| <!-- 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/core/CL/kernels/CLGEMMMatrixMultiplyKernel.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_matrix_multiply_kernel_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">CLGEMMMatrixMultiplyKernel.cpp</div> </div> |
| </div><!--header--> |
| <div class="contents"> |
| <a href="_c_l_g_e_m_m_matrix_multiply_kernel_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) 2017-2019 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_matrix_multiply_kernel_8h.xhtml">arm_compute/core/CL/kernels/CLGEMMMatrixMultiplyKernel.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="_access_window_static_8h.xhtml">arm_compute/core/AccessWindowStatic.h</a>"</span></div><div class="line"><a name="l00027"></a><span class="lineno"> 27</span> <span class="preprocessor">#include "<a class="code" href="core_2_c_l_2_c_l_helpers_8h.xhtml">arm_compute/core/CL/CLHelpers.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_kernel_library_8h.xhtml">arm_compute/core/CL/CLKernelLibrary.h</a>"</span></div><div class="line"><a name="l00029"></a><span class="lineno"> 29</span> <span class="preprocessor">#include "<a class="code" href="_c_l_validate_8h.xhtml">arm_compute/core/CL/CLValidate.h</a>"</span></div><div class="line"><a name="l00030"></a><span class="lineno"> 30</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="l00031"></a><span class="lineno"> 31</span> <span class="preprocessor">#include "<a class="code" href="_open_c_l_8h.xhtml">arm_compute/core/CL/OpenCL.h</a>"</span></div><div class="line"><a name="l00032"></a><span class="lineno"> 32</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="l00033"></a><span class="lineno"> 33</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="l00034"></a><span class="lineno"> 34</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="l00035"></a><span class="lineno"> 35</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="l00036"></a><span class="lineno"> 36</span> <span class="preprocessor">#include "<a class="code" href="arm__compute_2core_2_utils_8h.xhtml">arm_compute/core/Utils.h</a>"</span></div><div class="line"><a name="l00037"></a><span class="lineno"> 37</span> <span class="preprocessor">#include "<a class="code" href="_window_8h.xhtml">arm_compute/core/Window.h</a>"</span></div><div class="line"><a name="l00038"></a><span class="lineno"> 38</span> <span class="preprocessor">#include "<a class="code" href="float__ops_8h.xhtml">arm_compute/core/utils/helpers/float_ops.h</a>"</span></div><div class="line"><a name="l00039"></a><span class="lineno"> 39</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="l00040"></a><span class="lineno"> 40</span> </div><div class="line"><a name="l00041"></a><span class="lineno"> 41</span> <span class="preprocessor">#include <set></span></div><div class="line"><a name="l00042"></a><span class="lineno"> 42</span> <span class="preprocessor">#include <string></span></div><div class="line"><a name="l00043"></a><span class="lineno"> 43</span> </div><div class="line"><a name="l00044"></a><span class="lineno"> 44</span> <span class="keyword">namespace </span><a class="code" href="namespacearm__compute.xhtml">arm_compute</a></div><div class="line"><a name="l00045"></a><span class="lineno"> 45</span> {</div><div class="line"><a name="l00046"></a><span class="lineno"> 46</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="l00047"></a><span class="lineno"> 47</span> </div><div class="line"><a name="l00048"></a><span class="lineno"> 48</span> <span class="keyword">namespace</span></div><div class="line"><a name="l00049"></a><span class="lineno"> 49</span> {</div><div class="line"><a name="l00050"></a><span class="lineno"> 50</span> <span class="keyword">using</span> ElementsProcessed = Steps;</div><div class="line"><a name="l00051"></a><span class="lineno"> 51</span> </div><div class="line"><a name="l00052"></a><span class="lineno"> 52</span> <span class="keyword">inline</span> Status validate_arguments(<span class="keyword">const</span> ITensorInfo *input0, <span class="keyword">const</span> ITensorInfo *input1, <span class="keyword">const</span> ITensorInfo *input2, <span class="keyword">const</span> ITensorInfo *output, <span class="keywordtype">float</span> beta,</div><div class="line"><a name="l00053"></a><span class="lineno"> 53</span>  <span class="keywordtype">bool</span> is_interleaved_transposed, <span class="keyword">const</span> GEMMReshapeInfo &reshape_info, <span class="keywordtype">bool</span> fp_mixed_precision)</div><div class="line"><a name="l00054"></a><span class="lineno"> 54</span> {</div><div class="line"><a name="l00055"></a><span class="lineno"> 55</span>  <a class="code" href="_validate_8h.xhtml#aff911654521523937ff24372a870b89f">ARM_COMPUTE_RETURN_ERROR_ON_NULLPTR</a>(input0, input1, output);</div><div class="line"><a name="l00056"></a><span class="lineno"> 56</span>  <a class="code" href="_c_l_validate_8h.xhtml#ab82bd5de18ef067ae5d9ba4af8065dd6">ARM_COMPUTE_RETURN_ERROR_ON_F16_UNSUPPORTED</a>(input0);</div><div class="line"><a name="l00057"></a><span class="lineno"> 57</span>  <a class="code" href="_validate_8h.xhtml#ae7eed178dac535c6e727061b1f5bc6eb">ARM_COMPUTE_RETURN_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN</a>(input0, 1, <a class="code" href="namespacearm__compute.xhtml#ab4e88c89b3b7ea1735996cc4def22d58a56d8353718e6fdc78b8d69078a2cdb94">DataType::F16</a>, <a class="code" href="namespacearm__compute.xhtml#ab4e88c89b3b7ea1735996cc4def22d58a44ad4ef5a76e6aa6fb3e3fa079a54fda">DataType::F32</a>);</div><div class="line"><a name="l00058"></a><span class="lineno"> 58</span>  <a class="code" href="_validate_8h.xhtml#a8f3ff7da485ff7e75dab07baadf5b4bd">ARM_COMPUTE_RETURN_ERROR_ON_MISMATCHING_DATA_TYPES</a>(input0, input1);</div><div class="line"><a name="l00059"></a><span class="lineno"> 59</span>  <a class="code" href="_error_8h.xhtml#a1c69762a42ab8add645d0a949b6f4b1f">ARM_COMPUTE_RETURN_ERROR_ON_MSG</a>((fp_mixed_precision && (input0->data_type() != <a class="code" href="namespacearm__compute.xhtml#ab4e88c89b3b7ea1735996cc4def22d58a56d8353718e6fdc78b8d69078a2cdb94">DataType::F16</a>)), <span class="stringliteral">"Mixed precision floating point is supported only for F16 data"</span>);</div><div class="line"><a name="l00060"></a><span class="lineno"> 60</span>  <a class="code" href="_error_8h.xhtml#a1c69762a42ab8add645d0a949b6f4b1f">ARM_COMPUTE_RETURN_ERROR_ON_MSG</a>(input0->num_dimensions() > 4, <span class="stringliteral">"The number of dimensions for the matrix A must be <= 4"</span>);</div><div class="line"><a name="l00061"></a><span class="lineno"> 61</span>  <a class="code" href="_error_8h.xhtml#a1c69762a42ab8add645d0a949b6f4b1f">ARM_COMPUTE_RETURN_ERROR_ON_MSG</a>(input1->num_dimensions() > 3, <span class="stringliteral">"The number of dimensions for the matrix B must be <= 3"</span>);</div><div class="line"><a name="l00062"></a><span class="lineno"> 62</span>  <a class="code" href="_error_8h.xhtml#a1c69762a42ab8add645d0a949b6f4b1f">ARM_COMPUTE_RETURN_ERROR_ON_MSG</a>(is_interleaved_transposed && reshape_info.reinterpret_input_as_3d(), <span class="stringliteral">"The input tensor cannot be reinterpreted as 3D if is_interleaved_transposed is true"</span>);</div><div class="line"><a name="l00063"></a><span class="lineno"> 63</span>  <a class="code" href="_error_8h.xhtml#a1c69762a42ab8add645d0a949b6f4b1f">ARM_COMPUTE_RETURN_ERROR_ON_MSG</a>(input1->num_dimensions() > 2 && reshape_info.reinterpret_input_as_3d(), <span class="stringliteral">"The input1 tensor cannot have more than 2 dimensions if input0 has to be reinterpreted as 3D"</span>);</div><div class="line"><a name="l00064"></a><span class="lineno"> 64</span>  <a class="code" href="_error_8h.xhtml#a1c69762a42ab8add645d0a949b6f4b1f">ARM_COMPUTE_RETURN_ERROR_ON_MSG</a>((reshape_info.reinterpret_input_as_3d() || reshape_info.depth_output_gemm3d() != 0) && (input2 != <span class="keyword">nullptr</span>)</div><div class="line"><a name="l00065"></a><span class="lineno"> 65</span>  && (!reshape_info.broadcast_bias()), <span class="stringliteral">"Bias addition only supported with broadcast mode in case the input or output has to be reinterpreted as 3D"</span>);</div><div class="line"><a name="l00066"></a><span class="lineno"> 66</span> </div><div class="line"><a name="l00067"></a><span class="lineno"> 67</span>  <span class="keywordflow">if</span>(!is_interleaved_transposed)</div><div class="line"><a name="l00068"></a><span class="lineno"> 68</span>  {</div><div class="line"><a name="l00069"></a><span class="lineno"> 69</span>  <a class="code" href="_error_8h.xhtml#a206d6e247e0957ac3dee45d27756fc25">ARM_COMPUTE_RETURN_ERROR_ON</a>(input0->dimension(0) != input1->dimension(1));</div><div class="line"><a name="l00070"></a><span class="lineno"> 70</span> </div><div class="line"><a name="l00071"></a><span class="lineno"> 71</span>  <span class="keywordflow">if</span>(input2 != <span class="keyword">nullptr</span> && !(<a class="code" href="namespacearm__compute_1_1helpers_1_1float__ops.xhtml#a3bd19352aed7410633d1f9b95d74a809">helpers::float_ops::is_zero</a>(beta)))</div><div class="line"><a name="l00072"></a><span class="lineno"> 72</span>  {</div><div class="line"><a name="l00073"></a><span class="lineno"> 73</span>  <span class="keyword">const</span> <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> m = reshape_info.reinterpret_input_as_3d() ? input0->dimension(1) * input0->dimension(2) : input0->dimension(1);</div><div class="line"><a name="l00074"></a><span class="lineno"> 74</span>  <span class="keyword">const</span> <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> n = input1->dimension(0);</div><div class="line"><a name="l00075"></a><span class="lineno"> 75</span>  <span class="keyword">const</span> <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> input2_dim0 = input2->dimension(0);</div><div class="line"><a name="l00076"></a><span class="lineno"> 76</span>  <span class="keyword">const</span> <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> input2_dim1 = input2->dimension(1);</div><div class="line"><a name="l00077"></a><span class="lineno"> 77</span> </div><div class="line"><a name="l00078"></a><span class="lineno"> 78</span>  <a class="code" href="_validate_8h.xhtml#a8f3ff7da485ff7e75dab07baadf5b4bd">ARM_COMPUTE_RETURN_ERROR_ON_MISMATCHING_DATA_TYPES</a>(input2, input1);</div><div class="line"><a name="l00079"></a><span class="lineno"> 79</span>  <span class="keywordflow">if</span>(reshape_info.broadcast_bias())</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>  <a class="code" href="_error_8h.xhtml#a1c69762a42ab8add645d0a949b6f4b1f">ARM_COMPUTE_RETURN_ERROR_ON_MSG</a>((input2_dim1 != 1 || input2_dim0 != n), <span class="stringliteral">"Incorrect dimension of bias matrix which is to be broadcasted"</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"> 83</span>  <span class="keywordflow">else</span></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="_error_8h.xhtml#a1c69762a42ab8add645d0a949b6f4b1f">ARM_COMPUTE_RETURN_ERROR_ON_MSG</a>((input2_dim0 != n || input2_dim1 != m), <span class="stringliteral">"Incorrect dimension of bias matrix"</span>);</div><div class="line"><a name="l00086"></a><span class="lineno"> 86</span>  }</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>  }</div><div class="line"><a name="l00089"></a><span class="lineno"> 89</span>  <span class="keywordflow">else</span></div><div class="line"><a name="l00090"></a><span class="lineno"> 90</span>  {</div><div class="line"><a name="l00091"></a><span class="lineno"> 91</span>  GEMMRHSMatrixInfo rhs_info;</div><div class="line"><a name="l00092"></a><span class="lineno"> 92</span>  GEMMLHSMatrixInfo lhs_info;</div><div class="line"><a name="l00093"></a><span class="lineno"> 93</span>  <span class="keyword">const</span> <span class="keyword">auto</span> m = static_cast<unsigned int>(reshape_info.m());</div><div class="line"><a name="l00094"></a><span class="lineno"> 94</span>  <span class="keyword">const</span> <span class="keyword">auto</span> n = static_cast<unsigned int>(reshape_info.n());</div><div class="line"><a name="l00095"></a><span class="lineno"> 95</span>  <span class="keyword">const</span> <span class="keywordtype">int</span> k = reshape_info.k();</div><div class="line"><a name="l00096"></a><span class="lineno"> 96</span>  <span class="keyword">const</span> <span class="keywordtype">int</span> mult_transpose1xW_width = reshape_info.mult_transpose1xW_width();</div><div class="line"><a name="l00097"></a><span class="lineno"> 97</span>  <span class="keyword">const</span> <span class="keywordtype">int</span> mult_interleave4x4_height = reshape_info.mult_interleave4x4_height();</div><div class="line"><a name="l00098"></a><span class="lineno"> 98</span>  rhs_info.n0 = 16 / input1->element_size();</div><div class="line"><a name="l00099"></a><span class="lineno"> 99</span>  rhs_info.k0 = 1;</div><div class="line"><a name="l00100"></a><span class="lineno"> 100</span>  rhs_info.h0 = mult_transpose1xW_width;</div><div class="line"><a name="l00101"></a><span class="lineno"> 101</span>  rhs_info.interleave = <span class="keyword">false</span>;</div><div class="line"><a name="l00102"></a><span class="lineno"> 102</span>  rhs_info.transpose = <span class="keyword">false</span>;</div><div class="line"><a name="l00103"></a><span class="lineno"> 103</span>  lhs_info.m0 = 4;</div><div class="line"><a name="l00104"></a><span class="lineno"> 104</span>  lhs_info.k0 = 4;</div><div class="line"><a name="l00105"></a><span class="lineno"> 105</span>  lhs_info.v0 = mult_interleave4x4_height;</div><div class="line"><a name="l00106"></a><span class="lineno"> 106</span>  lhs_info.interleave = <span class="keyword">true</span>;</div><div class="line"><a name="l00107"></a><span class="lineno"> 107</span>  lhs_info.transpose = <span class="keyword">true</span>;</div><div class="line"><a name="l00108"></a><span class="lineno"> 108</span> </div><div class="line"><a name="l00109"></a><span class="lineno"> 109</span>  TensorShape tensor_shape0{ input0->tensor_shape() };</div><div class="line"><a name="l00110"></a><span class="lineno"> 110</span>  tensor_shape0.set(0, k);</div><div class="line"><a name="l00111"></a><span class="lineno"> 111</span>  tensor_shape0.set(1, m);</div><div class="line"><a name="l00112"></a><span class="lineno"> 112</span> </div><div class="line"><a name="l00113"></a><span class="lineno"> 113</span>  TensorShape tensor_shape1{ input1->tensor_shape() };</div><div class="line"><a name="l00114"></a><span class="lineno"> 114</span>  tensor_shape1.set(0, n);</div><div class="line"><a name="l00115"></a><span class="lineno"> 115</span>  tensor_shape1.set(1, k);</div><div class="line"><a name="l00116"></a><span class="lineno"> 116</span> </div><div class="line"><a name="l00117"></a><span class="lineno"> 117</span>  <span class="keyword">const</span> TensorInfo tensor_info0 = input0->clone()->set_tensor_shape(tensor_shape0);</div><div class="line"><a name="l00118"></a><span class="lineno"> 118</span>  <span class="keyword">const</span> TensorInfo tensor_info1 = input1->clone()->set_tensor_shape(tensor_shape1);</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="keyword">const</span> TensorInfo tensor_info_reshaped0 = input0->clone()->set_tensor_shape(<a class="code" href="namespacearm__compute_1_1misc_1_1shape__calculator.xhtml#a389f89ab5121dad0906d0b7324fbf73d">compute_lhs_reshaped_shape</a>(tensor_info0, lhs_info));</div><div class="line"><a name="l00121"></a><span class="lineno"> 121</span>  <span class="keyword">const</span> TensorInfo tensor_info_reshaped1 = input1->clone()->set_tensor_shape(<a class="code" href="namespacearm__compute_1_1misc_1_1shape__calculator.xhtml#a09ad10a110d947fd9c444b2ea5e4c127">compute_rhs_reshaped_shape</a>(tensor_info1, rhs_info));</div><div class="line"><a name="l00122"></a><span class="lineno"> 122</span> </div><div class="line"><a name="l00123"></a><span class="lineno"> 123</span>  <a class="code" href="_validate_8h.xhtml#a27e4638546c88b8916f967e6e54480a9">ARM_COMPUTE_RETURN_ERROR_ON_MISMATCHING_SHAPES</a>(input0, &tensor_info_reshaped0);</div><div class="line"><a name="l00124"></a><span class="lineno"> 124</span>  <a class="code" href="_validate_8h.xhtml#a27e4638546c88b8916f967e6e54480a9">ARM_COMPUTE_RETURN_ERROR_ON_MISMATCHING_SHAPES</a>(input1, &tensor_info_reshaped1);</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="keywordflow">if</span>(input2 != <span class="keyword">nullptr</span> && !(<a class="code" href="namespacearm__compute_1_1helpers_1_1float__ops.xhtml#a3bd19352aed7410633d1f9b95d74a809">helpers::float_ops::is_zero</a>(beta)))</div><div class="line"><a name="l00127"></a><span class="lineno"> 127</span>  {</div><div class="line"><a name="l00128"></a><span class="lineno"> 128</span>  <span class="keyword">const</span> <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> input2_dim0 = input2->dimension(0);</div><div class="line"><a name="l00129"></a><span class="lineno"> 129</span>  <span class="keyword">const</span> <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> input2_dim1 = input2->dimension(1);</div><div class="line"><a name="l00130"></a><span class="lineno"> 130</span> </div><div class="line"><a name="l00131"></a><span class="lineno"> 131</span>  <a class="code" href="_validate_8h.xhtml#a8f3ff7da485ff7e75dab07baadf5b4bd">ARM_COMPUTE_RETURN_ERROR_ON_MISMATCHING_DATA_TYPES</a>(input2, input1);</div><div class="line"><a name="l00132"></a><span class="lineno"> 132</span>  <span class="keywordflow">if</span>(reshape_info.broadcast_bias())</div><div class="line"><a name="l00133"></a><span class="lineno"> 133</span>  {</div><div class="line"><a name="l00134"></a><span class="lineno"> 134</span>  <a class="code" href="_error_8h.xhtml#a1c69762a42ab8add645d0a949b6f4b1f">ARM_COMPUTE_RETURN_ERROR_ON_MSG</a>((input2_dim1 != 1 || input2_dim0 != n), <span class="stringliteral">"Incorrect dimension of bias matrix which is to be broadcasted"</span>);</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>  <span class="keywordflow">else</span></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>  <a class="code" href="_error_8h.xhtml#a1c69762a42ab8add645d0a949b6f4b1f">ARM_COMPUTE_RETURN_ERROR_ON_MSG</a>((input2_dim0 != n || input2_dim1 != m), <span class="stringliteral">"Incorrect dimension of bias matrix"</span>);</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>  }</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="keywordflow">if</span>(output->total_size() != 0)</div><div class="line"><a name="l00144"></a><span class="lineno"> 144</span>  {</div><div class="line"><a name="l00145"></a><span class="lineno"> 145</span>  <span class="keyword">const</span> TensorInfo tensor_info_output = output->clone()->set_tensor_shape(<a class="code" href="namespacearm__compute_1_1misc_1_1shape__calculator.xhtml#adca241b012a5e00ddfcdc5a8db05a2a3">compute_mm_shape</a>(*input0, *input1, is_interleaved_transposed, reshape_info));</div><div class="line"><a name="l00146"></a><span class="lineno"> 146</span>  <a class="code" href="_validate_8h.xhtml#a27e4638546c88b8916f967e6e54480a9">ARM_COMPUTE_RETURN_ERROR_ON_MISMATCHING_SHAPES</a>(output, &tensor_info_output);</div><div class="line"><a name="l00147"></a><span class="lineno"> 147</span>  <a class="code" href="_validate_8h.xhtml#a8f3ff7da485ff7e75dab07baadf5b4bd">ARM_COMPUTE_RETURN_ERROR_ON_MISMATCHING_DATA_TYPES</a>(input0, output);</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="keywordflow">return</span> Status{};</div><div class="line"><a name="l00151"></a><span class="lineno"> 151</span> }</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> <span class="keyword">inline</span> std::pair<Status, Window> validate_and_configure_window(ITensorInfo *input0, ITensorInfo *input1, ITensorInfo *input2, ITensorInfo *output,</div><div class="line"><a name="l00154"></a><span class="lineno"> 154</span>  <span class="keywordtype">float</span> beta, <span class="keywordtype">bool</span> is_interleaved_transposed, <span class="keyword">const</span> GEMMReshapeInfo &reshape_info, <a class="code" href="namespacearm__compute.xhtml#a735ac6c2a02e320969625308810444f3">GPUTarget</a> gpu_target,</div><div class="line"><a name="l00155"></a><span class="lineno"> 155</span>  ElementsProcessed &num_elements_processed)</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>  <a class="code" href="_error_8h.xhtml#a6dc630a6ae9cc063b3924bcea8dee9d6">ARM_COMPUTE_UNUSED</a>(beta);</div><div class="line"><a name="l00158"></a><span class="lineno"> 158</span>  <span class="keywordtype">bool</span> window_changed = <span class="keyword">false</span>;</div><div class="line"><a name="l00159"></a><span class="lineno"> 159</span>  Window win{};</div><div class="line"><a name="l00160"></a><span class="lineno"> 160</span>  Window win_out{};</div><div class="line"><a name="l00161"></a><span class="lineno"> 161</span> </div><div class="line"><a name="l00162"></a><span class="lineno"> 162</span>  <span class="keyword">const</span> <a class="code" href="namespacearm__compute.xhtml#ad8ed01ff3ff33333d8e19db4d2818bb6">DataType</a> <a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a5f5b6c4337eac9e2e0046ca2304d80dc">data_type</a> = input0->data_type();</div><div class="line"><a name="l00163"></a><span class="lineno"> 163</span>  <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> &num_elems_processed_per_iteration_x = num_elements_processed[0];</div><div class="line"><a name="l00164"></a><span class="lineno"> 164</span>  <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> &num_elems_processed_per_iteration_y = num_elements_processed[1];</div><div class="line"><a name="l00165"></a><span class="lineno"> 165</span>  <span class="keywordtype">bool</span> reinterpret_input_as_3d = reshape_info.reinterpret_input_as_3d();</div><div class="line"><a name="l00166"></a><span class="lineno"> 166</span>  <span class="keywordtype">bool</span> reinterpret_output_as_3d = (reshape_info.depth_output_gemm3d() != 0);</div><div class="line"><a name="l00167"></a><span class="lineno"> 167</span> </div><div class="line"><a name="l00168"></a><span class="lineno"> 168</span>  <span class="comment">// In case both input and output have to be reinterpreted as 3D tensors,</span></div><div class="line"><a name="l00169"></a><span class="lineno"> 169</span>  <span class="comment">// force reinterpret_input_as_3d and reinterpret_output_as_3d to be false.</span></div><div class="line"><a name="l00170"></a><span class="lineno"> 170</span>  <span class="keywordflow">if</span>(reinterpret_input_as_3d == reinterpret_output_as_3d)</div><div class="line"><a name="l00171"></a><span class="lineno"> 171</span>  {</div><div class="line"><a name="l00172"></a><span class="lineno"> 172</span>  reinterpret_input_as_3d = <span class="keyword">false</span>;</div><div class="line"><a name="l00173"></a><span class="lineno"> 173</span>  reinterpret_output_as_3d = <span class="keyword">false</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> </div><div class="line"><a name="l00176"></a><span class="lineno"> 176</span>  <span class="comment">// Output tensor auto inizialitation if not yet initialized</span></div><div class="line"><a name="l00177"></a><span class="lineno"> 177</span>  <a class="code" href="namespacearm__compute.xhtml#a47be6fa38308d0003c25b60b7dbc45ce">auto_init_if_empty</a>(*output, input0->clone()->set_tensor_shape(<a class="code" href="namespacearm__compute_1_1misc_1_1shape__calculator.xhtml#adca241b012a5e00ddfcdc5a8db05a2a3">compute_mm_shape</a>(*input0, *input1, is_interleaved_transposed, reshape_info)));</div><div class="line"><a name="l00178"></a><span class="lineno"> 178</span> </div><div class="line"><a name="l00179"></a><span class="lineno"> 179</span>  TensorInfo tmp_info(*output);</div><div class="line"><a name="l00180"></a><span class="lineno"> 180</span> </div><div class="line"><a name="l00181"></a><span class="lineno"> 181</span>  <span class="keywordflow">if</span>(reinterpret_output_as_3d)</div><div class="line"><a name="l00182"></a><span class="lineno"> 182</span>  {</div><div class="line"><a name="l00183"></a><span class="lineno"> 183</span>  <span class="comment">// Since the output tensor has to be reinterpreted as 3D and the execute window is based on a 2D GEMM,</span></div><div class="line"><a name="l00184"></a><span class="lineno"> 184</span>  <span class="comment">// the window needs to be constructed on the 2D collapsed version of the tensor</span></div><div class="line"><a name="l00185"></a><span class="lineno"> 185</span>  TensorShape tmp_shape(output->tensor_shape());</div><div class="line"><a name="l00186"></a><span class="lineno"> 186</span>  tmp_shape.collapse(2U, 1U);</div><div class="line"><a name="l00187"></a><span class="lineno"> 187</span>  tmp_info.set_tensor_shape(tmp_shape);</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> </div><div class="line"><a name="l00190"></a><span class="lineno"> 190</span>  <span class="keywordflow">if</span>(is_interleaved_transposed)</div><div class="line"><a name="l00191"></a><span class="lineno"> 191</span>  {</div><div class="line"><a name="l00192"></a><span class="lineno"> 192</span>  <span class="comment">// reinterpret_input_as_3d is not supported if is_interleaved_transposed is set</span></div><div class="line"><a name="l00193"></a><span class="lineno"> 193</span>  <a class="code" href="_error_8h.xhtml#a54a6080c9f4df1f908e57a9bbb46f5da">ARM_COMPUTE_ERROR_ON</a>(reshape_info.reinterpret_input_as_3d());</div><div class="line"><a name="l00194"></a><span class="lineno"> 194</span> </div><div class="line"><a name="l00195"></a><span class="lineno"> 195</span>  <span class="comment">// Configure kernel window</span></div><div class="line"><a name="l00196"></a><span class="lineno"> 196</span>  num_elems_processed_per_iteration_x = max_cl_vector_width / <a class="code" href="namespacearm__compute.xhtml#abb7e0f23a4f2e63f39433f158dad47ab">data_size_from_type</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a5f5b6c4337eac9e2e0046ca2304d80dc">data_type</a>);</div><div class="line"><a name="l00197"></a><span class="lineno"> 197</span>  num_elems_processed_per_iteration_y = 4;</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">// Note: bottom paddings are calculated manually as the output can be reinterpreted as 3D tensor</span></div><div class="line"><a name="l00200"></a><span class="lineno"> 200</span>  <span class="comment">// The only way to set properly the paddings, it is to set those explicitly through the AccessWindowStatic</span></div><div class="line"><a name="l00201"></a><span class="lineno"> 201</span>  <span class="keyword">const</span> <span class="keywordtype">int</span> m = reshape_info.m();</div><div class="line"><a name="l00202"></a><span class="lineno"> 202</span>  <span class="keyword">const</span> <span class="keywordtype">int</span> bottom_pad = (num_elems_processed_per_iteration_y - (m % num_elems_processed_per_iteration_y)) % num_elems_processed_per_iteration_y;</div><div class="line"><a name="l00203"></a><span class="lineno"> 203</span> </div><div class="line"><a name="l00204"></a><span class="lineno"> 204</span>  win = <a class="code" href="namespacearm__compute.xhtml#ab7980fa5ee693e3282a76da047a1c3b5">calculate_max_window</a>(tmp_info, Steps(num_elems_processed_per_iteration_x, num_elems_processed_per_iteration_y));</div><div class="line"><a name="l00205"></a><span class="lineno"> 205</span>  win_out = <a class="code" href="namespacearm__compute.xhtml#ab7980fa5ee693e3282a76da047a1c3b5">calculate_max_window</a>(*output, Steps(num_elems_processed_per_iteration_x, num_elems_processed_per_iteration_y));</div><div class="line"><a name="l00206"></a><span class="lineno"> 206</span> </div><div class="line"><a name="l00207"></a><span class="lineno"> 207</span>  AccessWindowStatic input0_access(input0, 0, 0, input0->dimension(0), input0->dimension(1));</div><div class="line"><a name="l00208"></a><span class="lineno"> 208</span>  AccessWindowStatic input1_access(input1, 0, 0,</div><div class="line"><a name="l00209"></a><span class="lineno"> 209</span>  <a class="code" href="namespacearm__compute.xhtml#ab237a0a375cf382d52b61653248d3d4a">ceil_to_multiple</a>(input1->dimension(0), num_elems_processed_per_iteration_x),</div><div class="line"><a name="l00210"></a><span class="lineno"> 210</span>  <a class="code" href="namespacearm__compute.xhtml#ab237a0a375cf382d52b61653248d3d4a">ceil_to_multiple</a>(input1->dimension(1), num_elems_processed_per_iteration_y));</div><div class="line"><a name="l00211"></a><span class="lineno"> 211</span>  AccessWindowStatic output_access(output, 0, 0,</div><div class="line"><a name="l00212"></a><span class="lineno"> 212</span>  <a class="code" href="namespacearm__compute.xhtml#ab237a0a375cf382d52b61653248d3d4a">ceil_to_multiple</a>(output->dimension(0), num_elems_processed_per_iteration_x),</div><div class="line"><a name="l00213"></a><span class="lineno"> 213</span>  output->dimension(1) + bottom_pad);</div><div class="line"><a name="l00214"></a><span class="lineno"> 214</span> </div><div class="line"><a name="l00215"></a><span class="lineno"> 215</span>  <span class="keywordflow">if</span>(input2 != <span class="keyword">nullptr</span>)</div><div class="line"><a name="l00216"></a><span class="lineno"> 216</span>  {</div><div class="line"><a name="l00217"></a><span class="lineno"> 217</span>  <span class="keyword">const</span> <span class="keywordtype">int</span> bias_processed_per_iteration_x = num_elems_processed_per_iteration_x;</div><div class="line"><a name="l00218"></a><span class="lineno"> 218</span> </div><div class="line"><a name="l00219"></a><span class="lineno"> 219</span>  <span class="keyword">const</span> <span class="keywordtype">int</span> bias_processed_per_iteration_y = reshape_info.broadcast_bias() ? 1 : num_elems_processed_per_iteration_y;</div><div class="line"><a name="l00220"></a><span class="lineno"> 220</span> </div><div class="line"><a name="l00221"></a><span class="lineno"> 221</span>  AccessWindowStatic input2_access(input2, 0, 0,</div><div class="line"><a name="l00222"></a><span class="lineno"> 222</span>  <a class="code" href="namespacearm__compute.xhtml#ab237a0a375cf382d52b61653248d3d4a">ceil_to_multiple</a>(input2->dimension(0), bias_processed_per_iteration_x),</div><div class="line"><a name="l00223"></a><span class="lineno"> 223</span>  <a class="code" href="namespacearm__compute.xhtml#ab237a0a375cf382d52b61653248d3d4a">ceil_to_multiple</a>(input2->dimension(1), bias_processed_per_iteration_y));</div><div class="line"><a name="l00224"></a><span class="lineno"> 224</span> </div><div class="line"><a name="l00225"></a><span class="lineno"> 225</span>  window_changed = <a class="code" href="namespacearm__compute.xhtml#afc4bd8e872567d9c4c57d89eb0bb3da1">update_window_and_padding</a>(win, input0_access, input1_access, input2_access) || <span class="comment">// window used by the execute_window_loop</span></div><div class="line"><a name="l00226"></a><span class="lineno"> 226</span>  <a class="code" href="namespacearm__compute.xhtml#afc4bd8e872567d9c4c57d89eb0bb3da1">update_window_and_padding</a>(win_out, output_access); <span class="comment">// window used to update the padding requirements of output tensor</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">else</span></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>  window_changed = <a class="code" href="namespacearm__compute.xhtml#afc4bd8e872567d9c4c57d89eb0bb3da1">update_window_and_padding</a>(win, input0_access, input1_access) || <span class="comment">// window used by the execute_window_loop</span></div><div class="line"><a name="l00231"></a><span class="lineno"> 231</span>  <a class="code" href="namespacearm__compute.xhtml#afc4bd8e872567d9c4c57d89eb0bb3da1">update_window_and_padding</a>(win_out, output_access); <span class="comment">// window used to update the padding requirements of output tensor</span></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> </div><div class="line"><a name="l00234"></a><span class="lineno"> 234</span>  output_access.set_valid_region(win_out, ValidRegion(Coordinates(0, 0), output->tensor_shape()));</div><div class="line"><a name="l00235"></a><span class="lineno"> 235</span>  }</div><div class="line"><a name="l00236"></a><span class="lineno"> 236</span>  <span class="keywordflow">else</span> <span class="comment">// The input tensors have not been reshaped</span></div><div class="line"><a name="l00237"></a><span class="lineno"> 237</span>  {</div><div class="line"><a name="l00238"></a><span class="lineno"> 238</span>  <span class="comment">// Special case for 1xN, 2xN, 3xN and 4xN input0 tensor. num_elems_processed_per_iteration_x is set up for the default case.</span></div><div class="line"><a name="l00239"></a><span class="lineno"> 239</span>  num_elems_processed_per_iteration_x = max_cl_vector_width / <a class="code" href="namespacearm__compute.xhtml#abb7e0f23a4f2e63f39433f158dad47ab">data_size_from_type</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a5f5b6c4337eac9e2e0046ca2304d80dc">data_type</a>);</div><div class="line"><a name="l00240"></a><span class="lineno"> 240</span>  num_elems_processed_per_iteration_y = std::min(static_cast<int>(output->dimension(1)), 4);</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">// Note: bottom paddings are calculated manually as the output can be reinterpreted as 3D tensor</span></div><div class="line"><a name="l00243"></a><span class="lineno"> 243</span>  <span class="comment">// The only way to set properly the paddings, it is to set those explicitly through the AccessWindowStatic</span></div><div class="line"><a name="l00244"></a><span class="lineno"> 244</span>  <span class="keyword">const</span> <span class="keywordtype">int</span> m = reinterpret_input_as_3d ? input0->tensor_shape()[1] * input0->tensor_shape()[2] : input0->tensor_shape()[1];</div><div class="line"><a name="l00245"></a><span class="lineno"> 245</span>  <span class="keyword">const</span> <span class="keywordtype">int</span> bottom_pad = (num_elems_processed_per_iteration_y - (m % num_elems_processed_per_iteration_y)) % num_elems_processed_per_iteration_y;</div><div class="line"><a name="l00246"></a><span class="lineno"> 246</span> </div><div class="line"><a name="l00247"></a><span class="lineno"> 247</span>  <span class="comment">// Create kernels according to the architecture, data type and input size.</span></div><div class="line"><a name="l00248"></a><span class="lineno"> 248</span>  <a class="code" href="namespacearm__compute.xhtml#a735ac6c2a02e320969625308810444f3">GPUTarget</a> arch_target = <a class="code" href="namespacearm__compute.xhtml#a2355c2bf5d1950088937416baea24fe6">get_arch_from_target</a>(gpu_target);</div><div class="line"><a name="l00249"></a><span class="lineno"> 249</span>  <span class="keywordflow">if</span>(arch_target == <a class="code" href="namespacearm__compute.xhtml#a735ac6c2a02e320969625308810444f3aa78cc0fd1cab24af0fad71dc4c256f8e">GPUTarget::BIFROST</a> && <a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a5f5b6c4337eac9e2e0046ca2304d80dc">data_type</a> == <a class="code" href="namespacearm__compute.xhtml#ab4e88c89b3b7ea1735996cc4def22d58a44ad4ef5a76e6aa6fb3e3fa079a54fda">DataType::F32</a>)</div><div class="line"><a name="l00250"></a><span class="lineno"> 250</span>  {</div><div class="line"><a name="l00251"></a><span class="lineno"> 251</span>  num_elems_processed_per_iteration_x = (input1->dimension(0) <= 1000 && input0->num_dimensions() == 1) ? 2 : 4;</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">// Configure window</span></div><div class="line"><a name="l00255"></a><span class="lineno"> 255</span>  win = <a class="code" href="namespacearm__compute.xhtml#ab7980fa5ee693e3282a76da047a1c3b5">calculate_max_window</a>(tmp_info, Steps(num_elems_processed_per_iteration_x, num_elems_processed_per_iteration_y));</div><div class="line"><a name="l00256"></a><span class="lineno"> 256</span>  win_out = <a class="code" href="namespacearm__compute.xhtml#ab7980fa5ee693e3282a76da047a1c3b5">calculate_max_window</a>(*output, Steps(num_elems_processed_per_iteration_x, num_elems_processed_per_iteration_y));</div><div class="line"><a name="l00257"></a><span class="lineno"> 257</span> </div><div class="line"><a name="l00258"></a><span class="lineno"> 258</span>  AccessWindowStatic input0_access(input0, 0, 0, input0->dimension(0), input0->dimension(1) + bottom_pad);</div><div class="line"><a name="l00259"></a><span class="lineno"> 259</span>  AccessWindowStatic input1_access(input1, 0, 0, <a class="code" href="namespacearm__compute.xhtml#ab237a0a375cf382d52b61653248d3d4a">ceil_to_multiple</a>(input1->dimension(0), num_elems_processed_per_iteration_x), input1->dimension(1));</div><div class="line"><a name="l00260"></a><span class="lineno"> 260</span>  AccessWindowStatic output_access(output, 0, 0,</div><div class="line"><a name="l00261"></a><span class="lineno"> 261</span>  <a class="code" href="namespacearm__compute.xhtml#ab237a0a375cf382d52b61653248d3d4a">ceil_to_multiple</a>(output->dimension(0), num_elems_processed_per_iteration_x),</div><div class="line"><a name="l00262"></a><span class="lineno"> 262</span>  output->dimension(1) + bottom_pad);</div><div class="line"><a name="l00263"></a><span class="lineno"> 263</span> </div><div class="line"><a name="l00264"></a><span class="lineno"> 264</span>  <span class="keywordflow">if</span>(input2 != <span class="keyword">nullptr</span>)</div><div class="line"><a name="l00265"></a><span class="lineno"> 265</span>  {</div><div class="line"><a name="l00266"></a><span class="lineno"> 266</span>  <span class="keyword">const</span> <span class="keywordtype">int</span> bias_processed_per_iteration_x = num_elems_processed_per_iteration_x;</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="keyword">const</span> <span class="keywordtype">int</span> bias_processed_per_iteration_y = reshape_info.broadcast_bias() ? 1 : num_elems_processed_per_iteration_y;</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>  AccessWindowStatic input2_access(input2, 0, 0,</div><div class="line"><a name="l00271"></a><span class="lineno"> 271</span>  <a class="code" href="namespacearm__compute.xhtml#ab237a0a375cf382d52b61653248d3d4a">ceil_to_multiple</a>(input2->dimension(0), bias_processed_per_iteration_x),</div><div class="line"><a name="l00272"></a><span class="lineno"> 272</span>  <a class="code" href="namespacearm__compute.xhtml#ab237a0a375cf382d52b61653248d3d4a">ceil_to_multiple</a>(input2->dimension(1), bias_processed_per_iteration_y));</div><div class="line"><a name="l00273"></a><span class="lineno"> 273</span> </div><div class="line"><a name="l00274"></a><span class="lineno"> 274</span>  window_changed = <a class="code" href="namespacearm__compute.xhtml#afc4bd8e872567d9c4c57d89eb0bb3da1">update_window_and_padding</a>(win, input0_access, input1_access, input2_access) || <span class="comment">// window used by the execute_window_loop</span></div><div class="line"><a name="l00275"></a><span class="lineno"> 275</span>  <a class="code" href="namespacearm__compute.xhtml#afc4bd8e872567d9c4c57d89eb0bb3da1">update_window_and_padding</a>(win_out, output_access); <span class="comment">// window used to update the padding requirements of output tensor</span></div><div class="line"><a name="l00276"></a><span class="lineno"> 276</span>  }</div><div class="line"><a name="l00277"></a><span class="lineno"> 277</span>  <span class="keywordflow">else</span></div><div class="line"><a name="l00278"></a><span class="lineno"> 278</span>  {</div><div class="line"><a name="l00279"></a><span class="lineno"> 279</span>  window_changed = <a class="code" href="namespacearm__compute.xhtml#afc4bd8e872567d9c4c57d89eb0bb3da1">update_window_and_padding</a>(win, input0_access, input1_access) || <span class="comment">// window used by the execute_window_loop</span></div><div class="line"><a name="l00280"></a><span class="lineno"> 280</span>  <a class="code" href="namespacearm__compute.xhtml#afc4bd8e872567d9c4c57d89eb0bb3da1">update_window_and_padding</a>(win_out, output_access); <span class="comment">// window used to update the padding requirements of output tensor</span></div><div class="line"><a name="l00281"></a><span class="lineno"> 281</span>  }</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>  Coordinates coord;</div><div class="line"><a name="l00284"></a><span class="lineno"> 284</span>  coord.set_num_dimensions(output->num_dimensions());</div><div class="line"><a name="l00285"></a><span class="lineno"> 285</span>  output_access.set_valid_region(win_out, ValidRegion(coord, output->tensor_shape()));</div><div class="line"><a name="l00286"></a><span class="lineno"> 286</span>  }</div><div class="line"><a name="l00287"></a><span class="lineno"> 287</span> </div><div class="line"><a name="l00288"></a><span class="lineno"> 288</span>  <span class="comment">// Collapse along the Z direction</span></div><div class="line"><a name="l00289"></a><span class="lineno"> 289</span>  <span class="comment">// This collapse needs to be here in order to tune the Z dimension of LWS</span></div><div class="line"><a name="l00290"></a><span class="lineno"> 290</span>  Window collapsed = win;</div><div class="line"><a name="l00291"></a><span class="lineno"> 291</span>  <span class="keyword">const</span> <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> dimension_to_collapse = std::min(static_cast<unsigned int>(output->num_dimensions()), 2u);</div><div class="line"><a name="l00292"></a><span class="lineno"> 292</span>  collapsed = win.collapse(win, dimension_to_collapse);</div><div class="line"><a name="l00293"></a><span class="lineno"> 293</span> </div><div class="line"><a name="l00294"></a><span class="lineno"> 294</span>  Status err = (window_changed) ? <a class="code" href="_error_8h.xhtml#af1b8ff8eb557a2ad11272f1505f45d34">ARM_COMPUTE_CREATE_ERROR</a>(<a class="code" href="namespacearm__compute.xhtml#a59e56af19e754a6aa26a612ebf91d05fa62be47fdd89da032cf78dfce82239579">ErrorCode::RUNTIME_ERROR</a>, <span class="stringliteral">"Insufficient Padding!"</span>) : Status{};</div><div class="line"><a name="l00295"></a><span class="lineno"> 295</span>  <span class="keywordflow">return</span> std::make_pair(err, collapsed);</div><div class="line"><a name="l00296"></a><span class="lineno"> 296</span> }</div><div class="line"><a name="l00297"></a><span class="lineno"> 297</span> } <span class="comment">// namespace</span></div><div class="line"><a name="l00298"></a><span class="lineno"> 298</span> </div><div class="line"><a name="l00299"></a><span class="lineno"><a class="line" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#ac46a1c8a20b46838c9e894f703ddd3ee"> 299</a></span> <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#ac46a1c8a20b46838c9e894f703ddd3ee">CLGEMMMatrixMultiplyKernel::CLGEMMMatrixMultiplyKernel</a>()</div><div class="line"><a name="l00300"></a><span class="lineno"> 300</span>  : _input0(nullptr), _input1(nullptr), _input2(nullptr), _output(nullptr), _slide_matrix_b(true), _reinterpret_input_as_3d(false), _reinterpret_output_as_3d(false), _add_bias(false),</div><div class="line"><a name="l00301"></a><span class="lineno"> 301</span>  _broadcast_bias(false)</div><div class="line"><a name="l00302"></a><span class="lineno"> 302</span> {</div><div class="line"><a name="l00303"></a><span class="lineno"> 303</span> }</div><div class="line"><a name="l00304"></a><span class="lineno"> 304</span> </div><div class="line"><a name="l00305"></a><span class="lineno"><a class="line" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a46bee71bbf58053c85de3f5450566584"> 305</a></span> <span class="keywordtype">void</span> <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a46bee71bbf58053c85de3f5450566584">CLGEMMMatrixMultiplyKernel::configure</a>(<span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *input0, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *input1, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *input2, <a class="code" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *output, <span class="keywordtype">float</span> <a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a3161c2c93c655dd30953372064ec627b">alpha</a>, <span class="keywordtype">float</span> beta,</div><div class="line"><a name="l00306"></a><span class="lineno"> 306</span>  <span class="keywordtype">bool</span> is_interleaved_transposed, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_g_e_m_m_reshape_info.xhtml">GEMMReshapeInfo</a> &reshape_info, <span class="keywordtype">bool</span> fp_mixed_precision, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_activation_layer_info.xhtml">ActivationLayerInfo</a> &activation_info)</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>  <a class="code" href="_validate_8h.xhtml#a921b705e9e3e0fe928928447869e62a5">ARM_COMPUTE_ERROR_ON_NULLPTR</a>(input0, input1, output);</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="comment">// Perform validate step</span></div><div class="line"><a name="l00311"></a><span class="lineno"> 311</span>  <a class="code" href="_error_8h.xhtml#a938dcd406ce611ef5345ad2531cdb948">ARM_COMPUTE_ERROR_THROW_ON</a>(validate_arguments(input0-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>(), input1-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>(), (input2 != <span class="keyword">nullptr</span>) ? input2-><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>(), beta,</div><div class="line"><a name="l00312"></a><span class="lineno"> 312</span>  is_interleaved_transposed, reshape_info, fp_mixed_precision));</div><div class="line"><a name="l00313"></a><span class="lineno"> 313</span> </div><div class="line"><a name="l00314"></a><span class="lineno"> 314</span>  <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a1ab65df01f310bf054323607cd09956e">_input0</a> = input0;</div><div class="line"><a name="l00315"></a><span class="lineno"> 315</span>  <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a6590f81ae0c9f3e01546c73eb31a43c8">_input1</a> = input1;</div><div class="line"><a name="l00316"></a><span class="lineno"> 316</span>  <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a23faf35900f50c084fa1282511b7bd17">_input2</a> = <a class="code" href="namespacearm__compute_1_1helpers_1_1float__ops.xhtml#a3bd19352aed7410633d1f9b95d74a809">helpers::float_ops::is_zero</a>(beta) ? nullptr : input2;</div><div class="line"><a name="l00317"></a><span class="lineno"> 317</span>  <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a62d192d931002b4866443cd7fc71419b">_output</a> = output;</div><div class="line"><a name="l00318"></a><span class="lineno"> 318</span>  <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#aa71ec02f998e1dcfd49ef944ec8cf23e">_reinterpret_input_as_3d</a> = reshape_info.reinterpret_input_as_3d();</div><div class="line"><a name="l00319"></a><span class="lineno"> 319</span>  <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a7d222bcf0d803c0647a4b93061daa56c">_reinterpret_output_as_3d</a> = (reshape_info.depth_output_gemm3d() != 0);</div><div class="line"><a name="l00320"></a><span class="lineno"> 320</span>  <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a94e30ed1aed47fae8430cc4d3cd2b6c7">_add_bias</a> = <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a23faf35900f50c084fa1282511b7bd17">_input2</a> != <span class="keyword">nullptr</span>;</div><div class="line"><a name="l00321"></a><span class="lineno"> 321</span>  <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a42af734585418559f06f6ce9f7375910">_broadcast_bias</a> = reshape_info.broadcast_bias();</div><div class="line"><a name="l00322"></a><span class="lineno"> 322</span> </div><div class="line"><a name="l00323"></a><span class="lineno"> 323</span>  <span class="comment">// In case both input and output have to be reinterpreted as 3D tensors,</span></div><div class="line"><a name="l00324"></a><span class="lineno"> 324</span>  <span class="comment">// force reinterpret_input_as_3d and reinterpret_output_as_3d to be false.</span></div><div class="line"><a name="l00325"></a><span class="lineno"> 325</span>  <span class="keywordflow">if</span>(<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#aa71ec02f998e1dcfd49ef944ec8cf23e">_reinterpret_input_as_3d</a> == <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a7d222bcf0d803c0647a4b93061daa56c">_reinterpret_output_as_3d</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>  <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#aa71ec02f998e1dcfd49ef944ec8cf23e">_reinterpret_input_as_3d</a> = <span class="keyword">false</span>;</div><div class="line"><a name="l00328"></a><span class="lineno"> 328</span>  <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a7d222bcf0d803c0647a4b93061daa56c">_reinterpret_output_as_3d</a> = <span class="keyword">false</span>;</div><div class="line"><a name="l00329"></a><span class="lineno"> 329</span>  }</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>  <span class="comment">// Check if we need to slide the matrix B</span></div><div class="line"><a name="l00332"></a><span class="lineno"> 332</span>  <span class="keyword">const</span> <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> num_dimensions_input0 = <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#aa71ec02f998e1dcfd49ef944ec8cf23e">_reinterpret_input_as_3d</a> ? <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a1ab65df01f310bf054323607cd09956e">_input0</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#a1f4e725b8e1ea36b30e09dc08ae6961d">num_dimensions</a>() - 1 : <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a1ab65df01f310bf054323607cd09956e">_input0</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#a1f4e725b8e1ea36b30e09dc08ae6961d">num_dimensions</a>();</div><div class="line"><a name="l00333"></a><span class="lineno"> 333</span> </div><div class="line"><a name="l00334"></a><span class="lineno"> 334</span>  <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a170f236fd8751c4e1675873b496f7cf8">_slide_matrix_b</a> = (<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a6590f81ae0c9f3e01546c73eb31a43c8">_input1</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#a1f4e725b8e1ea36b30e09dc08ae6961d">num_dimensions</a>() >= num_dimensions_input0);</div><div class="line"><a name="l00335"></a><span class="lineno"> 335</span> </div><div class="line"><a name="l00336"></a><span class="lineno"> 336</span>  <span class="keyword">const</span> <a class="code" href="namespacearm__compute.xhtml#ad8ed01ff3ff33333d8e19db4d2818bb6">DataType</a> <a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a5f5b6c4337eac9e2e0046ca2304d80dc">data_type</a> = input0-><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="l00337"></a><span class="lineno"> 337</span> </div><div class="line"><a name="l00338"></a><span class="lineno"> 338</span>  <span class="comment">// Get target architecture</span></div><div class="line"><a name="l00339"></a><span class="lineno"> 339</span>  <a class="code" href="namespacearm__compute.xhtml#a735ac6c2a02e320969625308810444f3">GPUTarget</a> gpu_target = <a class="code" href="classarm__compute_1_1_i_c_l_kernel.xhtml#aa550ff0352ff2388e02f7b0a41bf5fe7">get_target</a>();</div><div class="line"><a name="l00340"></a><span class="lineno"> 340</span> </div><div class="line"><a name="l00341"></a><span class="lineno"> 341</span>  ElementsProcessed num_elements_processed{};</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>  <span class="comment">// Configure kernel window</span></div><div class="line"><a name="l00344"></a><span class="lineno"> 344</span>  <span class="keyword">auto</span> win_config = validate_and_configure_window(input0-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>(), input1-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>(), (input2 != <span class="keyword">nullptr</span>) ? input2-><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>(), beta, is_interleaved_transposed, reshape_info,</div><div class="line"><a name="l00345"></a><span class="lineno"> 345</span>  gpu_target, num_elements_processed);</div><div class="line"><a name="l00346"></a><span class="lineno"> 346</span>  <a class="code" href="_error_8h.xhtml#a938dcd406ce611ef5345ad2531cdb948">ARM_COMPUTE_ERROR_THROW_ON</a>(win_config.first);</div><div class="line"><a name="l00347"></a><span class="lineno"> 347</span>  ICLKernel::configure_internal(win_config.second);</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">// Create build options</span></div><div class="line"><a name="l00350"></a><span class="lineno"> 350</span>  <a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml">CLBuildOptions</a> build_opts;</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>  build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#a95b46e69297fad10b27a1baa000f92cc">add_option_if</a>(!(<a class="code" href="namespacearm__compute_1_1helpers_1_1float__ops.xhtml#ab2dcf325d146568ecc8d4a4bd36c02ac">helpers::float_ops::is_one</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a3161c2c93c655dd30953372064ec627b">alpha</a>)), <span class="stringliteral">"-DALPHA="</span> + <a class="code" href="namespacearm__compute.xhtml#a9cd394c15b73f79ca1d98f5328064be2">float_to_string_with_full_precision</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a3161c2c93c655dd30953372064ec627b">alpha</a>));</div><div class="line"><a name="l00353"></a><span class="lineno"> 353</span>  build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#a95b46e69297fad10b27a1baa000f92cc">add_option_if</a>(<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a23faf35900f50c084fa1282511b7bd17">_input2</a> != <span class="keyword">nullptr</span>, <span class="stringliteral">"-DBETA="</span> + <a class="code" href="namespacearm__compute.xhtml#a9cd394c15b73f79ca1d98f5328064be2">float_to_string_with_full_precision</a>(beta));</div><div class="line"><a name="l00354"></a><span class="lineno"> 354</span>  build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#a95b46e69297fad10b27a1baa000f92cc">add_option_if</a>(<a class="code" href="namespacearm__compute_1_1helpers_1_1float__ops.xhtml#ab2dcf325d146568ecc8d4a4bd36c02ac">helpers::float_ops::is_one</a>(beta), <span class="stringliteral">"-DUNIT_BETA"</span>);</div><div class="line"><a name="l00355"></a><span class="lineno"> 355</span>  build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#a95b46e69297fad10b27a1baa000f92cc">add_option_if</a>(reshape_info.broadcast_bias(), <span class="stringliteral">"-DBROADCAST_BIAS"</span>);</div><div class="line"><a name="l00356"></a><span class="lineno"> 356</span>  build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#a95b46e69297fad10b27a1baa000f92cc">add_option_if</a>(<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#aa71ec02f998e1dcfd49ef944ec8cf23e">_reinterpret_input_as_3d</a>, <span class="stringliteral">"-DREINTERPRET_INPUT_AS_3D"</span>);</div><div class="line"><a name="l00357"></a><span class="lineno"> 357</span>  build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#a95b46e69297fad10b27a1baa000f92cc">add_option_if</a>(<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a7d222bcf0d803c0647a4b93061daa56c">_reinterpret_output_as_3d</a>, <span class="stringliteral">"-DREINTERPRET_OUTPUT_AS_3D"</span>);</div><div class="line"><a name="l00358"></a><span class="lineno"> 358</span>  build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#a95b46e69297fad10b27a1baa000f92cc">add_option_if</a>(<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#aa71ec02f998e1dcfd49ef944ec8cf23e">_reinterpret_input_as_3d</a> || <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a7d222bcf0d803c0647a4b93061daa56c">_reinterpret_output_as_3d</a>, <span class="stringliteral">"-DHEIGHT_GEMM3D="</span> + <a class="code" href="namespacearm__compute_1_1support_1_1cpp11.xhtml#acc5dddee1cbe93a4eaf0a9f74ee96bb7">support::cpp11::to_string</a>(output-><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="l00359"></a><span class="lineno"> 359</span>  build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#a95b46e69297fad10b27a1baa000f92cc">add_option_if</a>(<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#aa71ec02f998e1dcfd49ef944ec8cf23e">_reinterpret_input_as_3d</a> || <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a7d222bcf0d803c0647a4b93061daa56c">_reinterpret_output_as_3d</a>, <span class="stringliteral">"-DDEPTH_GEMM3D="</span> + <a class="code" href="namespacearm__compute_1_1support_1_1cpp11.xhtml#acc5dddee1cbe93a4eaf0a9f74ee96bb7">support::cpp11::to_string</a>(output-><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="l00360"></a><span class="lineno"> 360</span>  build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#a95b46e69297fad10b27a1baa000f92cc">add_option_if</a>(!<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a170f236fd8751c4e1675873b496f7cf8">_slide_matrix_b</a>, <span class="stringliteral">"-DMATRIX_B_DEPTH="</span> + <a class="code" href="namespacearm__compute_1_1support_1_1cpp11.xhtml#acc5dddee1cbe93a4eaf0a9f74ee96bb7">support::cpp11::to_string</a>(input1-><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="l00361"></a><span class="lineno"> 361</span>  build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#a95b46e69297fad10b27a1baa000f92cc">add_option_if</a>(activation_info.<a class="code" href="classarm__compute_1_1_activation_layer_info.xhtml#af5a8385102f8f8dd6c5957eac08b04c2">enabled</a>(), <span class="stringliteral">"-DACTIVATION_TYPE="</span> + <a class="code" href="namespacearm__compute.xhtml#a0a7bb17a0a0414a7162f635776a02eb5">lower_string</a>(<a class="code" href="namespacearm__compute.xhtml#a635f1895d94050329b7da12850d1a056">string_from_activation_func</a>(activation_info.<a class="code" href="classarm__compute_1_1_activation_layer_info.xhtml#a9e0fb1d1462557f28966ae19988532c2">activation</a>())));</div><div class="line"><a name="l00362"></a><span class="lineno"> 362</span>  build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#a95b46e69297fad10b27a1baa000f92cc">add_option_if</a>(activation_info.<a class="code" href="classarm__compute_1_1_activation_layer_info.xhtml#af5a8385102f8f8dd6c5957eac08b04c2">enabled</a>(), <span class="stringliteral">"-DA_VAL="</span> + <a class="code" href="namespacearm__compute.xhtml#a9cd394c15b73f79ca1d98f5328064be2">float_to_string_with_full_precision</a>(activation_info.<a class="code" href="classarm__compute_1_1_activation_layer_info.xhtml#aec65e090c2276e8c8f8dffb6e3edd178">a</a>()));</div><div class="line"><a name="l00363"></a><span class="lineno"> 363</span>  build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#a95b46e69297fad10b27a1baa000f92cc">add_option_if</a>(activation_info.<a class="code" href="classarm__compute_1_1_activation_layer_info.xhtml#af5a8385102f8f8dd6c5957eac08b04c2">enabled</a>(), <span class="stringliteral">"-DB_VAL="</span> + <a class="code" href="namespacearm__compute.xhtml#a9cd394c15b73f79ca1d98f5328064be2">float_to_string_with_full_precision</a>(activation_info.<a class="code" href="classarm__compute_1_1_activation_layer_info.xhtml#a02a00a5d20986f3a7ab72b9c86be3a54">b</a>()));</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="keyword">const</span> <span class="keywordtype">bool</span> is_bifrost = <a class="code" href="namespacearm__compute.xhtml#a2355c2bf5d1950088937416baea24fe6">get_arch_from_target</a>(gpu_target) == <a class="code" href="namespacearm__compute.xhtml#a735ac6c2a02e320969625308810444f3aa78cc0fd1cab24af0fad71dc4c256f8e">GPUTarget::BIFROST</a>;</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>  std::string <a class="code" href="_c_l_im2_col_kernel_8cpp.xhtml#a624a24e6d361fe7b8b8b2f6b375683a4">kernel_name</a>;</div><div class="line"><a name="l00368"></a><span class="lineno"> 368</span>  <span class="keywordflow">if</span>(is_interleaved_transposed)</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="keyword">const</span> <span class="keywordtype">int</span> mult_transpose1xW_width = reshape_info.mult_transpose1xW_width();</div><div class="line"><a name="l00371"></a><span class="lineno"> 371</span>  <span class="keyword">const</span> <span class="keywordtype">int</span> mult_interleave4x4_height = reshape_info.mult_interleave4x4_height();</div><div class="line"><a name="l00372"></a><span class="lineno"> 372</span> </div><div class="line"><a name="l00373"></a><span class="lineno"> 373</span>  build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#a3e2b80ff5463b7d2017de847f5c32a30">add_option</a>(<span class="stringliteral">"-DCOLS_B="</span> + <a class="code" href="namespacearm__compute_1_1support_1_1cpp11.xhtml#acc5dddee1cbe93a4eaf0a9f74ee96bb7">support::cpp11::to_string</a>(input1-><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="l00374"></a><span class="lineno"> 374</span>  build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#a3e2b80ff5463b7d2017de847f5c32a30">add_option</a>(<span class="stringliteral">"-DMULT_TRANSPOSE1XW_WIDTH="</span> + <a class="code" href="namespacearm__compute_1_1support_1_1cpp11.xhtml#acc5dddee1cbe93a4eaf0a9f74ee96bb7">support::cpp11::to_string</a>(mult_transpose1xW_width));</div><div class="line"><a name="l00375"></a><span class="lineno"> 375</span>  build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#a3e2b80ff5463b7d2017de847f5c32a30">add_option</a>(<span class="stringliteral">"-DMULT_INTERLEAVE4X4_HEIGHT="</span> + <a class="code" href="namespacearm__compute_1_1support_1_1cpp11.xhtml#acc5dddee1cbe93a4eaf0a9f74ee96bb7">support::cpp11::to_string</a>(mult_interleave4x4_height));</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="keywordflow">if</span>(<a class="code" href="namespacearm__compute.xhtml#af5982a092e9eb743fce2d6392bdd8897">is_data_type_float</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a5f5b6c4337eac9e2e0046ca2304d80dc">data_type</a>) && is_bifrost)</div><div class="line"><a name="l00378"></a><span class="lineno"> 378</span>  {</div><div class="line"><a name="l00379"></a><span class="lineno"> 379</span>  <a class="code" href="_c_l_im2_col_kernel_8cpp.xhtml#a624a24e6d361fe7b8b8b2f6b375683a4">kernel_name</a> = <span class="stringliteral">"gemm_mm_interleaved_transposed_"</span> + <a class="code" href="namespacearm__compute.xhtml#a0a7bb17a0a0414a7162f635776a02eb5">lower_string</a>(<a class="code" href="namespacearm__compute.xhtml#a142b55a483cadf4e1068a1a09a55e8e9">string_from_data_type</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a5f5b6c4337eac9e2e0046ca2304d80dc">data_type</a>)) + <span class="stringliteral">"_bifrost"</span>;</div><div class="line"><a name="l00380"></a><span class="lineno"> 380</span>  }</div><div class="line"><a name="l00381"></a><span class="lineno"> 381</span>  <span class="keywordflow">else</span></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>  <a class="code" href="_c_l_im2_col_kernel_8cpp.xhtml#a624a24e6d361fe7b8b8b2f6b375683a4">kernel_name</a> = <span class="stringliteral">"gemm_mm_interleaved_transposed_"</span> + <a class="code" href="namespacearm__compute.xhtml#a0a7bb17a0a0414a7162f635776a02eb5">lower_string</a>(<a class="code" href="namespacearm__compute.xhtml#a142b55a483cadf4e1068a1a09a55e8e9">string_from_data_type</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a5f5b6c4337eac9e2e0046ca2304d80dc">data_type</a>));</div><div class="line"><a name="l00384"></a><span class="lineno"> 384</span>  <span class="keywordflow">if</span>(fp_mixed_precision && <a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a5f5b6c4337eac9e2e0046ca2304d80dc">data_type</a> == <a class="code" href="namespacearm__compute.xhtml#ab4e88c89b3b7ea1735996cc4def22d58a56d8353718e6fdc78b8d69078a2cdb94">DataType::F16</a>)</div><div class="line"><a name="l00385"></a><span class="lineno"> 385</span>  {</div><div class="line"><a name="l00386"></a><span class="lineno"> 386</span>  <span class="comment">// currently wider accumulator is only supported for fp16 kernels.</span></div><div class="line"><a name="l00387"></a><span class="lineno"> 387</span>  <a class="code" href="_c_l_im2_col_kernel_8cpp.xhtml#a624a24e6d361fe7b8b8b2f6b375683a4">kernel_name</a> += <span class="stringliteral">"_acc32"</span>;</div><div class="line"><a name="l00388"></a><span class="lineno"> 388</span>  }</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>  }</div><div class="line"><a name="l00391"></a><span class="lineno"> 391</span>  <span class="keywordflow">else</span> <span class="comment">// The input tensors have not been reshaped</span></div><div class="line"><a name="l00392"></a><span class="lineno"> 392</span>  {</div><div class="line"><a name="l00393"></a><span class="lineno"> 393</span>  build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#a3e2b80ff5463b7d2017de847f5c32a30">add_option</a>(<span class="stringliteral">"-DCOLS_A="</span> + <a class="code" href="namespacearm__compute_1_1support_1_1cpp11.xhtml#acc5dddee1cbe93a4eaf0a9f74ee96bb7">support::cpp11::to_string</a>(input0-><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="l00394"></a><span class="lineno"> 394</span>  build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#a3e2b80ff5463b7d2017de847f5c32a30">add_option</a>(<span class="stringliteral">"-DDATA_TYPE="</span> + <a class="code" href="namespacearm__compute.xhtml#a545eeda2eaa3f5a54345ce8169e21184">get_cl_type_from_data_type</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a5f5b6c4337eac9e2e0046ca2304d80dc">data_type</a>));</div><div class="line"><a name="l00395"></a><span class="lineno"> 395</span> </div><div class="line"><a name="l00396"></a><span class="lineno"> 396</span>  <span class="comment">// Create kernels according to the architecture, data type and input size.</span></div><div class="line"><a name="l00397"></a><span class="lineno"> 397</span>  <span class="keywordflow">if</span>(<a class="code" href="namespacearm__compute.xhtml#af5982a092e9eb743fce2d6392bdd8897">is_data_type_float</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a5f5b6c4337eac9e2e0046ca2304d80dc">data_type</a>) && is_bifrost)</div><div class="line"><a name="l00398"></a><span class="lineno"> 398</span>  {</div><div class="line"><a name="l00399"></a><span class="lineno"> 399</span>  <a class="code" href="_c_l_im2_col_kernel_8cpp.xhtml#a624a24e6d361fe7b8b8b2f6b375683a4">kernel_name</a> = <span class="stringliteral">"gemm_mm_floating_point"</span>;</div><div class="line"><a name="l00400"></a><span class="lineno"> 400</span> </div><div class="line"><a name="l00401"></a><span class="lineno"> 401</span>  <span class="keywordflow">if</span>(input0-><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#a1f4e725b8e1ea36b30e09dc08ae6961d">num_dimensions</a>() != 1)</div><div class="line"><a name="l00402"></a><span class="lineno"> 402</span>  {</div><div class="line"><a name="l00403"></a><span class="lineno"> 403</span>  <a class="code" href="_c_l_im2_col_kernel_8cpp.xhtml#a624a24e6d361fe7b8b8b2f6b375683a4">kernel_name</a> += <span class="stringliteral">"_"</span> + <a class="code" href="namespacearm__compute.xhtml#a0a7bb17a0a0414a7162f635776a02eb5">lower_string</a>(<a class="code" href="namespacearm__compute.xhtml#a142b55a483cadf4e1068a1a09a55e8e9">string_from_data_type</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a5f5b6c4337eac9e2e0046ca2304d80dc">data_type</a>)) + <span class="stringliteral">"_bifrost"</span>;</div><div class="line"><a name="l00404"></a><span class="lineno"> 404</span>  <span class="keywordflow">if</span>(fp_mixed_precision && <a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a5f5b6c4337eac9e2e0046ca2304d80dc">data_type</a> == <a class="code" href="namespacearm__compute.xhtml#ab4e88c89b3b7ea1735996cc4def22d58a56d8353718e6fdc78b8d69078a2cdb94">DataType::F16</a>)</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">// currently wider accumulator is only supported for fp16 kernels.</span></div><div class="line"><a name="l00407"></a><span class="lineno"> 407</span>  <a class="code" href="_c_l_im2_col_kernel_8cpp.xhtml#a624a24e6d361fe7b8b8b2f6b375683a4">kernel_name</a> += <span class="stringliteral">"_acc32"</span>;</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>  }</div><div class="line"><a name="l00410"></a><span class="lineno"> 410</span>  <span class="keywordflow">else</span> <span class="keywordflow">if</span>(input1-><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) <= 1000 && <a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a5f5b6c4337eac9e2e0046ca2304d80dc">data_type</a> == <a class="code" href="namespacearm__compute.xhtml#ab4e88c89b3b7ea1735996cc4def22d58a44ad4ef5a76e6aa6fb3e3fa079a54fda">DataType::F32</a>)</div><div class="line"><a name="l00411"></a><span class="lineno"> 411</span>  {</div><div class="line"><a name="l00412"></a><span class="lineno"> 412</span>  <span class="comment">// The first kernel is optimized for the case of 1000 or less output elements (e.g. FC8 of AlexNet and VGG-16, and</span></div><div class="line"><a name="l00413"></a><span class="lineno"> 413</span>  <span class="comment">// FC1 of Inception v3). The second kernel is optimized for the case of greater than 1000 output elements (e.g.</span></div><div class="line"><a name="l00414"></a><span class="lineno"> 414</span>  <span class="comment">// FC6 and FC7 of AlexNet and VGG-16).</span></div><div class="line"><a name="l00415"></a><span class="lineno"> 415</span>  <a class="code" href="_c_l_im2_col_kernel_8cpp.xhtml#a624a24e6d361fe7b8b8b2f6b375683a4">kernel_name</a> += <span class="stringliteral">"_"</span> + <a class="code" href="namespacearm__compute.xhtml#a0a7bb17a0a0414a7162f635776a02eb5">lower_string</a>(<a class="code" href="namespacearm__compute.xhtml#a142b55a483cadf4e1068a1a09a55e8e9">string_from_data_type</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a5f5b6c4337eac9e2e0046ca2304d80dc">data_type</a>)) + <span class="stringliteral">"_bifrost_1000"</span>;</div><div class="line"><a name="l00416"></a><span class="lineno"> 416</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">// The work-group size equal to the Bifrost quad size has been proved to be optimal for these kernels</span></div><div class="line"><a name="l00419"></a><span class="lineno"> 419</span>  <span class="comment">// via exhaustive autotuning over a range of representative layer configurations.</span></div><div class="line"><a name="l00420"></a><span class="lineno"> 420</span>  <a class="code" href="classarm__compute_1_1_i_c_l_kernel.xhtml#ad356b88c8f61267d593d9ed99835bde9">set_lws_hint</a>(cl::NDRange(4));</div><div class="line"><a name="l00421"></a><span class="lineno"> 421</span>  }</div><div class="line"><a name="l00422"></a><span class="lineno"> 422</span>  <span class="keywordflow">else</span> <span class="comment">// (MIDGARD and F32) or (F16)</span></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>  <a class="code" href="_c_l_im2_col_kernel_8cpp.xhtml#a624a24e6d361fe7b8b8b2f6b375683a4">kernel_name</a> = <span class="stringliteral">"gemm_mm_floating_point"</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>  build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#a3e2b80ff5463b7d2017de847f5c32a30">add_option</a>(<span class="stringliteral">"-DNUM_ELEMS_PROCESSED_PER_THREAD_Y="</span> + <a class="code" href="namespacearm__compute_1_1support_1_1cpp11.xhtml#acc5dddee1cbe93a4eaf0a9f74ee96bb7">support::cpp11::to_string</a>(num_elements_processed.y()));</div><div class="line"><a name="l00427"></a><span class="lineno"> 427</span>  build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#a3e2b80ff5463b7d2017de847f5c32a30">add_option</a>(<span class="stringliteral">"-DNUM_ELEMS_PROCESSED_PER_THREAD_X="</span> + <a class="code" href="namespacearm__compute_1_1support_1_1cpp11.xhtml#acc5dddee1cbe93a4eaf0a9f74ee96bb7">support::cpp11::to_string</a>(num_elements_processed.x()));</div><div class="line"><a name="l00428"></a><span class="lineno"> 428</span>  }</div><div class="line"><a name="l00429"></a><span class="lineno"> 429</span> </div><div class="line"><a name="l00430"></a><span class="lineno"> 430</span>  <span class="comment">// Create kernel</span></div><div class="line"><a name="l00431"></a><span class="lineno"> 431</span>  _kernel = static_cast<cl::Kernel>(<a class="code" href="classarm__compute_1_1_c_l_kernel_library.xhtml#acba005f5ce2c62cbf3f94d074d9007aa">CLKernelLibrary::get</a>().<a class="code" href="namespacearm__compute.xhtml#abc72c95941485d8a068fa38372308574">create_kernel</a>(<a class="code" href="_c_l_im2_col_kernel_8cpp.xhtml#a624a24e6d361fe7b8b8b2f6b375683a4">kernel_name</a>, build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#ae3b08139a1e57323c5d7dd93f30496c8">options</a>()));</div><div class="line"><a name="l00432"></a><span class="lineno"> 432</span> </div><div class="line"><a name="l00433"></a><span class="lineno"> 433</span>  <span class="comment">// Set config_id for enabling LWS tuning</span></div><div class="line"><a name="l00434"></a><span class="lineno"> 434</span>  _config_id = <span class="stringliteral">"gemm_"</span>;</div><div class="line"><a name="l00435"></a><span class="lineno"> 435</span>  _config_id += (is_interleaved_transposed ? <span class="stringliteral">"reshaped_"</span> : <span class="stringliteral">""</span>);</div><div class="line"><a name="l00436"></a><span class="lineno"> 436</span>  _config_id += (<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a94e30ed1aed47fae8430cc4d3cd2b6c7">_add_bias</a> ? <span class="stringliteral">"add_bias_"</span> : <span class="stringliteral">""</span>);</div><div class="line"><a name="l00437"></a><span class="lineno"> 437</span>  _config_id += (<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a42af734585418559f06f6ce9f7375910">_broadcast_bias</a> ? <span class="stringliteral">"broadcast_bias_"</span> : <span class="stringliteral">""</span>);</div><div class="line"><a name="l00438"></a><span class="lineno"> 438</span>  _config_id += (fp_mixed_precision ? <span class="stringliteral">"fp_mixed_"</span> : <span class="stringliteral">""</span>);</div><div class="line"><a name="l00439"></a><span class="lineno"> 439</span>  _config_id += (<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#aa71ec02f998e1dcfd49ef944ec8cf23e">_reinterpret_input_as_3d</a> ? <span class="stringliteral">"3di_"</span> : <span class="stringliteral">""</span>);</div><div class="line"><a name="l00440"></a><span class="lineno"> 440</span>  _config_id += (<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a7d222bcf0d803c0647a4b93061daa56c">_reinterpret_output_as_3d</a> ? <span class="stringliteral">"3do_"</span> : <span class="stringliteral">""</span>);</div><div class="line"><a name="l00441"></a><span class="lineno"> 441</span>  _config_id += <a class="code" href="namespacearm__compute.xhtml#a0a7bb17a0a0414a7162f635776a02eb5">lower_string</a>(<a class="code" href="namespacearm__compute.xhtml#a142b55a483cadf4e1068a1a09a55e8e9">string_from_data_type</a>(input0-><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="l00442"></a><span class="lineno"> 442</span>  _config_id += <span class="stringliteral">"_"</span>;</div><div class="line"><a name="l00443"></a><span class="lineno"> 443</span>  _config_id += <a class="code" href="namespacearm__compute_1_1support_1_1cpp11.xhtml#acc5dddee1cbe93a4eaf0a9f74ee96bb7">support::cpp11::to_string</a>(output-><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="l00444"></a><span class="lineno"> 444</span>  _config_id += <span class="stringliteral">"_"</span>;</div><div class="line"><a name="l00445"></a><span class="lineno"> 445</span>  _config_id += <a class="code" href="namespacearm__compute_1_1support_1_1cpp11.xhtml#acc5dddee1cbe93a4eaf0a9f74ee96bb7">support::cpp11::to_string</a>(output-><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="l00446"></a><span class="lineno"> 446</span>  _config_id += <span class="stringliteral">"_"</span>;</div><div class="line"><a name="l00447"></a><span class="lineno"> 447</span>  _config_id += <a class="code" href="namespacearm__compute_1_1support_1_1cpp11.xhtml#acc5dddee1cbe93a4eaf0a9f74ee96bb7">support::cpp11::to_string</a>(output-><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="l00448"></a><span class="lineno"> 448</span>  _config_id += <span class="stringliteral">"_"</span>;</div><div class="line"><a name="l00449"></a><span class="lineno"> 449</span>  _config_id += <a class="code" href="namespacearm__compute_1_1support_1_1cpp11.xhtml#acc5dddee1cbe93a4eaf0a9f74ee96bb7">support::cpp11::to_string</a>(output-><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));</div><div class="line"><a name="l00450"></a><span class="lineno"> 450</span>  _config_id += <span class="stringliteral">"_"</span>;</div><div class="line"><a name="l00451"></a><span class="lineno"> 451</span>  _config_id += (is_interleaved_transposed ? <a class="code" href="namespacearm__compute_1_1support_1_1cpp11.xhtml#acc5dddee1cbe93a4eaf0a9f74ee96bb7">support::cpp11::to_string</a>(input1-><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 class="code" href="namespacearm__compute_1_1support_1_1cpp11.xhtml#acc5dddee1cbe93a4eaf0a9f74ee96bb7">support::cpp11::to_string</a>(input1-><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="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"><a class="line" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#af45425674a854a3bb158b0b3d0ba9d3e"> 454</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_matrix_multiply_kernel.xhtml#af45425674a854a3bb158b0b3d0ba9d3e">CLGEMMMatrixMultiplyKernel::validate</a>(<span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *input0, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *input1, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *input2, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *output, <span class="keywordtype">float</span> <a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a3161c2c93c655dd30953372064ec627b">alpha</a>, <span class="keywordtype">float</span> beta,</div><div class="line"><a name="l00455"></a><span class="lineno"> 455</span>  <span class="keywordtype">bool</span> is_interleaved_transposed, <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="namespacearm__compute.xhtml#a735ac6c2a02e320969625308810444f3">GPUTarget</a> gpu_target, <span class="keywordtype">bool</span> fp_mixed_precision, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_activation_layer_info.xhtml">ActivationLayerInfo</a> &activation_info)</div><div class="line"><a name="l00456"></a><span class="lineno"> 456</span> {</div><div class="line"><a name="l00457"></a><span class="lineno"> 457</span>  <span class="comment">// Note: num_elements_processed will be set in validate_and_configure_window()</span></div><div class="line"><a name="l00458"></a><span class="lineno"> 458</span>  ElementsProcessed num_elements_processed{};</div><div class="line"><a name="l00459"></a><span class="lineno"> 459</span>  <a class="code" href="_error_8h.xhtml#a6dc630a6ae9cc063b3924bcea8dee9d6">ARM_COMPUTE_UNUSED</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a3161c2c93c655dd30953372064ec627b">alpha</a>);</div><div class="line"><a name="l00460"></a><span class="lineno"> 460</span>  <a class="code" href="_error_8h.xhtml#a6dc630a6ae9cc063b3924bcea8dee9d6">ARM_COMPUTE_UNUSED</a>(activation_info);</div><div class="line"><a name="l00461"></a><span class="lineno"> 461</span>  <a class="code" href="_error_8h.xhtml#a8a1e1c105f0bdaf37db408c7cfcb77a4">ARM_COMPUTE_RETURN_ON_ERROR</a>(validate_arguments(input0, input1, input2, output, beta, is_interleaved_transposed, reshape_info, fp_mixed_precision));</div><div class="line"><a name="l00462"></a><span class="lineno"> 462</span>  <a class="code" href="_error_8h.xhtml#a8a1e1c105f0bdaf37db408c7cfcb77a4">ARM_COMPUTE_RETURN_ON_ERROR</a>(validate_and_configure_window(input0-><a class="code" href="classarm__compute_1_1misc_1_1_i_cloneable.xhtml#a4d10e5012a872e7f78f2b539b673049d">clone</a>().get(),</div><div class="line"><a name="l00463"></a><span class="lineno"> 463</span>  input1-><a class="code" href="classarm__compute_1_1misc_1_1_i_cloneable.xhtml#a4d10e5012a872e7f78f2b539b673049d">clone</a>().get(),</div><div class="line"><a name="l00464"></a><span class="lineno"> 464</span>  (input2 != <span class="keyword">nullptr</span>) ? input2-><a class="code" href="classarm__compute_1_1misc_1_1_i_cloneable.xhtml#a4d10e5012a872e7f78f2b539b673049d">clone</a>().get() : <span class="keyword">nullptr</span>,</div><div class="line"><a name="l00465"></a><span class="lineno"> 465</span>  output-><a class="code" href="classarm__compute_1_1misc_1_1_i_cloneable.xhtml#a4d10e5012a872e7f78f2b539b673049d">clone</a>().get(),</div><div class="line"><a name="l00466"></a><span class="lineno"> 466</span>  beta,</div><div class="line"><a name="l00467"></a><span class="lineno"> 467</span>  is_interleaved_transposed,</div><div class="line"><a name="l00468"></a><span class="lineno"> 468</span>  reshape_info,</div><div class="line"><a name="l00469"></a><span class="lineno"> 469</span>  gpu_target,</div><div class="line"><a name="l00470"></a><span class="lineno"> 470</span>  num_elements_processed)</div><div class="line"><a name="l00471"></a><span class="lineno"> 471</span>  .first);</div><div class="line"><a name="l00472"></a><span class="lineno"> 472</span> </div><div class="line"><a name="l00473"></a><span class="lineno"> 473</span>  <span class="keywordflow">return</span> <a class="code" href="classarm__compute_1_1_status.xhtml">Status</a>{};</div><div class="line"><a name="l00474"></a><span class="lineno"> 474</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"><a class="line" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a493987e85723a8000eb26d1f00e2ad0e"> 476</a></span> <span class="keywordtype">void</span> <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a493987e85723a8000eb26d1f00e2ad0e">CLGEMMMatrixMultiplyKernel::run</a>(<span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_window.xhtml">Window</a> &window, cl::CommandQueue &queue)</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>  <a class="code" href="_validate_8h.xhtml#a1b35b0d258183cf9ef36adf684d0b88c">ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL</a>(<span class="keyword">this</span>);</div><div class="line"><a name="l00479"></a><span class="lineno"> 479</span>  <a class="code" href="_validate_8h.xhtml#a6eb9ce82815fe429250189da7592ba75">ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW</a>(<a class="code" href="classarm__compute_1_1_i_kernel.xhtml#ad34a46f53686c12a5c5e717cc9617fb6">ICLKernel::window</a>(), <a class="code" href="classarm__compute_1_1_i_kernel.xhtml#ad34a46f53686c12a5c5e717cc9617fb6">window</a>);</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>  <span class="keywordflow">if</span>(<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a6590f81ae0c9f3e01546c73eb31a43c8">_input1</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#a1f4e725b8e1ea36b30e09dc08ae6961d">num_dimensions</a>() < 3)</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>  <span class="comment">// The stride_z for matrix B must be zero if we do not slice</span></div><div class="line"><a name="l00484"></a><span class="lineno"> 484</span>  <a class="code" href="_error_8h.xhtml#a54a6080c9f4df1f908e57a9bbb46f5da">ARM_COMPUTE_ERROR_ON</a>(<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a6590f81ae0c9f3e01546c73eb31a43c8">_input1</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#a6b14f175bf5281f57b561e2d4e4b1f1f">strides_in_bytes</a>()[3] != 0);</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> </div><div class="line"><a name="l00487"></a><span class="lineno"> 487</span>  <a class="code" href="classarm__compute_1_1_window.xhtml">Window</a> <a class="code" href="namespacearm__compute_1_1test_1_1validation_1_1reference.xhtml#a548131b3d37da47a2e9d32111c88dfe1">slice</a> = <a class="code" href="classarm__compute_1_1_i_kernel.xhtml#ad34a46f53686c12a5c5e717cc9617fb6">window</a>.<a class="code" href="classarm__compute_1_1_window.xhtml#a30ca5bdbb60ee281d7f1ab34f7a4ee40">first_slice_window_3D</a>();</div><div class="line"><a name="l00488"></a><span class="lineno"> 488</span>  <a class="code" href="classarm__compute_1_1_window.xhtml">Window</a> slice_matrix_b = <a class="code" href="namespacearm__compute_1_1test_1_1validation_1_1reference.xhtml#a548131b3d37da47a2e9d32111c88dfe1">slice</a>;</div><div class="line"><a name="l00489"></a><span class="lineno"> 489</span> </div><div class="line"><a name="l00490"></a><span class="lineno"> 490</span>  slice_matrix_b.<a class="code" href="classarm__compute_1_1_window.xhtml#acd3d2bba51cb84d34dd7656ad2375a6e">set</a>(<a class="code" href="classarm__compute_1_1_window.xhtml#aa96e81276ee4f87ab386cd05a5539a7d">Window::DimX</a>, <a class="code" href="classarm__compute_1_1_window_1_1_dimension.xhtml">Window::Dimension</a>(0, 1, 1));</div><div class="line"><a name="l00491"></a><span class="lineno"> 491</span>  slice_matrix_b.<a class="code" href="classarm__compute_1_1_window.xhtml#acd3d2bba51cb84d34dd7656ad2375a6e">set</a>(<a class="code" href="classarm__compute_1_1_window.xhtml#ad2d402364fa822b0b7775081291eeca9">Window::DimY</a>, <a class="code" href="classarm__compute_1_1_window_1_1_dimension.xhtml">Window::Dimension</a>(0, 1, 1));</div><div class="line"><a name="l00492"></a><span class="lineno"> 492</span> </div><div class="line"><a name="l00493"></a><span class="lineno"> 493</span>  <span class="keyword">const</span> <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> num_arguments_bias = <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a94e30ed1aed47fae8430cc4d3cd2b6c7">_add_bias</a> ? <a class="code" href="classarm__compute_1_1_i_c_l_kernel.xhtml#a45601e0d46621a5b6f2e417d60e5c800">num_arguments_per_2D_tensor</a>() + 1 : 0;</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>  <span class="keywordflow">if</span>(<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#aa71ec02f998e1dcfd49ef944ec8cf23e">_reinterpret_input_as_3d</a>)</div><div class="line"><a name="l00496"></a><span class="lineno"> 496</span>  {</div><div class="line"><a name="l00497"></a><span class="lineno"> 497</span>  <span class="comment">// Pass bottom paddings to the kernel if the input has to be reinterpreted as 3D tensor</span></div><div class="line"><a name="l00498"></a><span class="lineno"> 498</span>  <span class="keyword">const</span> <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> idx0 = 3 * <a class="code" href="classarm__compute_1_1_i_c_l_kernel.xhtml#a45601e0d46621a5b6f2e417d60e5c800">num_arguments_per_2D_tensor</a>() + 3 + num_arguments_bias;</div><div class="line"><a name="l00499"></a><span class="lineno"> 499</span>  <span class="keyword">const</span> <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> total_cross_plane_pad = <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a1ab65df01f310bf054323607cd09956e">_input0</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#a07b929c34ad1dc823d8315876aa403ce">padding</a>().<a class="code" href="structarm__compute_1_1_border_size.xhtml#a3fdd42ea34070a54e696b3adc28c4be3">top</a> + <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a1ab65df01f310bf054323607cd09956e">_input0</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#a07b929c34ad1dc823d8315876aa403ce">padding</a>().<a class="code" href="structarm__compute_1_1_border_size.xhtml#a802ffcf1b49237efe5be8a314d3f3869">bottom</a>;</div><div class="line"><a name="l00500"></a><span class="lineno"> 500</span>  _kernel.setArg<cl_uint>(idx0, static_cast<unsigned int>(total_cross_plane_pad));</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> </div><div class="line"><a name="l00503"></a><span class="lineno"> 503</span>  <span class="keywordflow">if</span>(<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a7d222bcf0d803c0647a4b93061daa56c">_reinterpret_output_as_3d</a>)</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>  <span class="comment">// Pass bottom paddings to the kernel if the output has to be reinterpreted as 3D tensor</span></div><div class="line"><a name="l00506"></a><span class="lineno"> 506</span>  <span class="keyword">const</span> <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> idx0 = 3 * <a class="code" href="classarm__compute_1_1_i_c_l_kernel.xhtml#a45601e0d46621a5b6f2e417d60e5c800">num_arguments_per_2D_tensor</a>() + 3 + (<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#aa71ec02f998e1dcfd49ef944ec8cf23e">_reinterpret_input_as_3d</a> ? 1 : 0) + num_arguments_bias;</div><div class="line"><a name="l00507"></a><span class="lineno"> 507</span>  <span class="keyword">const</span> <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> total_cross_plane_pad = <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a62d192d931002b4866443cd7fc71419b">_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#a07b929c34ad1dc823d8315876aa403ce">padding</a>().<a class="code" href="structarm__compute_1_1_border_size.xhtml#a3fdd42ea34070a54e696b3adc28c4be3">top</a> + <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a62d192d931002b4866443cd7fc71419b">_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#a07b929c34ad1dc823d8315876aa403ce">padding</a>().<a class="code" href="structarm__compute_1_1_border_size.xhtml#a802ffcf1b49237efe5be8a314d3f3869">bottom</a>;</div><div class="line"><a name="l00508"></a><span class="lineno"> 508</span>  _kernel.setArg<cl_uint>(idx0, static_cast<unsigned int>(total_cross_plane_pad));</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> </div><div class="line"><a name="l00511"></a><span class="lineno"> 511</span>  <span class="keywordflow">do</span></div><div class="line"><a name="l00512"></a><span class="lineno"> 512</span>  {</div><div class="line"><a name="l00513"></a><span class="lineno"> 513</span>  <a class="code" href="classarm__compute_1_1_window.xhtml">Window</a> slice_b = <a class="code" href="namespacearm__compute_1_1test_1_1validation_1_1reference.xhtml#a548131b3d37da47a2e9d32111c88dfe1">slice</a>;</div><div class="line"><a name="l00514"></a><span class="lineno"> 514</span>  <span class="comment">// Don't slice matrix B along the z dimension if matrix B has just 2 dimensions and matrix A more than 2</span></div><div class="line"><a name="l00515"></a><span class="lineno"> 515</span>  <span class="comment">// This scenario can happen when the matrix multiplication is used to perform a convolution operation</span></div><div class="line"><a name="l00516"></a><span class="lineno"> 516</span>  <span class="keywordflow">if</span>(!<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a170f236fd8751c4e1675873b496f7cf8">_slide_matrix_b</a>)</div><div class="line"><a name="l00517"></a><span class="lineno"> 517</span>  {</div><div class="line"><a name="l00518"></a><span class="lineno"> 518</span>  slice_b = slice_matrix_b;</div><div class="line"><a name="l00519"></a><span class="lineno"> 519</span>  }</div><div class="line"><a name="l00520"></a><span class="lineno"> 520</span> </div><div class="line"><a name="l00521"></a><span class="lineno"> 521</span>  <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> idx = 0;</div><div class="line"><a name="l00522"></a><span class="lineno"> 522</span>  <a class="code" href="classarm__compute_1_1_i_c_l_kernel.xhtml#ac74dad3e61f79334f5e73f3c3ac603cb">add_2D_tensor_argument</a>(idx, <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a1ab65df01f310bf054323607cd09956e">_input0</a>, <a class="code" href="namespacearm__compute_1_1test_1_1validation_1_1reference.xhtml#a548131b3d37da47a2e9d32111c88dfe1">slice</a>);</div><div class="line"><a name="l00523"></a><span class="lineno"> 523</span>  <a class="code" href="classarm__compute_1_1_i_c_l_kernel.xhtml#ac74dad3e61f79334f5e73f3c3ac603cb">add_2D_tensor_argument</a>(idx, <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a6590f81ae0c9f3e01546c73eb31a43c8">_input1</a>, slice_b);</div><div class="line"><a name="l00524"></a><span class="lineno"> 524</span>  <span class="keywordflow">if</span>(<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a94e30ed1aed47fae8430cc4d3cd2b6c7">_add_bias</a>)</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>  <a class="code" href="classarm__compute_1_1_i_c_l_kernel.xhtml#ac74dad3e61f79334f5e73f3c3ac603cb">add_2D_tensor_argument</a>(idx, <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a23faf35900f50c084fa1282511b7bd17">_input2</a>, <a class="code" href="namespacearm__compute_1_1test_1_1validation_1_1reference.xhtml#a548131b3d37da47a2e9d32111c88dfe1">slice</a>);</div><div class="line"><a name="l00527"></a><span class="lineno"> 527</span>  }</div><div class="line"><a name="l00528"></a><span class="lineno"> 528</span>  <a class="code" href="classarm__compute_1_1_i_c_l_kernel.xhtml#ac74dad3e61f79334f5e73f3c3ac603cb">add_2D_tensor_argument</a>(idx, <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a62d192d931002b4866443cd7fc71419b">_output</a>, <a class="code" href="namespacearm__compute_1_1test_1_1validation_1_1reference.xhtml#a548131b3d37da47a2e9d32111c88dfe1">slice</a>);</div><div class="line"><a name="l00529"></a><span class="lineno"> 529</span>  _kernel.setArg<cl_uint>(idx++, static_cast<unsigned int>(<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a1ab65df01f310bf054323607cd09956e">_input0</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#a6b14f175bf5281f57b561e2d4e4b1f1f">strides_in_bytes</a>()[2]));</div><div class="line"><a name="l00530"></a><span class="lineno"> 530</span>  _kernel.setArg<cl_uint>(idx++, static_cast<unsigned int>(<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a6590f81ae0c9f3e01546c73eb31a43c8">_input1</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#a6b14f175bf5281f57b561e2d4e4b1f1f">strides_in_bytes</a>()[2]));</div><div class="line"><a name="l00531"></a><span class="lineno"> 531</span>  <span class="keywordflow">if</span>(<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a94e30ed1aed47fae8430cc4d3cd2b6c7">_add_bias</a>)</div><div class="line"><a name="l00532"></a><span class="lineno"> 532</span>  {</div><div class="line"><a name="l00533"></a><span class="lineno"> 533</span>  _kernel.setArg<cl_uint>(idx++, static_cast<unsigned int>(<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a23faf35900f50c084fa1282511b7bd17">_input2</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#a6b14f175bf5281f57b561e2d4e4b1f1f">strides_in_bytes</a>()[2]));</div><div class="line"><a name="l00534"></a><span class="lineno"> 534</span>  }</div><div class="line"><a name="l00535"></a><span class="lineno"> 535</span>  _kernel.setArg<cl_uint>(idx++, static_cast<unsigned int>(<a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a62d192d931002b4866443cd7fc71419b">_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#a6b14f175bf5281f57b561e2d4e4b1f1f">strides_in_bytes</a>()[2]));</div><div class="line"><a name="l00536"></a><span class="lineno"> 536</span>  <a class="code" href="namespacearm__compute.xhtml#a6e51ab3789678d3e0b0b72178dd6c4c6">enqueue</a>(queue, *<span class="keyword">this</span>, <a class="code" href="namespacearm__compute_1_1test_1_1validation_1_1reference.xhtml#a548131b3d37da47a2e9d32111c88dfe1">slice</a>, <a class="code" href="classarm__compute_1_1_i_c_l_kernel.xhtml#ab9f813c25ed75ea7b7ac2fa3926a8f55">lws_hint</a>());</div><div class="line"><a name="l00537"></a><span class="lineno"> 537</span>  }</div><div class="line"><a name="l00538"></a><span class="lineno"> 538</span>  <span class="keywordflow">while</span>(<a class="code" href="classarm__compute_1_1_i_kernel.xhtml#ad34a46f53686c12a5c5e717cc9617fb6">window</a>.<a class="code" href="classarm__compute_1_1_window.xhtml#aac792e3a11bc73bafafc4f4284c7f215">slide_window_slice_3D</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation_1_1reference.xhtml#a548131b3d37da47a2e9d32111c88dfe1">slice</a>));</div><div class="line"><a name="l00539"></a><span class="lineno"> 539</span> }</div><div class="line"><a name="l00540"></a><span class="lineno"> 540</span> } <span class="comment">// namespace arm_compute</span></div><div class="ttc" id="structarm__compute_1_1_border_size_xhtml_a3fdd42ea34070a54e696b3adc28c4be3"><div class="ttname"><a href="structarm__compute_1_1_border_size.xhtml#a3fdd42ea34070a54e696b3adc28c4be3">arm_compute::BorderSize::top</a></div><div class="ttdeci">unsigned int top</div><div class="ttdoc">top of the border</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_types_8h_source.xhtml#l00349">Types.h:349</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_tensor_info_xhtml_a1f4e725b8e1ea36b30e09dc08ae6961d"><div class="ttname"><a href="classarm__compute_1_1_i_tensor_info.xhtml#a1f4e725b8e1ea36b30e09dc08ae6961d">arm_compute::ITensorInfo::num_dimensions</a></div><div class="ttdeci">virtual size_t num_dimensions() const =0</div><div class="ttdoc">The number of dimensions of the tensor (rank)</div></div> |
| <div class="ttc" id="namespacearm__compute_1_1helpers_1_1float__ops_xhtml_ab2dcf325d146568ecc8d4a4bd36c02ac"><div class="ttname"><a href="namespacearm__compute_1_1helpers_1_1float__ops.xhtml#ab2dcf325d146568ecc8d4a4bd36c02ac">arm_compute::helpers::float_ops::is_one</a></div><div class="ttdeci">bool is_one(float a, float epsilon=0.00001f)</div><div class="ttdoc">Checks if the input floating point number is 1.0f checking if the difference is within a range define...</div><div class="ttdef"><b>Definition:</b> <a href="float__ops_8h_source.xhtml#l00097">float_ops.h:97</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel_xhtml_a493987e85723a8000eb26d1f00e2ad0e"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a493987e85723a8000eb26d1f00e2ad0e">arm_compute::CLGEMMMatrixMultiplyKernel::run</a></div><div class="ttdeci">void run(const Window &window, cl::CommandQueue &queue) override</div><div class="ttdoc">Enqueue the OpenCL kernel to process the given window on the passed OpenCL command queue.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_matrix_multiply_kernel_8cpp_source.xhtml#l00476">CLGEMMMatrixMultiplyKernel.cpp:476</a></div></div> |
| <div class="ttc" id="_c_l_validate_8h_xhtml_ab82bd5de18ef067ae5d9ba4af8065dd6"><div class="ttname"><a href="_c_l_validate_8h.xhtml#ab82bd5de18ef067ae5d9ba4af8065dd6">ARM_COMPUTE_RETURN_ERROR_ON_F16_UNSUPPORTED</a></div><div class="ttdeci">#define ARM_COMPUTE_RETURN_ERROR_ON_F16_UNSUPPORTED(tensor)</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_validate_8h_source.xhtml#l00034">CLValidate.h:34</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_kernel_xhtml_ad34a46f53686c12a5c5e717cc9617fb6"><div class="ttname"><a href="classarm__compute_1_1_i_kernel.xhtml#ad34a46f53686c12a5c5e717cc9617fb6">arm_compute::IKernel::window</a></div><div class="ttdeci">const Window & window() const</div><div class="ttdoc">The maximum window the kernel can be executed on.</div><div class="ttdef"><b>Definition:</b> <a href="_i_kernel_8cpp_source.xhtml#l00028">IKernel.cpp:28</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_activation_layer_info_xhtml_af5a8385102f8f8dd6c5957eac08b04c2"><div class="ttname"><a href="classarm__compute_1_1_activation_layer_info.xhtml#af5a8385102f8f8dd6c5957eac08b04c2">arm_compute::ActivationLayerInfo::enabled</a></div><div class="ttdeci">bool enabled() const</div><div class="ttdoc">Check if initialised.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_types_8h_source.xhtml#l01664">Types.h:1664</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="namespacearm__compute_xhtml_a6e51ab3789678d3e0b0b72178dd6c4c6"><div class="ttname"><a href="namespacearm__compute.xhtml#a6e51ab3789678d3e0b0b72178dd6c4c6">arm_compute::enqueue</a></div><div class="ttdeci">void enqueue(cl::CommandQueue &queue, ICLKernel &kernel, const Window &window, const cl::NDRange &lws_hint=CLKernelLibrary::get().default_ndrange(), bool use_dummy_work_items=false)</div><div class="ttdoc">Add the kernel to the command queue with the given window.</div><div class="ttdef"><b>Definition:</b> <a href="_i_c_l_kernel_8cpp_source.xhtml#l00039">ICLKernel.cpp:39</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_build_options_xhtml_ae3b08139a1e57323c5d7dd93f30496c8"><div class="ttname"><a href="classarm__compute_1_1_c_l_build_options.xhtml#ae3b08139a1e57323c5d7dd93f30496c8">arm_compute::CLBuildOptions::options</a></div><div class="ttdeci">const StringSet & options() const</div><div class="ttdoc">Gets the current options list set.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_kernel_library_8cpp_source.xhtml#l00074">CLKernelLibrary.cpp:74</a></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_i_c_l_kernel_xhtml_ab9f813c25ed75ea7b7ac2fa3926a8f55"><div class="ttname"><a href="classarm__compute_1_1_i_c_l_kernel.xhtml#ab9f813c25ed75ea7b7ac2fa3926a8f55">arm_compute::ICLKernel::lws_hint</a></div><div class="ttdeci">cl::NDRange lws_hint() const</div><div class="ttdoc">Return the Local-Workgroup-Size hint.</div><div class="ttdef"><b>Definition:</b> <a href="_i_c_l_kernel_8h_source.xhtml#l00247">ICLKernel.h:247</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_c_l_kernel_xhtml_ad356b88c8f61267d593d9ed99835bde9"><div class="ttname"><a href="classarm__compute_1_1_i_c_l_kernel.xhtml#ad356b88c8f61267d593d9ed99835bde9">arm_compute::ICLKernel::set_lws_hint</a></div><div class="ttdeci">void set_lws_hint(const cl::NDRange &lws_hint)</div><div class="ttdoc">Set the Local-Workgroup-Size hint.</div><div class="ttdef"><b>Definition:</b> <a href="_i_c_l_kernel_8h_source.xhtml#l00237">ICLKernel.h:237</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel_xhtml_a23faf35900f50c084fa1282511b7bd17"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a23faf35900f50c084fa1282511b7bd17">arm_compute::CLGEMMMatrixMultiplyKernel::_input2</a></div><div class="ttdeci">const ICLTensor * _input2</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_matrix_multiply_kernel_8h_source.xhtml#l00096">CLGEMMMatrixMultiplyKernel.h:96</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_activation_layer_info_xhtml_aec65e090c2276e8c8f8dffb6e3edd178"><div class="ttname"><a href="classarm__compute_1_1_activation_layer_info.xhtml#aec65e090c2276e8c8f8dffb6e3edd178">arm_compute::ActivationLayerInfo::a</a></div><div class="ttdeci">float a() const</div><div class="ttdoc">Get the alpha value.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_types_8h_source.xhtml#l01654">Types.h:1654</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="namespacearm__compute_1_1support_1_1cpp11_xhtml_acc5dddee1cbe93a4eaf0a9f74ee96bb7"><div class="ttname"><a href="namespacearm__compute_1_1support_1_1cpp11.xhtml#acc5dddee1cbe93a4eaf0a9f74ee96bb7">arm_compute::support::cpp11::to_string</a></div><div class="ttdeci">std::string to_string(T &&value)</div><div class="ttdoc">Convert integer and float values to string.</div><div class="ttdef"><b>Definition:</b> <a href="_toolchain_support_8h_source.xhtml#l00272">ToolchainSupport.h:272</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="_window_8h_xhtml"><div class="ttname"><a href="_window_8h.xhtml">Window.h</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="namespacearm__compute_xhtml_ab4e88c89b3b7ea1735996cc4def22d58a44ad4ef5a76e6aa6fb3e3fa079a54fda"><div class="ttname"><a href="namespacearm__compute.xhtml#ab4e88c89b3b7ea1735996cc4def22d58a44ad4ef5a76e6aa6fb3e3fa079a54fda">arm_compute::Format::F32</a></div><div class="ttdoc">1 channel, 1 F32 per channel</div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel_xhtml_a94e30ed1aed47fae8430cc4d3cd2b6c7"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a94e30ed1aed47fae8430cc4d3cd2b6c7">arm_compute::CLGEMMMatrixMultiplyKernel::_add_bias</a></div><div class="ttdeci">bool _add_bias</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_matrix_multiply_kernel_8h_source.xhtml#l00101">CLGEMMMatrixMultiplyKernel.h:101</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="namespacearm__compute_xhtml_a635f1895d94050329b7da12850d1a056"><div class="ttname"><a href="namespacearm__compute.xhtml#a635f1895d94050329b7da12850d1a056">arm_compute::string_from_activation_func</a></div><div class="ttdeci">const std::string & string_from_activation_func(ActivationLayerInfo::ActivationFunction act)</div><div class="ttdoc">Translates a given activation function to a string.</div><div class="ttdef"><b>Definition:</b> <a href="src_2core_2_utils_8cpp_source.xhtml#l00172">Utils.cpp:172</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel_xhtml_a42af734585418559f06f6ce9f7375910"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a42af734585418559f06f6ce9f7375910">arm_compute::CLGEMMMatrixMultiplyKernel::_broadcast_bias</a></div><div class="ttdeci">bool _broadcast_bias</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_matrix_multiply_kernel_8h_source.xhtml#l00102">CLGEMMMatrixMultiplyKernel.h:102</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_kernel_library_xhtml_acba005f5ce2c62cbf3f94d074d9007aa"><div class="ttname"><a href="classarm__compute_1_1_c_l_kernel_library.xhtml#acba005f5ce2c62cbf3f94d074d9007aa">arm_compute::CLKernelLibrary::get</a></div><div class="ttdeci">static CLKernelLibrary & get()</div><div class="ttdoc">Access the KernelLibrary singleton.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_kernel_library_8cpp_source.xhtml#l01072">CLKernelLibrary.cpp:1072</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="_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_window_1_1_dimension_xhtml"><div class="ttname"><a href="classarm__compute_1_1_window_1_1_dimension.xhtml">arm_compute::Window::Dimension</a></div><div class="ttdoc">Describe one of the image's dimensions with a start, end and step.</div><div class="ttdef"><b>Definition:</b> <a href="_window_8h_source.xhtml#l00075">Window.h:75</a></div></div> |
| <div class="ttc" id="structarm__compute_1_1_border_size_xhtml_a802ffcf1b49237efe5be8a314d3f3869"><div class="ttname"><a href="structarm__compute_1_1_border_size.xhtml#a802ffcf1b49237efe5be8a314d3f3869">arm_compute::BorderSize::bottom</a></div><div class="ttdeci">unsigned int bottom</div><div class="ttdoc">bottom of the border</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_types_8h_source.xhtml#l00351">Types.h:351</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="core_2_c_l_2_c_l_helpers_8h_xhtml"><div class="ttname"><a href="core_2_c_l_2_c_l_helpers_8h.xhtml">CLHelpers.h</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_a0a7bb17a0a0414a7162f635776a02eb5"><div class="ttname"><a href="namespacearm__compute.xhtml#a0a7bb17a0a0414a7162f635776a02eb5">arm_compute::lower_string</a></div><div class="ttdeci">std::string lower_string(const std::string &val)</div><div class="ttdoc">Lower a given string.</div><div class="ttdef"><b>Definition:</b> <a href="src_2core_2_utils_8cpp_source.xhtml#l00333">Utils.cpp:333</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="_error_8h_xhtml_a206d6e247e0957ac3dee45d27756fc25"><div class="ttname"><a href="_error_8h.xhtml#a206d6e247e0957ac3dee45d27756fc25">ARM_COMPUTE_RETURN_ERROR_ON</a></div><div class="ttdeci">#define ARM_COMPUTE_RETURN_ERROR_ON(cond)</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#l00296">Error.h:296</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_activation_layer_info_xhtml"><div class="ttname"><a href="classarm__compute_1_1_activation_layer_info.xhtml">arm_compute::ActivationLayerInfo</a></div><div class="ttdoc">Activation Layer Information class.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_types_8h_source.xhtml#l01615">Types.h:1615</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel_xhtml_ac46a1c8a20b46838c9e894f703ddd3ee"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#ac46a1c8a20b46838c9e894f703ddd3ee">arm_compute::CLGEMMMatrixMultiplyKernel::CLGEMMMatrixMultiplyKernel</a></div><div class="ttdeci">CLGEMMMatrixMultiplyKernel()</div><div class="ttdoc">Default constructor.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_matrix_multiply_kernel_8cpp_source.xhtml#l00299">CLGEMMMatrixMultiplyKernel.cpp:299</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_ab7980fa5ee693e3282a76da047a1c3b5"><div class="ttname"><a href="namespacearm__compute.xhtml#ab7980fa5ee693e3282a76da047a1c3b5">arm_compute::calculate_max_window</a></div><div class="ttdeci">Window calculate_max_window(const ValidRegion &valid_region, const Steps &steps=Steps(), bool skip_border=false, BorderSize border_size=BorderSize())</div><div class="ttdoc">Calculate the maximum window for a given tensor shape and border setting.</div><div class="ttdef"><b>Definition:</b> <a href="src_2core_2_helpers_8cpp_source.xhtml#l00028">Helpers.cpp:28</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel_xhtml_af45425674a854a3bb158b0b3d0ba9d3e"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#af45425674a854a3bb158b0b3d0ba9d3e">arm_compute::CLGEMMMatrixMultiplyKernel::validate</a></div><div class="ttdeci">static Status validate(const ITensorInfo *input0, const ITensorInfo *input1, const ITensorInfo *input2, const ITensorInfo *output, float alpha, float beta, bool is_interleaved_transposed, const GEMMReshapeInfo &reshape_info, GPUTarget gpu_target, bool fp_mixed_precision=false, const ActivationLayerInfo &activation_info=ActivationLayerInfo())</div><div class="ttdoc">Static function to check if given info will lead to a valid configuration of CLGEMMMatrixMultiplyKern...</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_matrix_multiply_kernel_8cpp_source.xhtml#l00454">CLGEMMMatrixMultiplyKernel.cpp:454</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="namespacearm__compute_xhtml_ab4e88c89b3b7ea1735996cc4def22d58a56d8353718e6fdc78b8d69078a2cdb94"><div class="ttname"><a href="namespacearm__compute.xhtml#ab4e88c89b3b7ea1735996cc4def22d58a56d8353718e6fdc78b8d69078a2cdb94">arm_compute::Format::F16</a></div><div class="ttdoc">1 channel, 1 F16 per channel</div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_build_options_xhtml_a3e2b80ff5463b7d2017de847f5c32a30"><div class="ttname"><a href="classarm__compute_1_1_c_l_build_options.xhtml#a3e2b80ff5463b7d2017de847f5c32a30">arm_compute::CLBuildOptions::add_option</a></div><div class="ttdeci">void add_option(std::string option)</div><div class="ttdoc">Adds option to the existing build option list.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_kernel_library_8cpp_source.xhtml#l00043">CLKernelLibrary.cpp:43</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel_xhtml_aa71ec02f998e1dcfd49ef944ec8cf23e"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#aa71ec02f998e1dcfd49ef944ec8cf23e">arm_compute::CLGEMMMatrixMultiplyKernel::_reinterpret_input_as_3d</a></div><div class="ttdeci">bool _reinterpret_input_as_3d</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_matrix_multiply_kernel_8h_source.xhtml#l00099">CLGEMMMatrixMultiplyKernel.h:99</a></div></div> |
| <div class="ttc" id="arm__compute_2core_2_utils_8h_xhtml"><div class="ttname"><a href="arm__compute_2core_2_utils_8h.xhtml">Utils.h</a></div></div> |
| <div class="ttc" id="_c_l_validate_8h_xhtml"><div class="ttname"><a href="_c_l_validate_8h.xhtml">CLValidate.h</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel_xhtml_a46bee71bbf58053c85de3f5450566584"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a46bee71bbf58053c85de3f5450566584">arm_compute::CLGEMMMatrixMultiplyKernel::configure</a></div><div class="ttdeci">void configure(const ICLTensor *input0, const ICLTensor *input1, const ICLTensor *input2, ICLTensor *output, float alpha, float beta=0.f, bool is_interleaved_transposed=true, const GEMMReshapeInfo &reshape_info=GEMMReshapeInfo(), bool fp_mixed_precision=false, const ActivationLayerInfo &activation_info=ActivationLayerInfo())</div><div class="ttdoc">Initialise the kernel's input, output and alpha.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_matrix_multiply_kernel_8cpp_source.xhtml#l00305">CLGEMMMatrixMultiplyKernel.cpp:305</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_a142b55a483cadf4e1068a1a09a55e8e9"><div class="ttname"><a href="namespacearm__compute.xhtml#a142b55a483cadf4e1068a1a09a55e8e9">arm_compute::string_from_data_type</a></div><div class="ttdeci">const std::string & string_from_data_type(DataType dt)</div><div class="ttdoc">Convert a data type identity into a string.</div><div class="ttdef"><b>Definition:</b> <a href="src_2core_2_utils_8cpp_source.xhtml#l00144">Utils.cpp:144</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_window_xhtml_aa96e81276ee4f87ab386cd05a5539a7d"><div class="ttname"><a href="classarm__compute_1_1_window.xhtml#aa96e81276ee4f87ab386cd05a5539a7d">arm_compute::Window::DimX</a></div><div class="ttdeci">static constexpr size_t DimX</div><div class="ttdoc">Alias for dimension 0 also known as X dimension.</div><div class="ttdef"><b>Definition:</b> <a href="_window_8h_source.xhtml#l00043">Window.h:43</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_afc4bd8e872567d9c4c57d89eb0bb3da1"><div class="ttname"><a href="namespacearm__compute.xhtml#afc4bd8e872567d9c4c57d89eb0bb3da1">arm_compute::update_window_and_padding</a></div><div class="ttdeci">bool update_window_and_padding(Window &win, Ts &&... patterns)</div><div class="ttdoc">Update window and padding size for each of the access patterns.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_helpers_8h_source.xhtml#l00402">Helpers.h:402</a></div></div> |
| <div class="ttc" id="_error_8h_xhtml_a6dc630a6ae9cc063b3924bcea8dee9d6"><div class="ttname"><a href="_error_8h.xhtml#a6dc630a6ae9cc063b3924bcea8dee9d6">ARM_COMPUTE_UNUSED</a></div><div class="ttdeci">#define ARM_COMPUTE_UNUSED(...)</div><div class="ttdoc">To avoid unused variables warnings.</div><div class="ttdef"><b>Definition:</b> <a href="_error_8h_source.xhtml#l00152">Error.h:152</a></div></div> |
| <div class="ttc" id="namespacearm__compute_1_1misc_1_1shape__calculator_xhtml_a389f89ab5121dad0906d0b7324fbf73d"><div class="ttname"><a href="namespacearm__compute_1_1misc_1_1shape__calculator.xhtml#a389f89ab5121dad0906d0b7324fbf73d">arm_compute::misc::shape_calculator::compute_lhs_reshaped_shape</a></div><div class="ttdeci">TensorShape compute_lhs_reshaped_shape(const ITensorInfo &a, const GEMMLHSMatrixInfo &lhs_info, bool reinterpret_input_as_3d=false)</div><div class="ttdoc">Calculate the Left Hand Side matrix reshaped shape.</div><div class="ttdef"><b>Definition:</b> <a href="_shape_calculator_8h_source.xhtml#l00180">ShapeCalculator.h:180</a></div></div> |
| <div class="ttc" id="_validate_8h_xhtml_a27e4638546c88b8916f967e6e54480a9"><div class="ttname"><a href="_validate_8h.xhtml#a27e4638546c88b8916f967e6e54480a9">ARM_COMPUTE_RETURN_ERROR_ON_MISMATCHING_SHAPES</a></div><div class="ttdeci">#define ARM_COMPUTE_RETURN_ERROR_ON_MISMATCHING_SHAPES(...)</div><div class="ttdef"><b>Definition:</b> <a href="_validate_8h_source.xhtml#l00443">Validate.h:443</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel_xhtml_a6590f81ae0c9f3e01546c73eb31a43c8"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a6590f81ae0c9f3e01546c73eb31a43c8">arm_compute::CLGEMMMatrixMultiplyKernel::_input1</a></div><div class="ttdeci">const ICLTensor * _input1</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_matrix_multiply_kernel_8h_source.xhtml#l00095">CLGEMMMatrixMultiplyKernel.h:95</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_a9cd394c15b73f79ca1d98f5328064be2"><div class="ttname"><a href="namespacearm__compute.xhtml#a9cd394c15b73f79ca1d98f5328064be2">arm_compute::float_to_string_with_full_precision</a></div><div class="ttdeci">std::string float_to_string_with_full_precision(float val)</div><div class="ttdoc">Create a string with the float in full precision.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_utils_8h_source.xhtml#l01211">Utils.h:1211</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_ab237a0a375cf382d52b61653248d3d4a"><div class="ttname"><a href="namespacearm__compute.xhtml#ab237a0a375cf382d52b61653248d3d4a">arm_compute::ceil_to_multiple</a></div><div class="ttdeci">auto ceil_to_multiple(S value, T divisor) -> decltype(((value+divisor - 1)/divisor) *divisor)</div><div class="ttdoc">Computes the smallest number larger or equal to value that is a multiple of divisor.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_utils_8h_source.xhtml#l00066">Utils.h:66</a></div></div> |
| <div class="ttc" id="float__ops_8h_xhtml"><div class="ttname"><a href="float__ops_8h.xhtml">float_ops.h</a></div></div> |
| <div class="ttc" id="_c_l_im2_col_kernel_8cpp_xhtml_a624a24e6d361fe7b8b8b2f6b375683a4"><div class="ttname"><a href="_c_l_im2_col_kernel_8cpp.xhtml#a624a24e6d361fe7b8b8b2f6b375683a4">kernel_name</a></div><div class="ttdeci">std::string kernel_name</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_im2_col_kernel_8cpp_source.xhtml#l00052">CLIm2ColKernel.cpp:52</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_c_l_kernel_xhtml_aa550ff0352ff2388e02f7b0a41bf5fe7"><div class="ttname"><a href="classarm__compute_1_1_i_c_l_kernel.xhtml#aa550ff0352ff2388e02f7b0a41bf5fe7">arm_compute::ICLKernel::get_target</a></div><div class="ttdeci">GPUTarget get_target() const</div><div class="ttdoc">Get the targeted GPU architecture.</div><div class="ttdef"><b>Definition:</b> <a href="_i_c_l_kernel_8h_source.xhtml#l00286">ICLKernel.h:286</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_a545eeda2eaa3f5a54345ce8169e21184"><div class="ttname"><a href="namespacearm__compute.xhtml#a545eeda2eaa3f5a54345ce8169e21184">arm_compute::get_cl_type_from_data_type</a></div><div class="ttdeci">std::string get_cl_type_from_data_type(const DataType &dt)</div><div class="ttdoc">Translates a tensor data type to the appropriate OpenCL type.</div><div class="ttdef"><b>Definition:</b> <a href="core_2_c_l_2_c_l_helpers_8cpp_source.xhtml#l00037">CLHelpers.cpp:37</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="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="namespacearm__compute_xhtml_abb7e0f23a4f2e63f39433f158dad47ab"><div class="ttname"><a href="namespacearm__compute.xhtml#abb7e0f23a4f2e63f39433f158dad47ab">arm_compute::data_size_from_type</a></div><div class="ttdeci">size_t data_size_from_type(DataType data_type)</div><div class="ttdoc">The size in bytes of the data type.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_utils_8h_source.xhtml#l00109">Utils.h:109</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_build_options_xhtml_a95b46e69297fad10b27a1baa000f92cc"><div class="ttname"><a href="classarm__compute_1_1_c_l_build_options.xhtml#a95b46e69297fad10b27a1baa000f92cc">arm_compute::CLBuildOptions::add_option_if</a></div><div class="ttdeci">void add_option_if(bool cond, std::string option)</div><div class="ttdoc">Adds option if a given condition is true;.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_kernel_library_8cpp_source.xhtml#l00048">CLKernelLibrary.cpp:48</a></div></div> |
| <div class="ttc" id="namespacearm__compute_1_1test_1_1validation_xhtml_a3161c2c93c655dd30953372064ec627b"><div class="ttname"><a href="namespacearm__compute_1_1test_1_1validation.xhtml#a3161c2c93c655dd30953372064ec627b">arm_compute::test::validation::alpha</a></div><div class="ttdeci">const float alpha</div><div class="ttdef"><b>Definition:</b> <a href="_n_e_o_n_2_accumulate_8cpp_source.xhtml#l00103">Accumulate.cpp:103</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_window_xhtml_acd3d2bba51cb84d34dd7656ad2375a6e"><div class="ttname"><a href="classarm__compute_1_1_window.xhtml#acd3d2bba51cb84d34dd7656ad2375a6e">arm_compute::Window::set</a></div><div class="ttdeci">void set(size_t dimension, const Dimension &dim)</div><div class="ttdoc">Set the values of a given dimension.</div><div class="ttdef"><b>Definition:</b> <a href="_window_8inl_source.xhtml#l00049">Window.inl:49</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_tensor_info_xhtml_a07b929c34ad1dc823d8315876aa403ce"><div class="ttname"><a href="classarm__compute_1_1_i_tensor_info.xhtml#a07b929c34ad1dc823d8315876aa403ce">arm_compute::ITensorInfo::padding</a></div><div class="ttdeci">virtual PaddingSize padding() const =0</div><div class="ttdoc">Padding of tensor.</div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_c_l_kernel_xhtml_a45601e0d46621a5b6f2e417d60e5c800"><div class="ttname"><a href="classarm__compute_1_1_i_c_l_kernel.xhtml#a45601e0d46621a5b6f2e417d60e5c800">arm_compute::ICLKernel::num_arguments_per_2D_tensor</a></div><div class="ttdeci">static constexpr unsigned int num_arguments_per_2D_tensor()</div><div class="ttdoc">Returns the number of arguments enqueued per 2D tensor object.</div><div class="ttdef"><b>Definition:</b> <a href="_i_c_l_kernel_8h_source.xhtml#l00192">ICLKernel.h:192</a></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_build_options_xhtml"><div class="ttname"><a href="classarm__compute_1_1_c_l_build_options.xhtml">arm_compute::CLBuildOptions</a></div><div class="ttdoc">Build options.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_kernel_library_8h_source.xhtml#l00037">CLKernelLibrary.h:37</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_abc72c95941485d8a068fa38372308574"><div class="ttname"><a href="namespacearm__compute.xhtml#abc72c95941485d8a068fa38372308574">arm_compute::create_kernel</a></div><div class="ttdeci">std::unique_ptr< Kernel > create_kernel()</div><div class="ttdoc">Helper function to create and return a unique_ptr pointed to a CL/GLES kernel object.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_helpers_8h_source.xhtml#l00086">Helpers.h:86</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_window_xhtml_aac792e3a11bc73bafafc4f4284c7f215"><div class="ttname"><a href="classarm__compute_1_1_window.xhtml#aac792e3a11bc73bafafc4f4284c7f215">arm_compute::Window::slide_window_slice_3D</a></div><div class="ttdeci">bool slide_window_slice_3D(Window &slice) const</div><div class="ttdoc">Slide the passed 3D window slice.</div><div class="ttdef"><b>Definition:</b> <a href="_window_8h_source.xhtml#l00333">Window.h:333</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_a59e56af19e754a6aa26a612ebf91d05fa62be47fdd89da032cf78dfce82239579"><div class="ttname"><a href="namespacearm__compute.xhtml#a59e56af19e754a6aa26a612ebf91d05fa62be47fdd89da032cf78dfce82239579">arm_compute::ErrorCode::RUNTIME_ERROR</a></div><div class="ttdoc">Generic runtime error.</div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel_xhtml_a1ab65df01f310bf054323607cd09956e"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a1ab65df01f310bf054323607cd09956e">arm_compute::CLGEMMMatrixMultiplyKernel::_input0</a></div><div class="ttdeci">const ICLTensor * _input0</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_matrix_multiply_kernel_8h_source.xhtml#l00094">CLGEMMMatrixMultiplyKernel.h:94</a></div></div> |
| <div class="ttc" id="_validate_8h_xhtml_aff911654521523937ff24372a870b89f"><div class="ttname"><a href="_validate_8h.xhtml#aff911654521523937ff24372a870b89f">ARM_COMPUTE_RETURN_ERROR_ON_NULLPTR</a></div><div class="ttdeci">#define ARM_COMPUTE_RETURN_ERROR_ON_NULLPTR(...)</div><div class="ttdef"><b>Definition:</b> <a href="_validate_8h_source.xhtml#l00163">Validate.h:163</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel_xhtml_a170f236fd8751c4e1675873b496f7cf8"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a170f236fd8751c4e1675873b496f7cf8">arm_compute::CLGEMMMatrixMultiplyKernel::_slide_matrix_b</a></div><div class="ttdeci">bool _slide_matrix_b</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_matrix_multiply_kernel_8h_source.xhtml#l00098">CLGEMMMatrixMultiplyKernel.h:98</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_window_xhtml_ad2d402364fa822b0b7775081291eeca9"><div class="ttname"><a href="classarm__compute_1_1_window.xhtml#ad2d402364fa822b0b7775081291eeca9">arm_compute::Window::DimY</a></div><div class="ttdeci">static constexpr size_t DimY</div><div class="ttdoc">Alias for dimension 1 also known as Y dimension.</div><div class="ttdef"><b>Definition:</b> <a href="_window_8h_source.xhtml#l00045">Window.h:45</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_c_l_kernel_xhtml_ac74dad3e61f79334f5e73f3c3ac603cb"><div class="ttname"><a href="classarm__compute_1_1_i_c_l_kernel.xhtml#ac74dad3e61f79334f5e73f3c3ac603cb">arm_compute::ICLKernel::add_2D_tensor_argument</a></div><div class="ttdeci">void add_2D_tensor_argument(unsigned int &idx, const ICLTensor *tensor, const Window &window)</div><div class="ttdoc">Add the passed 2D tensor's parameters to the object's kernel's arguments starting from the index idx.</div><div class="ttdef"><b>Definition:</b> <a href="_i_c_l_kernel_8h_source.xhtml#l00134">ICLKernel.h:134</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_c_l_g_e_m_m_matrix_multiply_kernel_xhtml_a62d192d931002b4866443cd7fc71419b"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a62d192d931002b4866443cd7fc71419b">arm_compute::CLGEMMMatrixMultiplyKernel::_output</a></div><div class="ttdeci">ICLTensor * _output</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_matrix_multiply_kernel_8h_source.xhtml#l00097">CLGEMMMatrixMultiplyKernel.h:97</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="_c_l_g_e_m_m_matrix_multiply_kernel_8h_xhtml"><div class="ttname"><a href="_c_l_g_e_m_m_matrix_multiply_kernel_8h.xhtml">CLGEMMMatrixMultiplyKernel.h</a></div></div> |
| <div class="ttc" id="_error_8h_xhtml_af1b8ff8eb557a2ad11272f1505f45d34"><div class="ttname"><a href="_error_8h.xhtml#af1b8ff8eb557a2ad11272f1505f45d34">ARM_COMPUTE_CREATE_ERROR</a></div><div class="ttdeci">#define ARM_COMPUTE_CREATE_ERROR(error_code, msg)</div><div class="ttdoc">Creates an error with a given message.</div><div class="ttdef"><b>Definition:</b> <a href="_error_8h_source.xhtml#l00159">Error.h:159</a></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="_c_l_kernel_library_8h_xhtml"><div class="ttname"><a href="_c_l_kernel_library_8h.xhtml">CLKernelLibrary.h</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel_xhtml_a7d222bcf0d803c0647a4b93061daa56c"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_matrix_multiply_kernel.xhtml#a7d222bcf0d803c0647a4b93061daa56c">arm_compute::CLGEMMMatrixMultiplyKernel::_reinterpret_output_as_3d</a></div><div class="ttdeci">bool _reinterpret_output_as_3d</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_matrix_multiply_kernel_8h_source.xhtml#l00100">CLGEMMMatrixMultiplyKernel.h:100</a></div></div> |
| <div class="ttc" id="_open_c_l_8h_xhtml"><div class="ttname"><a href="_open_c_l_8h.xhtml">OpenCL.h</a></div></div> |
| <div class="ttc" id="namespacearm__compute_1_1helpers_1_1float__ops_xhtml_a3bd19352aed7410633d1f9b95d74a809"><div class="ttname"><a href="namespacearm__compute_1_1helpers_1_1float__ops.xhtml#a3bd19352aed7410633d1f9b95d74a809">arm_compute::helpers::float_ops::is_zero</a></div><div class="ttdeci">bool is_zero(float a, float epsilon=0.00001f)</div><div class="ttdoc">Checks if the input floating point number is 0.0f checking if the difference is within a range define...</div><div class="ttdef"><b>Definition:</b> <a href="float__ops_8h_source.xhtml#l00109">float_ops.h:109</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="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="_access_window_static_8h_xhtml"><div class="ttname"><a href="_access_window_static_8h.xhtml">AccessWindowStatic.h</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_activation_layer_info_xhtml_a9e0fb1d1462557f28966ae19988532c2"><div class="ttname"><a href="classarm__compute_1_1_activation_layer_info.xhtml#a9e0fb1d1462557f28966ae19988532c2">arm_compute::ActivationLayerInfo::activation</a></div><div class="ttdeci">ActivationFunction activation() const</div><div class="ttdoc">Get the type of activation function.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_types_8h_source.xhtml#l01649">Types.h:1649</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_activation_layer_info_xhtml_a02a00a5d20986f3a7ab72b9c86be3a54"><div class="ttname"><a href="classarm__compute_1_1_activation_layer_info.xhtml#a02a00a5d20986f3a7ab72b9c86be3a54">arm_compute::ActivationLayerInfo::b</a></div><div class="ttdeci">float b() const</div><div class="ttdoc">Get the beta value.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_types_8h_source.xhtml#l01659">Types.h:1659</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_a735ac6c2a02e320969625308810444f3aa78cc0fd1cab24af0fad71dc4c256f8e"><div class="ttname"><a href="namespacearm__compute.xhtml#a735ac6c2a02e320969625308810444f3aa78cc0fd1cab24af0fad71dc4c256f8e">arm_compute::GPUTarget::BIFROST</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_tensor_info_xhtml_a6b14f175bf5281f57b561e2d4e4b1f1f"><div class="ttname"><a href="classarm__compute_1_1_i_tensor_info.xhtml#a6b14f175bf5281f57b561e2d4e4b1f1f">arm_compute::ITensorInfo::strides_in_bytes</a></div><div class="ttdeci">virtual const Strides & strides_in_bytes() const =0</div><div class="ttdoc">The strides in bytes for accessing each dimension of the tensor.</div></div> |
| <div class="ttc" id="namespacearm__compute_1_1test_1_1validation_xhtml_a5f5b6c4337eac9e2e0046ca2304d80dc"><div class="ttname"><a href="namespacearm__compute_1_1test_1_1validation.xhtml#a5f5b6c4337eac9e2e0046ca2304d80dc">arm_compute::test::validation::data_type</a></div><div class="ttdeci">data_type</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_2_arithmetic_addition_8cpp_source.xhtml#l00138">ArithmeticAddition.cpp:138</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_window_xhtml_a30ca5bdbb60ee281d7f1ab34f7a4ee40"><div class="ttname"><a href="classarm__compute_1_1_window.xhtml#a30ca5bdbb60ee281d7f1ab34f7a4ee40">arm_compute::Window::first_slice_window_3D</a></div><div class="ttdeci">Window first_slice_window_3D() const</div><div class="ttdoc">First 3D slice of the window.</div><div class="ttdef"><b>Definition:</b> <a href="_window_8h_source.xhtml#l00289">Window.h:289</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="namespacearm__compute_xhtml_ad8ed01ff3ff33333d8e19db4d2818bb6"><div class="ttname"><a href="namespacearm__compute.xhtml#ad8ed01ff3ff33333d8e19db4d2818bb6">arm_compute::DataType</a></div><div class="ttdeci">DataType</div><div class="ttdoc">Available data types.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_types_8h_source.xhtml#l00075">Types.h:75</a></div></div> |
| <div class="ttc" id="_validate_8h_xhtml_a6eb9ce82815fe429250189da7592ba75"><div class="ttname"><a href="_validate_8h.xhtml#a6eb9ce82815fe429250189da7592ba75">ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW</a></div><div class="ttdeci">#define ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW(f, s)</div><div class="ttdef"><b>Definition:</b> <a href="_validate_8h_source.xhtml#l00205">Validate.h:205</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_window_xhtml"><div class="ttname"><a href="classarm__compute_1_1_window.xhtml">arm_compute::Window</a></div><div class="ttdoc">Describe a multidimensional execution window.</div><div class="ttdef"><b>Definition:</b> <a href="_window_8h_source.xhtml#l00039">Window.h:39</a></div></div> |
| <div class="ttc" id="_validate_8h_xhtml_a1b35b0d258183cf9ef36adf684d0b88c"><div class="ttname"><a href="_validate_8h.xhtml#a1b35b0d258183cf9ef36adf684d0b88c">ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL</a></div><div class="ttdeci">#define ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL(k)</div><div class="ttdef"><b>Definition:</b> <a href="_validate_8h_source.xhtml#l00941">Validate.h:941</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_af5982a092e9eb743fce2d6392bdd8897"><div class="ttname"><a href="namespacearm__compute.xhtml#af5982a092e9eb743fce2d6392bdd8897">arm_compute::is_data_type_float</a></div><div class="ttdeci">bool is_data_type_float(DataType dt)</div><div class="ttdoc">Check if a given data type is of floating point type.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_utils_8h_source.xhtml#l01097">Utils.h:1097</a></div></div> |
| <div class="ttc" id="namespacearm__compute_1_1test_1_1validation_1_1reference_xhtml_a548131b3d37da47a2e9d32111c88dfe1"><div class="ttname"><a href="namespacearm__compute_1_1test_1_1validation_1_1reference.xhtml#a548131b3d37da47a2e9d32111c88dfe1">arm_compute::test::validation::reference::slice</a></div><div class="ttdeci">SimpleTensor< T > slice(const SimpleTensor< T > &src, Coordinates starts, Coordinates ends)</div><div class="ttdef"><b>Definition:</b> <a href="_slice_operations_8cpp_source.xhtml#l00038">SliceOperations.cpp:38</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_aebb8dcc11953d78e620bbef0b9e2183.xhtml">core</a></li><li class="navelem"><a class="el" href="dir_8c278f79c760e5c5fbd911f9870614c1.xhtml">CL</a></li><li class="navelem"><a class="el" href="dir_8a6a287e420ec4de95804825506d2fde.xhtml">kernels</a></li><li class="navelem"><a class="el" href="_c_l_g_e_m_m_matrix_multiply_kernel_8cpp.xhtml">CLGEMMMatrixMultiplyKernel.cpp</a></li> |
| <li class="footer">Generated on Thu Mar 5 2020 16:06:59 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> |
