zhangyiss/mlx
1
0
Fork 0
mirror of https://github.com/ml-explore/mlx.git synced 2025-11-04 18:48:15 +08:00
Code Issues Actions 1 Packages Projects Releases Wiki Activity
Files
bea7b910172e15b18d99b1982f2a6031750d2c5b
mlx/docs/build/html/reduce__all_8h_source.html
Awni Hannun 098d163a6e docs update
2025-10-17 19:14:56 +00:00

238 lines
27 KiB
HTML
Raw Blame History

<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "https://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml" lang="en-US">
<head>
<meta http-equiv="Content-Type" content="text/xhtml;charset=UTF-8"/>
<meta http-equiv="X-UA-Compatible" content="IE=11"/>
<meta name="generator" content="Doxygen 1.10.0"/>
<meta name="viewport" content="width=device-width, initial-scale=1"/>
<title>MLX: mlx/backend/metal/kernels/reduction/reduce_all.h 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>
<script type="text/javascript" src="clipboard.js"></script>
<script type="text/javascript" src="cookie.js"></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>
<link href="doxygen.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 id="projectrow">
<td id="projectalign">
<div id="projectname">MLX
</div>
</td>
</tr>
</tbody>
</table>
</div>
<!-- end header part -->
<!-- Generated by Doxygen 1.10.0 -->
<script type="text/javascript">
/* @license magnet:?xt=urn:btih:d3d9a9a6595521f9666a5e94cc830dab83b65699&amp;dn=expat.txt MIT */
var searchBox = new SearchBox("searchBox", "search/",'.html');
/* @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:d3d9a9a6595521f9666a5e94cc830dab83b65699&amp;dn=expat.txt MIT */
$(function() {
initMenu('',true,false,'search.php','Search');
$(function() { init_search(); });
});
/* @license-end */
</script>
<div id="main-nav"></div>
<script type="text/javascript">
/* @license magnet:?xt=urn:btih:d3d9a9a6595521f9666a5e94cc830dab83b65699&amp;dn=expat.txt MIT */
$(function() { codefold.init(0); });
/* @license-end */
</script>
<!-- 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">
<div id="MSearchResults">
<div class="SRPage">
<div id="SRIndex">
<div id="SRResults"></div>
<div class="SRStatus" id="Loading">Loading...</div>
<div class="SRStatus" id="Searching">Searching...</div>
<div class="SRStatus" id="NoMatches">No Matches</div>
</div>
</div>
</div>
</div>
<div id="nav-path" class="navpath">
<ul>
<li class="navelem"><a class="el" href="dir_938ab0ecf10b8b860ff766c820f665fd.html">mlx</a></li><li class="navelem"><a class="el" href="dir_1d446c9bd3c99228254c9484e0bc5c06.html">backend</a></li><li class="navelem"><a class="el" href="dir_d0c977ea65824390717cdb7efc36c157.html">metal</a></li><li class="navelem"><a class="el" href="dir_70a37effa88bcbd6b791977fa1e64356.html">kernels</a></li><li class="navelem"><a class="el" href="dir_f60cd69d27fd3faa641c79056fff0e2d.html">reduction</a></li> </ul>
</div>
</div><!-- top -->
<div class="header">
<div class="headertitle"><div class="title">reduce_all.h</div></div>
</div><!--header-->
<div class="contents">
<a href="reduce__all_8h.html">Go to the documentation of this file.</a><div class="fragment"><div class="line"><a id="l00001" name="l00001"></a><span class="lineno"> 1</span><span class="comment">// Copyright © 2023-2024 Apple Inc.</span></div>
<div class="line"><a id="l00002" name="l00002"></a><span class="lineno"> 2</span> </div>
<div class="line"><a id="l00004" name="l00004"></a><span class="lineno"> 4</span><span class="comment">// All reduce helper</span></div>
<div class="line"><a id="l00006" name="l00006"></a><span class="lineno"> 6</span> </div>
<div class="line"><a id="l00007" name="l00007"></a><span class="lineno"> 7</span><span class="keyword">template</span> &lt;<span class="keyword">typename</span> T, <span class="keyword">typename</span> U, <span class="keyword">typename</span> Op, <span class="keywordtype">int</span> N_READS = REDUCE_N_READS&gt;</div>
<div class="foldopen" id="foldopen00008" data-start="{" data-end="}">
<div class="line"><a id="l00008" name="l00008"></a><span class="lineno"><a class="line" href="reduce__all_8h.html#a0895975942131768f280decc854d0ac1"> 8</a></span>METAL_FUNC U <a class="code hl_function" href="reduce__all_8h.html#a0895975942131768f280decc854d0ac1">per_thread_all_reduce</a>(</div>
<div class="line"><a id="l00009" name="l00009"></a><span class="lineno"> 9</span> <span class="keyword">const</span> device T* in,</div>
<div class="line"><a id="l00010" name="l00010"></a><span class="lineno"> 10</span> <span class="keyword">const</span> device <span class="keywordtype">size_t</span>&amp; in_size,</div>
<div class="line"><a id="l00011" name="l00011"></a><span class="lineno"> 11</span> uint gid,</div>
<div class="line"><a id="l00012" name="l00012"></a><span class="lineno"> 12</span> uint grid_size) {</div>
<div class="line"><a id="l00013" name="l00013"></a><span class="lineno"> 13</span> Op <a class="code hl_variable" href="common_2binary_8h.html#a70228731d29946574b238d21fb4b360c">op</a>;</div>
<div class="line"><a id="l00014" name="l00014"></a><span class="lineno"> 14</span> U total_val = Op::init;</div>
<div class="line"><a id="l00015" name="l00015"></a><span class="lineno"> 15</span> </div>
<div class="line"><a id="l00016" name="l00016"></a><span class="lineno"> 16</span> <span class="keywordflow">if</span> (gid * N_READS &lt; in_size) {</div>
<div class="line"><a id="l00017" name="l00017"></a><span class="lineno"> 17</span> in += gid * N_READS;</div>
<div class="line"><a id="l00018" name="l00018"></a><span class="lineno"> 18</span> </div>
<div class="line"><a id="l00019" name="l00019"></a><span class="lineno"> 19</span> <span class="keywordtype">int</span> r = 0;</div>
<div class="line"><a id="l00020" name="l00020"></a><span class="lineno"> 20</span> <span class="keywordflow">for</span> (; r &lt; (int)<a class="code hl_function" href="backend_2metal_2kernels_2utils_8h.html#a51c19db777f43943e4b35f25dd88d49d">ceildiv</a>(in_size, grid_size * N_READS) - 1; r++) {</div>
<div class="line"><a id="l00021" name="l00021"></a><span class="lineno"> 21</span> U vals[N_READS] = {<a class="code hl_variable" href="common_2binary_8h.html#a70228731d29946574b238d21fb4b360c">op</a>.init};</div>
<div class="line"><a id="l00022" name="l00022"></a><span class="lineno"> 22</span> </div>
<div class="line"><a id="l00023" name="l00023"></a><span class="lineno"> 23</span> <span class="keywordflow">for</span> (<span class="keywordtype">int</span> i = 0; i &lt; N_READS; i++) {</div>
<div class="line"><a id="l00024" name="l00024"></a><span class="lineno"> 24</span> vals[i] = <span class="keyword">static_cast&lt;</span>U<span class="keyword">&gt;</span>(in[i]);</div>
<div class="line"><a id="l00025" name="l00025"></a><span class="lineno"> 25</span> }</div>
<div class="line"><a id="l00026" name="l00026"></a><span class="lineno"> 26</span> <span class="keywordflow">for</span> (<span class="keywordtype">int</span> i = 0; i &lt; N_READS; i++) {</div>
<div class="line"><a id="l00027" name="l00027"></a><span class="lineno"> 27</span> total_val = <a class="code hl_variable" href="common_2binary_8h.html#a70228731d29946574b238d21fb4b360c">op</a>(vals[i], total_val);</div>
<div class="line"><a id="l00028" name="l00028"></a><span class="lineno"> 28</span> }</div>
<div class="line"><a id="l00029" name="l00029"></a><span class="lineno"> 29</span> </div>
<div class="line"><a id="l00030" name="l00030"></a><span class="lineno"> 30</span> in += grid_size * N_READS;</div>
<div class="line"><a id="l00031" name="l00031"></a><span class="lineno"> 31</span> }</div>
<div class="line"><a id="l00032" name="l00032"></a><span class="lineno"> 32</span> </div>
<div class="line"><a id="l00033" name="l00033"></a><span class="lineno"> 33</span> <span class="comment">// Separate case for the last set as we close the reduction size</span></div>
<div class="line"><a id="l00034" name="l00034"></a><span class="lineno"> 34</span> <span class="keywordtype">size_t</span> curr_idx = (gid + r * (size_t)grid_size) * N_READS;</div>
<div class="line"><a id="l00035" name="l00035"></a><span class="lineno"> 35</span> <span class="keywordflow">if</span> (curr_idx &lt; in_size) {</div>
<div class="line"><a id="l00036" name="l00036"></a><span class="lineno"> 36</span> <span class="keywordtype">int</span> max_reads = in_size - curr_idx;</div>
<div class="line"><a id="l00037" name="l00037"></a><span class="lineno"> 37</span> T vals[N_READS];</div>
<div class="line"><a id="l00038" name="l00038"></a><span class="lineno"> 38</span> </div>
<div class="line"><a id="l00039" name="l00039"></a><span class="lineno"> 39</span> <span class="keywordflow">for</span> (<span class="keywordtype">int</span> i = 0, idx = 0; i &lt; N_READS; i++, idx++) {</div>
<div class="line"><a id="l00040" name="l00040"></a><span class="lineno"> 40</span> idx = idx &lt; max_reads ? idx : max_reads - 1;</div>
<div class="line"><a id="l00041" name="l00041"></a><span class="lineno"> 41</span> vals[i] = in[idx];</div>
<div class="line"><a id="l00042" name="l00042"></a><span class="lineno"> 42</span> }</div>
<div class="line"><a id="l00043" name="l00043"></a><span class="lineno"> 43</span> <span class="keywordflow">for</span> (<span class="keywordtype">int</span> i = 0; i &lt; N_READS; i++) {</div>
<div class="line"><a id="l00044" name="l00044"></a><span class="lineno"> 44</span> U val = i &lt; max_reads ? vals[i] : Op::init;</div>
<div class="line"><a id="l00045" name="l00045"></a><span class="lineno"> 45</span> total_val = <a class="code hl_variable" href="common_2binary_8h.html#a70228731d29946574b238d21fb4b360c">op</a>(<span class="keyword">static_cast&lt;</span>U<span class="keyword">&gt;</span>(val), total_val);</div>
<div class="line"><a id="l00046" name="l00046"></a><span class="lineno"> 46</span> }</div>
<div class="line"><a id="l00047" name="l00047"></a><span class="lineno"> 47</span> }</div>
<div class="line"><a id="l00048" name="l00048"></a><span class="lineno"> 48</span> }</div>
<div class="line"><a id="l00049" name="l00049"></a><span class="lineno"> 49</span> </div>
<div class="line"><a id="l00050" name="l00050"></a><span class="lineno"> 50</span> <span class="keywordflow">return</span> total_val;</div>
<div class="line"><a id="l00051" name="l00051"></a><span class="lineno"> 51</span>}</div>
</div>
<div class="line"><a id="l00052" name="l00052"></a><span class="lineno"> 52</span> </div>
<div class="line"><a id="l00054" name="l00054"></a><span class="lineno"> 54</span><span class="comment">// All reduce kernel</span></div>
<div class="line"><a id="l00056" name="l00056"></a><span class="lineno"> 56</span> </div>
<div class="line"><a id="l00057" name="l00057"></a><span class="lineno"> 57</span><span class="comment">// NB: This kernel assumes threads_per_threadgroup is at most</span></div>
<div class="line"><a id="l00058" name="l00058"></a><span class="lineno"> 58</span><span class="comment">// 1024. This way with a simd_size of 32, we are guaranteed to</span></div>
<div class="line"><a id="l00059" name="l00059"></a><span class="lineno"> 59</span><span class="comment">// complete the reduction in two steps of simd-level reductions.</span></div>
<div class="line"><a id="l00060" name="l00060"></a><span class="lineno"> 60</span><span class="keyword">template</span> &lt;<span class="keyword">typename</span> T, <span class="keyword">typename</span> U, <span class="keyword">typename</span> Op, <span class="keywordtype">int</span> N_READS = REDUCE_N_READS&gt;</div>
<div class="foldopen" id="foldopen00061" data-start="{" data-end="}">
<div class="line"><a id="l00061" name="l00061"></a><span class="lineno"><a class="line" href="reduce__all_8h.html#a8d143d2fea044e4808b69cc3723e33c7"> 61</a></span>[[kernel]] <span class="keywordtype">void</span> <a class="code hl_function" href="reduce__all_8h.html#a8d143d2fea044e4808b69cc3723e33c7">all_reduce</a>(</div>
<div class="line"><a id="l00062" name="l00062"></a><span class="lineno"> 62</span> <span class="keyword">const</span> device T* in [[buffer(0)]],</div>
<div class="line"><a id="l00063" name="l00063"></a><span class="lineno"> 63</span> device <a class="code hl_struct" href="structmlx__atomic.html">mlx_atomic&lt;U&gt;</a>* out [[buffer(1)]],</div>
<div class="line"><a id="l00064" name="l00064"></a><span class="lineno"> 64</span> <span class="keyword">const</span> device <span class="keywordtype">size_t</span>&amp; in_size [[buffer(2)]],</div>
<div class="line"><a id="l00065" name="l00065"></a><span class="lineno"> 65</span> uint gid [[thread_position_in_grid]],</div>
<div class="line"><a id="l00066" name="l00066"></a><span class="lineno"> 66</span> uint lid [[thread_position_in_threadgroup]],</div>
<div class="line"><a id="l00067" name="l00067"></a><span class="lineno"> 67</span> uint grid_size [[threads_per_grid]],</div>
<div class="line"><a id="l00068" name="l00068"></a><span class="lineno"> 68</span> uint simd_per_group [[simdgroups_per_threadgroup]],</div>
<div class="line"><a id="l00069" name="l00069"></a><span class="lineno"> 69</span> uint simd_lane_id [[thread_index_in_simdgroup]],</div>
<div class="line"><a id="l00070" name="l00070"></a><span class="lineno"> 70</span> uint simd_group_id [[simdgroup_index_in_threadgroup]]) {</div>
<div class="line"><a id="l00071" name="l00071"></a><span class="lineno"> 71</span> Op <a class="code hl_variable" href="common_2binary_8h.html#a70228731d29946574b238d21fb4b360c">op</a>;</div>
<div class="line"><a id="l00072" name="l00072"></a><span class="lineno"> 72</span> threadgroup U local_vals[<a class="code hl_variable" href="backend_2metal_2kernels_2reduction_2ops_8h.html#a515b75d563a93d3c09ee677948dc83e3">simd_size</a>];</div>
<div class="line"><a id="l00073" name="l00073"></a><span class="lineno"> 73</span> </div>
<div class="line"><a id="l00074" name="l00074"></a><span class="lineno"> 74</span> U total_val =</div>
<div class="line"><a id="l00075" name="l00075"></a><span class="lineno"> 75</span> per_thread_all_reduce&lt;T, U, Op, N_READS&gt;(in, in_size, gid, grid_size);</div>
<div class="line"><a id="l00076" name="l00076"></a><span class="lineno"> 76</span> </div>
<div class="line"><a id="l00077" name="l00077"></a><span class="lineno"> 77</span> <span class="comment">// Reduction within simd group</span></div>
<div class="line"><a id="l00078" name="l00078"></a><span class="lineno"> 78</span> total_val = <a class="code hl_variable" href="common_2binary_8h.html#a70228731d29946574b238d21fb4b360c">op</a>.simd_reduce(total_val);</div>
<div class="line"><a id="l00079" name="l00079"></a><span class="lineno"> 79</span> <span class="keywordflow">if</span> (simd_lane_id == 0) {</div>
<div class="line"><a id="l00080" name="l00080"></a><span class="lineno"> 80</span> local_vals[simd_group_id] = total_val;</div>
<div class="line"><a id="l00081" name="l00081"></a><span class="lineno"> 81</span> }</div>
<div class="line"><a id="l00082" name="l00082"></a><span class="lineno"> 82</span> </div>
<div class="line"><a id="l00083" name="l00083"></a><span class="lineno"> 83</span> <span class="comment">// Reduction within thread group</span></div>
<div class="line"><a id="l00084" name="l00084"></a><span class="lineno"> 84</span> threadgroup_barrier(mem_flags::mem_threadgroup);</div>
<div class="line"><a id="l00085" name="l00085"></a><span class="lineno"> 85</span> total_val = lid &lt; simd_per_group ? local_vals[lid] : <a class="code hl_variable" href="common_2binary_8h.html#a70228731d29946574b238d21fb4b360c">op</a>.init;</div>
<div class="line"><a id="l00086" name="l00086"></a><span class="lineno"> 86</span> total_val = <a class="code hl_variable" href="common_2binary_8h.html#a70228731d29946574b238d21fb4b360c">op</a>.simd_reduce(total_val);</div>
<div class="line"><a id="l00087" name="l00087"></a><span class="lineno"> 87</span> </div>
<div class="line"><a id="l00088" name="l00088"></a><span class="lineno"> 88</span> <span class="comment">// Reduction across threadgroups</span></div>
<div class="line"><a id="l00089" name="l00089"></a><span class="lineno"> 89</span> <span class="keywordflow">if</span> (lid == 0) {</div>
<div class="line"><a id="l00090" name="l00090"></a><span class="lineno"> 90</span> <a class="code hl_variable" href="common_2binary_8h.html#a70228731d29946574b238d21fb4b360c">op</a>.atomic_update(out, total_val);</div>
<div class="line"><a id="l00091" name="l00091"></a><span class="lineno"> 91</span> }</div>
<div class="line"><a id="l00092" name="l00092"></a><span class="lineno"> 92</span>}</div>
</div>
<div class="line"><a id="l00093" name="l00093"></a><span class="lineno"> 93</span> </div>
<div class="line"><a id="l00094" name="l00094"></a><span class="lineno"> 94</span><span class="keyword">template</span> &lt;<span class="keyword">typename</span> T, <span class="keyword">typename</span> U, <span class="keyword">typename</span> Op, <span class="keywordtype">int</span> N_READS = REDUCE_N_READS&gt;</div>
<div class="foldopen" id="foldopen00095" data-start="{" data-end="}">
<div class="line"><a id="l00095" name="l00095"></a><span class="lineno"><a class="line" href="reduce__all_8h.html#a09d140566cb3e6418a821be1e42db282"> 95</a></span>[[kernel]] <span class="keywordtype">void</span> <a class="code hl_function" href="reduce__all_8h.html#a09d140566cb3e6418a821be1e42db282">all_reduce_no_atomics</a>(</div>
<div class="line"><a id="l00096" name="l00096"></a><span class="lineno"> 96</span> <span class="keyword">const</span> device T* in [[buffer(0)]],</div>
<div class="line"><a id="l00097" name="l00097"></a><span class="lineno"> 97</span> device U* out [[buffer(1)]],</div>
<div class="line"><a id="l00098" name="l00098"></a><span class="lineno"> 98</span> <span class="keyword">const</span> device <span class="keywordtype">size_t</span>&amp; in_size [[buffer(2)]],</div>
<div class="line"><a id="l00099" name="l00099"></a><span class="lineno"> 99</span> uint gid [[thread_position_in_grid]],</div>
<div class="line"><a id="l00100" name="l00100"></a><span class="lineno"> 100</span> uint lid [[thread_position_in_threadgroup]],</div>
<div class="line"><a id="l00101" name="l00101"></a><span class="lineno"> 101</span> uint grid_size [[threads_per_grid]],</div>
<div class="line"><a id="l00102" name="l00102"></a><span class="lineno"> 102</span> uint simd_per_group [[simdgroups_per_threadgroup]],</div>
<div class="line"><a id="l00103" name="l00103"></a><span class="lineno"> 103</span> uint simd_lane_id [[thread_index_in_simdgroup]],</div>
<div class="line"><a id="l00104" name="l00104"></a><span class="lineno"> 104</span> uint simd_group_id [[simdgroup_index_in_threadgroup]],</div>
<div class="line"><a id="l00105" name="l00105"></a><span class="lineno"> 105</span> uint thread_group_id [[threadgroup_position_in_grid]]) {</div>
<div class="line"><a id="l00106" name="l00106"></a><span class="lineno"> 106</span> Op <a class="code hl_variable" href="common_2binary_8h.html#a70228731d29946574b238d21fb4b360c">op</a>;</div>
<div class="line"><a id="l00107" name="l00107"></a><span class="lineno"> 107</span> threadgroup U local_vals[<a class="code hl_variable" href="backend_2metal_2kernels_2reduction_2ops_8h.html#a515b75d563a93d3c09ee677948dc83e3">simd_size</a>];</div>
<div class="line"><a id="l00108" name="l00108"></a><span class="lineno"> 108</span> </div>
<div class="line"><a id="l00109" name="l00109"></a><span class="lineno"> 109</span> U total_val =</div>
<div class="line"><a id="l00110" name="l00110"></a><span class="lineno"> 110</span> per_thread_all_reduce&lt;T, U, Op, N_READS&gt;(in, in_size, gid, grid_size);</div>
<div class="line"><a id="l00111" name="l00111"></a><span class="lineno"> 111</span> </div>
<div class="line"><a id="l00112" name="l00112"></a><span class="lineno"> 112</span> <span class="comment">// Reduction within simd group (simd_add isn&#39;t supported for uint64/int64</span></div>
<div class="line"><a id="l00113" name="l00113"></a><span class="lineno"> 113</span> <span class="comment">// types)</span></div>
<div class="line"><a id="l00114" name="l00114"></a><span class="lineno"> 114</span> <span class="keywordflow">for</span> (uint16_t lane_offset = <a class="code hl_variable" href="backend_2metal_2kernels_2reduction_2ops_8h.html#a515b75d563a93d3c09ee677948dc83e3">simd_size</a> / 2; lane_offset &gt; 0;</div>
<div class="line"><a id="l00115" name="l00115"></a><span class="lineno"> 115</span> lane_offset /= 2) {</div>
<div class="line"><a id="l00116" name="l00116"></a><span class="lineno"> 116</span> total_val = <a class="code hl_variable" href="common_2binary_8h.html#a70228731d29946574b238d21fb4b360c">op</a>(total_val, <a class="code hl_function" href="backend_2metal_2kernels_2utils_8h.html#aba6279624b1d30c525efee856a222b5c">simd_shuffle_down</a>(total_val, lane_offset));</div>
<div class="line"><a id="l00117" name="l00117"></a><span class="lineno"> 117</span> }</div>
<div class="line"><a id="l00118" name="l00118"></a><span class="lineno"> 118</span> <span class="comment">// Write simd group reduction results to local memory</span></div>
<div class="line"><a id="l00119" name="l00119"></a><span class="lineno"> 119</span> <span class="keywordflow">if</span> (simd_lane_id == 0) {</div>
<div class="line"><a id="l00120" name="l00120"></a><span class="lineno"> 120</span> local_vals[simd_group_id] = total_val;</div>
<div class="line"><a id="l00121" name="l00121"></a><span class="lineno"> 121</span> }</div>
<div class="line"><a id="l00122" name="l00122"></a><span class="lineno"> 122</span> threadgroup_barrier(mem_flags::mem_threadgroup);</div>
<div class="line"><a id="l00123" name="l00123"></a><span class="lineno"> 123</span> </div>
<div class="line"><a id="l00124" name="l00124"></a><span class="lineno"> 124</span> <span class="comment">// Reduction of simdgroup reduction results within threadgroup.</span></div>
<div class="line"><a id="l00125" name="l00125"></a><span class="lineno"> 125</span> total_val = lid &lt; simd_per_group ? local_vals[lid] : <a class="code hl_variable" href="common_2binary_8h.html#a70228731d29946574b238d21fb4b360c">op</a>.init;</div>
<div class="line"><a id="l00126" name="l00126"></a><span class="lineno"> 126</span> <span class="keywordflow">for</span> (uint16_t lane_offset = <a class="code hl_variable" href="backend_2metal_2kernels_2reduction_2ops_8h.html#a515b75d563a93d3c09ee677948dc83e3">simd_size</a> / 2; lane_offset &gt; 0;</div>
<div class="line"><a id="l00127" name="l00127"></a><span class="lineno"> 127</span> lane_offset /= 2) {</div>
<div class="line"><a id="l00128" name="l00128"></a><span class="lineno"> 128</span> total_val = <a class="code hl_variable" href="common_2binary_8h.html#a70228731d29946574b238d21fb4b360c">op</a>(total_val, <a class="code hl_function" href="backend_2metal_2kernels_2utils_8h.html#aba6279624b1d30c525efee856a222b5c">simd_shuffle_down</a>(total_val, lane_offset));</div>
<div class="line"><a id="l00129" name="l00129"></a><span class="lineno"> 129</span> }</div>
<div class="line"><a id="l00130" name="l00130"></a><span class="lineno"> 130</span> </div>
<div class="line"><a id="l00131" name="l00131"></a><span class="lineno"> 131</span> <span class="comment">// Reduction across threadgroups</span></div>
<div class="line"><a id="l00132" name="l00132"></a><span class="lineno"> 132</span> <span class="keywordflow">if</span> (lid == 0) {</div>
<div class="line"><a id="l00133" name="l00133"></a><span class="lineno"> 133</span> out[thread_group_id] = total_val;</div>
<div class="line"><a id="l00134" name="l00134"></a><span class="lineno"> 134</span> }</div>
<div class="line"><a id="l00135" name="l00135"></a><span class="lineno"> 135</span>}</div>
</div>
<div class="ttc" id="abackend_2metal_2kernels_2reduction_2ops_8h_html_a515b75d563a93d3c09ee677948dc83e3"><div class="ttname"><a href="backend_2metal_2kernels_2reduction_2ops_8h.html#a515b75d563a93d3c09ee677948dc83e3">simd_size</a></div><div class="ttdeci">static constant constexpr const uint8_t simd_size</div><div class="ttdef"><b>Definition</b> ops.h:8</div></div>
<div class="ttc" id="abackend_2metal_2kernels_2utils_8h_html_a51c19db777f43943e4b35f25dd88d49d"><div class="ttname"><a href="backend_2metal_2kernels_2utils_8h.html#a51c19db777f43943e4b35f25dd88d49d">ceildiv</a></div><div class="ttdeci">size_t ceildiv(size_t N, size_t M)</div><div class="ttdoc">Compute ceil((float)N/(float)M)</div><div class="ttdef"><b>Definition</b> utils.h:296</div></div>
<div class="ttc" id="abackend_2metal_2kernels_2utils_8h_html_aba6279624b1d30c525efee856a222b5c"><div class="ttname"><a href="backend_2metal_2kernels_2utils_8h.html#aba6279624b1d30c525efee856a222b5c">simd_shuffle_down</a></div><div class="ttdeci">uint64_t simd_shuffle_down(uint64_t data, uint16_t delta)</div><div class="ttdef"><b>Definition</b> utils.h:329</div></div>
<div class="ttc" id="acommon_2binary_8h_html_a70228731d29946574b238d21fb4b360c"><div class="ttname"><a href="common_2binary_8h.html#a70228731d29946574b238d21fb4b360c">op</a></div><div class="ttdeci">Op op</div><div class="ttdef"><b>Definition</b> binary.h:141</div></div>
<div class="ttc" id="areduce__all_8h_html_a0895975942131768f280decc854d0ac1"><div class="ttname"><a href="reduce__all_8h.html#a0895975942131768f280decc854d0ac1">per_thread_all_reduce</a></div><div class="ttdeci">METAL_FUNC U per_thread_all_reduce(const device T *in, const device size_t &amp;in_size, uint gid, uint grid_size)</div><div class="ttdef"><b>Definition</b> reduce_all.h:8</div></div>
<div class="ttc" id="areduce__all_8h_html_a09d140566cb3e6418a821be1e42db282"><div class="ttname"><a href="reduce__all_8h.html#a09d140566cb3e6418a821be1e42db282">all_reduce_no_atomics</a></div><div class="ttdeci">void all_reduce_no_atomics(const device T *in, device U *out, const device size_t &amp;in_size, uint gid, uint lid, uint grid_size, uint simd_per_group, uint simd_lane_id, uint simd_group_id, uint thread_group_id)</div><div class="ttdef"><b>Definition</b> reduce_all.h:95</div></div>
<div class="ttc" id="areduce__all_8h_html_a8d143d2fea044e4808b69cc3723e33c7"><div class="ttname"><a href="reduce__all_8h.html#a8d143d2fea044e4808b69cc3723e33c7">all_reduce</a></div><div class="ttdeci">void all_reduce(const device T *in, device mlx_atomic&lt; U &gt; *out, const device size_t &amp;in_size, uint gid, uint lid, uint grid_size, uint simd_per_group, uint simd_lane_id, uint simd_group_id)</div><div class="ttdef"><b>Definition</b> reduce_all.h:61</div></div>
<div class="ttc" id="astructmlx__atomic_html"><div class="ttname"><a href="structmlx__atomic.html">mlx_atomic</a></div><div class="ttdef"><b>Definition</b> atomic.h:25</div></div>
</div><!-- fragment --></div><!-- contents -->
<!-- start footer part -->
<hr class="footer"/><address class="footer"><small>
Generated by&#160;<a href="https://www.doxygen.org/index.html"><img class="footer" src="doxygen.svg" width="104" height="31" alt="doxygen"/></a> 1.10.0
</small></address>
</body>
</html>
Reference in New Issue View Git Blame Copy Permalink