tristan/ZealOS
1
0
Fork 0
mirror of https://github.com/Zeal-Operating-System/ZealOS.git synced 2024-12-28 16:26:35 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions
ZealOS/docs/Demo/MultiCore/MPRadix.CC.html
TomAwezome 293db76ddb Update documentation.
2021-10-27 03:28:16 -04:00

189 lines
15 KiB
HTML
Executable file
Raw Blame History

<!DOCTYPE HTML>
<html>
<head>
<meta http-equiv="Content-Type" content="text/html;charset=US-ASCII">
<meta name="generator" content="ZealOS V1.04">
<style type="text/css">
body {background-color:#1f1f1f;}
.cF0{color:#e3e3e3;background-color:#1f1f1f;}
.cF1{color:#4f84a6;background-color:#1f1f1f;}
.cF2{color:#73a255;background-color:#1f1f1f;}
.cF3{color:#297582;background-color:#1f1f1f;}
.cF4{color:#b34f4b;background-color:#1f1f1f;}
.cF5{color:#8a52c3;background-color:#1f1f1f;}
.cF6{color:#b7822f;background-color:#1f1f1f;}
.cF7{color:#444444;background-color:#1f1f1f;}
.cF8{color:#6d6d6d;background-color:#1f1f1f;}
.cF9{color:#94bfde;background-color:#1f1f1f;}
.cFA{color:#a1ce97;background-color:#1f1f1f;}
.cFB{color:#6db4be;background-color:#1f1f1f;}
.cFC{color:#e88e88;background-color:#1f1f1f;}
.cFD{color:#ca94e8;background-color:#1f1f1f;}
.cFE{color:#d4b475;background-color:#1f1f1f;}
.cFF{color:#1f1f1f;background-color:#1f1f1f;}
</style>
</head>
<body>
<pre style="font-family:monospace;font-size:12pt">
<a name="l1"></a><span class=cF2>/*</span><span class=cF0>
<a name="l2"></a></span><span class=cF2>On an 8-core machine, this takes the top 3-bits</span><span class=cF0>
<a name="l3"></a></span><span class=cF2>of random numbers and distributes them to the 8 cores</span><span class=cF0>
<a name="l4"></a></span><span class=cF2>for sorting.</span><span class=cF0> </span><span class=cF2>Then, it merge sorts them.</span><span class=cF0>
<a name="l5"></a></span><span class=cF2>*/</span><span class=cF0>
<a name="l6"></a>
<a name="l7"></a>#</span><span class=cF1>define</span><span class=cF0> NUM </span><span class=cFE>1000000</span><span class=cF0>
<a name="l8"></a>
<a name="l9"></a></span><span class=cF9>I64</span><span class=cF0> my_mp_count = </span><span class=cFE>1</span><span class=cF0> &lt;&lt; </span><span class=cF5>Bsr</span><span class=cF0>(</span><span class=cFB>mp_count</span><span class=cF0>);</span><span class=cF2>//Power of 2</span><span class=cF0>
<a name="l10"></a>
<a name="l11"></a></span><span class=cF9>I32</span><span class=cF0> *arg1, *arg2;
<a name="l12"></a></span><span class=cF9>I32</span><span class=cF0> *b[my_mp_count], bn[my_mp_count];
<a name="l13"></a></span><span class=cF9>I64</span><span class=cF0> mp_not_done_flags;
<a name="l14"></a>
<a name="l15"></a></span><span class=cF9>I64</span><span class=cF0> Compare(</span><span class=cF9>I32</span><span class=cF0> *e1, </span><span class=cF9>I32</span><span class=cF0> *e2)
<a name="l16"></a>{
<a name="l17"></a> </span><span class=cF1>return</span><span class=cF0> *e1 - *e2;
<a name="l18"></a>}
<a name="l19"></a>
<a name="l20"></a></span><span class=cF1>U0</span><span class=cF0> QuickSortU32(</span><span class=cF9>I32</span><span class=cF0> *base, </span><span class=cF9>I64</span><span class=cF0> num)
<a name="l21"></a>{</span><span class=cF2>//By customizing, we dramatically improve it!</span><span class=cF0>
<a name="l22"></a></span><span class=cF2>//Cut and paste from </span><a href="https://zeal-operating-system.github.io/ZealOS/Kernel/QuickSort.CC.html#l1"><span class=cF4>QuickSortI64</span></a><span class=cF2>().</span><span class=cF0>
<a name="l23"></a> </span><span class=cF9>I64</span><span class=cF0> i;
<a name="l24"></a> </span><span class=cF9>I32</span><span class=cF0> *less, *greater, pivot;
<a name="l25"></a>
<a name="l26"></a> </span><span class=cF1>if</span><span class=cF0> (num &gt; </span><span class=cFE>1</span><span class=cF0>)
<a name="l27"></a> </span><span class=cF7>{</span><span class=cF0>
<a name="l28"></a> </span><span class=cF1>do</span><span class=cF0>
<a name="l29"></a> {
<a name="l30"></a> less = base;
<a name="l31"></a> greater = base + num;
<a name="l32"></a> pivot = base[num / </span><span class=cFE>2</span><span class=cF0>];
<a name="l33"></a> </span><span class=cF1>while</span><span class=cF0> (less &lt; greater)
<a name="l34"></a> </span><span class=cF7>{</span><span class=cF0>
<a name="l35"></a> </span><span class=cF1>if</span><span class=cF0> (*less &lt;= pivot)
<a name="l36"></a> less++;
<a name="l37"></a> </span><span class=cF1>else</span><span class=cF0>
<a name="l38"></a> {
<a name="l39"></a> greater--;
<a name="l40"></a> </span><span class=cF5>SwapU32</span><span class=cF0>(less, greater);
<a name="l41"></a> }
<a name="l42"></a> </span><span class=cF7>}</span><span class=cF0>
<a name="l43"></a> i = less - base;
<a name="l44"></a> </span><span class=cF1>if</span><span class=cF0> (i == num)
<a name="l45"></a> </span><span class=cF7>{</span><span class=cF2>//All less or equ to pivot</span><span class=cF0>
<a name="l46"></a>
<a name="l47"></a> </span><span class=cF2>//Point greater to first less</span><span class=cF0>
<a name="l48"></a> </span><span class=cF1>do</span><span class=cF0> greater--;
<a name="l49"></a> </span><span class=cF1>while</span><span class=cF0> (--i &amp;&amp; *greater == pivot);
<a name="l50"></a>
<a name="l51"></a> </span><span class=cF1>if</span><span class=cF0> (i)
<a name="l52"></a> {
<a name="l53"></a> less = base +num / </span><span class=cFE>2</span><span class=cF0>; </span><span class=cF2>//Pivot was not moved, point to it</span><span class=cF0>
<a name="l54"></a> </span><span class=cF1>if</span><span class=cF0> (less &lt; greater)
<a name="l55"></a> </span><span class=cF5>SwapU32</span><span class=cF0>(less, greater);
<a name="l56"></a> num = i;
<a name="l57"></a> }
<a name="l58"></a> </span><span class=cF1>else</span><span class=cF0> </span><span class=cF2>//All equ</span><span class=cF0>
<a name="l59"></a> </span><span class=cF1>break</span><span class=cF0>;
<a name="l60"></a> </span><span class=cF7>}</span><span class=cF0>
<a name="l61"></a> </span><span class=cF1>else</span><span class=cF0> </span><span class=cF1>if</span><span class=cF0> (i &lt; num / </span><span class=cFE>2</span><span class=cF0>)
<a name="l62"></a> </span><span class=cF7>{</span><span class=cF0>
<a name="l63"></a> QuickSortU32(base, i);
<a name="l64"></a> num -= i;
<a name="l65"></a> base = greater;
<a name="l66"></a> </span><span class=cF7>}</span><span class=cF0>
<a name="l67"></a> </span><span class=cF1>else</span><span class=cF0>
<a name="l68"></a> </span><span class=cF7>{</span><span class=cF0>
<a name="l69"></a> QuickSortU32(greater, num - i);
<a name="l70"></a> num = i;
<a name="l71"></a> </span><span class=cF7>}</span><span class=cF0>
<a name="l72"></a> }
<a name="l73"></a> </span><span class=cF1>while</span><span class=cF0> (num &gt; </span><span class=cFE>1</span><span class=cF0>);
<a name="l74"></a> </span><span class=cF7>}</span><span class=cF0>
<a name="l75"></a>}
<a name="l76"></a>
<a name="l77"></a></span><span class=cF1>U0</span><span class=cF0> MPSort(</span><span class=cF9>I64</span><span class=cF0> dummy=</span><span class=cFE>0</span><span class=cF0>)
<a name="l78"></a>{
<a name="l79"></a> </span><span class=cF1>no_warn</span><span class=cF0> dummy;
<a name="l80"></a> QuickSortU32(b[</span><span class=cF5>Gs</span><span class=cF0>-&gt;num], bn[</span><span class=cF5>Gs</span><span class=cF0>-&gt;num]);
<a name="l81"></a> </span><span class=cF5>LBtr</span><span class=cF0>(&amp;mp_not_done_flags, </span><span class=cF5>Gs</span><span class=cF0>-&gt;num);
<a name="l82"></a>}
<a name="l83"></a>
<a name="l84"></a></span><span class=cF1>U0</span><span class=cF0> MPRadixSortDemo(</span><span class=cF9>I64</span><span class=cF0> dummy=</span><span class=cFE>0</span><span class=cF0>)
<a name="l85"></a>{
<a name="l86"></a> </span><span class=cF1>no_warn</span><span class=cF0> dummy;
<a name="l87"></a> </span><span class=cF9>I64</span><span class=cF0> i, j, k1, k2;
<a name="l88"></a> </span><span class=cF1>F64</span><span class=cF0> t0;
<a name="l89"></a>
<a name="l90"></a> arg1 = </span><span class=cF5>MAlloc</span><span class=cF0>(NUM * </span><span class=cF1>sizeof</span><span class=cF7>(</span><span class=cF9>I32</span><span class=cF7>)</span><span class=cF0>);
<a name="l91"></a> </span><span class=cF1>for</span><span class=cF0> (i = </span><span class=cFE>0</span><span class=cF0>; i &lt; NUM; i++)
<a name="l92"></a> arg1[i] = </span><span class=cF5>RandI32</span><span class=cF0>;
<a name="l93"></a>
<a name="l94"></a> arg2 = </span><span class=cF5>MAlloc</span><span class=cF0>(NUM * </span><span class=cF1>sizeof</span><span class=cF7>(</span><span class=cF9>I32</span><span class=cF7>)</span><span class=cF0>);
<a name="l95"></a>
<a name="l96"></a> </span><span class=cF6>&quot;$GREEN$QuickSort$FG$\n&quot;</span><span class=cF0>;
<a name="l97"></a> t0 = </span><span class=cF5>tS</span><span class=cF0>;
<a name="l98"></a> </span><span class=cF5>MemCopy</span><span class=cF0>(arg2, arg1, </span><span class=cF1>sizeof</span><span class=cF7>(</span><span class=cF9>I32</span><span class=cF7>)</span><span class=cF0> * NUM);
<a name="l99"></a> </span><span class=cF5>QuickSort</span><span class=cF0>(arg2, NUM, </span><span class=cF1>sizeof</span><span class=cF7>(</span><span class=cF9>I32</span><span class=cF7>)</span><span class=cF0>, &amp;Compare);
<a name="l100"></a> </span><span class=cF6>&quot;Time:%9.6f\n&quot;</span><span class=cF0>, </span><span class=cF5>tS</span><span class=cF0> - t0;
<a name="l101"></a> </span><span class=cF5>D</span><span class=cF0>(arg2 + NUM / </span><span class=cFE>4</span><span class=cF0>);
<a name="l102"></a>
<a name="l103"></a> </span><span class=cF6>&quot;$GREEN$QuickSortU32$FG$\n&quot;</span><span class=cF0>;
<a name="l104"></a> t0 = </span><span class=cF5>tS</span><span class=cF0>;
<a name="l105"></a> </span><span class=cF5>MemCopy</span><span class=cF0>(arg2, arg1, </span><span class=cF1>sizeof</span><span class=cF7>(</span><span class=cF9>I32</span><span class=cF7>)</span><span class=cF0>*NUM);
<a name="l106"></a> QuickSortU32(arg2, NUM);
<a name="l107"></a> </span><span class=cF6>&quot;Time:%9.6f\n&quot;</span><span class=cF0>, </span><span class=cF5>tS</span><span class=cF0> - t0;
<a name="l108"></a> </span><span class=cF5>D</span><span class=cF0>(arg2 + NUM / </span><span class=cFE>4</span><span class=cF0>);
<a name="l109"></a>
<a name="l110"></a> </span><span class=cF1>for</span><span class=cF0> (i = </span><span class=cFE>0</span><span class=cF0>; i &lt; my_mp_count; i++)
<a name="l111"></a> </span><span class=cF7>{</span><span class=cF0>
<a name="l112"></a> </span><span class=cF2>//We must do full size, just in case.</span><span class=cF0>
<a name="l113"></a> </span><span class=cF2>//There will be uneven split between cores</span><span class=cF0>
<a name="l114"></a> </span><span class=cF2>//depending on the distribution of rand numbers.</span><span class=cF0>
<a name="l115"></a> b[i] = </span><span class=cF5>MAlloc</span><span class=cF0>(NUM * </span><span class=cF1>sizeof</span><span class=cF7>(</span><span class=cF9>I32</span><span class=cF7>)</span><span class=cF0>);
<a name="l116"></a> bn[i] = </span><span class=cFE>0</span><span class=cF0>;
<a name="l117"></a> </span><span class=cF7>}</span><span class=cF0>
<a name="l118"></a>
<a name="l119"></a> </span><span class=cF1>if</span><span class=cF0> (my_mp_count &lt; </span><span class=cFE>2</span><span class=cF0>)
<a name="l120"></a> </span><span class=cF5>throw</span><span class=cF0>(</span><span class=cF6>'MultCore'</span><span class=cF0>);
<a name="l121"></a>
<a name="l122"></a> </span><span class=cF6>&quot;$GREEN$MP Radix QuickSortU32$FG$\n&quot;</span><span class=cF0>;
<a name="l123"></a> t0 = </span><span class=cF5>tS</span><span class=cF0>;
<a name="l124"></a> k1 = </span><span class=cFE>32</span><span class=cF0> - </span><span class=cF5>Bsr</span><span class=cF0>(my_mp_count);
<a name="l125"></a> k2 = my_mp_count / </span><span class=cFE>2</span><span class=cF0>;
<a name="l126"></a> </span><span class=cF1>for</span><span class=cF0> (i = </span><span class=cFE>0</span><span class=cF0>; i &lt; NUM; i++)
<a name="l127"></a> </span><span class=cF7>{</span><span class=cF0>
<a name="l128"></a> j = arg1[i] &gt;&gt; k1 + k2; </span><span class=cF2>//This is a preliminary radix sort.</span><span class=cF0>
<a name="l129"></a> b[j][bn[j]++] = arg1[i];
<a name="l130"></a> </span><span class=cF7>}</span><span class=cF0>
<a name="l131"></a> mp_not_done_flags = </span><span class=cFE>1</span><span class=cF0> &lt;&lt; my_mp_count - </span><span class=cFE>1</span><span class=cF0>;
<a name="l132"></a> </span><span class=cF1>for</span><span class=cF0> (i = </span><span class=cFE>0</span><span class=cF0>; i &lt; my_mp_count; i++)
<a name="l133"></a> </span><span class=cF5>Spawn</span><span class=cF0>(&amp;MPSort, </span><span class=cF3>NULL</span><span class=cF0>, </span><span class=cF3>NULL</span><span class=cF0>, i);
<a name="l134"></a> </span><span class=cF1>while</span><span class=cF0> (mp_not_done_flags)
<a name="l135"></a> </span><span class=cF5>Yield</span><span class=cF0>;
<a name="l136"></a> j = </span><span class=cFE>0</span><span class=cF0>;
<a name="l137"></a> </span><span class=cF1>for</span><span class=cF0> (i = </span><span class=cFE>0</span><span class=cF0>; i &lt; my_mp_count; i++)
<a name="l138"></a> </span><span class=cF7>{</span><span class=cF0>
<a name="l139"></a> </span><span class=cF5>MemCopy</span><span class=cF0>(&amp;arg2[j], b[i], bn[i] * </span><span class=cF1>sizeof</span><span class=cF7>(</span><span class=cF9>I32</span><span class=cF7>)</span><span class=cF0>);
<a name="l140"></a> j += bn[i];
<a name="l141"></a> </span><span class=cF7>}</span><span class=cF0>
<a name="l142"></a> </span><span class=cF6>&quot;Time:%9.6f\n&quot;</span><span class=cF0>, </span><span class=cF5>tS</span><span class=cF0> - t0;
<a name="l143"></a> </span><span class=cF5>D</span><span class=cF0>(arg2 + NUM / </span><span class=cFE>4</span><span class=cF0>);
<a name="l144"></a>
<a name="l145"></a> </span><span class=cF5>Free</span><span class=cF0>(arg1);
<a name="l146"></a> </span><span class=cF5>Free</span><span class=cF0>(arg2);
<a name="l147"></a> </span><span class=cF1>for</span><span class=cF0> (i = </span><span class=cFE>0</span><span class=cF0>; i &lt; my_mp_count; i++)
<a name="l148"></a> </span><span class=cF5>Free</span><span class=cF0>(b[i]);
<a name="l149"></a>}
<a name="l150"></a>
<a name="l151"></a>MPRadixSortDemo;
<a name="l152"></a>
<a name="l153"></a></span><span class=cF2>/*</span><span class=cF0> Results on 8 Cores 3.397GHz Core i7:
<a name="l154"></a></span><span class=cF2>QuickSort</span><span class=cF0>
<a name="l155"></a>Time: 0.759998
<a name="l156"></a></span><span class=cF2>QuickSortU32</span><span class=cF0>
<a name="l157"></a>Time: 0.093684
<a name="l158"></a></span><span class=cF2>MP Radix QuickSortU32</span><span class=cF0>
<a name="l159"></a>Time: 0.045450
<a name="l160"></a></span><span class=cF2>*/</span><span class=cF0>
</span></pre></body>
</html>
Reference in a new issue View git blame Copy permalink