arm_compute v19.11
diff --git a/documentation/channel__shuffle_8cl_source.xhtml b/documentation/channel__shuffle_8cl_source.xhtml
index 449a01f..b2b008d 100644
--- a/documentation/channel__shuffle_8cl_source.xhtml
+++ b/documentation/channel__shuffle_8cl_source.xhtml
@@ -40,7 +40,7 @@
<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">19.08</span>
+  <span id="projectnumber">19.11</span>
</div>
</td>
</tr>
@@ -100,13 +100,14 @@
<div class="title">channel_shuffle.cl</div> </div>
</div><!--header-->
<div class="contents">
-<a href="channel__shuffle_8cl.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) 2018 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="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml">helpers.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">#if defined(DATA_TYPE) && defined(VEC_SIZE) && defined(NUM_GROUPS) && defined(K) && defined(SRC_DIM_Z)</span></div><div class="line"><a name="l00027"></a><span class="lineno"> 27</span> </div><div class="line"><a name="l00028"></a><span class="lineno"> 28</span> <span class="comment">// Check valid VEC_SIZES</span></div><div class="line"><a name="l00029"></a><span class="lineno"> 29</span> <span class="preprocessor">#if VEC_SIZE != 4 && VEC_SIZE != 8 && VEC_SIZE != 16</span></div><div class="line"><a name="l00030"></a><span class="lineno"> 30</span> <span class="preprocessor">#error "Only vector sizes 4, 8 and 16 are supported"</span></div><div class="line"><a name="l00031"></a><span class="lineno"> 31</span> <span class="preprocessor">#endif // VEC_SIZE != 4 && VEC_SIZE != 8 && VEC_SIZE != 16</span></div><div class="line"><a name="l00032"></a><span class="lineno"> 32</span> </div><div class="line"><a name="l00033"></a><span class="lineno"> 33</span> <span class="preprocessor">#define TYPE VEC_DATA_TYPE(DATA_TYPE, VEC_SIZE)</span></div><div class="line"><a name="l00034"></a><span class="lineno"> 34</span> </div><div class="line"><a name="l00035"></a><span class="lineno"> 35</span> <span class="preprocessor">#define DIV_MOD_UINT(x, y, div_res, mod_res) \</span></div><div class="line"><a name="l00036"></a><span class="lineno"> 36</span> <span class="preprocessor"> ({ \</span></div><div class="line"><a name="l00037"></a><span class="lineno"> 37</span> <span class="preprocessor"> div_res = (uint)((x) * (float)(1.0f / (float)(y))); \</span></div><div class="line"><a name="l00038"></a><span class="lineno"> 38</span> <span class="preprocessor"> uint r = div_res * (y); \</span></div><div class="line"><a name="l00039"></a><span class="lineno"> 39</span> <span class="preprocessor"> mod_res = (x)-r; \</span></div><div class="line"><a name="l00040"></a><span class="lineno"> 40</span> <span class="preprocessor"> })</span></div><div class="line"><a name="l00041"></a><span class="lineno"> 41</span> </div><div class="line"><a name="l00071"></a><span class="lineno"> 71</span> __kernel <span class="keywordtype">void</span> channel_shuffle_nchw(<a class="code" href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml#a481bdc6d61b3df9dcdbdb244f0f97790">TENSOR4D_DECLARATION</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a989ab3e96426615bb98e04e0235088ca">src</a>),</div><div class="line"><a name="l00072"></a><span class="lineno"> 72</span>  <a class="code" href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml#a481bdc6d61b3df9dcdbdb244f0f97790">TENSOR4D_DECLARATION</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#adbf67dcee294e673cf796f1ed8aeb6a4">dst</a>))</div><div class="line"><a name="l00073"></a><span class="lineno"> 73</span> {</div><div class="line"><a name="l00074"></a><span class="lineno"> 74</span>  uint curr_channel = 0; <span class="comment">// channel id of input</span></div><div class="line"><a name="l00075"></a><span class="lineno"> 75</span>  uint batch_id = 0; <span class="comment">// batch id</span></div><div class="line"><a name="l00076"></a><span class="lineno"> 76</span>  uint group_id = 0; <span class="comment">// group id</span></div><div class="line"><a name="l00077"></a><span class="lineno"> 77</span>  uint channel_id = 0; <span class="comment">// channel id within the group</span></div><div class="line"><a name="l00078"></a><span class="lineno"> 78</span> </div><div class="line"><a name="l00079"></a><span class="lineno"> 79</span>  <span class="comment">// Compute curr_channel and batch_id</span></div><div class="line"><a name="l00080"></a><span class="lineno"> 80</span>  DIV_MOD_UINT(get_global_id(2), SRC_DIM_Z, batch_id, curr_channel);</div><div class="line"><a name="l00081"></a><span class="lineno"> 81</span> </div><div class="line"><a name="l00082"></a><span class="lineno"> 82</span>  <span class="comment">// Compute group_id and channel_id</span></div><div class="line"><a name="l00083"></a><span class="lineno"> 83</span>  DIV_MOD_UINT(curr_channel, K, group_id, channel_id);</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>  <span class="keyword">const</span> uint x = get_global_id(0) * VEC_SIZE;</div><div class="line"><a name="l00086"></a><span class="lineno"> 86</span>  <span class="keyword">const</span> uint y = get_global_id(1) * 2;</div><div class="line"><a name="l00087"></a><span class="lineno"> 87</span>  <span class="keyword">const</span> uint z = channel_id * NUM_GROUPS + group_id;</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="comment">// Load the Nx2 block</span></div><div class="line"><a name="l00090"></a><span class="lineno"> 90</span>  <span class="keyword">const</span> __global uchar *input_ptr = src_ptr + src_offset_first_element_in_bytes + x * <span class="keyword">sizeof</span>(<a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a>) + y * src_stride_y + curr_channel * src_stride_z + batch_id * src_stride_w;</div><div class="line"><a name="l00091"></a><span class="lineno"> 91</span>  <a class="code" href="activation__quant__helpers_8h.xhtml#a5a392548f2df67370cb15d2a5d75cd7b">TYPE</a> u0 = <a class="code" href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml#a287e2fc366c312b468382c95bb90f91f">VLOAD</a>(VEC_SIZE)(0, (__global <a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a> *)(input_ptr + 0 * src_stride_y));</div><div class="line"><a name="l00092"></a><span class="lineno"> 92</span>  <a class="code" href="activation__quant__helpers_8h.xhtml#a5a392548f2df67370cb15d2a5d75cd7b">TYPE</a> u1 = <a class="code" href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml#a287e2fc366c312b468382c95bb90f91f">VLOAD</a>(VEC_SIZE)(0, (__global <a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a> *)(input_ptr + 1 * src_stride_y));</div><div class="line"><a name="l00093"></a><span class="lineno"> 93</span> </div><div class="line"><a name="l00094"></a><span class="lineno"> 94</span>  <span class="comment">// Store blocks</span></div><div class="line"><a name="l00095"></a><span class="lineno"> 95</span>  __global uchar *output_ptr = dst_ptr + dst_offset_first_element_in_bytes + x * <span class="keyword">sizeof</span>(<a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a>) + y * dst_stride_y + z * dst_stride_z + batch_id * dst_stride_w;</div><div class="line"><a name="l00096"></a><span class="lineno"> 96</span>  <a class="code" href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml#acb282042d1edeeaa3cc979a206f78b54">VSTORE</a>(VEC_SIZE)</div><div class="line"><a name="l00097"></a><span class="lineno"> 97</span>  (u0, 0, (__global <a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a> *)(output_ptr + 0 * dst_stride_y));</div><div class="line"><a name="l00098"></a><span class="lineno"> 98</span>  <a class="code" href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml#acb282042d1edeeaa3cc979a206f78b54">VSTORE</a>(VEC_SIZE)</div><div class="line"><a name="l00099"></a><span class="lineno"> 99</span>  (u1, 0, (__global <a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a> *)(output_ptr + 1 * dst_stride_y));</div><div class="line"><a name="l00100"></a><span class="lineno"> 100</span> }</div><div class="line"><a name="l00101"></a><span class="lineno"> 101</span> </div><div class="line"><a name="l00102"></a><span class="lineno"> 102</span> <span class="preprocessor">#if VEC_SIZE == 4 && defined(LAST_ACCESSED)</span></div><div class="line"><a name="l00103"></a><span class="lineno"> 103</span> </div><div class="line"><a name="l00134"></a><span class="lineno"> 134</span> __kernel <span class="keywordtype">void</span> channel_shuffle_nhwc(<a class="code" href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml#a481bdc6d61b3df9dcdbdb244f0f97790">TENSOR4D_DECLARATION</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a989ab3e96426615bb98e04e0235088ca">src</a>),</div><div class="line"><a name="l00135"></a><span class="lineno"> 135</span>  <a class="code" href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml#a481bdc6d61b3df9dcdbdb244f0f97790">TENSOR4D_DECLARATION</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#adbf67dcee294e673cf796f1ed8aeb6a4">dst</a>))</div><div class="line"><a name="l00136"></a><span class="lineno"> 136</span> {</div><div class="line"><a name="l00137"></a><span class="lineno"> 137</span>  <span class="keyword">const</span> uint curr_channel = min((uint)(get_global_id(0) * VEC_SIZE), (uint)LAST_ACCESSED); <span class="comment">// input feature map</span></div><div class="line"><a name="l00138"></a><span class="lineno"> 138</span>  uint channel_id0 = 0;</div><div class="line"><a name="l00139"></a><span class="lineno"> 139</span>  uint channel_id1 = 0;</div><div class="line"><a name="l00140"></a><span class="lineno"> 140</span>  uint channel_id2 = 0;</div><div class="line"><a name="l00141"></a><span class="lineno"> 141</span>  uint channel_id3 = 0;</div><div class="line"><a name="l00142"></a><span class="lineno"> 142</span>  uint group_id0 = 0;</div><div class="line"><a name="l00143"></a><span class="lineno"> 143</span>  uint group_id1 = 0;</div><div class="line"><a name="l00144"></a><span class="lineno"> 144</span>  uint group_id2 = 0;</div><div class="line"><a name="l00145"></a><span class="lineno"> 145</span>  uint group_id3 = 0;</div><div class="line"><a name="l00146"></a><span class="lineno"> 146</span>  uint y = 0;</div><div class="line"><a name="l00147"></a><span class="lineno"> 147</span>  uint batch_id = 0;</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>  <span class="comment">// Compute curr_channel and batch_id</span></div><div class="line"><a name="l00150"></a><span class="lineno"> 150</span>  DIV_MOD_UINT(get_global_id(2), (uint)SRC_DIM_Z, batch_id, y);</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>  <span class="comment">// Compute group_id and channel_id</span></div><div class="line"><a name="l00153"></a><span class="lineno"> 153</span>  DIV_MOD_UINT(curr_channel + (uint)0, K, group_id0, channel_id0);</div><div class="line"><a name="l00154"></a><span class="lineno"> 154</span>  DIV_MOD_UINT(curr_channel + (uint)1, K, group_id1, channel_id1);</div><div class="line"><a name="l00155"></a><span class="lineno"> 155</span>  DIV_MOD_UINT(curr_channel + (uint)2, K, group_id2, channel_id2);</div><div class="line"><a name="l00156"></a><span class="lineno"> 156</span>  DIV_MOD_UINT(curr_channel + (uint)3, K, group_id3, channel_id3);</div><div class="line"><a name="l00157"></a><span class="lineno"> 157</span> </div><div class="line"><a name="l00158"></a><span class="lineno"> 158</span>  <span class="keyword">const</span> uint x = get_global_id(1) * 2;</div><div class="line"><a name="l00159"></a><span class="lineno"> 159</span>  <span class="keyword">const</span> uint z0 = channel_id0 * (uint)NUM_GROUPS + group_id0;</div><div class="line"><a name="l00160"></a><span class="lineno"> 160</span>  <span class="keyword">const</span> uint z1 = channel_id1 * (uint)NUM_GROUPS + group_id1;</div><div class="line"><a name="l00161"></a><span class="lineno"> 161</span>  <span class="keyword">const</span> uint z2 = channel_id2 * (uint)NUM_GROUPS + group_id2;</div><div class="line"><a name="l00162"></a><span class="lineno"> 162</span>  <span class="keyword">const</span> uint z3 = channel_id3 * (uint)NUM_GROUPS + group_id3;</div><div class="line"><a name="l00163"></a><span class="lineno"> 163</span> </div><div class="line"><a name="l00164"></a><span class="lineno"> 164</span>  <span class="comment">// Load the Nx2 block</span></div><div class="line"><a name="l00165"></a><span class="lineno"> 165</span>  <span class="keyword">const</span> __global uchar *input_ptr = src_ptr + src_offset_first_element_in_bytes + curr_channel * <span class="keyword">sizeof</span>(<a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a>) + x * src_stride_y + y * src_stride_z + batch_id * src_stride_w;</div><div class="line"><a name="l00166"></a><span class="lineno"> 166</span>  <a class="code" href="activation__quant__helpers_8h.xhtml#a5a392548f2df67370cb15d2a5d75cd7b">TYPE</a> u0 = <a class="code" href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml#a287e2fc366c312b468382c95bb90f91f">VLOAD</a>(VEC_SIZE)(0, (__global <a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a> *)(input_ptr + 0 * src_stride_y));</div><div class="line"><a name="l00167"></a><span class="lineno"> 167</span>  <a class="code" href="activation__quant__helpers_8h.xhtml#a5a392548f2df67370cb15d2a5d75cd7b">TYPE</a> u1 = <a class="code" href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml#a287e2fc366c312b468382c95bb90f91f">VLOAD</a>(VEC_SIZE)(0, (__global <a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a> *)(input_ptr + 1 * src_stride_y));</div><div class="line"><a name="l00168"></a><span class="lineno"> 168</span> </div><div class="line"><a name="l00169"></a><span class="lineno"> 169</span>  <span class="comment">// Store blocks</span></div><div class="line"><a name="l00170"></a><span class="lineno"> 170</span>  __global uchar *output_ptr = dst_ptr + dst_offset_first_element_in_bytes + x * dst_stride_y + y * dst_stride_z + batch_id * dst_stride_w;</div><div class="line"><a name="l00171"></a><span class="lineno"> 171</span>  *((__global <a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a> *)(output_ptr + (uint)0 * dst_stride_y + z0 * <span class="keyword">sizeof</span>(<a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a>))) = u0.s0;</div><div class="line"><a name="l00172"></a><span class="lineno"> 172</span>  *((__global <a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a> *)(output_ptr + (uint)0 * dst_stride_y + z1 * <span class="keyword">sizeof</span>(<a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a>))) = u0.s1;</div><div class="line"><a name="l00173"></a><span class="lineno"> 173</span>  *((__global <a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a> *)(output_ptr + (uint)0 * dst_stride_y + z2 * <span class="keyword">sizeof</span>(<a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a>))) = u0.s2;</div><div class="line"><a name="l00174"></a><span class="lineno"> 174</span>  *((__global <a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a> *)(output_ptr + (uint)0 * dst_stride_y + z3 * <span class="keyword">sizeof</span>(<a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a>))) = u0.s3;</div><div class="line"><a name="l00175"></a><span class="lineno"> 175</span>  *((__global <a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a> *)(output_ptr + (uint)1 * dst_stride_y + z0 * <span class="keyword">sizeof</span>(<a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a>))) = u1.s0;</div><div class="line"><a name="l00176"></a><span class="lineno"> 176</span>  *((__global <a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a> *)(output_ptr + (uint)1 * dst_stride_y + z1 * <span class="keyword">sizeof</span>(<a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a>))) = u1.s1;</div><div class="line"><a name="l00177"></a><span class="lineno"> 177</span>  *((__global <a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a> *)(output_ptr + (uint)1 * dst_stride_y + z2 * <span class="keyword">sizeof</span>(<a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a>))) = u1.s2;</div><div class="line"><a name="l00178"></a><span class="lineno"> 178</span>  *((__global <a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a> *)(output_ptr + (uint)1 * dst_stride_y + z3 * <span class="keyword">sizeof</span>(<a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a>))) = u1.s3;</div><div class="line"><a name="l00179"></a><span class="lineno"> 179</span> }</div><div class="line"><a name="l00180"></a><span class="lineno"> 180</span> <span class="preprocessor">#endif // VEC_SIZE == 4 && defined(LAST_ACCESSED)</span></div><div class="line"><a name="l00181"></a><span class="lineno"> 181</span> <span class="preprocessor">#endif // defined(DATA_TYPE) && defined(VEC_SIZE) && defined(NUM_GROUPS) && defined(K) && defined(SRC_DIM_Z)</span></div><div class="ttc" id="convolution3x3_8cl_xhtml_afb8c72ce35c4a1f4a2588d6573e54aa1"><div class="ttname"><a href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a></div><div class="ttdeci">#define DATA_TYPE</div><div class="ttdef"><b>Definition:</b> <a href="convolution3x3_8cl_source.xhtml#l00027">convolution3x3.cl:27</a></div></div>
+<a href="channel__shuffle_8cl.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) 2018 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="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml">helpers.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">#if defined(DATA_TYPE) && defined(VEC_SIZE) && defined(NUM_GROUPS) && defined(K) && defined(SRC_DIM_Z)</span></div><div class="line"><a name="l00027"></a><span class="lineno"> 27</span> </div><div class="line"><a name="l00028"></a><span class="lineno"> 28</span> <span class="comment">// Check valid VEC_SIZES</span></div><div class="line"><a name="l00029"></a><span class="lineno"> 29</span> <span class="preprocessor">#if VEC_SIZE != 4 && VEC_SIZE != 8 && VEC_SIZE != 16</span></div><div class="line"><a name="l00030"></a><span class="lineno"> 30</span> <span class="preprocessor">#error "Only vector sizes 4, 8 and 16 are supported"</span></div><div class="line"><a name="l00031"></a><span class="lineno"> 31</span> <span class="preprocessor">#endif // VEC_SIZE != 4 && VEC_SIZE != 8 && VEC_SIZE != 16</span></div><div class="line"><a name="l00032"></a><span class="lineno"> 32</span> </div><div class="line"><a name="l00033"></a><span class="lineno"> 33</span> <span class="preprocessor">#define TYPE VEC_DATA_TYPE(DATA_TYPE, VEC_SIZE)</span></div><div class="line"><a name="l00034"></a><span class="lineno"> 34</span> </div><div class="line"><a name="l00035"></a><span class="lineno"> 35</span> <span class="preprocessor">#define DIV_MOD_UINT(x, y, div_res, mod_res) \</span></div><div class="line"><a name="l00036"></a><span class="lineno"> 36</span> <span class="preprocessor"> ({ \</span></div><div class="line"><a name="l00037"></a><span class="lineno"> 37</span> <span class="preprocessor"> div_res = (uint)((x) * (float)(1.0f / (float)(y))); \</span></div><div class="line"><a name="l00038"></a><span class="lineno"> 38</span> <span class="preprocessor"> uint r = div_res * (y); \</span></div><div class="line"><a name="l00039"></a><span class="lineno"> 39</span> <span class="preprocessor"> mod_res = (x)-r; \</span></div><div class="line"><a name="l00040"></a><span class="lineno"> 40</span> <span class="preprocessor"> })</span></div><div class="line"><a name="l00041"></a><span class="lineno"> 41</span> <span class="comment"></span></div><div class="line"><a name="l00042"></a><span class="lineno"> 42</span> <span class="comment">/** Performs channel shuffle when the data layout is NCHW. See https://arxiv.org/pdf/1707.01083.pdf for details.</span></div><div class="line"><a name="l00043"></a><span class="lineno"> 43</span> <span class="comment"> *</span></div><div class="line"><a name="l00044"></a><span class="lineno"> 44</span> <span class="comment"> * @note The vector size must be given as a preprocessor argument using -DVEC_SIZE=num. e.g. -DVEC_SIZE=4</span></div><div class="line"><a name="l00045"></a><span class="lineno"> 45</span> <span class="comment"> * @note The depth of the tensor must be given as a preprocessor argument using -DSRC_DIM_Z=num. e.g. -DSRC_DIM_Z=64</span></div><div class="line"><a name="l00046"></a><span class="lineno"> 46</span> <span class="comment"> * @note The number of groups must be given as a preprocessor argument using -DNUM_GROUPS=num_groups. e.g. -DNUM_GROUPS=2</span></div><div class="line"><a name="l00047"></a><span class="lineno"> 47</span> <span class="comment"> * @note The number of channels in each group must be given as a preprocessor argument using -DK=num. e.g. -DK=1</span></div><div class="line"><a name="l00048"></a><span class="lineno"> 48</span> <span class="comment"> * K is equal to num_channels / num_groups.</span></div><div class="line"><a name="l00049"></a><span class="lineno"> 49</span> <span class="comment"> *</span></div><div class="line"><a name="l00050"></a><span class="lineno"> 50</span> <span class="comment"> * @param[in] src_ptr Pointer to the source matrix. Supported data types: U8/S8/QASYMM8/U16/S16/F16/U32/S32/F32</span></div><div class="line"><a name="l00051"></a><span class="lineno"> 51</span> <span class="comment"> * @param[in] src_stride_x Stride of the first source tensor in X dimension (in bytes)</span></div><div class="line"><a name="l00052"></a><span class="lineno"> 52</span> <span class="comment"> * @param[in] src_step_x src_stride_x * number of elements along X processed per workitem(in bytes)</span></div><div class="line"><a name="l00053"></a><span class="lineno"> 53</span> <span class="comment"> * @param[in] src_stride_y Stride of the first source tensor in Y dimension (in bytes)</span></div><div class="line"><a name="l00054"></a><span class="lineno"> 54</span> <span class="comment"> * @param[in] src_step_y src_stride_y * number of elements along Y processed per workitem(in bytes)</span></div><div class="line"><a name="l00055"></a><span class="lineno"> 55</span> <span class="comment"> * @param[in] src_stride_z Stride of the first source tensor in Z dimension (in bytes)</span></div><div class="line"><a name="l00056"></a><span class="lineno"> 56</span> <span class="comment"> * @param[in] src_step_z src_stride_z * number of elements along Z processed per workitem(in bytes)</span></div><div class="line"><a name="l00057"></a><span class="lineno"> 57</span> <span class="comment"> * @param[in] src_stride_w Stride of the first source tensor in Z dimension (in bytes)</span></div><div class="line"><a name="l00058"></a><span class="lineno"> 58</span> <span class="comment"> * @param[in] src_step_w src_stride_z * number of elements along Z processed per workitem(in bytes)</span></div><div class="line"><a name="l00059"></a><span class="lineno"> 59</span> <span class="comment"> * @param[in] src_offset_first_element_in_bytes The offset of the first element in the first source tensor</span></div><div class="line"><a name="l00060"></a><span class="lineno"> 60</span> <span class="comment"> * @param[out] dst_ptr Pointer to the destination tensor. Supported data types: same as @p src_ptr</span></div><div class="line"><a name="l00061"></a><span class="lineno"> 61</span> <span class="comment"> * @param[in] dst_stride_x Stride of the destination tensor in X dimension (in bytes)</span></div><div class="line"><a name="l00062"></a><span class="lineno"> 62</span> <span class="comment"> * @param[in] dst_step_x output_stride_x * number of elements along X processed per workitem(in bytes)</span></div><div class="line"><a name="l00063"></a><span class="lineno"> 63</span> <span class="comment"> * @param[in] dst_stride_y Stride of the destination tensor in Y dimension (in bytes)</span></div><div class="line"><a name="l00064"></a><span class="lineno"> 64</span> <span class="comment"> * @param[in] dst_step_y output_stride_y * number of elements along Y processed per workitem(in bytes)</span></div><div class="line"><a name="l00065"></a><span class="lineno"> 65</span> <span class="comment"> * @param[in] dst_stride_z Stride of the destination tensor in Z dimension (in bytes)</span></div><div class="line"><a name="l00066"></a><span class="lineno"> 66</span> <span class="comment"> * @param[in] dst_step_z output_stride_z * number of elements along Z processed per workitem(in bytes)</span></div><div class="line"><a name="l00067"></a><span class="lineno"> 67</span> <span class="comment"> * @param[in] dst_stride_w Stride of the destination tensor in Z dimension (in bytes)</span></div><div class="line"><a name="l00068"></a><span class="lineno"> 68</span> <span class="comment"> * @param[in] dst_step_w output_stride_z * number of elements along Z processed per workitem(in bytes)</span></div><div class="line"><a name="l00069"></a><span class="lineno"> 69</span> <span class="comment"> * @param[in] dst_offset_first_element_in_bytes The offset of the first element in the destination tensor</span></div><div class="line"><a name="l00070"></a><span class="lineno"> 70</span> <span class="comment"> */</span></div><div class="line"><a name="l00071"></a><span class="lineno"> 71</span> __kernel <span class="keywordtype">void</span> channel_shuffle_nchw(<a class="code" href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml#a481bdc6d61b3df9dcdbdb244f0f97790">TENSOR4D_DECLARATION</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a989ab3e96426615bb98e04e0235088ca">src</a>),</div><div class="line"><a name="l00072"></a><span class="lineno"> 72</span>  <a class="code" href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml#a481bdc6d61b3df9dcdbdb244f0f97790">TENSOR4D_DECLARATION</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#adbf67dcee294e673cf796f1ed8aeb6a4">dst</a>))</div><div class="line"><a name="l00073"></a><span class="lineno"> 73</span> {</div><div class="line"><a name="l00074"></a><span class="lineno"> 74</span>  uint curr_channel = 0; <span class="comment">// channel id of input</span></div><div class="line"><a name="l00075"></a><span class="lineno"> 75</span>  uint batch_id = 0; <span class="comment">// batch id</span></div><div class="line"><a name="l00076"></a><span class="lineno"> 76</span>  uint group_id = 0; <span class="comment">// group id</span></div><div class="line"><a name="l00077"></a><span class="lineno"> 77</span>  uint channel_id = 0; <span class="comment">// channel id within the group</span></div><div class="line"><a name="l00078"></a><span class="lineno"> 78</span> </div><div class="line"><a name="l00079"></a><span class="lineno"> 79</span>  <span class="comment">// Compute curr_channel and batch_id</span></div><div class="line"><a name="l00080"></a><span class="lineno"> 80</span>  DIV_MOD_UINT(get_global_id(2), SRC_DIM_Z, batch_id, curr_channel);</div><div class="line"><a name="l00081"></a><span class="lineno"> 81</span> </div><div class="line"><a name="l00082"></a><span class="lineno"> 82</span>  <span class="comment">// Compute group_id and channel_id</span></div><div class="line"><a name="l00083"></a><span class="lineno"> 83</span>  DIV_MOD_UINT(curr_channel, K, group_id, channel_id);</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>  <span class="keyword">const</span> uint x = get_global_id(0) * <a class="code" href="depthwise__convolution__quantized_8cl.xhtml#a3fffea119c04c7680f2e9cf3fadf63b4">VEC_SIZE</a>;</div><div class="line"><a name="l00086"></a><span class="lineno"> 86</span>  <span class="keyword">const</span> uint y = get_global_id(1) * 2;</div><div class="line"><a name="l00087"></a><span class="lineno"> 87</span>  <span class="keyword">const</span> uint z = channel_id * NUM_GROUPS + group_id;</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="comment">// Load the Nx2 block</span></div><div class="line"><a name="l00090"></a><span class="lineno"> 90</span>  <span class="keyword">const</span> __global uchar *input_ptr = src_ptr + src_offset_first_element_in_bytes + x * <span class="keyword">sizeof</span>(<a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a>) + y * src_stride_y + curr_channel * src_stride_z + batch_id * src_stride_w;</div><div class="line"><a name="l00091"></a><span class="lineno"> 91</span>  <a class="code" href="activation__quant__helpers_8h.xhtml#a5a392548f2df67370cb15d2a5d75cd7b">TYPE</a> u0 = <a class="code" href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml#a287e2fc366c312b468382c95bb90f91f">VLOAD</a>(<a class="code" href="depthwise__convolution__quantized_8cl.xhtml#a3fffea119c04c7680f2e9cf3fadf63b4">VEC_SIZE</a>)(0, (__global <a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a> *)(input_ptr + 0 * src_stride_y));</div><div class="line"><a name="l00092"></a><span class="lineno"> 92</span>  <a class="code" href="activation__quant__helpers_8h.xhtml#a5a392548f2df67370cb15d2a5d75cd7b">TYPE</a> u1 = <a class="code" href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml#a287e2fc366c312b468382c95bb90f91f">VLOAD</a>(<a class="code" href="depthwise__convolution__quantized_8cl.xhtml#a3fffea119c04c7680f2e9cf3fadf63b4">VEC_SIZE</a>)(0, (__global <a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a> *)(input_ptr + 1 * src_stride_y));</div><div class="line"><a name="l00093"></a><span class="lineno"> 93</span> </div><div class="line"><a name="l00094"></a><span class="lineno"> 94</span>  <span class="comment">// Store blocks</span></div><div class="line"><a name="l00095"></a><span class="lineno"> 95</span>  __global uchar *output_ptr = dst_ptr + dst_offset_first_element_in_bytes + x * <span class="keyword">sizeof</span>(<a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a>) + y * dst_stride_y + z * dst_stride_z + batch_id * dst_stride_w;</div><div class="line"><a name="l00096"></a><span class="lineno"> 96</span>  <a class="code" href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml#acb282042d1edeeaa3cc979a206f78b54">VSTORE</a>(<a class="code" href="depthwise__convolution__quantized_8cl.xhtml#a3fffea119c04c7680f2e9cf3fadf63b4">VEC_SIZE</a>)</div><div class="line"><a name="l00097"></a><span class="lineno"> 97</span>  (u0, 0, (__global <a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a> *)(output_ptr + 0 * dst_stride_y));</div><div class="line"><a name="l00098"></a><span class="lineno"> 98</span>  <a class="code" href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml#acb282042d1edeeaa3cc979a206f78b54">VSTORE</a>(<a class="code" href="depthwise__convolution__quantized_8cl.xhtml#a3fffea119c04c7680f2e9cf3fadf63b4">VEC_SIZE</a>)</div><div class="line"><a name="l00099"></a><span class="lineno"> 99</span>  (u1, 0, (__global <a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a> *)(output_ptr + 1 * dst_stride_y));</div><div class="line"><a name="l00100"></a><span class="lineno"> 100</span> }</div><div class="line"><a name="l00101"></a><span class="lineno"> 101</span> </div><div class="line"><a name="l00102"></a><span class="lineno"> 102</span> <span class="preprocessor">#if VEC_SIZE == 4 && defined(LAST_ACCESSED)</span></div><div class="line"><a name="l00103"></a><span class="lineno"> 103</span> <span class="comment">/** Performs channel shuffle when the data layout is NHWC. See https://arxiv.org/pdf/1707.01083.pdf for details.</span></div><div class="line"><a name="l00104"></a><span class="lineno"> 104</span> <span class="comment"> *</span></div><div class="line"><a name="l00105"></a><span class="lineno"> 105</span> <span class="comment"> * @note This implementation is only defined for VEC_SIZE = 4</span></div><div class="line"><a name="l00106"></a><span class="lineno"> 106</span> <span class="comment"> * @note This last element accessed along the first dimension must be given as a preprocessor argument using -DLAST_ACCESSED=num. e.g. -DLAST_ACCESSED=64 in order to prevent out-of-bound writes.</span></div><div class="line"><a name="l00107"></a><span class="lineno"> 107</span> <span class="comment"> * @note The vector size must be given as a preprocessor argument using -DVEC_SIZE=num. e.g. -DVEC_SIZE=4</span></div><div class="line"><a name="l00108"></a><span class="lineno"> 108</span> <span class="comment"> * @note The height of the tensor must be given as a preprocessor argument using -DSRC_DIM_Z=num. e.g. -DSRC_DIM_Z=64</span></div><div class="line"><a name="l00109"></a><span class="lineno"> 109</span> <span class="comment"> * @note The number of groups must be given as a preprocessor argument using -DNUM_GROUPS=num_groups. e.g. -DNUM_GROUPS=2</span></div><div class="line"><a name="l00110"></a><span class="lineno"> 110</span> <span class="comment"> * @note The number of channels in each group must be given as a preprocessor argument using -DK=num. e.g. -DK=1</span></div><div class="line"><a name="l00111"></a><span class="lineno"> 111</span> <span class="comment"> * K is equal to num_channels / num_groups.</span></div><div class="line"><a name="l00112"></a><span class="lineno"> 112</span> <span class="comment"> *</span></div><div class="line"><a name="l00113"></a><span class="lineno"> 113</span> <span class="comment"> * @param[in] src_ptr Pointer to the source matrix. Supported data types: U8/S8/QASYMM8/U16/S16/F16/U32/S32/F32</span></div><div class="line"><a name="l00114"></a><span class="lineno"> 114</span> <span class="comment"> * @param[in] src_stride_x Stride of the first source tensor in X dimension (in bytes)</span></div><div class="line"><a name="l00115"></a><span class="lineno"> 115</span> <span class="comment"> * @param[in] src_step_x src_stride_x * number of elements along X processed per workitem(in bytes)</span></div><div class="line"><a name="l00116"></a><span class="lineno"> 116</span> <span class="comment"> * @param[in] src_stride_y Stride of the first source tensor in Y dimension (in bytes)</span></div><div class="line"><a name="l00117"></a><span class="lineno"> 117</span> <span class="comment"> * @param[in] src_step_y src_stride_y * number of elements along Y processed per workitem(in bytes)</span></div><div class="line"><a name="l00118"></a><span class="lineno"> 118</span> <span class="comment"> * @param[in] src_stride_z Stride of the first source tensor in Z dimension (in bytes)</span></div><div class="line"><a name="l00119"></a><span class="lineno"> 119</span> <span class="comment"> * @param[in] src_step_z src_stride_z * number of elements along Z processed per workitem(in bytes)</span></div><div class="line"><a name="l00120"></a><span class="lineno"> 120</span> <span class="comment"> * @param[in] src_stride_w Stride of the first source tensor in Z dimension (in bytes)</span></div><div class="line"><a name="l00121"></a><span class="lineno"> 121</span> <span class="comment"> * @param[in] src_step_w src_stride_z * number of elements along Z processed per workitem(in bytes)</span></div><div class="line"><a name="l00122"></a><span class="lineno"> 122</span> <span class="comment"> * @param[in] src_offset_first_element_in_bytes The offset of the first element in the first source tensor</span></div><div class="line"><a name="l00123"></a><span class="lineno"> 123</span> <span class="comment"> * @param[out] dst_ptr Pointer to the destination tensor. Supported data types: same as @p src_ptr</span></div><div class="line"><a name="l00124"></a><span class="lineno"> 124</span> <span class="comment"> * @param[in] dst_stride_x Stride of the destination tensor in X dimension (in bytes)</span></div><div class="line"><a name="l00125"></a><span class="lineno"> 125</span> <span class="comment"> * @param[in] dst_step_x output_stride_x * number of elements along X processed per workitem(in bytes)</span></div><div class="line"><a name="l00126"></a><span class="lineno"> 126</span> <span class="comment"> * @param[in] dst_stride_y Stride of the destination tensor in Y dimension (in bytes)</span></div><div class="line"><a name="l00127"></a><span class="lineno"> 127</span> <span class="comment"> * @param[in] dst_step_y output_stride_y * number of elements along Y processed per workitem(in bytes)</span></div><div class="line"><a name="l00128"></a><span class="lineno"> 128</span> <span class="comment"> * @param[in] dst_stride_z Stride of the destination tensor in Z dimension (in bytes)</span></div><div class="line"><a name="l00129"></a><span class="lineno"> 129</span> <span class="comment"> * @param[in] dst_step_z output_stride_z * number of elements along Z processed per workitem(in bytes)</span></div><div class="line"><a name="l00130"></a><span class="lineno"> 130</span> <span class="comment"> * @param[in] dst_stride_w Stride of the destination tensor in Z dimension (in bytes)</span></div><div class="line"><a name="l00131"></a><span class="lineno"> 131</span> <span class="comment"> * @param[in] dst_step_w output_stride_z * number of elements along Z processed per workitem(in bytes)</span></div><div class="line"><a name="l00132"></a><span class="lineno"> 132</span> <span class="comment"> * @param[in] dst_offset_first_element_in_bytes The offset of the first element in the destination tensor</span></div><div class="line"><a name="l00133"></a><span class="lineno"> 133</span> <span class="comment"> */</span></div><div class="line"><a name="l00134"></a><span class="lineno"> 134</span> __kernel <span class="keywordtype">void</span> channel_shuffle_nhwc(<a class="code" href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml#a481bdc6d61b3df9dcdbdb244f0f97790">TENSOR4D_DECLARATION</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a989ab3e96426615bb98e04e0235088ca">src</a>),</div><div class="line"><a name="l00135"></a><span class="lineno"> 135</span>  <a class="code" href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml#a481bdc6d61b3df9dcdbdb244f0f97790">TENSOR4D_DECLARATION</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#adbf67dcee294e673cf796f1ed8aeb6a4">dst</a>))</div><div class="line"><a name="l00136"></a><span class="lineno"> 136</span> {</div><div class="line"><a name="l00137"></a><span class="lineno"> 137</span>  <span class="keyword">const</span> uint curr_channel = min((uint)(get_global_id(0) * <a class="code" href="depthwise__convolution__quantized_8cl.xhtml#a3fffea119c04c7680f2e9cf3fadf63b4">VEC_SIZE</a>), (uint)LAST_ACCESSED); <span class="comment">// input feature map</span></div><div class="line"><a name="l00138"></a><span class="lineno"> 138</span>  uint channel_id0 = 0;</div><div class="line"><a name="l00139"></a><span class="lineno"> 139</span>  uint channel_id1 = 0;</div><div class="line"><a name="l00140"></a><span class="lineno"> 140</span>  uint channel_id2 = 0;</div><div class="line"><a name="l00141"></a><span class="lineno"> 141</span>  uint channel_id3 = 0;</div><div class="line"><a name="l00142"></a><span class="lineno"> 142</span>  uint group_id0 = 0;</div><div class="line"><a name="l00143"></a><span class="lineno"> 143</span>  uint group_id1 = 0;</div><div class="line"><a name="l00144"></a><span class="lineno"> 144</span>  uint group_id2 = 0;</div><div class="line"><a name="l00145"></a><span class="lineno"> 145</span>  uint group_id3 = 0;</div><div class="line"><a name="l00146"></a><span class="lineno"> 146</span>  uint y = 0;</div><div class="line"><a name="l00147"></a><span class="lineno"> 147</span>  uint batch_id = 0;</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>  <span class="comment">// Compute curr_channel and batch_id</span></div><div class="line"><a name="l00150"></a><span class="lineno"> 150</span>  DIV_MOD_UINT(get_global_id(2), (uint)SRC_DIM_Z, batch_id, y);</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>  <span class="comment">// Compute group_id and channel_id</span></div><div class="line"><a name="l00153"></a><span class="lineno"> 153</span>  DIV_MOD_UINT(curr_channel + (uint)0, K, group_id0, channel_id0);</div><div class="line"><a name="l00154"></a><span class="lineno"> 154</span>  DIV_MOD_UINT(curr_channel + (uint)1, K, group_id1, channel_id1);</div><div class="line"><a name="l00155"></a><span class="lineno"> 155</span>  DIV_MOD_UINT(curr_channel + (uint)2, K, group_id2, channel_id2);</div><div class="line"><a name="l00156"></a><span class="lineno"> 156</span>  DIV_MOD_UINT(curr_channel + (uint)3, K, group_id3, channel_id3);</div><div class="line"><a name="l00157"></a><span class="lineno"> 157</span> </div><div class="line"><a name="l00158"></a><span class="lineno"> 158</span>  <span class="keyword">const</span> uint x = get_global_id(1) * 2;</div><div class="line"><a name="l00159"></a><span class="lineno"> 159</span>  <span class="keyword">const</span> uint z0 = channel_id0 * (uint)NUM_GROUPS + group_id0;</div><div class="line"><a name="l00160"></a><span class="lineno"> 160</span>  <span class="keyword">const</span> uint z1 = channel_id1 * (uint)NUM_GROUPS + group_id1;</div><div class="line"><a name="l00161"></a><span class="lineno"> 161</span>  <span class="keyword">const</span> uint z2 = channel_id2 * (uint)NUM_GROUPS + group_id2;</div><div class="line"><a name="l00162"></a><span class="lineno"> 162</span>  <span class="keyword">const</span> uint z3 = channel_id3 * (uint)NUM_GROUPS + group_id3;</div><div class="line"><a name="l00163"></a><span class="lineno"> 163</span> </div><div class="line"><a name="l00164"></a><span class="lineno"> 164</span>  <span class="comment">// Load the Nx2 block</span></div><div class="line"><a name="l00165"></a><span class="lineno"> 165</span>  <span class="keyword">const</span> __global uchar *input_ptr = src_ptr + src_offset_first_element_in_bytes + curr_channel * <span class="keyword">sizeof</span>(<a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a>) + x * src_stride_y + y * src_stride_z + batch_id * src_stride_w;</div><div class="line"><a name="l00166"></a><span class="lineno"> 166</span>  <a class="code" href="activation__quant__helpers_8h.xhtml#a5a392548f2df67370cb15d2a5d75cd7b">TYPE</a> u0 = <a class="code" href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml#a287e2fc366c312b468382c95bb90f91f">VLOAD</a>(<a class="code" href="depthwise__convolution__quantized_8cl.xhtml#a3fffea119c04c7680f2e9cf3fadf63b4">VEC_SIZE</a>)(0, (__global <a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a> *)(input_ptr + 0 * src_stride_y));</div><div class="line"><a name="l00167"></a><span class="lineno"> 167</span>  <a class="code" href="activation__quant__helpers_8h.xhtml#a5a392548f2df67370cb15d2a5d75cd7b">TYPE</a> u1 = <a class="code" href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml#a287e2fc366c312b468382c95bb90f91f">VLOAD</a>(<a class="code" href="depthwise__convolution__quantized_8cl.xhtml#a3fffea119c04c7680f2e9cf3fadf63b4">VEC_SIZE</a>)(0, (__global <a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a> *)(input_ptr + 1 * src_stride_y));</div><div class="line"><a name="l00168"></a><span class="lineno"> 168</span> </div><div class="line"><a name="l00169"></a><span class="lineno"> 169</span>  <span class="comment">// Store blocks</span></div><div class="line"><a name="l00170"></a><span class="lineno"> 170</span>  __global uchar *output_ptr = dst_ptr + dst_offset_first_element_in_bytes + x * dst_stride_y + y * dst_stride_z + batch_id * dst_stride_w;</div><div class="line"><a name="l00171"></a><span class="lineno"> 171</span>  *((__global <a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a> *)(output_ptr + (uint)0 * dst_stride_y + z0 * <span class="keyword">sizeof</span>(<a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a>))) = u0.s0;</div><div class="line"><a name="l00172"></a><span class="lineno"> 172</span>  *((__global <a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a> *)(output_ptr + (uint)0 * dst_stride_y + z1 * <span class="keyword">sizeof</span>(<a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a>))) = u0.s1;</div><div class="line"><a name="l00173"></a><span class="lineno"> 173</span>  *((__global <a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a> *)(output_ptr + (uint)0 * dst_stride_y + z2 * <span class="keyword">sizeof</span>(<a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a>))) = u0.s2;</div><div class="line"><a name="l00174"></a><span class="lineno"> 174</span>  *((__global <a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a> *)(output_ptr + (uint)0 * dst_stride_y + z3 * <span class="keyword">sizeof</span>(<a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a>))) = u0.s3;</div><div class="line"><a name="l00175"></a><span class="lineno"> 175</span>  *((__global <a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a> *)(output_ptr + (uint)1 * dst_stride_y + z0 * <span class="keyword">sizeof</span>(<a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a>))) = u1.s0;</div><div class="line"><a name="l00176"></a><span class="lineno"> 176</span>  *((__global <a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a> *)(output_ptr + (uint)1 * dst_stride_y + z1 * <span class="keyword">sizeof</span>(<a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a>))) = u1.s1;</div><div class="line"><a name="l00177"></a><span class="lineno"> 177</span>  *((__global <a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a> *)(output_ptr + (uint)1 * dst_stride_y + z2 * <span class="keyword">sizeof</span>(<a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a>))) = u1.s2;</div><div class="line"><a name="l00178"></a><span class="lineno"> 178</span>  *((__global <a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a> *)(output_ptr + (uint)1 * dst_stride_y + z3 * <span class="keyword">sizeof</span>(<a class="code" href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a>))) = u1.s3;</div><div class="line"><a name="l00179"></a><span class="lineno"> 179</span> }</div><div class="line"><a name="l00180"></a><span class="lineno"> 180</span> <span class="preprocessor">#endif // VEC_SIZE == 4 && defined(LAST_ACCESSED)</span></div><div class="line"><a name="l00181"></a><span class="lineno"> 181</span> <span class="preprocessor">#endif // defined(DATA_TYPE) && defined(VEC_SIZE) && defined(NUM_GROUPS) && defined(K) && defined(SRC_DIM_Z)</span></div><div class="ttc" id="depthwise__convolution__quantized_8cl_xhtml_a3fffea119c04c7680f2e9cf3fadf63b4"><div class="ttname"><a href="depthwise__convolution__quantized_8cl.xhtml#a3fffea119c04c7680f2e9cf3fadf63b4">VEC_SIZE</a></div><div class="ttdeci">#define VEC_SIZE</div><div class="ttdef"><b>Definition:</b> <a href="depthwise__convolution__quantized_8cl_source.xhtml#l00031">depthwise_convolution_quantized.cl:31</a></div></div>
+<div class="ttc" id="convolution3x3_8cl_xhtml_afb8c72ce35c4a1f4a2588d6573e54aa1"><div class="ttname"><a href="convolution3x3_8cl.xhtml#afb8c72ce35c4a1f4a2588d6573e54aa1">DATA_TYPE</a></div><div class="ttdeci">#define DATA_TYPE</div><div class="ttdef"><b>Definition:</b> <a href="convolution3x3_8cl_source.xhtml#l00027">convolution3x3.cl:27</a></div></div>
<div class="ttc" id="activation__quant__helpers_8h_xhtml_a5a392548f2df67370cb15d2a5d75cd7b"><div class="ttname"><a href="activation__quant__helpers_8h.xhtml#a5a392548f2df67370cb15d2a5d75cd7b">TYPE</a></div><div class="ttdeci">#define TYPE</div><div class="ttdef"><b>Definition:</b> <a href="activation__quant__helpers_8h_source.xhtml#l00027">activation_quant_helpers.h:27</a></div></div>
<div class="ttc" id="namespacearm__compute_1_1test_1_1validation_xhtml_adbf67dcee294e673cf796f1ed8aeb6a4"><div class="ttname"><a href="namespacearm__compute_1_1test_1_1validation.xhtml#adbf67dcee294e673cf796f1ed8aeb6a4">arm_compute::test::validation::dst</a></div><div class="ttdeci">CLTensor dst</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_2_absolute_difference_8cpp_source.xhtml#l00102">AbsoluteDifference.cpp:102</a></div></div>
<div class="ttc" id="src_2core_2_c_l_2cl__kernels_2_helpers_8h_xhtml"><div class="ttname"><a href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml">helpers.h</a></div></div>
-<div class="ttc" id="src_2core_2_c_l_2cl__kernels_2_helpers_8h_xhtml_acb282042d1edeeaa3cc979a206f78b54"><div class="ttname"><a href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml#acb282042d1edeeaa3cc979a206f78b54">VSTORE</a></div><div class="ttdeci">#define VSTORE(size)</div><div class="ttdef"><b>Definition:</b> <a href="src_2core_2_c_l_2cl__kernels_2_helpers_8h_source.xhtml#l00056">helpers.h:56</a></div></div>
-<div class="ttc" id="src_2core_2_c_l_2cl__kernels_2_helpers_8h_xhtml_a481bdc6d61b3df9dcdbdb244f0f97790"><div class="ttname"><a href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml#a481bdc6d61b3df9dcdbdb244f0f97790">TENSOR4D_DECLARATION</a></div><div class="ttdeci">#define TENSOR4D_DECLARATION(name)</div><div class="ttdef"><b>Definition:</b> <a href="src_2core_2_c_l_2cl__kernels_2_helpers_8h_source.xhtml#l00111">helpers.h:111</a></div></div>
-<div class="ttc" id="src_2core_2_c_l_2cl__kernels_2_helpers_8h_xhtml_a287e2fc366c312b468382c95bb90f91f"><div class="ttname"><a href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml#a287e2fc366c312b468382c95bb90f91f">VLOAD</a></div><div class="ttdeci">#define VLOAD(size)</div><div class="ttdef"><b>Definition:</b> <a href="src_2core_2_c_l_2cl__kernels_2_helpers_8h_source.xhtml#l00053">helpers.h:53</a></div></div>
+<div class="ttc" id="src_2core_2_c_l_2cl__kernels_2_helpers_8h_xhtml_acb282042d1edeeaa3cc979a206f78b54"><div class="ttname"><a href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml#acb282042d1edeeaa3cc979a206f78b54">VSTORE</a></div><div class="ttdeci">#define VSTORE(size)</div><div class="ttdef"><b>Definition:</b> <a href="src_2core_2_c_l_2cl__kernels_2_helpers_8h_source.xhtml#l00198">helpers.h:198</a></div></div>
+<div class="ttc" id="src_2core_2_c_l_2cl__kernels_2_helpers_8h_xhtml_a481bdc6d61b3df9dcdbdb244f0f97790"><div class="ttname"><a href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml#a481bdc6d61b3df9dcdbdb244f0f97790">TENSOR4D_DECLARATION</a></div><div class="ttdeci">#define TENSOR4D_DECLARATION(name)</div><div class="ttdef"><b>Definition:</b> <a href="src_2core_2_c_l_2cl__kernels_2_helpers_8h_source.xhtml#l00293">helpers.h:293</a></div></div>
+<div class="ttc" id="src_2core_2_c_l_2cl__kernels_2_helpers_8h_xhtml_a287e2fc366c312b468382c95bb90f91f"><div class="ttname"><a href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml#a287e2fc366c312b468382c95bb90f91f">VLOAD</a></div><div class="ttdeci">#define VLOAD(size)</div><div class="ttdef"><b>Definition:</b> <a href="src_2core_2_c_l_2cl__kernels_2_helpers_8h_source.xhtml#l00195">helpers.h:195</a></div></div>
<div class="ttc" id="namespacearm__compute_1_1test_1_1validation_xhtml_a989ab3e96426615bb98e04e0235088ca"><div class="ttname"><a href="namespacearm__compute_1_1test_1_1validation.xhtml#a989ab3e96426615bb98e04e0235088ca">arm_compute::test::validation::src</a></div><div class="ttdeci">cast configure & src</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_2_cast_8cpp_source.xhtml#l00169">Cast.cpp:169</a></div></div>
</div><!-- fragment --></div><!-- contents -->
</div><!-- doc-content -->
@@ -114,7 +115,7 @@
<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_25885286e9dad4fa105b7b25a8031bbf.xhtml">cl_kernels</a></li><li class="navelem"><a class="el" href="channel__shuffle_8cl.xhtml">channel_shuffle.cl</a></li>
- <li class="footer">Generated on Mon Sep 2 2019 11:47:20 for Compute Library by
+ <li class="footer">Generated on Thu Nov 28 2019 16:52:57 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>