> rhs_info;</div><div class="line"><a name="l00108"></a><span class="lineno"> 108</span>  <a class="code" href="structarm__compute_1_1_g_e_m_m_l_h_s_matrix_info.xhtml">GEMMLHSMatrixInfo</a> lhs_info;</div><div class="line"><a name="l00109"></a><span class="lineno"> 109</span> </div><div class="line"><a name="l00110"></a><span class="lineno"> 110</span>  <span class="comment">// Arguments used by GEMMReshapeInfo</span></div><div class="line"><a name="l00111"></a><span class="lineno"> 111</span>  <span class="comment">// If we pass the matrix A and matrix B reshaped to CLGEMMMatrixMultiplyKernel, we need to pass m, n, k, mult_transpose1xW_width and mult_interleave4x4_height to CLGEMMReshapeInfo</span></div><div class="line"><a name="l00112"></a><span class="lineno"> 112</span>  <span class="comment">// in order to know how the matrices have been reshaped</span></div><div class="line"><a name="l00113"></a><span class="lineno"> 113</span>  <span class="keywordtype">bool</span> reinterpret_input_as_3d = gemm_info.<a class="code" href="classarm__compute_1_1_g_e_m_m_info.xhtml#a00330b8913cac3b07029ac0c3350e806">reinterpret_input_as_3d</a>();</div><div class="line"><a name="l00114"></a><span class="lineno"> 114</span>  <span class="keyword">const</span> <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> m = reinterpret_input_as_3d ? (a-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>()-><a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml#a178f0d3d87f959e00a743328d95359d2">dimension</a>(1) * a-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>()-><a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml#a178f0d3d87f959e00a743328d95359d2">dimension</a>(2)) : a-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>()-><a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml#a178f0d3d87f959e00a743328d95359d2">dimension</a>(1);</div><div class="line"><a name="l00115"></a><span class="lineno"> 115</span>  <span class="keyword">const</span> <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> n = <a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#aa76b4a6e74940dabc5b7fc6b2dab3545">b</a>->info()->dimension(0);</div><div class="line"><a name="l00116"></a><span class="lineno"> 116</span>  <span class="keyword">const</span> <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> k = a-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>()-><a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml#a178f0d3d87f959e00a743328d95359d2">dimension</a>(0);</div><div class="line"><a name="l00117"></a><span class="lineno"> 117</span>  <span class="keyword">const</span> <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> batch_size = reinterpret_input_as_3d ? a-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>()-><a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml#a178f0d3d87f959e00a743328d95359d2">dimension</a>(3) : a-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>()-><a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml#a178f0d3d87f959e00a743328d95359d2">dimension</a>(2);</div><div class="line"><a name="l00118"></a><span class="lineno"> 118</span>  <span class="keyword">const</span> <span class="keywordtype">int</span> depth_output_gemm3d = gemm_info.<a class="code" href="classarm__compute_1_1_g_e_m_m_info.xhtml#abbd888f118c2209bf7578eb4f8942a07">depth_output_gemm3d</a>();</div><div class="line"><a name="l00119"></a><span class="lineno"> 119</span> </div><div class="line"><a name="l00120"></a><span class="lineno"> 120</span>  <span class="comment">// Check if we need to reshape the matrix A and matrix B</span></div><div class="line"><a name="l00121"></a><span class="lineno"> 121</span>  _is_gemm_reshaped = is_gemm_reshaped(_reshape_b_only_on_first_run, gpu_target);</div><div class="line"><a name="l00122"></a><span class="lineno"> 122</span>  _is_midgard = gpu_target == <a class="code" href="namespacearm__compute.xhtml#a735ac6c2a02e320969625308810444f3afa362f4daec88442a387ff7cda411a38">GPUTarget::MIDGARD</a>;</div><div class="line"><a name="l00123"></a><span class="lineno"> 123</span> </div><div class="line"><a name="l00124"></a><span class="lineno"> 124</span>  <span class="keywordflow">if</span>(_convert_to_qasymm8)</div><div class="line"><a name="l00125"></a><span class="lineno"> 125</span>  {</div><div class="line"><a name="l00126"></a><span class="lineno"> 126</span>  <span class="comment">// Set data type for converted weights</span></div><div class="line"><a name="l00127"></a><span class="lineno"> 127</span>  <a class="code" href="classarm__compute_1_1_tensor_info.xhtml">TensorInfo</a> <a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a7cb842ebfe255726066039853a4322f0">weights_info</a>(*<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#aa76b4a6e74940dabc5b7fc6b2dab3545">b</a>->info());</div><div class="line"><a name="l00128"></a><span class="lineno"> 128</span>  <a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a7cb842ebfe255726066039853a4322f0">weights_info</a>.set_data_type(<a class="code" href="namespacearm__compute.xhtml#ad8ed01ff3ff33333d8e19db4d2818bb6af14462d71aa842202c3e4b272c7ec924">DataType::QASYMM8</a>);</div><div class="line"><a name="l00129"></a><span class="lineno"> 129</span>  _qasymm8_weights.<a class="code" href="classarm__compute_1_1_c_l_tensor.xhtml#a4083de30daebd6bdee6b35d9c8262108">allocator</a>()-><a class="code" href="classarm__compute_1_1_i_tensor_allocator.xhtml#af36143939a43fa124312e395975091ed">init</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a7cb842ebfe255726066039853a4322f0">weights_info</a>);</div><div class="line"><a name="l00130"></a><span class="lineno"> 130</span>  _weights_to_qasymm8.<a class="code" href="classarm__compute_1_1_c_l_depth_convert_layer_kernel.xhtml#a536f0de1a70ea80d40b0a24657ac5290">configure</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#aa76b4a6e74940dabc5b7fc6b2dab3545">b</a>, &_qasymm8_weights, <a class="code" href="namespacearm__compute.xhtml#a82b8ac759c804bc1fb4e2d21e178fb6fae1c8555fcf0ea2bb648a6fd527d658c0">ConvertPolicy::WRAP</a>, 0);</div><div class="line"><a name="l00131"></a><span class="lineno"> 131</span>  }</div><div class="line"><a name="l00132"></a><span class="lineno"> 132</span> </div><div class="line"><a name="l00133"></a><span class="lineno"> 133</span>  <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *matrix_b = _convert_to_qasymm8 ? &_qasymm8_weights : <a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#aa76b4a6e74940dabc5b7fc6b2dab3545">b</a>;</div><div class="line"><a name="l00134"></a><span class="lineno"> 134</span>  <span class="keywordflow">if</span>(_is_gemm_reshaped)</div><div class="line"><a name="l00135"></a><span class="lineno"> 135</span>  {</div><div class="line"><a name="l00136"></a><span class="lineno"> 136</span>  matrix_b = &_tmp_b;</div><div class="line"><a name="l00137"></a><span class="lineno"> 137</span> </div><div class="line"><a name="l00138"></a><span class="lineno"> 138</span>  <span class="keywordflow">if</span>(!_reshape_b_only_on_first_run)</div><div class="line"><a name="l00139"></a><span class="lineno"> 139</span>  {</div><div class="line"><a name="l00140"></a><span class="lineno"> 140</span>  _memory_group.<a class="code" href="classarm__compute_1_1_memory_group.xhtml#a6fc0a49304c152c20a0f6df0634fb3cd">manage</a>(&_tmp_b);</div><div class="line"><a name="l00141"></a><span class="lineno"> 141</span>  }</div><div class="line"><a name="l00142"></a><span class="lineno"> 142</span> </div><div class="line"><a name="l00143"></a><span class="lineno"> 143</span>  <span class="comment">// Pick up the GEMM configuration</span></div><div class="line"><a name="l00144"></a><span class="lineno"> 144</span>  std::tie(lhs_info, rhs_info) = <a class="code" href="classarm__compute_1_1cl__gemm_1_1_c_l_g_e_m_m_reshaped_only_r_h_s_kernel_configuration_factory.xhtml#ab2b50810acd1c74616ee97dee3cbdb6b">CLGEMMReshapedOnlyRHSKernelConfigurationFactory::create</a>(gpu_target)->configure(m, n, k, batch_size, <a class="code" href="namespacearm__compute.xhtml#ad8ed01ff3ff33333d8e19db4d2818bb6af14462d71aa842202c3e4b272c7ec924">DataType::QASYMM8</a>);</div><div class="line"><a name="l00145"></a><span class="lineno"> 145</span> </div><div class="line"><a name="l00146"></a><span class="lineno"> 146</span>  <span class="comment">// Configure reshape RHS kernel</span></div><div class="line"><a name="l00147"></a><span class="lineno"> 147</span>  _mtx_b_reshape_kernel.<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_reshape_r_h_s_matrix_kernel.xhtml#aa0b0eaffc9a83e1a6495d957b63bbccc">configure</a>(_convert_to_qasymm8 ? &_qasymm8_weights : <a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#aa76b4a6e74940dabc5b7fc6b2dab3545">b</a>, &_tmp_b, rhs_info);</div><div class="line"><a name="l00148"></a><span class="lineno"> 148</span>  }</div><div class="line"><a name="l00149"></a><span class="lineno"> 149</span> </div><div class="line"><a name="l00150"></a><span class="lineno"> 150</span>  <span class="comment">// Initialize matrix B reduction kernel only if _a_offset is not equal to 0</span></div><div class="line"><a name="l00151"></a><span class="lineno"> 151</span>  <span class="keywordflow">if</span>(_a_offset != 0)</div><div class="line"><a name="l00152"></a><span class="lineno"> 152</span>  {</div><div class="line"><a name="l00153"></a><span class="lineno"> 153</span>  <a class="code" href="classarm__compute_1_1_tensor_info.xhtml">TensorInfo</a> info_vector_sum_col(<a class="code" href="namespacearm__compute_1_1misc_1_1shape__calculator.xhtml#a60ce6c017f70d978b48b101ce314969e">compute_reductionA_shape</a>(*<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#aa76b4a6e74940dabc5b7fc6b2dab3545">b</a>->info()), 1, <a class="code" href="namespacearm__compute.xhtml#ab4e88c89b3b7ea1735996cc4def22d58aa1e28eee0339658d39a8b4d325b56e9c">DataType::S32</a>);</div><div class="line"><a name="l00154"></a><span class="lineno"> 154</span>  _vector_sum_col.<a class="code" href="classarm__compute_1_1_c_l_tensor.xhtml#a4083de30daebd6bdee6b35d9c8262108">allocator</a>()-><a class="code" href="classarm__compute_1_1_i_tensor_allocator.xhtml#af36143939a43fa124312e395975091ed">init</a>(info_vector_sum_col);</div><div class="line"><a name="l00155"></a><span class="lineno"> 155</span>  <span class="keywordflow">if</span>(!_reshape_b_only_on_first_run)</div><div class="line"><a name="l00156"></a><span class="lineno"> 156</span>  {</div><div class="line"><a name="l00157"></a><span class="lineno"> 157</span>  _memory_group.<a class="code" href="classarm__compute_1_1_memory_group.xhtml#a6fc0a49304c152c20a0f6df0634fb3cd">manage</a>(&_vector_sum_col);</div><div class="line"><a name="l00158"></a><span class="lineno"> 158</span>  }</div><div class="line"><a name="l00159"></a><span class="lineno"> 159</span> </div><div class="line"><a name="l00160"></a><span class="lineno"> 160</span>  <span class="comment">// Configure Matrix B reduction kernel</span></div><div class="line"><a name="l00161"></a><span class="lineno"> 161</span>  _mtx_b_reduction_kernel.<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_b_reduction_kernel.xhtml#aa526095c6ddce38f2ef22d45e18e2fe2">configure</a>(_convert_to_qasymm8 ? &_qasymm8_weights : <a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#aa76b4a6e74940dabc5b7fc6b2dab3545">b</a>, &_vector_sum_col);</div><div class="line"><a name="l00162"></a><span class="lineno"> 162</span>  }</div><div class="line"><a name="l00163"></a><span class="lineno"> 163</span> </div><div class="line"><a name="l00164"></a><span class="lineno"> 164</span>  <span class="comment">// Initialize Matrix A reduction kernel only if _b_offset is not equal to 0</span></div><div class="line"><a name="l00165"></a><span class="lineno"> 165</span>  <span class="keywordflow">if</span>(_b_offset != 0)</div><div class="line"><a name="l00166"></a><span class="lineno"> 166</span>  {</div><div class="line"><a name="l00167"></a><span class="lineno"> 167</span>  <a class="code" href="classarm__compute_1_1_tensor_info.xhtml">TensorInfo</a> info_vector_sum_row(<a class="code" href="namespacearm__compute_1_1misc_1_1shape__calculator.xhtml#a69f9b3191aafc4905f9d029ff9d48fea">compute_reductionB_shape</a>(*a-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>()), 1, <a class="code" href="namespacearm__compute.xhtml#ab4e88c89b3b7ea1735996cc4def22d58aa1e28eee0339658d39a8b4d325b56e9c">DataType::S32</a>);</div><div class="line"><a name="l00168"></a><span class="lineno"> 168</span>  _vector_sum_row.<a class="code" href="classarm__compute_1_1_c_l_tensor.xhtml#a4083de30daebd6bdee6b35d9c8262108">allocator</a>()-><a class="code" href="classarm__compute_1_1_i_tensor_allocator.xhtml#af36143939a43fa124312e395975091ed">init</a>(info_vector_sum_row);</div><div class="line"><a name="l00169"></a><span class="lineno"> 169</span>  _memory_group.<a class="code" href="classarm__compute_1_1_memory_group.xhtml#a6fc0a49304c152c20a0f6df0634fb3cd">manage</a>(&_vector_sum_row);</div><div class="line"><a name="l00170"></a><span class="lineno"> 170</span> </div><div class="line"><a name="l00171"></a><span class="lineno"> 171</span>  <span class="comment">// Configure matrix A reduction kernel</span></div><div class="line"><a name="l00172"></a><span class="lineno"> 172</span>  _mtx_a_reduction_kernel.<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_a_reduction_kernel.xhtml#a613f25e524dbabbff55ea9d7df53c7f6">configure</a>(a, &_vector_sum_row);</div><div class="line"><a name="l00173"></a><span class="lineno"> 173</span>  }</div><div class="line"><a name="l00174"></a><span class="lineno"> 174</span> </div><div class="line"><a name="l00175"></a><span class="lineno"> 175</span>  <span class="comment">// If GEMMLowpOutputStage != NONE, fuse the offset contribution with the output stage</span></div><div class="line"><a name="l00176"></a><span class="lineno"> 176</span>  <span class="keywordflow">if</span>(gemm_info.<a class="code" href="classarm__compute_1_1_g_e_m_m_info.xhtml#a11d8f855e323a8396fe6944edcef4238">gemmlowp_output_stage</a>().<a class="code" href="structarm__compute_1_1_g_e_m_m_lowp_output_stage_info.xhtml#a6e019ad85979fd73c74f97e5483faf35">type</a> != <a class="code" href="namespacearm__compute.xhtml#a5558e2cc22f7f4771653d992c8ad8864ab50339a10e1de285ac99d4c3990b8693">GEMMLowpOutputStageType::NONE</a>)</div><div class="line"><a name="l00177"></a><span class="lineno"> 177</span>  {</div><div class="line"><a name="l00178"></a><span class="lineno"> 178</span>  _fuse_output_stage = <span class="keyword">true</span>;</div><div class="line"><a name="l00179"></a><span class="lineno"> 179</span> </div><div class="line"><a name="l00180"></a><span class="lineno"> 180</span>  _memory_group.<a class="code" href="classarm__compute_1_1_memory_group.xhtml#a6fc0a49304c152c20a0f6df0634fb3cd">manage</a>(&_mm_result_s32);</div><div class="line"><a name="l00181"></a><span class="lineno"> 181</span> </div><div class="line"><a name="l00182"></a><span class="lineno"> 182</span>  <span class="keywordflow">if</span>(_is_gemm_reshaped)</div><div class="line"><a name="l00183"></a><span class="lineno"> 183</span>  {</div><div class="line"><a name="l00184"></a><span class="lineno"> 184</span>  <span class="comment">// Configure and tune matrix multiply kernel</span></div><div class="line"><a name="l00185"></a><span class="lineno"> 185</span>  _mm_reshaped_only_rhs_kernel.<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_reshaped_only_r_h_s_kernel.xhtml#aa7967bee001197b8642e30d4efaa96cc">configure</a>(_matrix_a, matrix_b, &_mm_result_s32, lhs_info, rhs_info, <a class="code" href="classarm__compute_1_1_g_e_m_m_reshape_info.xhtml">GEMMReshapeInfo</a>(m, n, k, 1, 1, depth_output_gemm3d, reinterpret_input_as_3d));</div><div class="line"><a name="l00186"></a><span class="lineno"> 186</span>  }</div><div class="line"><a name="l00187"></a><span class="lineno"> 187</span>  <span class="keywordflow">else</span></div><div class="line"><a name="l00188"></a><span class="lineno"> 188</span>  {</div><div class="line"><a name="l00189"></a><span class="lineno"> 189</span>  <span class="keywordflow">if</span>(_is_midgard)</div><div class="line"><a name="l00190"></a><span class="lineno"> 190</span>  {</div><div class="line"><a name="l00191"></a><span class="lineno"> 191</span>  <span class="comment">// Configure matrix multiply kernel</span></div><div class="line"><a name="l00192"></a><span class="lineno"> 192</span>  _mm_midgard_kernel.<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_kernel.xhtml#a8aa57bd9c0703378b82393e8f440e9c8">configure</a>(_matrix_a, matrix_b, &_mm_result_s32, <a class="code" href="classarm__compute_1_1_g_e_m_m_reshape_info.xhtml">GEMMReshapeInfo</a>(m, n, k, 1, 1, depth_output_gemm3d, reinterpret_input_as_3d));</div><div class="line"><a name="l00193"></a><span class="lineno"> 193</span>  }</div><div class="line"><a name="l00194"></a><span class="lineno"> 194</span>  <span class="keywordflow">else</span></div><div class="line"><a name="l00195"></a><span class="lineno"> 195</span>  {</div><div class="line"><a name="l00196"></a><span class="lineno"> 196</span>  <span class="comment">// Pick up the GEMM configuration</span></div><div class="line"><a name="l00197"></a><span class="lineno"> 197</span>  std::tie(lhs_info, rhs_info) = <a class="code" href="classarm__compute_1_1cl__gemm_1_1_c_l_g_e_m_m_native_kernel_configuration_factory.xhtml#ab2b50810acd1c74616ee97dee3cbdb6b">CLGEMMNativeKernelConfigurationFactory::create</a>(gpu_target)->configure(m, n, k, batch_size, <a class="code" href="namespacearm__compute.xhtml#ad8ed01ff3ff33333d8e19db4d2818bb6af14462d71aa842202c3e4b272c7ec924">DataType::QASYMM8</a>);</div><div class="line"><a name="l00198"></a><span class="lineno"> 198</span> </div><div class="line"><a name="l00199"></a><span class="lineno"> 199</span>  <span class="comment">// Configure matrix multiply kernel</span></div><div class="line"><a name="l00200"></a><span class="lineno"> 200</span>  _mm_native_kernel.<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_native_kernel.xhtml#aa7967bee001197b8642e30d4efaa96cc">configure</a>(_matrix_a, matrix_b, &_mm_result_s32, lhs_info, rhs_info, <a class="code" href="classarm__compute_1_1_g_e_m_m_reshape_info.xhtml">GEMMReshapeInfo</a>(m, n, k, 1, 1, depth_output_gemm3d, reinterpret_input_as_3d));</div><div class="line"><a name="l00201"></a><span class="lineno"> 201</span>  }</div><div class="line"><a name="l00202"></a><span class="lineno"> 202</span>  }</div><div class="line"><a name="l00203"></a><span class="lineno"> 203</span>  <span class="comment">// Configure offset contribution kernel</span></div><div class="line"><a name="l00204"></a><span class="lineno"> 204</span>  <span class="keyword">const</span> <span class="keywordtype">size_t</span> num_filters = (gemm_info.<a class="code" href="classarm__compute_1_1_g_e_m_m_info.xhtml#a11d8f855e323a8396fe6944edcef4238">gemmlowp_output_stage</a>().<a class="code" href="structarm__compute_1_1_g_e_m_m_lowp_output_stage_info.xhtml#a94e1801be6c3d9d6645c694d7e280cda">is_quantized_per_channel</a>) ? gemm_info.<a class="code" href="classarm__compute_1_1_g_e_m_m_info.xhtml#a11d8f855e323a8396fe6944edcef4238">gemmlowp_output_stage</a>().<a class="code" href="structarm__compute_1_1_g_e_m_m_lowp_output_stage_info.xhtml#ae5bd6bebbc0c7ebd9e7dbfd47d939c2a">gemmlowp_multipliers</a>.size() : 1;</div><div class="line"><a name="l00205"></a><span class="lineno"> 205</span> </div><div class="line"><a name="l00206"></a><span class="lineno"> 206</span>  _gemm_output_stage_multipliers.<a class="code" href="classarm__compute_1_1_c_l_tensor.xhtml#a4083de30daebd6bdee6b35d9c8262108">allocator</a>()-><a class="code" href="classarm__compute_1_1_i_tensor_allocator.xhtml#af36143939a43fa124312e395975091ed">init</a>(<a class="code" href="classarm__compute_1_1_tensor_info.xhtml">TensorInfo</a>(<a class="code" href="classarm__compute_1_1_tensor_shape.xhtml">TensorShape</a>(num_filters), 1, <a class="code" href="namespacearm__compute.xhtml#ab4e88c89b3b7ea1735996cc4def22d58aa1e28eee0339658d39a8b4d325b56e9c">DataType::S32</a>));</div><div class="line"><a name="l00207"></a><span class="lineno"> 207</span>  _gemm_output_stage_shifts.<a class="code" href="classarm__compute_1_1_c_l_tensor.xhtml#a4083de30daebd6bdee6b35d9c8262108">allocator</a>()-><a class="code" href="classarm__compute_1_1_i_tensor_allocator.xhtml#af36143939a43fa124312e395975091ed">init</a>(<a class="code" href="classarm__compute_1_1_tensor_info.xhtml">TensorInfo</a>(<a class="code" href="classarm__compute_1_1_tensor_shape.xhtml">TensorShape</a>(num_filters), 1, <a class="code" href="namespacearm__compute.xhtml#ab4e88c89b3b7ea1735996cc4def22d58aa1e28eee0339658d39a8b4d325b56e9c">DataType::S32</a>));</div><div class="line"><a name="l00208"></a><span class="lineno"> 208</span> </div><div class="line"><a name="l00209"></a><span class="lineno"> 209</span>  <a class="code" href="structarm__compute_1_1_g_e_m_m_lowp_output_stage_info.xhtml">GEMMLowpOutputStageInfo</a> gemmlowp_output_stage = gemm_info.<a class="code" href="classarm__compute_1_1_g_e_m_m_info.xhtml#a11d8f855e323a8396fe6944edcef4238">gemmlowp_output_stage</a>();</div><div class="line"><a name="l00210"></a><span class="lineno"> 210</span>  gemmlowp_output_stage.<a class="code" href="structarm__compute_1_1_g_e_m_m_lowp_output_stage_info.xhtml#ab233758aca2751c6e71a2f79baf7b92a">output_data_type</a> = _matrix_a-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>()-><a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml#a7cfb31af63202568efef5214acfbf3ba">data_type</a>();</div><div class="line"><a name="l00211"></a><span class="lineno"> 211</span>  _offset_contribution_output_stage_kernel.<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_offset_contribution_output_stage_kernel.xhtml#ac727ebddb913babf19047a6444b89fba">configure</a>(&_mm_result_s32, _a_offset == 0 ? <span class="keyword">nullptr</span> : &_vector_sum_col, _b_offset == 0 ? <span class="keyword">nullptr</span> : &_vector_sum_row, c, output, a-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>()-><a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml#a178f0d3d87f959e00a743328d95359d2">dimension</a>(0),</div><div class="line"><a name="l00212"></a><span class="lineno"> 212</span>  _a_offset, _b_offset, gemmlowp_output_stage, &_gemm_output_stage_multipliers, &_gemm_output_stage_shifts);</div><div class="line"><a name="l00213"></a><span class="lineno"> 213</span> </div><div class="line"><a name="l00214"></a><span class="lineno"> 214</span>  _gemm_output_stage_multipliers.allocator()->allocate();</div><div class="line"><a name="l00215"></a><span class="lineno"> 215</span>  _gemm_output_stage_shifts.<a class="code" href="classarm__compute_1_1_c_l_tensor.xhtml#a4083de30daebd6bdee6b35d9c8262108">allocator</a>()-><a class="code" href="classarm__compute_1_1_c_l_tensor_allocator.xhtml#a6e509c2a177b0b29e9e2369535094dee">allocate</a>();</div><div class="line"><a name="l00216"></a><span class="lineno"> 216</span>  <span class="comment">// Compute GEMM output multipliers and shifts for output stage</span></div><div class="line"><a name="l00217"></a><span class="lineno"> 217</span>  _gemm_output_stage_multipliers.map();</div><div class="line"><a name="l00218"></a><span class="lineno"> 218</span>  _gemm_output_stage_shifts.<a class="code" href="classarm__compute_1_1_c_l_tensor.xhtml#a14c53d2d17be6fa8a2c9861527c7b002">map</a>();</div><div class="line"><a name="l00219"></a><span class="lineno"> 219</span>  std::memcpy(_gemm_output_stage_multipliers.ptr_to_element(<a class="code" href="classarm__compute_1_1_coordinates.xhtml">Coordinates</a>(0)), gemm_info.<a class="code" href="classarm__compute_1_1_g_e_m_m_info.xhtml#a11d8f855e323a8396fe6944edcef4238">gemmlowp_output_stage</a>().<a class="code" href="structarm__compute_1_1_g_e_m_m_lowp_output_stage_info.xhtml#ae5bd6bebbc0c7ebd9e7dbfd47d939c2a">gemmlowp_multipliers</a>.data(), num_filters * <span class="keyword">sizeof</span>(int32_t));</div><div class="line"><a name="l00220"></a><span class="lineno"> 220</span>  std::memcpy(_gemm_output_stage_shifts.<a class="code" href="classarm__compute_1_1_i_tensor.xhtml#adbd73147d41e8a640bc299d12613c31e">ptr_to_element</a>(<a class="code" href="classarm__compute_1_1_coordinates.xhtml">Coordinates</a>(0)), gemm_info.<a class="code" href="classarm__compute_1_1_g_e_m_m_info.xhtml#a11d8f855e323a8396fe6944edcef4238">gemmlowp_output_stage</a>().<a class="code" href="structarm__compute_1_1_g_e_m_m_lowp_output_stage_info.xhtml#ab269b182588a158cd256f9d4bb2a00dd">gemmlowp_shifts</a>.data(), num_filters * <span class="keyword">sizeof</span>(int32_t));</div><div class="line"><a name="l00221"></a><span class="lineno"> 221</span>  _gemm_output_stage_multipliers.unmap();</div><div class="line"><a name="l00222"></a><span class="lineno"> 222</span>  _gemm_output_stage_shifts.<a class="code" href="classarm__compute_1_1_c_l_tensor.xhtml#a1ffeb3b5abb3d61f62b58a391816201c">unmap</a>();</div><div class="line"><a name="l00223"></a><span class="lineno"> 223</span> </div><div class="line"><a name="l00224"></a><span class="lineno"> 224</span>  _mm_result_s32.<a class="code" href="classarm__compute_1_1_c_l_tensor.xhtml#a4083de30daebd6bdee6b35d9c8262108">allocator</a>()-><a class="code" href="classarm__compute_1_1_c_l_tensor_allocator.xhtml#a6e509c2a177b0b29e9e2369535094dee">allocate</a>();</div><div class="line"><a name="l00225"></a><span class="lineno"> 225</span>  }</div><div class="line"><a name="l00226"></a><span class="lineno"> 226</span>  <span class="keywordflow">else</span></div><div class="line"><a name="l00227"></a><span class="lineno"> 227</span>  {</div><div class="line"><a name="l00228"></a><span class="lineno"> 228</span>  <span class="keywordflow">if</span>(_is_gemm_reshaped)</div><div class="line"><a name="l00229"></a><span class="lineno"> 229</span>  {</div><div class="line"><a name="l00230"></a><span class="lineno"> 230</span>  <span class="comment">// Configure and tune matrix multiply kernel</span></div><div class="line"><a name="l00231"></a><span class="lineno"> 231</span>  _mm_reshaped_only_rhs_kernel.<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_reshaped_only_r_h_s_kernel.xhtml#aa7967bee001197b8642e30d4efaa96cc">configure</a>(_matrix_a, matrix_b, output, lhs_info, rhs_info, <a class="code" href="classarm__compute_1_1_g_e_m_m_reshape_info.xhtml">GEMMReshapeInfo</a>(m, n, k, 1, 1, depth_output_gemm3d, reinterpret_input_as_3d));</div><div class="line"><a name="l00232"></a><span class="lineno"> 232</span>  }</div><div class="line"><a name="l00233"></a><span class="lineno"> 233</span>  <span class="keywordflow">else</span></div><div class="line"><a name="l00234"></a><span class="lineno"> 234</span>  {</div><div class="line"><a name="l00235"></a><span class="lineno"> 235</span>  <span class="keywordflow">if</span>(_is_midgard)</div><div class="line"><a name="l00236"></a><span class="lineno"> 236</span>  {</div><div class="line"><a name="l00237"></a><span class="lineno"> 237</span>  <span class="comment">// Configure matrix multiply kernel</span></div><div class="line"><a name="l00238"></a><span class="lineno"> 238</span>  _mm_midgard_kernel.<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_kernel.xhtml#a8aa57bd9c0703378b82393e8f440e9c8">configure</a>(_matrix_a, matrix_b, output, <a class="code" href="classarm__compute_1_1_g_e_m_m_reshape_info.xhtml">GEMMReshapeInfo</a>(m, n, k, 1, 1, depth_output_gemm3d, reinterpret_input_as_3d));</div><div class="line"><a name="l00239"></a><span class="lineno"> 239</span>  }</div><div class="line"><a name="l00240"></a><span class="lineno"> 240</span>  <span class="keywordflow">else</span></div><div class="line"><a name="l00241"></a><span class="lineno"> 241</span>  {</div><div class="line"><a name="l00242"></a><span class="lineno"> 242</span>  <span class="comment">// Pick up the GEMM configuration</span></div><div class="line"><a name="l00243"></a><span class="lineno"> 243</span>  std::tie(lhs_info, rhs_info) = <a class="code" href="classarm__compute_1_1cl__gemm_1_1_c_l_g_e_m_m_native_kernel_configuration_factory.xhtml#ab2b50810acd1c74616ee97dee3cbdb6b">CLGEMMNativeKernelConfigurationFactory::create</a>(gpu_target)->configure(m, n, k, batch_size, <a class="code" href="namespacearm__compute.xhtml#ad8ed01ff3ff33333d8e19db4d2818bb6af14462d71aa842202c3e4b272c7ec924">DataType::QASYMM8</a>);</div><div class="line"><a name="l00244"></a><span class="lineno"> 244</span> </div><div class="line"><a name="l00245"></a><span class="lineno"> 245</span>  <span class="comment">// Configure matrix multiply kernel</span></div><div class="line"><a name="l00246"></a><span class="lineno"> 246</span>  _mm_native_kernel.<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_native_kernel.xhtml#aa7967bee001197b8642e30d4efaa96cc">configure</a>(_matrix_a, matrix_b, output, lhs_info, rhs_info, <a class="code" href="classarm__compute_1_1_g_e_m_m_reshape_info.xhtml">GEMMReshapeInfo</a>(m, n, k, 1, 1, depth_output_gemm3d, reinterpret_input_as_3d));</div><div class="line"><a name="l00247"></a><span class="lineno"> 247</span>  }</div><div class="line"><a name="l00248"></a><span class="lineno"> 248</span>  }</div><div class="line"><a name="l00249"></a><span class="lineno"> 249</span> </div><div class="line"><a name="l00250"></a><span class="lineno"> 250</span>  <span class="comment">// Configure offset contribution kernel</span></div><div class="line"><a name="l00251"></a><span class="lineno"> 251</span>  _offset_contribution_kernel.<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_offset_contribution_kernel.xhtml#a094cc43211df267051139e766f002d77">configure</a>(output, _a_offset == 0 ? <span class="keyword">nullptr</span> : &_vector_sum_col, _b_offset == 0 ? <span class="keyword">nullptr</span> : &_vector_sum_row, c, a-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>()-><a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml#a178f0d3d87f959e00a743328d95359d2">dimension</a>(0), _a_offset, _b_offset);</div><div class="line"><a name="l00252"></a><span class="lineno"> 252</span>  }</div><div class="line"><a name="l00253"></a><span class="lineno"> 253</span> </div><div class="line"><a name="l00254"></a><span class="lineno"> 254</span>  <span class="comment">// Allocate tensors</span></div><div class="line"><a name="l00255"></a><span class="lineno"> 255</span>  <span class="keywordflow">if</span>(_is_gemm_reshaped)</div><div class="line"><a name="l00256"></a><span class="lineno"> 256</span>  {</div><div class="line"><a name="l00257"></a><span class="lineno"> 257</span>  <span class="keywordflow">if</span>(!_reshape_b_only_on_first_run)</div><div class="line"><a name="l00258"></a><span class="lineno"> 258</span>  {</div><div class="line"><a name="l00259"></a><span class="lineno"> 259</span>  _tmp_b.<a class="code" href="classarm__compute_1_1_c_l_tensor.xhtml#a4083de30daebd6bdee6b35d9c8262108">allocator</a>()-><a class="code" href="classarm__compute_1_1_c_l_tensor_allocator.xhtml#a6e509c2a177b0b29e9e2369535094dee">allocate</a>();</div><div class="line"><a name="l00260"></a><span class="lineno"> 260</span>  }</div><div class="line"><a name="l00261"></a><span class="lineno"> 261</span>  }</div><div class="line"><a name="l00262"></a><span class="lineno"> 262</span> </div><div class="line"><a name="l00263"></a><span class="lineno"> 263</span>  <span class="keywordflow">if</span>(_a_offset != 0 && !_reshape_b_only_on_first_run)</div><div class="line"><a name="l00264"></a><span class="lineno"> 264</span>  {</div><div class="line"><a name="l00265"></a><span class="lineno"> 265</span>  _vector_sum_col.<a class="code" href="classarm__compute_1_1_c_l_tensor.xhtml#a4083de30daebd6bdee6b35d9c8262108">allocator</a>()-><a class="code" href="classarm__compute_1_1_c_l_tensor_allocator.xhtml#a6e509c2a177b0b29e9e2369535094dee">allocate</a>();</div><div class="line"><a name="l00266"></a><span class="lineno"> 266</span>  }</div><div class="line"><a name="l00267"></a><span class="lineno"> 267</span> </div><div class="line"><a name="l00268"></a><span class="lineno"> 268</span>  <span class="keywordflow">if</span>(_b_offset != 0)</div><div class="line"><a name="l00269"></a><span class="lineno"> 269</span>  {</div><div class="line"><a name="l00270"></a><span class="lineno"> 270</span>  _vector_sum_row.<a class="code" href="classarm__compute_1_1_c_l_tensor.xhtml#a4083de30daebd6bdee6b35d9c8262108">allocator</a>()-><a class="code" href="classarm__compute_1_1_c_l_tensor_allocator.xhtml#a6e509c2a177b0b29e9e2369535094dee">allocate</a>();</div><div class="line"><a name="l00271"></a><span class="lineno"> 271</span>  }</div><div class="line"><a name="l00272"></a><span class="lineno"> 272</span> }</div><div class="line"><a name="l00273"></a><span class="lineno"> 273</span> </div><div class="line"><a name="l00274"></a><span class="lineno"><a class="line" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_core.xhtml#a8c3cf2d65afb288e39909171ada19566"> 274</a></span> <a class="code" href="classarm__compute_1_1_status.xhtml">Status</a> <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_core.xhtml#a8c3cf2d65afb288e39909171ada19566">CLGEMMLowpMatrixMultiplyCore::validate</a>(<span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *a, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#aa76b4a6e74940dabc5b7fc6b2dab3545">b</a>, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *c, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *output, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_g_e_m_m_info.xhtml">GEMMInfo</a> &gemm_info)</div><div class="line"><a name="l00275"></a><span class="lineno"> 275</span> {</div><div class="line"><a name="l00276"></a><span class="lineno"> 276</span>  <a class="code" href="_validate_8h.xhtml#a921b705e9e3e0fe928928447869e62a5">ARM_COMPUTE_ERROR_ON_NULLPTR</a>(a, <a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#aa76b4a6e74940dabc5b7fc6b2dab3545">b</a>, output);</div><div class="line"><a name="l00277"></a><span class="lineno"> 277</span>  <a class="code" href="_validate_8h.xhtml#ae7eed178dac535c6e727061b1f5bc6eb">ARM_COMPUTE_RETURN_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN</a>(a, 1, <a class="code" href="namespacearm__compute.xhtml#ad8ed01ff3ff33333d8e19db4d2818bb6af14462d71aa842202c3e4b272c7ec924">DataType::QASYMM8</a>, <a class="code" href="namespacearm__compute.xhtml#ad8ed01ff3ff33333d8e19db4d2818bb6a329f5d0c4b0c80e3474951d2c4435dd9">DataType::QASYMM8_SIGNED</a>);</div><div class="line"><a name="l00278"></a><span class="lineno"> 278</span>  <span class="comment">//DataType::QSYMM8_PER_CHANNEL supported only for weights</span></div><div class="line"><a name="l00279"></a><span class="lineno"> 279</span>  <span class="keywordflow">if</span>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#aa76b4a6e74940dabc5b7fc6b2dab3545">b</a>->data_type() != <a class="code" href="namespacearm__compute.xhtml#ad8ed01ff3ff33333d8e19db4d2818bb6a34f500e941c4df30b870126ec868ebd5">DataType::QSYMM8_PER_CHANNEL</a>)</div><div class="line"><a name="l00280"></a><span class="lineno"> 280</span>  {</div><div class="line"><a name="l00281"></a><span class="lineno"> 281</span>  <a class="code" href="_validate_8h.xhtml#a8f3ff7da485ff7e75dab07baadf5b4bd">ARM_COMPUTE_RETURN_ERROR_ON_MISMATCHING_DATA_TYPES</a>(a, <a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#aa76b4a6e74940dabc5b7fc6b2dab3545">b</a>);</div><div class="line"><a name="l00282"></a><span class="lineno"> 282</span>  }</div><div class="line"><a name="l00283"></a><span class="lineno"> 283</span>  <a class="code" href="_error_8h.xhtml#a1c69762a42ab8add645d0a949b6f4b1f">ARM_COMPUTE_RETURN_ERROR_ON_MSG</a>(gemm_info.<a class="code" href="classarm__compute_1_1_g_e_m_m_info.xhtml#aa7e9584d7080ca6442cec62afaff6cad">is_a_reshaped</a>(), <span class="stringliteral">"Matrix A already reshaped is not supported"</span>);</div><div class="line"><a name="l00284"></a><span class="lineno"> 284</span>  <a class="code" href="_error_8h.xhtml#a1c69762a42ab8add645d0a949b6f4b1f">ARM_COMPUTE_RETURN_ERROR_ON_MSG</a>(gemm_info.<a class="code" href="classarm__compute_1_1_g_e_m_m_info.xhtml#a77964edb8d16bb8ec14ddd8985e03cb0">is_b_reshaped</a>(), <span class="stringliteral">"Matrix B already reshaped is not supported"</span>);</div><div class="line"><a name="l00285"></a><span class="lineno"> 285</span> </div><div class="line"><a name="l00286"></a><span class="lineno"> 286</span>  int32_t a_offset = a-><a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml#a3f3e1a3200223e6a304a533b1016e749">quantization_info</a>().<a class="code" href="classarm__compute_1_1_quantization_info.xhtml#a706fc156bcd4c45441bcaad05884b57d">uniform</a>().<a class="code" href="structarm__compute_1_1_uniform_quantization_info.xhtml#a97bd6c077f3c7769f575b82988b9b668">offset</a>;</div><div class="line"><a name="l00287"></a><span class="lineno"> 287</span>  int32_t b_offset = <a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#aa76b4a6e74940dabc5b7fc6b2dab3545">b</a>->quantization_info().uniform().offset;</div><div class="line"><a name="l00288"></a><span class="lineno"> 288</span> </div><div class="line"><a name="l00289"></a><span class="lineno"> 289</span>  <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *matrix_a_info = a;</div><div class="line"><a name="l00290"></a><span class="lineno"> 290</span> </div><div class="line"><a name="l00291"></a><span class="lineno"> 291</span>  <a class="code" href="classarm__compute_1_1_tensor_info.xhtml">TensorInfo</a> tmp_b_info{};</div><div class="line"><a name="l00292"></a><span class="lineno"> 292</span>  <a class="code" href="structarm__compute_1_1_g_e_m_m_r_h_s_matrix_info.xhtml">GEMMRHSMatrixInfo</a> rhs_info;</div><div class="line"><a name="l00293"></a><span class="lineno"> 293</span>  <a class="code" href="structarm__compute_1_1_g_e_m_m_l_h_s_matrix_info.xhtml">GEMMLHSMatrixInfo</a> lhs_info;</div><div class="line"><a name="l00294"></a><span class="lineno"> 294</span> </div><div class="line"><a name="l00295"></a><span class="lineno"> 295</span>  <span class="comment">// Get the GPU target</span></div><div class="line"><a name="l00296"></a><span class="lineno"> 296</span>  <span class="keyword">const</span> <a class="code" href="namespacearm__compute.xhtml#a735ac6c2a02e320969625308810444f3">GPUTarget</a> gpu_target = <a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#a9b58d0eb9a2af8e6d7908695e1557d6c">CLScheduler::get</a>().<a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#a27561688e2fc60176608ef725a4ecb30">target</a>();</div><div class="line"><a name="l00297"></a><span class="lineno"> 297</span> </div><div class="line"><a name="l00298"></a><span class="lineno"> 298</span>  <span class="keywordtype">bool</span> reinterpret_input_as_3d = gemm_info.<a class="code" href="classarm__compute_1_1_g_e_m_m_info.xhtml#a00330b8913cac3b07029ac0c3350e806">reinterpret_input_as_3d</a>();</div><div class="line"><a name="l00299"></a><span class="lineno"> 299</span>  <span class="keyword">const</span> <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> m = reinterpret_input_as_3d ? (a-><a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml#a178f0d3d87f959e00a743328d95359d2">dimension</a>(1) * a-><a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml#a178f0d3d87f959e00a743328d95359d2">dimension</a>(2)) : a-><a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml#a178f0d3d87f959e00a743328d95359d2">dimension</a>(1);</div><div class="line"><a name="l00300"></a><span class="lineno"> 300</span>  <span class="keyword">const</span> <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> n = <a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#aa76b4a6e74940dabc5b7fc6b2dab3545">b</a>->dimension(0);</div><div class="line"><a name="l00301"></a><span class="lineno"> 301</span>  <span class="keyword">const</span> <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> k = a-><a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml#a178f0d3d87f959e00a743328d95359d2">dimension</a>(0);</div><div class="line"><a name="l00302"></a><span class="lineno"> 302</span>  <span class="keyword">const</span> <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> batch_size = reinterpret_input_as_3d ? a-><a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml#a178f0d3d87f959e00a743328d95359d2">dimension</a>(3) : a-><a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml#a178f0d3d87f959e00a743328d95359d2">dimension</a>(2);</div><div class="line"><a name="l00303"></a><span class="lineno"> 303</span>  <span class="keyword">const</span> <span class="keywordtype">int</span> depth_output_gemm3d = gemm_info.<a class="code" href="classarm__compute_1_1_g_e_m_m_info.xhtml#abbd888f118c2209bf7578eb4f8942a07">depth_output_gemm3d</a>();</div><div class="line"><a name="l00304"></a><span class="lineno"> 304</span>  <span class="keyword">const</span> <span class="keywordtype">bool</span> is_midgard = gpu_target == <a class="code" href="namespacearm__compute.xhtml#a735ac6c2a02e320969625308810444f3afa362f4daec88442a387ff7cda411a38">GPUTarget::MIDGARD</a>;</div><div class="line"><a name="l00305"></a><span class="lineno"> 305</span> </div><div class="line"><a name="l00306"></a><span class="lineno"> 306</span>  <span class="keywordtype">bool</span> reshape_matrix_b = is_gemm_reshaped(gemm_info.<a class="code" href="classarm__compute_1_1_g_e_m_m_info.xhtml#a4c8f9fa843de1086c27c86a6b8cf4582">reshape_b_only_on_first_run</a>(), <a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#a9b58d0eb9a2af8e6d7908695e1557d6c">CLScheduler::get</a>().<a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#a27561688e2fc60176608ef725a4ecb30">target</a>());</div><div class="line"><a name="l00307"></a><span class="lineno"> 307</span> </div><div class="line"><a name="l00308"></a><span class="lineno"> 308</span>  <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_g_e_m_m_reshape_info.xhtml">GEMMReshapeInfo</a> reshape_info = <a class="code" href="classarm__compute_1_1_g_e_m_m_reshape_info.xhtml">GEMMReshapeInfo</a>(m, n, k, 1, 1, depth_output_gemm3d, reinterpret_input_as_3d);</div><div class="line"><a name="l00309"></a><span class="lineno"> 309</span> </div><div class="line"><a name="l00310"></a><span class="lineno"> 310</span>  <span class="keywordtype">bool</span> convert_to_qasymm8 = <a class="code" href="namespacearm__compute.xhtml#a84437d80241f6a31e1a07c231ee8e3ac">is_data_type_quantized_per_channel</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#aa76b4a6e74940dabc5b7fc6b2dab3545">b</a>->data_type()) && <a class="code" href="namespacearm__compute.xhtml#a922cd85c475c0d725fc030116ea57979">is_data_type_quantized_symmetric</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#aa76b4a6e74940dabc5b7fc6b2dab3545">b</a>->data_type())</div><div class="line"><a name="l00311"></a><span class="lineno"> 311</span>  && <a class="code" href="namespacearm__compute.xhtml#a14f46283f316e7f0fad301d5c1507e9f">is_data_type_quantized_asymmetric</a>(a-><a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml#a7cfb31af63202568efef5214acfbf3ba">data_type</a>());</div><div class="line"><a name="l00312"></a><span class="lineno"> 312</span>  <a class="code" href="classarm__compute_1_1_tensor_info.xhtml">TensorInfo</a> <a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a7cb842ebfe255726066039853a4322f0">weights_info</a>(*<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#aa76b4a6e74940dabc5b7fc6b2dab3545">b</a>);</div><div class="line"><a name="l00313"></a><span class="lineno"> 313</span>  <span class="keywordflow">if</span>(convert_to_qasymm8)</div><div class="line"><a name="l00314"></a><span class="lineno"> 314</span>  {</div><div class="line"><a name="l00315"></a><span class="lineno"> 315</span>  b_offset = -128;</div><div class="line"><a name="l00316"></a><span class="lineno"> 316</span>  <a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a7cb842ebfe255726066039853a4322f0">weights_info</a>.set_data_type(<a class="code" href="namespacearm__compute.xhtml#ad8ed01ff3ff33333d8e19db4d2818bb6af14462d71aa842202c3e4b272c7ec924">DataType::QASYMM8</a>);</div><div class="line"><a name="l00317"></a><span class="lineno"> 317</span>  <a class="code" href="_error_8h.xhtml#a8a1e1c105f0bdaf37db408c7cfcb77a4">ARM_COMPUTE_RETURN_ON_ERROR</a>(<a class="code" href="classarm__compute_1_1_c_l_depth_convert_layer_kernel.xhtml#a31fdffa6cbd16d74f8c2da415e6f4912">CLDepthConvertLayerKernel::validate</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#aa76b4a6e74940dabc5b7fc6b2dab3545">b</a>, &<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a7cb842ebfe255726066039853a4322f0">weights_info</a>, <a class="code" href="namespacearm__compute.xhtml#a82b8ac759c804bc1fb4e2d21e178fb6fae1c8555fcf0ea2bb648a6fd527d658c0">ConvertPolicy::WRAP</a>, 0));</div><div class="line"><a name="l00318"></a><span class="lineno"> 318</span>  }</div><div class="line"><a name="l00319"></a><span class="lineno"> 319</span>  <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *matrix_b_info = &<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a7cb842ebfe255726066039853a4322f0">weights_info</a>;</div><div class="line"><a name="l00320"></a><span class="lineno"> 320</span>  <span class="keywordflow">if</span>(reshape_matrix_b)</div><div class="line"><a name="l00321"></a><span class="lineno"> 321</span>  {</div><div class="line"><a name="l00322"></a><span class="lineno"> 322</span>  matrix_b_info = &tmp_b_info;</div><div class="line"><a name="l00323"></a><span class="lineno"> 323</span> </div><div class="line"><a name="l00324"></a><span class="lineno"> 324</span>  <span class="comment">// Pick up the GEMM configuration</span></div><div class="line"><a name="l00325"></a><span class="lineno"> 325</span>  std::tie(lhs_info, rhs_info) = <a class="code" href="classarm__compute_1_1cl__gemm_1_1_c_l_g_e_m_m_reshaped_only_r_h_s_kernel_configuration_factory.xhtml#ab2b50810acd1c74616ee97dee3cbdb6b">CLGEMMReshapedOnlyRHSKernelConfigurationFactory::create</a>(gpu_target)->configure(m, n, k, batch_size, <a class="code" href="namespacearm__compute.xhtml#ad8ed01ff3ff33333d8e19db4d2818bb6af14462d71aa842202c3e4b272c7ec924">DataType::QASYMM8</a>);</div><div class="line"><a name="l00326"></a><span class="lineno"> 326</span> </div><div class="line"><a name="l00327"></a><span class="lineno"> 327</span>  <span class="comment">// Validate reshape RHS kernel</span></div><div class="line"><a name="l00328"></a><span class="lineno"> 328</span>  <a class="code" href="namespacearm__compute.xhtml#a47be6fa38308d0003c25b60b7dbc45ce">auto_init_if_empty</a>(tmp_b_info, <a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a7cb842ebfe255726066039853a4322f0">weights_info</a>.clone()->set_tensor_shape(<a class="code" href="namespacearm__compute_1_1misc_1_1shape__calculator.xhtml#a09ad10a110d947fd9c444b2ea5e4c127">compute_rhs_reshaped_shape</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a7cb842ebfe255726066039853a4322f0">weights_info</a>, rhs_info)));</div><div class="line"><a name="l00329"></a><span class="lineno"> 329</span>  <a class="code" href="_error_8h.xhtml#a8a1e1c105f0bdaf37db408c7cfcb77a4">ARM_COMPUTE_RETURN_ON_ERROR</a>(<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_reshape_r_h_s_matrix_kernel.xhtml#aca0da52b749b54d37ff996f4b28f6762">CLGEMMReshapeRHSMatrixKernel::validate</a>(&<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a7cb842ebfe255726066039853a4322f0">weights_info</a>, &tmp_b_info, rhs_info));</div><div class="line"><a name="l00330"></a><span class="lineno"> 330</span>  }</div><div class="line"><a name="l00331"></a><span class="lineno"> 331</span> </div><div class="line"><a name="l00332"></a><span class="lineno"> 332</span>  <a class="code" href="classarm__compute_1_1_tensor_info.xhtml">TensorInfo</a> info_vector_sum_col{};</div><div class="line"><a name="l00333"></a><span class="lineno"> 333</span>  <a class="code" href="classarm__compute_1_1_tensor_info.xhtml">TensorInfo</a> info_vector_sum_row{};</div><div class="line"><a name="l00334"></a><span class="lineno"> 334</span> </div><div class="line"><a name="l00335"></a><span class="lineno"> 335</span>  <span class="comment">// Validate matrix B reduction kernel only if _a_offset is not equal to 0</span></div><div class="line"><a name="l00336"></a><span class="lineno"> 336</span>  <span class="keywordflow">if</span>(a_offset != 0)</div><div class="line"><a name="l00337"></a><span class="lineno"> 337</span>  {</div><div class="line"><a name="l00338"></a><span class="lineno"> 338</span>  info_vector_sum_col = <a class="code" href="classarm__compute_1_1_tensor_info.xhtml">TensorInfo</a>(<a class="code" href="namespacearm__compute_1_1misc_1_1shape__calculator.xhtml#a60ce6c017f70d978b48b101ce314969e">compute_reductionA_shape</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a7cb842ebfe255726066039853a4322f0">weights_info</a>), 1, <a class="code" href="namespacearm__compute.xhtml#ab4e88c89b3b7ea1735996cc4def22d58aa1e28eee0339658d39a8b4d325b56e9c">DataType::S32</a>);</div><div class="line"><a name="l00339"></a><span class="lineno"> 339</span> </div><div class="line"><a name="l00340"></a><span class="lineno"> 340</span>  <span class="comment">// Configure Matrix B reduction kernel</span></div><div class="line"><a name="l00341"></a><span class="lineno"> 341</span>  <a class="code" href="_error_8h.xhtml#a8a1e1c105f0bdaf37db408c7cfcb77a4">ARM_COMPUTE_RETURN_ON_ERROR</a>(<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_b_reduction_kernel.xhtml#a21f628daa45675076fd25e46a052a53a">CLGEMMLowpMatrixBReductionKernel::validate</a>(&<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a7cb842ebfe255726066039853a4322f0">weights_info</a>, &info_vector_sum_col));</div><div class="line"><a name="l00342"></a><span class="lineno"> 342</span>  }</div><div class="line"><a name="l00343"></a><span class="lineno"> 343</span> </div><div class="line"><a name="l00344"></a><span class="lineno"> 344</span>  <span class="comment">// Validate Matrix A reduction kernel only if _b_offset is not equal to 0</span></div><div class="line"><a name="l00345"></a><span class="lineno"> 345</span>  <span class="keywordflow">if</span>(b_offset != 0)</div><div class="line"><a name="l00346"></a><span class="lineno"> 346</span>  {</div><div class="line"><a name="l00347"></a><span class="lineno"> 347</span>  info_vector_sum_row = <a class="code" href="classarm__compute_1_1_tensor_info.xhtml">TensorInfo</a>(<a class="code" href="namespacearm__compute_1_1misc_1_1shape__calculator.xhtml#a69f9b3191aafc4905f9d029ff9d48fea">compute_reductionB_shape</a>(*a), 1, <a class="code" href="namespacearm__compute.xhtml#ab4e88c89b3b7ea1735996cc4def22d58aa1e28eee0339658d39a8b4d325b56e9c">DataType::S32</a>);</div><div class="line"><a name="l00348"></a><span class="lineno"> 348</span> </div><div class="line"><a name="l00349"></a><span class="lineno"> 349</span>  <span class="comment">// Configure matrix A reduction kernel</span></div><div class="line"><a name="l00350"></a><span class="lineno"> 350</span>  <a class="code" href="_error_8h.xhtml#a8a1e1c105f0bdaf37db408c7cfcb77a4">ARM_COMPUTE_RETURN_ON_ERROR</a>(<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_a_reduction_kernel.xhtml#a46e5929d20bb4cc0678590dfa251df9e">CLGEMMLowpMatrixAReductionKernel::validate</a>(a, &info_vector_sum_row));</div><div class="line"><a name="l00351"></a><span class="lineno"> 351</span>  }</div><div class="line"><a name="l00352"></a><span class="lineno"> 352</span> </div><div class="line"><a name="l00353"></a><span class="lineno"> 353</span>  <span class="keywordflow">if</span>(gemm_info.<a class="code" href="classarm__compute_1_1_g_e_m_m_info.xhtml#a11d8f855e323a8396fe6944edcef4238">gemmlowp_output_stage</a>().<a class="code" href="structarm__compute_1_1_g_e_m_m_lowp_output_stage_info.xhtml#a6e019ad85979fd73c74f97e5483faf35">type</a> != <a class="code" href="namespacearm__compute.xhtml#a5558e2cc22f7f4771653d992c8ad8864ab50339a10e1de285ac99d4c3990b8693">GEMMLowpOutputStageType::NONE</a>)</div><div class="line"><a name="l00354"></a><span class="lineno"> 354</span>  {</div><div class="line"><a name="l00355"></a><span class="lineno"> 355</span>  <a class="code" href="classarm__compute_1_1_tensor_info.xhtml">TensorInfo</a> mm_result_s32_info{};</div><div class="line"><a name="l00356"></a><span class="lineno"> 356</span> </div><div class="line"><a name="l00357"></a><span class="lineno"> 357</span>  <span class="keywordflow">if</span>(reshape_matrix_b)</div><div class="line"><a name="l00358"></a><span class="lineno"> 358</span>  {</div><div class="line"><a name="l00359"></a><span class="lineno"> 359</span>  <span class="comment">// Output tensor auto inizialitation if not yet initialized</span></div><div class="line"><a name="l00360"></a><span class="lineno"> 360</span>  <a class="code" href="namespacearm__compute.xhtml#a47be6fa38308d0003c25b60b7dbc45ce">auto_init_if_empty</a>(mm_result_s32_info, a-><a class="code" href="classarm__compute_1_1misc_1_1_i_cloneable.xhtml#a4d10e5012a872e7f78f2b539b673049d">clone</a>()->set_tensor_shape(<a class="code" href="namespacearm__compute_1_1misc_1_1shape__calculator.xhtml#adca241b012a5e00ddfcdc5a8db05a2a3">compute_mm_shape</a>(*matrix_a_info, *matrix_b_info, reshape_info)).set_data_type(<a class="code" href="namespacearm__compute.xhtml#ab4e88c89b3b7ea1735996cc4def22d58aa1e28eee0339658d39a8b4d325b56e9c">DataType::S32</a>));</div><div class="line"><a name="l00361"></a><span class="lineno"> 361</span> </div><div class="line"><a name="l00362"></a><span class="lineno"> 362</span>  <span class="comment">// Validate matrix multiply</span></div><div class="line"><a name="l00363"></a><span class="lineno"> 363</span>  <a class="code" href="_error_8h.xhtml#a8a1e1c105f0bdaf37db408c7cfcb77a4">ARM_COMPUTE_RETURN_ON_ERROR</a>(<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_reshaped_only_r_h_s_kernel.xhtml#a96f40eda5c40e254a6729e8bc0cb8195">CLGEMMLowpMatrixMultiplyReshapedOnlyRHSKernel::validate</a>(matrix_a_info, matrix_b_info, &mm_result_s32_info, lhs_info, rhs_info, reshape_info));</div><div class="line"><a name="l00364"></a><span class="lineno"> 364</span>  }</div><div class="line"><a name="l00365"></a><span class="lineno"> 365</span>  <span class="keywordflow">else</span></div><div class="line"><a name="l00366"></a><span class="lineno"> 366</span>  {</div><div class="line"><a name="l00367"></a><span class="lineno"> 367</span>  <span class="comment">// Output tensor auto inizialitation if not yet initialized</span></div><div class="line"><a name="l00368"></a><span class="lineno"> 368</span>  <a class="code" href="namespacearm__compute.xhtml#a47be6fa38308d0003c25b60b7dbc45ce">auto_init_if_empty</a>(mm_result_s32_info, a-><a class="code" href="classarm__compute_1_1misc_1_1_i_cloneable.xhtml#a4d10e5012a872e7f78f2b539b673049d">clone</a>()->set_tensor_shape(<a class="code" href="namespacearm__compute_1_1misc_1_1shape__calculator.xhtml#adca241b012a5e00ddfcdc5a8db05a2a3">compute_mm_shape</a>(*matrix_a_info, *matrix_b_info, <span class="keyword">false</span>, reshape_info)).set_data_type(<a class="code" href="namespacearm__compute.xhtml#ab4e88c89b3b7ea1735996cc4def22d58aa1e28eee0339658d39a8b4d325b56e9c">DataType::S32</a>));</div><div class="line"><a name="l00369"></a><span class="lineno"> 369</span> </div><div class="line"><a name="l00370"></a><span class="lineno"> 370</span>  <span class="keywordflow">if</span>(is_midgard)</div><div class="line"><a name="l00371"></a><span class="lineno"> 371</span>  {</div><div class="line"><a name="l00372"></a><span class="lineno"> 372</span>  <span class="comment">// Validate matrix multiply</span></div><div class="line"><a name="l00373"></a><span class="lineno"> 373</span>  <a class="code" href="_error_8h.xhtml#a8a1e1c105f0bdaf37db408c7cfcb77a4">ARM_COMPUTE_RETURN_ON_ERROR</a>(<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_kernel.xhtml#a9c52840c4fadda388888e26f655de61d">CLGEMMLowpMatrixMultiplyKernel::validate</a>(matrix_a_info, matrix_b_info, &mm_result_s32_info, reshape_info));</div><div class="line"><a name="l00374"></a><span class="lineno"> 374</span>  }</div><div class="line"><a name="l00375"></a><span class="lineno"> 375</span>  <span class="keywordflow">else</span></div><div class="line"><a name="l00376"></a><span class="lineno"> 376</span>  {</div><div class="line"><a name="l00377"></a><span class="lineno"> 377</span>  <span class="comment">// Pick up the GEMM configuration</span></div><div class="line"><a name="l00378"></a><span class="lineno"> 378</span>  std::tie(lhs_info, rhs_info) = <a class="code" href="classarm__compute_1_1cl__gemm_1_1_c_l_g_e_m_m_native_kernel_configuration_factory.xhtml#ab2b50810acd1c74616ee97dee3cbdb6b">CLGEMMNativeKernelConfigurationFactory::create</a>(gpu_target)->configure(m, n, k, batch_size, <a class="code" href="namespacearm__compute.xhtml#ad8ed01ff3ff33333d8e19db4d2818bb6af14462d71aa842202c3e4b272c7ec924">DataType::QASYMM8</a>);</div><div class="line"><a name="l00379"></a><span class="lineno"> 379</span> </div><div class="line"><a name="l00380"></a><span class="lineno"> 380</span>  <span class="comment">// Validate matrix multiply</span></div><div class="line"><a name="l00381"></a><span class="lineno"> 381</span>  <a class="code" href="_error_8h.xhtml#a8a1e1c105f0bdaf37db408c7cfcb77a4">ARM_COMPUTE_RETURN_ON_ERROR</a>(<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_native_kernel.xhtml#a96f40eda5c40e254a6729e8bc0cb8195">CLGEMMLowpMatrixMultiplyNativeKernel::validate</a>(matrix_a_info, matrix_b_info, &mm_result_s32_info, lhs_info, rhs_info, reshape_info));</div><div class="line"><a name="l00382"></a><span class="lineno"> 382</span>  }</div><div class="line"><a name="l00383"></a><span class="lineno"> 383</span>  }</div><div class="line"><a name="l00384"></a><span class="lineno"> 384</span> </div><div class="line"><a name="l00385"></a><span class="lineno"> 385</span>  <span class="comment">// Validate offset contribution kernel</span></div><div class="line"><a name="l00386"></a><span class="lineno"> 386</span>  <span class="keyword">const</span> <span class="keywordtype">size_t</span> num_filters = (gemm_info.<a class="code" href="classarm__compute_1_1_g_e_m_m_info.xhtml#a11d8f855e323a8396fe6944edcef4238">gemmlowp_output_stage</a>().<a class="code" href="structarm__compute_1_1_g_e_m_m_lowp_output_stage_info.xhtml#a94e1801be6c3d9d6645c694d7e280cda">is_quantized_per_channel</a>) ? gemm_info.<a class="code" href="classarm__compute_1_1_g_e_m_m_info.xhtml#a11d8f855e323a8396fe6944edcef4238">gemmlowp_output_stage</a>().<a class="code" href="structarm__compute_1_1_g_e_m_m_lowp_output_stage_info.xhtml#ae5bd6bebbc0c7ebd9e7dbfd47d939c2a">gemmlowp_multipliers</a>.size() : 1;</div><div class="line"><a name="l00387"></a><span class="lineno"> 387</span> </div><div class="line"><a name="l00388"></a><span class="lineno"> 388</span>  <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_tensor_info.xhtml">TensorInfo</a> gemm_output_stage_multipliers_shifts_info(<a class="code" href="classarm__compute_1_1_tensor_info.xhtml">TensorInfo</a>(<a class="code" href="classarm__compute_1_1_tensor_shape.xhtml">TensorShape</a>(num_filters), 1, <a class="code" href="namespacearm__compute.xhtml#ab4e88c89b3b7ea1735996cc4def22d58aa1e28eee0339658d39a8b4d325b56e9c">DataType::S32</a>));</div><div class="line"><a name="l00389"></a><span class="lineno"> 389</span> </div><div class="line"><a name="l00390"></a><span class="lineno"> 390</span>  <a class="code" href="structarm__compute_1_1_g_e_m_m_lowp_output_stage_info.xhtml">GEMMLowpOutputStageInfo</a> gemmlowp_output_stage = gemm_info.<a class="code" href="classarm__compute_1_1_g_e_m_m_info.xhtml#a11d8f855e323a8396fe6944edcef4238">gemmlowp_output_stage</a>();</div><div class="line"><a name="l00391"></a><span class="lineno"> 391</span>  gemmlowp_output_stage.<a class="code" href="structarm__compute_1_1_g_e_m_m_lowp_output_stage_info.xhtml#ab233758aca2751c6e71a2f79baf7b92a">output_data_type</a> = a-><a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml#a7cfb31af63202568efef5214acfbf3ba">data_type</a>();</div><div class="line"><a name="l00392"></a><span class="lineno"> 392</span>  <a class="code" href="_error_8h.xhtml#a8a1e1c105f0bdaf37db408c7cfcb77a4">ARM_COMPUTE_RETURN_ON_ERROR</a>(<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_offset_contribution_output_stage_kernel.xhtml#aad155f02e1a76694b771f82953c258c4">CLGEMMLowpOffsetContributionOutputStageKernel::validate</a>(&mm_result_s32_info,</div><div class="line"><a name="l00393"></a><span class="lineno"> 393</span>  a_offset == 0 ? <span class="keyword">nullptr</span> : &info_vector_sum_col,</div><div class="line"><a name="l00394"></a><span class="lineno"> 394</span>  b_offset == 0 ? <span class="keyword">nullptr</span> : &info_vector_sum_row,</div><div class="line"><a name="l00395"></a><span class="lineno"> 395</span>  c,</div><div class="line"><a name="l00396"></a><span class="lineno"> 396</span>  output,</div><div class="line"><a name="l00397"></a><span class="lineno"> 397</span>  a_offset, b_offset,</div><div class="line"><a name="l00398"></a><span class="lineno"> 398</span>  gemmlowp_output_stage,</div><div class="line"><a name="l00399"></a><span class="lineno"> 399</span>  &gemm_output_stage_multipliers_shifts_info,</div><div class="line"><a name="l00400"></a><span class="lineno"> 400</span>  &gemm_output_stage_multipliers_shifts_info));</div><div class="line"><a name="l00401"></a><span class="lineno"> 401</span>  }</div><div class="line"><a name="l00402"></a><span class="lineno"> 402</span>  <span class="keywordflow">else</span></div><div class="line"><a name="l00403"></a><span class="lineno"> 403</span>  {</div><div class="line"><a name="l00404"></a><span class="lineno"> 404</span>  <span class="keywordflow">if</span>(reshape_matrix_b)</div><div class="line"><a name="l00405"></a><span class="lineno"> 405</span>  {</div><div class="line"><a name="l00406"></a><span class="lineno"> 406</span>  <span class="comment">// Validate matrix multiply</span></div><div class="line"><a name="l00407"></a><span class="lineno"> 407</span>  <a class="code" href="_error_8h.xhtml#a8a1e1c105f0bdaf37db408c7cfcb77a4">ARM_COMPUTE_RETURN_ON_ERROR</a>(<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_reshaped_only_r_h_s_kernel.xhtml#a96f40eda5c40e254a6729e8bc0cb8195">CLGEMMLowpMatrixMultiplyReshapedOnlyRHSKernel::validate</a>(matrix_a_info, matrix_b_info, output, lhs_info, rhs_info, reshape_info));</div><div class="line"><a name="l00408"></a><span class="lineno"> 408</span>  }</div><div class="line"><a name="l00409"></a><span class="lineno"> 409</span>  <span class="keywordflow">else</span></div><div class="line"><a name="l00410"></a><span class="lineno"> 410</span>  {</div><div class="line"><a name="l00411"></a><span class="lineno"> 411</span>  <span class="keywordflow">if</span>(is_midgard)</div><div class="line"><a name="l00412"></a><span class="lineno"> 412</span>  {</div><div class="line"><a name="l00413"></a><span class="lineno"> 413</span>  <span class="comment">// Validate matrix multiply</span></div><div class="line"><a name="l00414"></a><span class="lineno"> 414</span>  <a class="code" href="_error_8h.xhtml#a8a1e1c105f0bdaf37db408c7cfcb77a4">ARM_COMPUTE_RETURN_ON_ERROR</a>(<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_kernel.xhtml#a9c52840c4fadda388888e26f655de61d">CLGEMMLowpMatrixMultiplyKernel::validate</a>(matrix_a_info, matrix_b_info, output, reshape_info));</div><div class="line"><a name="l00415"></a><span class="lineno"> 415</span>  }</div><div class="line"><a name="l00416"></a><span class="lineno"> 416</span>  <span class="keywordflow">else</span></div><div class="line"><a name="l00417"></a><span class="lineno"> 417</span>  {</div><div class="line"><a name="l00418"></a><span class="lineno"> 418</span>  <span class="comment">// Pick up the GEMM configuration</span></div><div class="line"><a name="l00419"></a><span class="lineno"> 419</span>  std::tie(lhs_info, rhs_info) = <a class="code" href="classarm__compute_1_1cl__gemm_1_1_c_l_g_e_m_m_native_kernel_configuration_factory.xhtml#ab2b50810acd1c74616ee97dee3cbdb6b">CLGEMMNativeKernelConfigurationFactory::create</a>(gpu_target)->configure(m, n, k, batch_size, <a class="code" href="namespacearm__compute.xhtml#ad8ed01ff3ff33333d8e19db4d2818bb6af14462d71aa842202c3e4b272c7ec924">DataType::QASYMM8</a>);</div><div class="line"><a name="l00420"></a><span class="lineno"> 420</span> </div><div class="line"><a name="l00421"></a><span class="lineno"> 421</span>  <span class="comment">// Validate matrix multiply</span></div><div class="line"><a name="l00422"></a><span class="lineno"> 422</span>  <a class="code" href="_error_8h.xhtml#a8a1e1c105f0bdaf37db408c7cfcb77a4">ARM_COMPUTE_RETURN_ON_ERROR</a>(<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_native_kernel.xhtml#a96f40eda5c40e254a6729e8bc0cb8195">CLGEMMLowpMatrixMultiplyNativeKernel::validate</a>(matrix_a_info, matrix_b_info, output, lhs_info, rhs_info, reshape_info));</div><div class="line"><a name="l00423"></a><span class="lineno"> 423</span>  }</div><div class="line"><a name="l00424"></a><span class="lineno"> 424</span>  }</div><div class="line"><a name="l00425"></a><span class="lineno"> 425</span> </div><div class="line"><a name="l00426"></a><span class="lineno"> 426</span>  <span class="keywordflow">if</span>(output-><a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml#a18064e0011c3869d884653e9e7c47b66">total_size</a>() != 0)</div><div class="line"><a name="l00427"></a><span class="lineno"> 427</span>  {</div><div class="line"><a name="l00428"></a><span class="lineno"> 428</span>  <span class="comment">// Validate offset contribution kernel</span></div><div class="line"><a name="l00429"></a><span class="lineno"> 429</span>  <a class="code" href="_error_8h.xhtml#a8a1e1c105f0bdaf37db408c7cfcb77a4">ARM_COMPUTE_RETURN_ON_ERROR</a>(<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_offset_contribution_kernel.xhtml#a0c8e9fe3b2d6411e9f75bfcfa8b8e7d2">CLGEMMLowpOffsetContributionKernel::validate</a>(output,</div><div class="line"><a name="l00430"></a><span class="lineno"> 430</span>  a_offset == 0 ? <span class="keyword">nullptr</span> : &info_vector_sum_col,</div><div class="line"><a name="l00431"></a><span class="lineno"> 431</span>  b_offset == 0 ? <span class="keyword">nullptr</span> : &info_vector_sum_row,</div><div class="line"><a name="l00432"></a><span class="lineno"> 432</span>  c,</div><div class="line"><a name="l00433"></a><span class="lineno"> 433</span>  a_offset, b_offset));</div><div class="line"><a name="l00434"></a><span class="lineno"> 434</span>  }</div><div class="line"><a name="l00435"></a><span class="lineno"> 435</span>  }</div><div class="line"><a name="l00436"></a><span class="lineno"> 436</span> </div><div class="line"><a name="l00437"></a><span class="lineno"> 437</span>  <span class="keywordflow">return</span> <a class="code" href="classarm__compute_1_1_status.xhtml">Status</a>{};</div><div class="line"><a name="l00438"></a><span class="lineno"> 438</span> }</div><div class="line"><a name="l00439"></a><span class="lineno"> 439</span> </div><div class="line"><a name="l00440"></a><span class="lineno"><a class="line" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_core.xhtml#ad1717410afd0be936c6213a63c8005fb"> 440</a></span> <span class="keywordtype">void</span> <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_core.xhtml#ad1717410afd0be936c6213a63c8005fb">CLGEMMLowpMatrixMultiplyCore::run</a>()</div><div class="line"><a name="l00441"></a><span class="lineno"> 441</span> {</div><div class="line"><a name="l00442"></a><span class="lineno"> 442</span>  <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_core.xhtml#aa9b93ef660fc3c5b4b19d3fc7b891b77">prepare</a>();</div><div class="line"><a name="l00443"></a><span class="lineno"> 443</span> </div><div class="line"><a name="l00444"></a><span class="lineno"> 444</span>  <a class="code" href="classarm__compute_1_1_memory_group_resource_scope.xhtml">MemoryGroupResourceScope</a> scope_mg(_memory_group);</div><div class="line"><a name="l00445"></a><span class="lineno"> 445</span> </div><div class="line"><a name="l00446"></a><span class="lineno"> 446</span>  <span class="keywordflow">if</span>(_is_gemm_reshaped)</div><div class="line"><a name="l00447"></a><span class="lineno"> 447</span>  {</div><div class="line"><a name="l00448"></a><span class="lineno"> 448</span>  <span class="keywordflow">if</span>(!_reshape_b_only_on_first_run)</div><div class="line"><a name="l00449"></a><span class="lineno"> 449</span>  {</div><div class="line"><a name="l00450"></a><span class="lineno"> 450</span>  <span class="comment">// Run reshape matrix B</span></div><div class="line"><a name="l00451"></a><span class="lineno"> 451</span>  <a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#a9b58d0eb9a2af8e6d7908695e1557d6c">CLScheduler::get</a>().<a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#ae1a643e517f50bf0392fb6516dd7cf67">enqueue</a>(_mtx_b_reshape_kernel, <span class="keyword">false</span>);</div><div class="line"><a name="l00452"></a><span class="lineno"> 452</span>  }</div><div class="line"><a name="l00453"></a><span class="lineno"> 453</span>  }</div><div class="line"><a name="l00454"></a><span class="lineno"> 454</span> </div><div class="line"><a name="l00455"></a><span class="lineno"> 455</span>  <span class="comment">// Run matrix B reduction kernel only if _a_offset is not equal to 0</span></div><div class="line"><a name="l00456"></a><span class="lineno"> 456</span>  <span class="keywordflow">if</span>(_a_offset != 0 && !_reshape_b_only_on_first_run)</div><div class="line"><a name="l00457"></a><span class="lineno"> 457</span>  {</div><div class="line"><a name="l00458"></a><span class="lineno"> 458</span>  <a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#a9b58d0eb9a2af8e6d7908695e1557d6c">CLScheduler::get</a>().<a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#ae1a643e517f50bf0392fb6516dd7cf67">enqueue</a>(_mtx_b_reduction_kernel, <span class="keyword">false</span>);</div><div class="line"><a name="l00459"></a><span class="lineno"> 459</span>  }</div><div class="line"><a name="l00460"></a><span class="lineno"> 460</span> </div><div class="line"><a name="l00461"></a><span class="lineno"> 461</span>  <span class="comment">// Run matrix multiply</span></div><div class="line"><a name="l00462"></a><span class="lineno"> 462</span>  <span class="keywordflow">if</span>(_is_gemm_reshaped)</div><div class="line"><a name="l00463"></a><span class="lineno"> 463</span>  {</div><div class="line"><a name="l00464"></a><span class="lineno"> 464</span>  <a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#a9b58d0eb9a2af8e6d7908695e1557d6c">CLScheduler::get</a>().<a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#ae1a643e517f50bf0392fb6516dd7cf67">enqueue</a>(_mm_reshaped_only_rhs_kernel, <span class="keyword">false</span>);</div><div class="line"><a name="l00465"></a><span class="lineno"> 465</span>  }</div><div class="line"><a name="l00466"></a><span class="lineno"> 466</span>  <span class="keywordflow">else</span></div><div class="line"><a name="l00467"></a><span class="lineno"> 467</span>  {</div><div class="line"><a name="l00468"></a><span class="lineno"> 468</span>  <span class="keywordflow">if</span>(_is_midgard)</div><div class="line"><a name="l00469"></a><span class="lineno"> 469</span>  {</div><div class="line"><a name="l00470"></a><span class="lineno"> 470</span>  <a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#a9b58d0eb9a2af8e6d7908695e1557d6c">CLScheduler::get</a>().<a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#ae1a643e517f50bf0392fb6516dd7cf67">enqueue</a>(_mm_midgard_kernel, <span class="keyword">false</span>);</div><div class="line"><a name="l00471"></a><span class="lineno"> 471</span>  }</div><div class="line"><a name="l00472"></a><span class="lineno"> 472</span>  <span class="keywordflow">else</span></div><div class="line"><a name="l00473"></a><span class="lineno"> 473</span>  {</div><div class="line"><a name="l00474"></a><span class="lineno"> 474</span>  <a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#a9b58d0eb9a2af8e6d7908695e1557d6c">CLScheduler::get</a>().<a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#ae1a643e517f50bf0392fb6516dd7cf67">enqueue</a>(_mm_native_kernel, <span class="keyword">false</span>);</div><div class="line"><a name="l00475"></a><span class="lineno"> 475</span>  }</div><div class="line"><a name="l00476"></a><span class="lineno"> 476</span>  }</div><div class="line"><a name="l00477"></a><span class="lineno"> 477</span> </div><div class="line"><a name="l00478"></a><span class="lineno"> 478</span>  <span class="comment">// Run matrix A reduction kernel only if _b_offset is not equal to 0</span></div><div class="line"><a name="l00479"></a><span class="lineno"> 479</span>  <span class="keywordflow">if</span>(_b_offset != 0)</div><div class="line"><a name="l00480"></a><span class="lineno"> 480</span>  {</div><div class="line"><a name="l00481"></a><span class="lineno"> 481</span>  <a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#a9b58d0eb9a2af8e6d7908695e1557d6c">CLScheduler::get</a>().<a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#ae1a643e517f50bf0392fb6516dd7cf67">enqueue</a>(_mtx_a_reduction_kernel, <span class="keyword">false</span>);</div><div class="line"><a name="l00482"></a><span class="lineno"> 482</span>  }</div><div class="line"><a name="l00483"></a><span class="lineno"> 483</span> </div><div class="line"><a name="l00484"></a><span class="lineno"> 484</span>  <span class="keywordflow">if</span>(_fuse_output_stage)</div><div class="line"><a name="l00485"></a><span class="lineno"> 485</span>  {</div><div class="line"><a name="l00486"></a><span class="lineno"> 486</span>  <span class="comment">// Run offset contribution/output stage kernel</span></div><div class="line"><a name="l00487"></a><span class="lineno"> 487</span>  <a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#a9b58d0eb9a2af8e6d7908695e1557d6c">CLScheduler::get</a>().<a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#ae1a643e517f50bf0392fb6516dd7cf67">enqueue</a>(_offset_contribution_output_stage_kernel, <span class="keyword">true</span>);</div><div class="line"><a name="l00488"></a><span class="lineno"> 488</span>  }</div><div class="line"><a name="l00489"></a><span class="lineno"> 489</span>  <span class="keywordflow">else</span></div><div class="line"><a name="l00490"></a><span class="lineno"> 490</span>  {</div><div class="line"><a name="l00491"></a><span class="lineno"> 491</span>  <span class="comment">// Run offset contribution kernel</span></div><div class="line"><a name="l00492"></a><span class="lineno"> 492</span>  <a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#a9b58d0eb9a2af8e6d7908695e1557d6c">CLScheduler::get</a>().<a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#ae1a643e517f50bf0392fb6516dd7cf67">enqueue</a>(_offset_contribution_kernel, <span class="keyword">true</span>);</div><div class="line"><a name="l00493"></a><span class="lineno"> 493</span>  }</div><div class="line"><a name="l00494"></a><span class="lineno"> 494</span> }</div><div class="line"><a name="l00495"></a><span class="lineno"> 495</span> </div><div class="line"><a name="l00496"></a><span class="lineno"><a class="line" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_core.xhtml#aa9b93ef660fc3c5b4b19d3fc7b891b77"> 496</a></span> <span class="keywordtype">void</span> <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_core.xhtml#aa9b93ef660fc3c5b4b19d3fc7b891b77">CLGEMMLowpMatrixMultiplyCore::prepare</a>()</div><div class="line"><a name="l00497"></a><span class="lineno"> 497</span> {</div><div class="line"><a name="l00498"></a><span class="lineno"> 498</span>  <span class="keywordflow">if</span>(!_is_prepared)</div><div class="line"><a name="l00499"></a><span class="lineno"> 499</span>  {</div><div class="line"><a name="l00500"></a><span class="lineno"> 500</span>  <span class="keywordflow">if</span>(_convert_to_qasymm8)</div><div class="line"><a name="l00501"></a><span class="lineno"> 501</span>  {</div><div class="line"><a name="l00502"></a><span class="lineno"> 502</span>  _qasymm8_weights.<a class="code" href="classarm__compute_1_1_c_l_tensor.xhtml#a4083de30daebd6bdee6b35d9c8262108">allocator</a>()-><a class="code" href="classarm__compute_1_1_c_l_tensor_allocator.xhtml#a6e509c2a177b0b29e9e2369535094dee">allocate</a>();</div><div class="line"><a name="l00503"></a><span class="lineno"> 503</span>  <a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#a9b58d0eb9a2af8e6d7908695e1557d6c">CLScheduler::get</a>().<a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#ae1a643e517f50bf0392fb6516dd7cf67">enqueue</a>(_weights_to_qasymm8, <span class="keyword">false</span>);</div><div class="line"><a name="l00504"></a><span class="lineno"> 504</span>  }</div><div class="line"><a name="l00505"></a><span class="lineno"> 505</span> </div><div class="line"><a name="l00506"></a><span class="lineno"> 506</span>  <span class="keywordflow">if</span>(_is_gemm_reshaped && _reshape_b_only_on_first_run)</div><div class="line"><a name="l00507"></a><span class="lineno"> 507</span>  {</div><div class="line"><a name="l00508"></a><span class="lineno"> 508</span>  <a class="code" href="_error_8h.xhtml#a54a6080c9f4df1f908e57a9bbb46f5da">ARM_COMPUTE_ERROR_ON</a>(!_original_b-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a209ea2ddfdfa80703799c92da8beb643">is_used</a>());</div><div class="line"><a name="l00509"></a><span class="lineno"> 509</span> </div><div class="line"><a name="l00510"></a><span class="lineno"> 510</span>  <span class="comment">// Run reshape kernel and mark original weights tensor as unused</span></div><div class="line"><a name="l00511"></a><span class="lineno"> 511</span>  _tmp_b.<a class="code" href="classarm__compute_1_1_c_l_tensor.xhtml#a4083de30daebd6bdee6b35d9c8262108">allocator</a>()-><a class="code" href="classarm__compute_1_1_c_l_tensor_allocator.xhtml#a6e509c2a177b0b29e9e2369535094dee">allocate</a>();</div><div class="line"><a name="l00512"></a><span class="lineno"> 512</span>  <a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#a9b58d0eb9a2af8e6d7908695e1557d6c">CLScheduler::get</a>().<a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#ae1a643e517f50bf0392fb6516dd7cf67">enqueue</a>(_mtx_b_reshape_kernel, <span class="keyword">false</span>);</div><div class="line"><a name="l00513"></a><span class="lineno"> 513</span>  _original_b-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a9bc00234de9adf8c99a21eb1d7d494c2">mark_as_unused</a>();</div><div class="line"><a name="l00514"></a><span class="lineno"> 514</span>  }</div><div class="line"><a name="l00515"></a><span class="lineno"> 515</span> </div><div class="line"><a name="l00516"></a><span class="lineno"> 516</span>  <span class="comment">// Run matrix B reduction kernel only if _a_offset is not equal to 0</span></div><div class="line"><a name="l00517"></a><span class="lineno"> 517</span>  <span class="keywordflow">if</span>(_a_offset != 0 && _reshape_b_only_on_first_run)</div><div class="line"><a name="l00518"></a><span class="lineno"> 518</span>  {</div><div class="line"><a name="l00519"></a><span class="lineno"> 519</span>  _vector_sum_col.<a class="code" href="classarm__compute_1_1_c_l_tensor.xhtml#a4083de30daebd6bdee6b35d9c8262108">allocator</a>()-><a class="code" href="classarm__compute_1_1_c_l_tensor_allocator.xhtml#a6e509c2a177b0b29e9e2369535094dee">allocate</a>();</div><div class="line"><a name="l00520"></a><span class="lineno"> 520</span>  <a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#a9b58d0eb9a2af8e6d7908695e1557d6c">CLScheduler::get</a>().<a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#ae1a643e517f50bf0392fb6516dd7cf67">enqueue</a>(_mtx_b_reduction_kernel, <span class="keyword">false</span>);</div><div class="line"><a name="l00521"></a><span class="lineno"> 521</span>  }</div><div class="line"><a name="l00522"></a><span class="lineno"> 522</span> </div><div class="line"><a name="l00523"></a><span class="lineno"> 523</span>  <a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#a9b58d0eb9a2af8e6d7908695e1557d6c">CLScheduler::get</a>().<a class="code" href="classarm__compute_1_1_c_l_scheduler.xhtml#a398a2582c746d28fc125487a44c9ed74">queue</a>().finish();</div><div class="line"><a name="l00524"></a><span class="lineno"> 524</span>  _is_prepared = <span class="keyword">true</span>;</div><div class="line"><a name="l00525"></a><span class="lineno"> 525</span>  }</div><div class="line"><a name="l00526"></a><span class="lineno"> 526</span> }</div><div class="line"><a name="l00527"></a><span class="lineno"> 527</span> } <span class="comment">// namespace arm_compute</span></div><div class="ttc" id="classarm__compute_1_1_i_tensor_xhtml_adbd73147d41e8a640bc299d12613c31e"><div class="ttname"><a href="classarm__compute_1_1_i_tensor.xhtml#adbd73147d41e8a640bc299d12613c31e">arm_compute::ITensor::ptr_to_element</a></div><div class="ttdeci">uint8_t * ptr_to_element(const Coordinates &id) const</div><div class="ttdoc">Return a pointer to the element at the passed coordinates.</div><div class="ttdef"><b>Definition:</b> <a href="_i_tensor_8h_source.xhtml#l00063">ITensor.h:63</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_tensor_shape_xhtml"><div class="ttname"><a href="classarm__compute_1_1_tensor_shape.xhtml">arm_compute::TensorShape</a></div><div class="ttdoc">Shape of a tensor.</div><div class="ttdef"><b>Definition:</b> <a href="_tensor_shape_8h_source.xhtml#l00039">TensorShape.h:39</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_offset_contribution_kernel_xhtml_a094cc43211df267051139e766f002d77"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_offset_contribution_kernel.xhtml#a094cc43211df267051139e766f002d77">arm_compute::CLGEMMLowpOffsetContributionKernel::configure</a></div><div class="ttdeci">void configure(ICLTensor *mm_result, const ICLTensor *vector_sum_col, const ICLTensor *vector_sum_row, const ICLTensor *bias, int32_t k, int32_t a_offset, int32_t b_offset)</div><div class="ttdoc">Initialise the kernel's input and output.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_offset_contribution_kernel_8cpp_source.xhtml#l00145">CLGEMMLowpOffsetContributionKernel.cpp:145</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_core_xhtml_aa9b93ef660fc3c5b4b19d3fc7b891b77"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_core.xhtml#aa9b93ef660fc3c5b4b19d3fc7b891b77">arm_compute::CLGEMMLowpMatrixMultiplyCore::prepare</a></div><div class="ttdeci">void prepare() override</div><div class="ttdoc">Prepare the function for executing.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_matrix_multiply_core_8cpp_source.xhtml#l00496">CLGEMMLowpMatrixMultiplyCore.cpp:496</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_tensor_info_xhtml_a178f0d3d87f959e00a743328d95359d2"><div class="ttname"><a href="classarm__compute_1_1_i_tensor_info.xhtml#a178f0d3d87f959e00a743328d95359d2">arm_compute::ITensorInfo::dimension</a></div><div class="ttdeci">virtual size_t dimension(size_t index) const =0</div><div class="ttdoc">Return the size of the requested dimension.</div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_offset_contribution_kernel_xhtml_a0c8e9fe3b2d6411e9f75bfcfa8b8e7d2"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_offset_contribution_kernel.xhtml#a0c8e9fe3b2d6411e9f75bfcfa8b8e7d2">arm_compute::CLGEMMLowpOffsetContributionKernel::validate</a></div><div class="ttdeci">static Status validate(const ITensorInfo *mm_result, const ITensorInfo *vector_sum_col, const ITensorInfo *vector_sum_row, const ITensorInfo *bias, int32_t a_offset, int32_t b_offset)</div><div class="ttdoc">Static function to check if given info will lead to a valid configuration of CLGEMMLowpOffsetContribu...</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_offset_contribution_kernel_8cpp_source.xhtml#l00205">CLGEMMLowpOffsetContributionKernel.cpp:205</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_depth_convert_layer_kernel_xhtml_a31fdffa6cbd16d74f8c2da415e6f4912"><div class="ttname"><a href="classarm__compute_1_1_c_l_depth_convert_layer_kernel.xhtml#a31fdffa6cbd16d74f8c2da415e6f4912">arm_compute::CLDepthConvertLayerKernel::validate</a></div><div class="ttdeci">static Status validate(const ITensorInfo *input, const ITensorInfo *output, ConvertPolicy policy, uint32_t shift)</div><div class="ttdoc">Static function to check if given info will lead to a valid configuration of CLDepthConvertLayerKerne...</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_depth_convert_layer_kernel_8cpp_source.xhtml#l00130">CLDepthConvertLayerKernel.cpp:130</a></div></div> |
| <div class="ttc" id="namespacearm__compute_1_1test_1_1validation_xhtml_aa76b4a6e74940dabc5b7fc6b2dab3545"><div class="ttname"><a href="namespacearm__compute_1_1test_1_1validation.xhtml#aa76b4a6e74940dabc5b7fc6b2dab3545">arm_compute::test::validation::b</a></div><div class="ttdeci">SimpleTensor< float > b</div><div class="ttdef"><b>Definition:</b> <a href="_c_p_p_2_d_f_t_8cpp_source.xhtml#l00157">DFT.cpp:157</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_scheduler_xhtml_a9b58d0eb9a2af8e6d7908695e1557d6c"><div class="ttname"><a href="classarm__compute_1_1_c_l_scheduler.xhtml#a9b58d0eb9a2af8e6d7908695e1557d6c">arm_compute::CLScheduler::get</a></div><div class="ttdeci">static CLScheduler & get()</div><div class="ttdoc">Access the scheduler singleton.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_scheduler_8cpp_source.xhtml#l00099">CLScheduler.cpp:99</a></div></div> |
| <div class="ttc" id="namespacearm__compute_1_1test_1_1validation_xhtml_a7cb842ebfe255726066039853a4322f0"><div class="ttname"><a href="namespacearm__compute_1_1test_1_1validation.xhtml#a7cb842ebfe255726066039853a4322f0">arm_compute::test::validation::weights_info</a></div><div class="ttdeci">weights_info</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_2_batch_normalization_layer_8cpp_source.xhtml#l00196">BatchNormalizationLayer.cpp:196</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_a5558e2cc22f7f4771653d992c8ad8864ab50339a10e1de285ac99d4c3990b8693"><div class="ttname"><a href="namespacearm__compute.xhtml#a5558e2cc22f7f4771653d992c8ad8864ab50339a10e1de285ac99d4c3990b8693">arm_compute::GEMMLowpOutputStageType::NONE</a></div><div class="ttdoc">No quantization.</div></div> |
| <div class="ttc" id="_validate_8h_xhtml_a8f3ff7da485ff7e75dab07baadf5b4bd"><div class="ttname"><a href="_validate_8h.xhtml#a8f3ff7da485ff7e75dab07baadf5b4bd">ARM_COMPUTE_RETURN_ERROR_ON_MISMATCHING_DATA_TYPES</a></div><div class="ttdeci">#define ARM_COMPUTE_RETURN_ERROR_ON_MISMATCHING_DATA_TYPES(...)</div><div class="ttdef"><b>Definition:</b> <a href="_validate_8h_source.xhtml#l00545">Validate.h:545</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_scheduler_xhtml_a27561688e2fc60176608ef725a4ecb30"><div class="ttname"><a href="classarm__compute_1_1_c_l_scheduler.xhtml#a27561688e2fc60176608ef725a4ecb30">arm_compute::CLScheduler::target</a></div><div class="ttdeci">GPUTarget target() const</div><div class="ttdoc">Get the target GPU.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_scheduler_8cpp_source.xhtml#l00047">CLScheduler.cpp:47</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_g_e_m_m_reshape_info_xhtml"><div class="ttname"><a href="classarm__compute_1_1_g_e_m_m_reshape_info.xhtml">arm_compute::GEMMReshapeInfo</a></div><div class="ttdoc">GEMM reshape information class.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_types_8h_source.xhtml#l01823">Types.h:1823</a></div></div> |
| <div class="ttc" id="_error_8h_xhtml_a8a1e1c105f0bdaf37db408c7cfcb77a4"><div class="ttname"><a href="_error_8h.xhtml#a8a1e1c105f0bdaf37db408c7cfcb77a4">ARM_COMPUTE_RETURN_ON_ERROR</a></div><div class="ttdeci">#define ARM_COMPUTE_RETURN_ON_ERROR(status)</div><div class="ttdoc">Checks if a status contains an error and returns it.</div><div class="ttdef"><b>Definition:</b> <a href="_error_8h_source.xhtml#l00204">Error.h:204</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_tensor_info_xhtml_a7cfb31af63202568efef5214acfbf3ba"><div class="ttname"><a href="classarm__compute_1_1_i_tensor_info.xhtml#a7cfb31af63202568efef5214acfbf3ba">arm_compute::ITensorInfo::data_type</a></div><div class="ttdeci">virtual DataType data_type() const =0</div><div class="ttdoc">Data type used for each element of the tensor.</div></div> |
| <div class="ttc" id="namespacearm__compute_1_1misc_1_1shape__calculator_xhtml_adca241b012a5e00ddfcdc5a8db05a2a3"><div class="ttname"><a href="namespacearm__compute_1_1misc_1_1shape__calculator.xhtml#adca241b012a5e00ddfcdc5a8db05a2a3">arm_compute::misc::shape_calculator::compute_mm_shape</a></div><div class="ttdeci">TensorShape compute_mm_shape(const ITensorInfo &input0, const ITensorInfo &input1, bool is_interleaved_transposed, const GEMMReshapeInfo &reshape_info)</div><div class="ttdoc">Calculate the matrix multiplication output shape of two tensors.</div><div class="ttdef"><b>Definition:</b> <a href="_shape_calculator_8h_source.xhtml#l00858">ShapeCalculator.h:858</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_tensor_xhtml_a209ea2ddfdfa80703799c92da8beb643"><div class="ttname"><a href="classarm__compute_1_1_i_tensor.xhtml#a209ea2ddfdfa80703799c92da8beb643">arm_compute::ITensor::is_used</a></div><div class="ttdeci">bool is_used() const</div><div class="ttdoc">Flags if the tensor is used or not.</div><div class="ttdef"><b>Definition:</b> <a href="_i_tensor_8cpp_source.xhtml#l00162">ITensor.cpp:162</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_g_e_m_m_info_xhtml_a11d8f855e323a8396fe6944edcef4238"><div class="ttname"><a href="classarm__compute_1_1_g_e_m_m_info.xhtml#a11d8f855e323a8396fe6944edcef4238">arm_compute::GEMMInfo::gemmlowp_output_stage</a></div><div class="ttdeci">GEMMLowpOutputStageInfo gemmlowp_output_stage() const</div><div class="ttdoc">GEMMLowp output stage.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_types_8h_source.xhtml#l02086">Types.h:2086</a></div></div> |
| <div class="ttc" id="namespacearm__compute_1_1misc_1_1shape__calculator_xhtml_a60ce6c017f70d978b48b101ce314969e"><div class="ttname"><a href="namespacearm__compute_1_1misc_1_1shape__calculator.xhtml#a60ce6c017f70d978b48b101ce314969e">arm_compute::misc::shape_calculator::compute_reductionA_shape</a></div><div class="ttdeci">TensorShape compute_reductionA_shape(const ITensorInfo &b)</div><div class="ttdoc">Calculate the reductionA shape used in GEMMLowp.</div><div class="ttdef"><b>Definition:</b> <a href="_shape_calculator_8h_source.xhtml#l00357">ShapeCalculator.h:357</a></div></div> |
| <div class="ttc" id="_validate_8h_xhtml_ae7eed178dac535c6e727061b1f5bc6eb"><div class="ttname"><a href="_validate_8h.xhtml#ae7eed178dac535c6e727061b1f5bc6eb">ARM_COMPUTE_RETURN_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN</a></div><div class="ttdeci">#define ARM_COMPUTE_RETURN_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(t, c,...)</div><div class="ttdef"><b>Definition:</b> <a href="_validate_8h_source.xhtml#l00792">Validate.h:792</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_offset_contribution_output_stage_kernel_xhtml_ac727ebddb913babf19047a6444b89fba"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_offset_contribution_output_stage_kernel.xhtml#ac727ebddb913babf19047a6444b89fba">arm_compute::CLGEMMLowpOffsetContributionOutputStageKernel::configure</a></div><div class="ttdeci">void configure(const ICLTensor *mm_result, const ICLTensor *vector_sum_col, const ICLTensor *vector_sum_row, const ICLTensor *bias, ICLTensor *output, int32_t k, int32_t a_offset, int32_t b_offset, const GEMMLowpOutputStageInfo &output_stage, const ICLTensor *output_multipliers, const ICLTensor *output_shifts)</div><div class="ttdoc">Initialise the kernel's input and output.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_offset_contribution_output_stage_kernel_8cpp_source.xhtml#l00196">CLGEMMLowpOffsetContributionOutputStageKernel.cpp:196</a></div></div> |
| <div class="ttc" id="namespacearm__compute_1_1cl__gemm_xhtml"><div class="ttname"><a href="namespacearm__compute_1_1cl__gemm.xhtml">arm_compute::cl_gemm</a></div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_helpers_8h_source.xhtml#l00031">CLGEMMHelpers.h:31</a></div></div> |
| <div class="ttc" id="_error_8h_xhtml_a54a6080c9f4df1f908e57a9bbb46f5da"><div class="ttname"><a href="_error_8h.xhtml#a54a6080c9f4df1f908e57a9bbb46f5da">ARM_COMPUTE_ERROR_ON</a></div><div class="ttdeci">#define ARM_COMPUTE_ERROR_ON(cond)</div><div class="ttdoc">If the condition is true then an error message is printed and an exception thrown.</div><div class="ttdef"><b>Definition:</b> <a href="_error_8h_source.xhtml#l00466">Error.h:466</a></div></div> |
| <div class="ttc" id="structarm__compute_1_1_g_e_m_m_l_h_s_matrix_info_xhtml"><div class="ttname"><a href="structarm__compute_1_1_g_e_m_m_l_h_s_matrix_info.xhtml">arm_compute::GEMMLHSMatrixInfo</a></div><div class="ttdoc">GEMM LHS (Left Hand Side) matrix information.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_types_8h_source.xhtml#l01959">Types.h:1959</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_tensor_info_xhtml"><div class="ttname"><a href="classarm__compute_1_1_i_tensor_info.xhtml">arm_compute::ITensorInfo</a></div><div class="ttdoc">Store the tensor's metadata.</div><div class="ttdef"><b>Definition:</b> <a href="_i_tensor_info_8h_source.xhtml#l00040">ITensorInfo.h:40</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_a_reduction_kernel_xhtml_a46e5929d20bb4cc0678590dfa251df9e"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_a_reduction_kernel.xhtml#a46e5929d20bb4cc0678590dfa251df9e">arm_compute::CLGEMMLowpMatrixAReductionKernel::validate</a></div><div class="ttdeci">static Status validate(const ITensorInfo *mtx_a, const ITensorInfo *vector_sum_row)</div><div class="ttdoc">Static function to check if given info will lead to a valid configuration of CLGEMMLowpMatrixAReducti...</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_reduction_kernel_8cpp_source.xhtml#l00139">CLGEMMLowpReductionKernel.cpp:139</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_tensor_xhtml_a4083de30daebd6bdee6b35d9c8262108"><div class="ttname"><a href="classarm__compute_1_1_c_l_tensor.xhtml#a4083de30daebd6bdee6b35d9c8262108">arm_compute::CLTensor::allocator</a></div><div class="ttdeci">CLTensorAllocator * allocator()</div><div class="ttdoc">Return a pointer to the tensor's allocator.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_tensor_8cpp_source.xhtml#l00061">CLTensor.cpp:61</a></div></div> |
| <div class="ttc" id="_error_8h_xhtml_a938dcd406ce611ef5345ad2531cdb948"><div class="ttname"><a href="_error_8h.xhtml#a938dcd406ce611ef5345ad2531cdb948">ARM_COMPUTE_ERROR_THROW_ON</a></div><div class="ttdeci">#define ARM_COMPUTE_ERROR_THROW_ON(status)</div><div class="ttdef"><b>Definition:</b> <a href="_error_8h_source.xhtml#l00455">Error.h:455</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_core_xhtml_a0ae54876c8c3394f2e207f6b884f2b58"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_core.xhtml#a0ae54876c8c3394f2e207f6b884f2b58">arm_compute::CLGEMMLowpMatrixMultiplyCore::configure</a></div><div class="ttdeci">void configure(const ICLTensor *a, const ICLTensor *b, const ICLTensor *c, ICLTensor *output, const GEMMInfo &gemm_info=GEMMInfo())</div><div class="ttdoc">Initialise the kernel's inputs, output.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_matrix_multiply_core_8cpp_source.xhtml#l00083">CLGEMMLowpMatrixMultiplyCore.cpp:83</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_g_e_m_m_info_xhtml_abbd888f118c2209bf7578eb4f8942a07"><div class="ttname"><a href="classarm__compute_1_1_g_e_m_m_info.xhtml#abbd888f118c2209bf7578eb4f8942a07">arm_compute::GEMMInfo::depth_output_gemm3d</a></div><div class="ttdeci">int depth_output_gemm3d() const</div><div class="ttdoc">Depth of the output when GEMM output is reinterpreted as 3D tensor.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_types_8h_source.xhtml#l02062">Types.h:2062</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_status_xhtml"><div class="ttname"><a href="classarm__compute_1_1_status.xhtml">arm_compute::Status</a></div><div class="ttdoc">Status class.</div><div class="ttdef"><b>Definition:</b> <a href="_error_8h_source.xhtml#l00052">Error.h:52</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_a2355c2bf5d1950088937416baea24fe6"><div class="ttname"><a href="namespacearm__compute.xhtml#a2355c2bf5d1950088937416baea24fe6">arm_compute::get_arch_from_target</a></div><div class="ttdeci">GPUTarget get_arch_from_target(GPUTarget target)</div><div class="ttdoc">Helper function to get the GPU arch.</div><div class="ttdef"><b>Definition:</b> <a href="src_2core_2_g_p_u_target_8cpp_source.xhtml#l00189">GPUTarget.cpp:189</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_core_xhtml_ad1717410afd0be936c6213a63c8005fb"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_core.xhtml#ad1717410afd0be936c6213a63c8005fb">arm_compute::CLGEMMLowpMatrixMultiplyCore::run</a></div><div class="ttdeci">void run() override</div><div class="ttdoc">Run the kernels contained in the function.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_matrix_multiply_core_8cpp_source.xhtml#l00440">CLGEMMLowpMatrixMultiplyCore.cpp:440</a></div></div> |
| <div class="ttc" id="_tensor_info_8h_xhtml"><div class="ttname"><a href="_tensor_info_8h.xhtml">TensorInfo.h</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_a922cd85c475c0d725fc030116ea57979"><div class="ttname"><a href="namespacearm__compute.xhtml#a922cd85c475c0d725fc030116ea57979">arm_compute::is_data_type_quantized_symmetric</a></div><div class="ttdeci">bool is_data_type_quantized_symmetric(DataType dt)</div><div class="ttdoc">Check if a given data type is of symmetric quantized type.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_utils_8h_source.xhtml#l01175">Utils.h:1175</a></div></div> |
| <div class="ttc" id="structarm__compute_1_1_g_e_m_m_lowp_output_stage_info_xhtml_a6e019ad85979fd73c74f97e5483faf35"><div class="ttname"><a href="structarm__compute_1_1_g_e_m_m_lowp_output_stage_info.xhtml#a6e019ad85979fd73c74f97e5483faf35">arm_compute::GEMMLowpOutputStageInfo::type</a></div><div class="ttdeci">GEMMLowpOutputStageType type</div><div class="ttdoc">GEMMLowp output stage type.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_types_8h_source.xhtml#l01946">Types.h:1946</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_tensor_allocator_xhtml_af36143939a43fa124312e395975091ed"><div class="ttname"><a href="classarm__compute_1_1_i_tensor_allocator.xhtml#af36143939a43fa124312e395975091ed">arm_compute::ITensorAllocator::init</a></div><div class="ttdeci">void init(const TensorInfo &input, size_t alignment=0)</div><div class="ttdoc">Initialize a tensor based on the passed TensorInfo.</div><div class="ttdef"><b>Definition:</b> <a href="_i_tensor_allocator_8cpp_source.xhtml#l00038">ITensorAllocator.cpp:38</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml"><div class="ttname"><a href="namespacearm__compute.xhtml">arm_compute</a></div><div class="ttdoc">Copyright (c) 2017-2020 ARM Limited.</div><div class="ttdef"><b>Definition:</b> <a href="00__introduction_8dox_source.xhtml#l00024">00_introduction.dox:24</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_a47be6fa38308d0003c25b60b7dbc45ce"><div class="ttname"><a href="namespacearm__compute.xhtml#a47be6fa38308d0003c25b60b7dbc45ce">arm_compute::auto_init_if_empty</a></div><div class="ttdeci">bool auto_init_if_empty(ITensorInfo &info, const TensorShape &shape, int num_channels, DataType data_type, QuantizationInfo quantization_info=QuantizationInfo())</div><div class="ttdoc">Auto initialize the tensor info (shape, number of channels and data type) if the current assignment i...</div><div class="ttdef"><b>Definition:</b> <a href="_helpers_8inl_source.xhtml#l00202">Helpers.inl:202</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_g_e_m_m_info_xhtml_a77964edb8d16bb8ec14ddd8985e03cb0"><div class="ttname"><a href="classarm__compute_1_1_g_e_m_m_info.xhtml#a77964edb8d16bb8ec14ddd8985e03cb0">arm_compute::GEMMInfo::is_b_reshaped</a></div><div class="ttdeci">bool is_b_reshaped() const</div><div class="ttdoc">Flag which specifies if the matrix B has been reshaped.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_types_8h_source.xhtml#l02044">Types.h:2044</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_tensor_xhtml_a14c53d2d17be6fa8a2c9861527c7b002"><div class="ttname"><a href="classarm__compute_1_1_c_l_tensor.xhtml#a14c53d2d17be6fa8a2c9861527c7b002">arm_compute::CLTensor::map</a></div><div class="ttdeci">void map(bool blocking=true)</div><div class="ttdoc">Enqueue a map operation of the allocated buffer.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_tensor_8cpp_source.xhtml#l00066">CLTensor.cpp:66</a></div></div> |
| <div class="ttc" id="structarm__compute_1_1_g_e_m_m_lowp_output_stage_info_xhtml_a94e1801be6c3d9d6645c694d7e280cda"><div class="ttname"><a href="structarm__compute_1_1_g_e_m_m_lowp_output_stage_info.xhtml#a94e1801be6c3d9d6645c694d7e280cda">arm_compute::GEMMLowpOutputStageInfo::is_quantized_per_channel</a></div><div class="ttdeci">bool is_quantized_per_channel</div><div class="ttdoc">GEMMLowp quantized per-channel flag.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_types_8h_source.xhtml#l01954">Types.h:1954</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_depth_convert_layer_kernel_xhtml_a536f0de1a70ea80d40b0a24657ac5290"><div class="ttname"><a href="classarm__compute_1_1_c_l_depth_convert_layer_kernel.xhtml#a536f0de1a70ea80d40b0a24657ac5290">arm_compute::CLDepthConvertLayerKernel::configure</a></div><div class="ttdeci">void configure(const ICLTensor *input, ICLTensor *output, ConvertPolicy policy, uint32_t shift)</div><div class="ttdoc">Set the input and output of the kernel.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_depth_convert_layer_kernel_8cpp_source.xhtml#l00074">CLDepthConvertLayerKernel.cpp:74</a></div></div> |
| <div class="ttc" id="structarm__compute_1_1_g_e_m_m_lowp_output_stage_info_xhtml_ab269b182588a158cd256f9d4bb2a00dd"><div class="ttname"><a href="structarm__compute_1_1_g_e_m_m_lowp_output_stage_info.xhtml#ab269b182588a158cd256f9d4bb2a00dd">arm_compute::GEMMLowpOutputStageInfo::gemmlowp_shifts</a></div><div class="ttdeci">std::vector< int32_t > gemmlowp_shifts</div><div class="ttdoc">GEMMLowp output stage multiplier used for quantizing to QASYMM8.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_types_8h_source.xhtml#l01953">Types.h:1953</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_tensor_xhtml_a9bc00234de9adf8c99a21eb1d7d494c2"><div class="ttname"><a href="classarm__compute_1_1_i_tensor.xhtml#a9bc00234de9adf8c99a21eb1d7d494c2">arm_compute::ITensor::mark_as_unused</a></div><div class="ttdeci">void mark_as_unused() const</div><div class="ttdoc">Marks a tensor as unused.</div><div class="ttdef"><b>Definition:</b> <a href="_i_tensor_8cpp_source.xhtml#l00167">ITensor.cpp:167</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_ab4e88c89b3b7ea1735996cc4def22d58aa1e28eee0339658d39a8b4d325b56e9c"><div class="ttname"><a href="namespacearm__compute.xhtml#ab4e88c89b3b7ea1735996cc4def22d58aa1e28eee0339658d39a8b4d325b56e9c">arm_compute::Format::S32</a></div><div class="ttdoc">1 channel, 1 S32 per channel</div></div> |
| <div class="ttc" id="classarm__compute_1_1_memory_group_xhtml_a6fc0a49304c152c20a0f6df0634fb3cd"><div class="ttname"><a href="classarm__compute_1_1_memory_group.xhtml#a6fc0a49304c152c20a0f6df0634fb3cd">arm_compute::MemoryGroup::manage</a></div><div class="ttdeci">void manage(IMemoryManageable *obj) override</div><div class="ttdoc">Sets a object to be managed by the given memory group.</div><div class="ttdef"><b>Definition:</b> <a href="_memory_group_8h_source.xhtml#l00079">MemoryGroup.h:79</a></div></div> |
| <div class="ttc" id="_c_l_scheduler_8h_xhtml"><div class="ttname"><a href="_c_l_scheduler_8h.xhtml">CLScheduler.h</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_offset_contribution_output_stage_kernel_xhtml_aad155f02e1a76694b771f82953c258c4"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_offset_contribution_output_stage_kernel.xhtml#aad155f02e1a76694b771f82953c258c4">arm_compute::CLGEMMLowpOffsetContributionOutputStageKernel::validate</a></div><div class="ttdeci">static Status validate(const ITensorInfo *mm_result, const ITensorInfo *vector_sum_col, const ITensorInfo *vector_sum_row, const ITensorInfo *bias, const ITensorInfo *output, int32_t a_offset, int32_t b_offset, const GEMMLowpOutputStageInfo &output_stage, const ITensorInfo *output_multipliers, const ITensorInfo *output_shifts)</div><div class="ttdoc">Static function to check if given info will lead to a valid configuration of CLGEMMLowpOffsetContribu...</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_offset_contribution_output_stage_kernel_8cpp_source.xhtml#l00280">CLGEMMLowpOffsetContributionOutputStageKernel.cpp:280</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_core_xhtml_a8c3cf2d65afb288e39909171ada19566"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_core.xhtml#a8c3cf2d65afb288e39909171ada19566">arm_compute::CLGEMMLowpMatrixMultiplyCore::validate</a></div><div class="ttdeci">static Status validate(const ITensorInfo *a, const ITensorInfo *b, const ITensorInfo *c, const ITensorInfo *output, const GEMMInfo &gemm_info=GEMMInfo())</div><div class="ttdoc">Static function to check if given info will lead to a valid configuration of CLGEMMLowpMatrixMultiply...</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_matrix_multiply_core_8cpp_source.xhtml#l00274">CLGEMMLowpMatrixMultiplyCore.cpp:274</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1cl__gemm_1_1_c_l_g_e_m_m_native_kernel_configuration_factory_xhtml_ab2b50810acd1c74616ee97dee3cbdb6b"><div class="ttname"><a href="classarm__compute_1_1cl__gemm_1_1_c_l_g_e_m_m_native_kernel_configuration_factory.xhtml#ab2b50810acd1c74616ee97dee3cbdb6b">arm_compute::cl_gemm::CLGEMMNativeKernelConfigurationFactory::create</a></div><div class="ttdeci">static std::unique_ptr< ICLGEMMKernelConfiguration > create(GPUTarget arch)</div><div class="ttdoc">Static method to construct CLGEMMNative kernel object accordingly with the GPU architecture.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_native_kernel_configuration_8h_source.xhtml#l00046">CLGEMMNativeKernelConfiguration.h:46</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_a84437d80241f6a31e1a07c231ee8e3ac"><div class="ttname"><a href="namespacearm__compute.xhtml#a84437d80241f6a31e1a07c231ee8e3ac">arm_compute::is_data_type_quantized_per_channel</a></div><div class="ttdeci">bool is_data_type_quantized_per_channel(DataType dt)</div><div class="ttdoc">Check if a given data type is of per channel type.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_utils_8h_source.xhtml#l01194">Utils.h:1194</a></div></div> |
| <div class="ttc" id="structarm__compute_1_1_g_e_m_m_r_h_s_matrix_info_xhtml"><div class="ttname"><a href="structarm__compute_1_1_g_e_m_m_r_h_s_matrix_info.xhtml">arm_compute::GEMMRHSMatrixInfo</a></div><div class="ttdoc">GEMM RHS (Right Hand Side) matrix information.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_types_8h_source.xhtml#l01969">Types.h:1969</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_kernel_xhtml_a9c52840c4fadda388888e26f655de61d"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_kernel.xhtml#a9c52840c4fadda388888e26f655de61d">arm_compute::CLGEMMLowpMatrixMultiplyKernel::validate</a></div><div class="ttdeci">static Status validate(const ITensorInfo *input0, const ITensorInfo *input1, const ITensorInfo *output, const GEMMReshapeInfo &gemm_info=GEMMReshapeInfo())</div><div class="ttdoc">Static function to check if given info will lead to a valid configuration of CLGEMMLowpMatrixMultiply...</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_matrix_multiply_kernel_8cpp_source.xhtml#l00235">CLGEMMLowpMatrixMultiplyKernel.cpp:235</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_ad8ed01ff3ff33333d8e19db4d2818bb6af14462d71aa842202c3e4b272c7ec924"><div class="ttname"><a href="namespacearm__compute.xhtml#ad8ed01ff3ff33333d8e19db4d2818bb6af14462d71aa842202c3e4b272c7ec924">arm_compute::DataType::QASYMM8</a></div><div class="ttdoc">quantized, asymmetric fixed-point 8-bit number unsigned</div></div> |
| <div class="ttc" id="_c_l_g_e_m_m_native_kernel_configuration_8h_xhtml"><div class="ttname"><a href="_c_l_g_e_m_m_native_kernel_configuration_8h.xhtml">CLGEMMNativeKernelConfiguration.h</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_coordinates_xhtml"><div class="ttname"><a href="classarm__compute_1_1_coordinates.xhtml">arm_compute::Coordinates</a></div><div class="ttdoc">Coordinates of an item.</div><div class="ttdef"><b>Definition:</b> <a href="_coordinates_8h_source.xhtml#l00037">Coordinates.h:37</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_native_kernel_xhtml_aa7967bee001197b8642e30d4efaa96cc"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_native_kernel.xhtml#aa7967bee001197b8642e30d4efaa96cc">arm_compute::CLGEMMLowpMatrixMultiplyNativeKernel::configure</a></div><div class="ttdeci">void configure(const ICLTensor *input0, const ICLTensor *input1, ICLTensor *output, const GEMMLHSMatrixInfo &lhs_info, const GEMMRHSMatrixInfo &rhs_info, const GEMMReshapeInfo &gemm_info)</div><div class="ttdoc">Initialise the kernel's input and output.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_matrix_multiply_native_kernel_8cpp_source.xhtml#l00173">CLGEMMLowpMatrixMultiplyNativeKernel.cpp:173</a></div></div> |
| <div class="ttc" id="structarm__compute_1_1_g_e_m_m_lowp_output_stage_info_xhtml_ae5bd6bebbc0c7ebd9e7dbfd47d939c2a"><div class="ttname"><a href="structarm__compute_1_1_g_e_m_m_lowp_output_stage_info.xhtml#ae5bd6bebbc0c7ebd9e7dbfd47d939c2a">arm_compute::GEMMLowpOutputStageInfo::gemmlowp_multipliers</a></div><div class="ttdeci">std::vector< int32_t > gemmlowp_multipliers</div><div class="ttdoc">GEMMLowp output stage multiplier used for quantizing to QASYMM8.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_types_8h_source.xhtml#l01952">Types.h:1952</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_quantization_info_xhtml_a706fc156bcd4c45441bcaad05884b57d"><div class="ttname"><a href="classarm__compute_1_1_quantization_info.xhtml#a706fc156bcd4c45441bcaad05884b57d">arm_compute::QuantizationInfo::uniform</a></div><div class="ttdeci">UniformQuantizationInfo uniform() const</div><div class="ttdoc">Return per layer quantization info.</div><div class="ttdef"><b>Definition:</b> <a href="_quantization_info_8h_source.xhtml#l00148">QuantizationInfo.h:148</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_b_reduction_kernel_xhtml_aa526095c6ddce38f2ef22d45e18e2fe2"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_b_reduction_kernel.xhtml#aa526095c6ddce38f2ef22d45e18e2fe2">arm_compute::CLGEMMLowpMatrixBReductionKernel::configure</a></div><div class="ttdeci">void configure(const ICLTensor *mtx_b, ICLTensor *vector_sum_col) override</div><div class="ttdoc">Initialise the kernel's input and output.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_reduction_kernel_8cpp_source.xhtml#l00171">CLGEMMLowpReductionKernel.cpp:171</a></div></div> |
| <div class="ttc" id="namespacearm__compute_1_1misc_1_1shape__calculator_xhtml_a09ad10a110d947fd9c444b2ea5e4c127"><div class="ttname"><a href="namespacearm__compute_1_1misc_1_1shape__calculator.xhtml#a09ad10a110d947fd9c444b2ea5e4c127">arm_compute::misc::shape_calculator::compute_rhs_reshaped_shape</a></div><div class="ttdeci">TensorShape compute_rhs_reshaped_shape(const ITensorInfo &a, const GEMMRHSMatrixInfo &rhs_info)</div><div class="ttdoc">Calculate the Right Hand Side matrix reshaped shape.</div><div class="ttdef"><b>Definition:</b> <a href="_shape_calculator_8h_source.xhtml#l00224">ShapeCalculator.h:224</a></div></div> |
| <div class="ttc" id="_shape_calculator_8h_xhtml"><div class="ttname"><a href="_shape_calculator_8h.xhtml">ShapeCalculator.h</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1misc_1_1_i_cloneable_xhtml_a4d10e5012a872e7f78f2b539b673049d"><div class="ttname"><a href="classarm__compute_1_1misc_1_1_i_cloneable.xhtml#a4d10e5012a872e7f78f2b539b673049d">arm_compute::misc::ICloneable::clone</a></div><div class="ttdeci">virtual std::unique_ptr< T > clone() const =0</div><div class="ttdoc">Provide a clone of the current object of class T.</div></div> |
| <div class="ttc" id="structarm__compute_1_1_g_e_m_m_lowp_output_stage_info_xhtml"><div class="ttname"><a href="structarm__compute_1_1_g_e_m_m_lowp_output_stage_info.xhtml">arm_compute::GEMMLowpOutputStageInfo</a></div><div class="ttdoc">GEMMLowp output stage info.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_types_8h_source.xhtml#l01944">Types.h:1944</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_tensor_xhtml_a0e95dc1e53c361348314873b168ae237"><div class="ttname"><a href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">arm_compute::ITensor::info</a></div><div class="ttdeci">virtual ITensorInfo * info() const =0</div><div class="ttdoc">Interface to be implemented by the child class to return the tensor's metadata.</div></div> |
| <div class="ttc" id="_error_8h_xhtml"><div class="ttname"><a href="_error_8h.xhtml">Error.h</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_scheduler_xhtml_a398a2582c746d28fc125487a44c9ed74"><div class="ttname"><a href="classarm__compute_1_1_c_l_scheduler.xhtml#a398a2582c746d28fc125487a44c9ed74">arm_compute::CLScheduler::queue</a></div><div class="ttdeci">cl::CommandQueue & queue()</div><div class="ttdoc">Accessor for the associated CL command queue.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_scheduler_8cpp_source.xhtml#l00041">CLScheduler.cpp:41</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_g_e_m_m_info_xhtml_a00330b8913cac3b07029ac0c3350e806"><div class="ttname"><a href="classarm__compute_1_1_g_e_m_m_info.xhtml#a00330b8913cac3b07029ac0c3350e806">arm_compute::GEMMInfo::reinterpret_input_as_3d</a></div><div class="ttdeci">bool reinterpret_input_as_3d() const</div><div class="ttdoc">Flag which specifies if the input tensor has to be reinterpreted as 3D.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_types_8h_source.xhtml#l02070">Types.h:2070</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_tensor_info_xhtml_a3f3e1a3200223e6a304a533b1016e749"><div class="ttname"><a href="classarm__compute_1_1_i_tensor_info.xhtml#a3f3e1a3200223e6a304a533b1016e749">arm_compute::ITensorInfo::quantization_info</a></div><div class="ttdeci">virtual QuantizationInfo quantization_info() const =0</div><div class="ttdoc">Get the quantization settings (scale and offset) of the tensor.</div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_scheduler_xhtml_ae1a643e517f50bf0392fb6516dd7cf67"><div class="ttname"><a href="classarm__compute_1_1_c_l_scheduler.xhtml#ae1a643e517f50bf0392fb6516dd7cf67">arm_compute::CLScheduler::enqueue</a></div><div class="ttdeci">void enqueue(ICLKernel &kernel, bool flush=true)</div><div class="ttdoc">Schedule the execution of the passed kernel if possible.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_scheduler_8cpp_source.xhtml#l00154">CLScheduler.cpp:154</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1cl__gemm_1_1_c_l_g_e_m_m_reshaped_only_r_h_s_kernel_configuration_factory_xhtml_ab2b50810acd1c74616ee97dee3cbdb6b"><div class="ttname"><a href="classarm__compute_1_1cl__gemm_1_1_c_l_g_e_m_m_reshaped_only_r_h_s_kernel_configuration_factory.xhtml#ab2b50810acd1c74616ee97dee3cbdb6b">arm_compute::cl_gemm::CLGEMMReshapedOnlyRHSKernelConfigurationFactory::create</a></div><div class="ttdeci">static std::unique_ptr< ICLGEMMKernelConfiguration > create(GPUTarget arch)</div><div class="ttdoc">Static method to call the CLGEMMReshapedOnlyRHS kernel configuration class accordingly with the GPU a...</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_reshaped_only_r_h_s_kernel_configuration_8h_source.xhtml#l00046">CLGEMMReshapedOnlyRHSKernelConfiguration.h:46</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_a14f46283f316e7f0fad301d5c1507e9f"><div class="ttname"><a href="namespacearm__compute.xhtml#a14f46283f316e7f0fad301d5c1507e9f">arm_compute::is_data_type_quantized_asymmetric</a></div><div class="ttdeci">bool is_data_type_quantized_asymmetric(DataType dt)</div><div class="ttdoc">Check if a given data type is of asymmetric quantized type.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_utils_8h_source.xhtml#l01139">Utils.h:1139</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_g_e_m_m_info_xhtml_aa7e9584d7080ca6442cec62afaff6cad"><div class="ttname"><a href="classarm__compute_1_1_g_e_m_m_info.xhtml#aa7e9584d7080ca6442cec62afaff6cad">arm_compute::GEMMInfo::is_a_reshaped</a></div><div class="ttdeci">bool is_a_reshaped() const</div><div class="ttdoc">Flag which specifies if the matrix A has been reshaped.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_types_8h_source.xhtml#l02036">Types.h:2036</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_native_kernel_xhtml_a96f40eda5c40e254a6729e8bc0cb8195"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_native_kernel.xhtml#a96f40eda5c40e254a6729e8bc0cb8195">arm_compute::CLGEMMLowpMatrixMultiplyNativeKernel::validate</a></div><div class="ttdeci">static Status validate(const ITensorInfo *input0, const ITensorInfo *input1, const ITensorInfo *output, const GEMMLHSMatrixInfo &lhs_info, const GEMMRHSMatrixInfo &rhs_info, const GEMMReshapeInfo &gemm_info)</div><div class="ttdoc">Static function to check if given info will lead to a valid configuration of CLGEMMLowpMatrixMultiply...</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_matrix_multiply_native_kernel_8cpp_source.xhtml#l00250">CLGEMMLowpMatrixMultiplyNativeKernel.cpp:250</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_ad8ed01ff3ff33333d8e19db4d2818bb6a34f500e941c4df30b870126ec868ebd5"><div class="ttname"><a href="namespacearm__compute.xhtml#ad8ed01ff3ff33333d8e19db4d2818bb6a34f500e941c4df30b870126ec868ebd5">arm_compute::DataType::QSYMM8_PER_CHANNEL</a></div><div class="ttdoc">quantized, symmetric per channel fixed-point 8-bit number</div></div> |
| <div class="ttc" id="namespacearm__compute_1_1misc_1_1shape__calculator_xhtml_a69f9b3191aafc4905f9d029ff9d48fea"><div class="ttname"><a href="namespacearm__compute_1_1misc_1_1shape__calculator.xhtml#a69f9b3191aafc4905f9d029ff9d48fea">arm_compute::misc::shape_calculator::compute_reductionB_shape</a></div><div class="ttdeci">TensorShape compute_reductionB_shape(const ITensorInfo &a)</div><div class="ttdoc">Calculate the reductionB shape used in GEMMLowp.</div><div class="ttdef"><b>Definition:</b> <a href="_shape_calculator_8h_source.xhtml#l00374">ShapeCalculator.h:374</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_core_xhtml_a1c3a177d5a3188691c9946e1a78839d0"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_core.xhtml#a1c3a177d5a3188691c9946e1a78839d0">arm_compute::CLGEMMLowpMatrixMultiplyCore::CLGEMMLowpMatrixMultiplyCore</a></div><div class="ttdeci">CLGEMMLowpMatrixMultiplyCore(std::shared_ptr< IMemoryManager > memory_manager=nullptr)</div><div class="ttdoc">Constructor.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_matrix_multiply_core_8cpp_source.xhtml#l00051">CLGEMMLowpMatrixMultiplyCore.cpp:51</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_tensor_allocator_xhtml_a6e509c2a177b0b29e9e2369535094dee"><div class="ttname"><a href="classarm__compute_1_1_c_l_tensor_allocator.xhtml#a6e509c2a177b0b29e9e2369535094dee">arm_compute::CLTensorAllocator::allocate</a></div><div class="ttdeci">void allocate() override</div><div class="ttdoc">Allocate size specified by TensorInfo of OpenCL memory.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_tensor_allocator_8cpp_source.xhtml#l00127">CLTensorAllocator.cpp:127</a></div></div> |
| <div class="ttc" id="_c_l_g_e_m_m_lowp_matrix_multiply_core_8h_xhtml"><div class="ttname"><a href="_c_l_g_e_m_m_lowp_matrix_multiply_core_8h.xhtml">CLGEMMLowpMatrixMultiplyCore.h</a></div></div> |
| <div class="ttc" id="_validate_8h_xhtml_a921b705e9e3e0fe928928447869e62a5"><div class="ttname"><a href="_validate_8h.xhtml#a921b705e9e3e0fe928928447869e62a5">ARM_COMPUTE_ERROR_ON_NULLPTR</a></div><div class="ttdeci">#define ARM_COMPUTE_ERROR_ON_NULLPTR(...)</div><div class="ttdef"><b>Definition:</b> <a href="_validate_8h_source.xhtml#l00161">Validate.h:161</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_memory_group_resource_scope_xhtml"><div class="ttname"><a href="classarm__compute_1_1_memory_group_resource_scope.xhtml">arm_compute::MemoryGroupResourceScope</a></div><div class="ttdoc">Memory group resources scope handling class.</div><div class="ttdef"><b>Definition:</b> <a href="_i_memory_group_8h_source.xhtml#l00082">IMemoryGroup.h:82</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_c_l_tensor_xhtml"><div class="ttname"><a href="classarm__compute_1_1_i_c_l_tensor.xhtml">arm_compute::ICLTensor</a></div><div class="ttdoc">Interface for OpenCL tensor.</div><div class="ttdef"><b>Definition:</b> <a href="_i_c_l_tensor_8h_source.xhtml#l00042">ICLTensor.h:42</a></div></div> |
| <div class="ttc" id="_i_c_l_tensor_8h_xhtml"><div class="ttname"><a href="_i_c_l_tensor_8h.xhtml">ICLTensor.h</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_tensor_info_xhtml_a18064e0011c3869d884653e9e7c47b66"><div class="ttname"><a href="classarm__compute_1_1_i_tensor_info.xhtml#a18064e0011c3869d884653e9e7c47b66">arm_compute::ITensorInfo::total_size</a></div><div class="ttdeci">virtual size_t total_size() const =0</div><div class="ttdoc">Returns the total size of the tensor in bytes.</div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_a735ac6c2a02e320969625308810444f3"><div class="ttname"><a href="namespacearm__compute.xhtml#a735ac6c2a02e320969625308810444f3">arm_compute::GPUTarget</a></div><div class="ttdeci">GPUTarget</div><div class="ttdoc">Available GPU Targets.</div><div class="ttdef"><b>Definition:</b> <a href="_g_p_u_target_8h_source.xhtml#l00034">GPUTarget.h:34</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_reshaped_only_r_h_s_kernel_xhtml_a96f40eda5c40e254a6729e8bc0cb8195"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_reshaped_only_r_h_s_kernel.xhtml#a96f40eda5c40e254a6729e8bc0cb8195">arm_compute::CLGEMMLowpMatrixMultiplyReshapedOnlyRHSKernel::validate</a></div><div class="ttdeci">static Status validate(const ITensorInfo *input0, const ITensorInfo *input1, const ITensorInfo *output, const GEMMLHSMatrixInfo &lhs_info, const GEMMRHSMatrixInfo &rhs_info, const GEMMReshapeInfo &gemm_info)</div><div class="ttdoc">Static function to check if given info will lead to a valid configuration of CLGEMMLowpMatrixMultiply...</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_matrix_multiply_reshaped_only_r_h_s_kernel_8cpp_source.xhtml#l00257">CLGEMMLowpMatrixMultiplyReshapedOnlyRHSKernel.cpp:257</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_kernel_xhtml_a8aa57bd9c0703378b82393e8f440e9c8"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_kernel.xhtml#a8aa57bd9c0703378b82393e8f440e9c8">arm_compute::CLGEMMLowpMatrixMultiplyKernel::configure</a></div><div class="ttdeci">void configure(const ICLTensor *input0, const ICLTensor *input1, ICLTensor *output, const GEMMReshapeInfo &gemm_info=GEMMReshapeInfo())</div><div class="ttdoc">Initialise the kernel's input and output.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_matrix_multiply_kernel_8cpp_source.xhtml#l00173">CLGEMMLowpMatrixMultiplyKernel.cpp:173</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_a735ac6c2a02e320969625308810444f3afa362f4daec88442a387ff7cda411a38"><div class="ttname"><a href="namespacearm__compute.xhtml#a735ac6c2a02e320969625308810444f3afa362f4daec88442a387ff7cda411a38">arm_compute::GPUTarget::MIDGARD</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_a82b8ac759c804bc1fb4e2d21e178fb6fae1c8555fcf0ea2bb648a6fd527d658c0"><div class="ttname"><a href="namespacearm__compute.xhtml#a82b8ac759c804bc1fb4e2d21e178fb6fae1c8555fcf0ea2bb648a6fd527d658c0">arm_compute::ConvertPolicy::WRAP</a></div><div class="ttdoc">Wrap around.</div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_reshape_r_h_s_matrix_kernel_xhtml_aa0b0eaffc9a83e1a6495d957b63bbccc"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_reshape_r_h_s_matrix_kernel.xhtml#aa0b0eaffc9a83e1a6495d957b63bbccc">arm_compute::CLGEMMReshapeRHSMatrixKernel::configure</a></div><div class="ttdeci">void configure(const ICLTensor *input, ICLTensor *output, const GEMMRHSMatrixInfo &rhs_info)</div><div class="ttdoc">Initialise the kernel's input and output.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_reshape_r_h_s_matrix_kernel_8cpp_source.xhtml#l00102">CLGEMMReshapeRHSMatrixKernel.cpp:102</a></div></div> |
| <div class="ttc" id="structarm__compute_1_1_uniform_quantization_info_xhtml_a97bd6c077f3c7769f575b82988b9b668"><div class="ttname"><a href="structarm__compute_1_1_uniform_quantization_info.xhtml#a97bd6c077f3c7769f575b82988b9b668">arm_compute::UniformQuantizationInfo::offset</a></div><div class="ttdeci">int32_t offset</div><div class="ttdef"><b>Definition:</b> <a href="_quantization_info_8h_source.xhtml#l00065">QuantizationInfo.h:65</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_reshaped_only_r_h_s_kernel_xhtml_aa7967bee001197b8642e30d4efaa96cc"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_multiply_reshaped_only_r_h_s_kernel.xhtml#aa7967bee001197b8642e30d4efaa96cc">arm_compute::CLGEMMLowpMatrixMultiplyReshapedOnlyRHSKernel::configure</a></div><div class="ttdeci">void configure(const ICLTensor *input0, const ICLTensor *input1, ICLTensor *output, const GEMMLHSMatrixInfo &lhs_info, const GEMMRHSMatrixInfo &rhs_info, const GEMMReshapeInfo &gemm_info)</div><div class="ttdoc">Initialise the kernel's input and output.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_matrix_multiply_reshaped_only_r_h_s_kernel_8cpp_source.xhtml#l00173">CLGEMMLowpMatrixMultiplyReshapedOnlyRHSKernel.cpp:173</a></div></div> |
| <div class="ttc" id="_error_8h_xhtml_a1c69762a42ab8add645d0a949b6f4b1f"><div class="ttname"><a href="_error_8h.xhtml#a1c69762a42ab8add645d0a949b6f4b1f">ARM_COMPUTE_RETURN_ERROR_ON_MSG</a></div><div class="ttdeci">#define ARM_COMPUTE_RETURN_ERROR_ON_MSG(cond, msg)</div><div class="ttdoc">If the condition is true, an error is returned.</div><div class="ttdef"><b>Definition:</b> <a href="_error_8h_source.xhtml#l00244">Error.h:244</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_tensor_info_xhtml"><div class="ttname"><a href="classarm__compute_1_1_tensor_info.xhtml">arm_compute::TensorInfo</a></div><div class="ttdoc">Store the tensor's metadata.</div><div class="ttdef"><b>Definition:</b> <a href="_tensor_info_8h_source.xhtml#l00045">TensorInfo.h:45</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_g_e_m_m_info_xhtml_a4c8f9fa843de1086c27c86a6b8cf4582"><div class="ttname"><a href="classarm__compute_1_1_g_e_m_m_info.xhtml#a4c8f9fa843de1086c27c86a6b8cf4582">arm_compute::GEMMInfo::reshape_b_only_on_first_run</a></div><div class="ttdeci">bool reshape_b_only_on_first_run() const</div><div class="ttdoc">Flag which specifies if the reshape of matrix B should executed only for the first.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_types_8h_source.xhtml#l02054">Types.h:2054</a></div></div> |
| <div class="ttc" id="arm__compute_2core_2_helpers_8h_xhtml"><div class="ttname"><a href="arm__compute_2core_2_helpers_8h.xhtml">Helpers.h</a></div></div> |
| <div class="ttc" id="namespacearm__compute_1_1misc_1_1shape__calculator_xhtml"><div class="ttname"><a href="namespacearm__compute_1_1misc_1_1shape__calculator.xhtml">arm_compute::misc::shape_calculator</a></div><div class="ttdef"><b>Definition:</b> <a href="_shape_calculator_8h_source.xhtml#l00040">ShapeCalculator.h:40</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_c_l_kernel_xhtml_ad5ba9d34a3a855bf1dd2e36316ff550a"><div class="ttname"><a href="classarm__compute_1_1_i_c_l_kernel.xhtml#ad5ba9d34a3a855bf1dd2e36316ff550a">arm_compute::ICLKernel::set_target</a></div><div class="ttdeci">void set_target(GPUTarget target)</div><div class="ttdoc">Set the targeted GPU architecture.</div><div class="ttdef"><b>Definition:</b> <a href="_i_c_l_kernel_8h_source.xhtml#l00271">ICLKernel.h:271</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_g_e_m_m_info_xhtml"><div class="ttname"><a href="classarm__compute_1_1_g_e_m_m_info.xhtml">arm_compute::GEMMInfo</a></div><div class="ttdoc">GEMM information class.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_types_8h_source.xhtml#l01983">Types.h:1983</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_ad8ed01ff3ff33333d8e19db4d2818bb6a329f5d0c4b0c80e3474951d2c4435dd9"><div class="ttname"><a href="namespacearm__compute.xhtml#ad8ed01ff3ff33333d8e19db4d2818bb6a329f5d0c4b0c80e3474951d2c4435dd9">arm_compute::DataType::QASYMM8_SIGNED</a></div><div class="ttdoc">quantized, asymmetric fixed-point 8-bit number signed</div></div> |
| <div class="ttc" id="_asymm_helpers_8h_xhtml"><div class="ttname"><a href="_asymm_helpers_8h.xhtml">AsymmHelpers.h</a></div></div> |
| <div class="ttc" id="structarm__compute_1_1_g_e_m_m_lowp_output_stage_info_xhtml_ab233758aca2751c6e71a2f79baf7b92a"><div class="ttname"><a href="structarm__compute_1_1_g_e_m_m_lowp_output_stage_info.xhtml#ab233758aca2751c6e71a2f79baf7b92a">arm_compute::GEMMLowpOutputStageInfo::output_data_type</a></div><div class="ttdeci">DataType output_data_type</div><div class="ttdoc">Output tensor data type to use if the output is not initialized.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_types_8h_source.xhtml#l01955">Types.h:1955</a></div></div> |
| <div class="ttc" id="arm__compute_2core_2_types_8h_xhtml"><div class="ttname"><a href="arm__compute_2core_2_types_8h.xhtml">Types.h</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_tensor_xhtml_a1ffeb3b5abb3d61f62b58a391816201c"><div class="ttname"><a href="classarm__compute_1_1_c_l_tensor.xhtml#a1ffeb3b5abb3d61f62b58a391816201c">arm_compute::CLTensor::unmap</a></div><div class="ttdeci">void unmap()</div><div class="ttdoc">Enqueue an unmap operation of the allocated and mapped buffer.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_tensor_8cpp_source.xhtml#l00071">CLTensor.cpp:71</a></div></div> |
| <div class="ttc" id="_validate_8h_xhtml"><div class="ttname"><a href="_validate_8h.xhtml">Validate.h</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_b_reduction_kernel_xhtml_a21f628daa45675076fd25e46a052a53a"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_b_reduction_kernel.xhtml#a21f628daa45675076fd25e46a052a53a">arm_compute::CLGEMMLowpMatrixBReductionKernel::validate</a></div><div class="ttdeci">static Status validate(const ITensorInfo *mtx_b, const ITensorInfo *vector_sum_col)</div><div class="ttdoc">Static function to check if given info will lead to a valid configuration of CLGEMMLowpMatrixBReducti...</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_reduction_kernel_8cpp_source.xhtml#l00195">CLGEMMLowpReductionKernel.cpp:195</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_reshape_r_h_s_matrix_kernel_xhtml_aca0da52b749b54d37ff996f4b28f6762"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_reshape_r_h_s_matrix_kernel.xhtml#aca0da52b749b54d37ff996f4b28f6762">arm_compute::CLGEMMReshapeRHSMatrixKernel::validate</a></div><div class="ttdeci">static Status validate(const ITensorInfo *input, const ITensorInfo *output, const GEMMRHSMatrixInfo &rhs_info)</div><div class="ttdoc">Static function to check if given info will lead to a valid configuration of CLGEMMReshapeRHSMatrixKe...</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_reshape_r_h_s_matrix_kernel_8cpp_source.xhtml#l00134">CLGEMMReshapeRHSMatrixKernel.cpp:134</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_a_reduction_kernel_xhtml_a613f25e524dbabbff55ea9d7df53c7f6"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_matrix_a_reduction_kernel.xhtml#a613f25e524dbabbff55ea9d7df53c7f6">arm_compute::CLGEMMLowpMatrixAReductionKernel::configure</a></div><div class="ttdeci">void configure(const ICLTensor *mtx_a, ICLTensor *vector_sum_row) override</div><div class="ttdoc">Initialise the kernel's input and output.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_reduction_kernel_8cpp_source.xhtml#l00103">CLGEMMLowpReductionKernel.cpp:103</a></div></div> |
| <div class="ttc" id="_c_l_g_e_m_m_reshaped_only_r_h_s_kernel_configuration_8h_xhtml"><div class="ttname"><a href="_c_l_g_e_m_m_reshaped_only_r_h_s_kernel_configuration_8h.xhtml">CLGEMMReshapedOnlyRHSKernelConfiguration.h</a></div></div> |
| </div><!-- fragment --></div><!-- contents --> |
| </div><!-- doc-content --> |
| <!-- start footer part --> |
| <div id="nav-path" class="navpath"><!-- id is needed for treeview function! --> |
| <ul> |
| <li class="navelem"><a class="el" href="dir_68267d1309a1af8e8297ef4c3efbcdba.xhtml">src</a></li><li class="navelem"><a class="el" href="dir_bf9f26469d00835ba20ff8d80ee5a804.xhtml">runtime</a></li><li class="navelem"><a class="el" href="dir_43c3fdbf778d1fd99e2e38f09fddd920.xhtml">CL</a></li><li class="navelem"><a class="el" href="dir_0304d3529340c629ae0050036d07056a.xhtml">functions</a></li><li class="navelem"><a class="el" href="_c_l_g_e_m_m_lowp_matrix_multiply_core_8cpp.xhtml">CLGEMMLowpMatrixMultiplyCore.cpp</a></li> |
| <li class="footer">Generated on Thu Mar 5 2020 16:07:02 for Compute Library by |
| <a href="http://www.doxygen.org/index.html"> |
| <img class="footer" src="doxygen.png" alt="doxygen"/></a> 1.8.15 </li> |
| </ul> |
| </div> |
| </body> |
| </html> |
