snapraid/cov/raid/x86z.c.gcov.html
Andrea Mazzoleni e01ca1e0fa Auto update
2016-11-21 19:48:21 +01:00

331 lines
29 KiB
HTML

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html lang="en">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
<title>LCOV - lcov.info - raid/x86z.c</title>
<link rel="stylesheet" type="text/css" href="../gcov.css">
</head>
<body>
<table width="100%" border=0 cellspacing=0 cellpadding=0>
<tr><td class="title">LCOV - code coverage report</td></tr>
<tr><td class="ruler"><img src="../glass.png" width=3 height=3 alt=""></td></tr>
<tr>
<td width="100%">
<table cellpadding=1 border=0 width="100%">
<tr>
<td width="10%" class="headerItem">Current view:</td>
<td width="35%" class="headerValue"><a href="../index.html">top level</a> - <a href="index.html">raid</a> - x86z.c<span style="font-size: 80%;"> (source / <a href="x86z.c.func.html">functions</a>)</span></td>
<td width="5%"></td>
<td width="15%"></td>
<td width="10%" class="headerCovTableHead">Hit</td>
<td width="10%" class="headerCovTableHead">Total</td>
<td width="15%" class="headerCovTableHead">Coverage</td>
</tr>
<tr>
<td class="headerItem">Test:</td>
<td class="headerValue">lcov.info</td>
<td></td>
<td class="headerItem">Lines:</td>
<td class="headerCovTableEntry">35</td>
<td class="headerCovTableEntry">35</td>
<td class="headerCovTableEntryHi">100.0 %</td>
</tr>
<tr>
<td class="headerItem">Date:</td>
<td class="headerValue">2016-11-21</td>
<td></td>
<td class="headerItem">Functions:</td>
<td class="headerCovTableEntry">1</td>
<td class="headerCovTableEntry">1</td>
<td class="headerCovTableEntryHi">100.0 %</td>
</tr>
<tr><td><img src="../glass.png" width=3 height=3 alt=""></td></tr>
</table>
</td>
</tr>
<tr><td class="ruler"><img src="../glass.png" width=3 height=3 alt=""></td></tr>
</table>
<table cellpadding=0 cellspacing=0 border=0>
<tr>
<td><br></td>
</tr>
<tr>
<td>
<pre class="sourceHeading"> Line data Source code</pre>
<pre class="source">
<a name="1"><span class="lineNum"> 1 </span> : /*</a>
<span class="lineNum"> 2 </span> : * Copyright (C) 2013 Andrea Mazzoleni
<span class="lineNum"> 3 </span> : *
<span class="lineNum"> 4 </span> : * This program is free software: you can redistribute it and/or modify
<span class="lineNum"> 5 </span> : * it under the terms of the GNU General Public License as published by
<span class="lineNum"> 6 </span> : * the Free Software Foundation, either version 2 of the License, or
<span class="lineNum"> 7 </span> : * (at your option) any later version.
<span class="lineNum"> 8 </span> : *
<span class="lineNum"> 9 </span> : * This program is distributed in the hope that it will be useful,
<span class="lineNum"> 10 </span> : * but WITHOUT ANY WARRANTY; without even the implied warranty of
<span class="lineNum"> 11 </span> : * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
<span class="lineNum"> 12 </span> : * GNU General Public License for more details.
<span class="lineNum"> 13 </span> : */
<span class="lineNum"> 14 </span> :
<span class="lineNum"> 15 </span> : #include &quot;internal.h&quot;
<span class="lineNum"> 16 </span> :
<span class="lineNum"> 17 </span> : #if defined(CONFIG_X86) &amp;&amp; defined(CONFIG_SSE2)
<span class="lineNum"> 18 </span> : static const struct gfzconst16 {
<span class="lineNum"> 19 </span> : uint8_t poly[16];
<span class="lineNum"> 20 </span> : uint8_t half[16];
<span class="lineNum"> 21 </span> : uint8_t low7[16];
<span class="lineNum"> 22 </span> : } gfzconst16 __aligned(64) =
<span class="lineNum"> 23 </span> : {
<span class="lineNum"> 24 </span> : {
<span class="lineNum"> 25 </span> : 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d,
<span class="lineNum"> 26 </span> : 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d
<span class="lineNum"> 27 </span> : },
<span class="lineNum"> 28 </span> : {
<span class="lineNum"> 29 </span> : 0x8e, 0x8e, 0x8e, 0x8e, 0x8e, 0x8e, 0x8e, 0x8e,
<span class="lineNum"> 30 </span> : 0x8e, 0x8e, 0x8e, 0x8e, 0x8e, 0x8e, 0x8e, 0x8e
<span class="lineNum"> 31 </span> : },
<span class="lineNum"> 32 </span> : {
<span class="lineNum"> 33 </span> : 0x7f, 0x7f, 0x7f, 0x7f, 0x7f, 0x7f, 0x7f, 0x7f,
<span class="lineNum"> 34 </span> : 0x7f, 0x7f, 0x7f, 0x7f, 0x7f, 0x7f, 0x7f, 0x7f
<span class="lineNum"> 35 </span> : }
<span class="lineNum"> 36 </span> : };
<span class="lineNum"> 37 </span> : #endif
<span class="lineNum"> 38 </span> :
<span class="lineNum"> 39 </span> : #if defined(CONFIG_X86) &amp;&amp; defined(CONFIG_SSE2)
<span class="lineNum"> 40 </span> : /*
<a name="41"><span class="lineNum"> 41 </span> : * GENz (triple parity with powers of 2^-1) SSE2 implementation</a>
<span class="lineNum"> 42 </span> : */
<span class="lineNum"> 43 </span><span class="lineCov"> 1176 : void raid_genz_sse2(int nd, size_t size, void **vv)</span>
<span class="lineNum"> 44 </span> : {
<span class="lineNum"> 45 </span><span class="lineCov"> 1176 : uint8_t **v = (uint8_t**)vv;</span>
<span class="lineNum"> 46 </span> : uint8_t *p;
<span class="lineNum"> 47 </span> : uint8_t *q;
<span class="lineNum"> 48 </span> : uint8_t *r;
<span class="lineNum"> 49 </span> : int d, l;
<span class="lineNum"> 50 </span> : size_t i;
<span class="lineNum"> 51 </span> :
<span class="lineNum"> 52 </span><span class="lineCov"> 1176 : l = nd - 1;</span>
<span class="lineNum"> 53 </span><span class="lineCov"> 1176 : p = v[nd];</span>
<span class="lineNum"> 54 </span><span class="lineCov"> 1176 : q = v[nd + 1];</span>
<span class="lineNum"> 55 </span><span class="lineCov"> 1176 : r = v[nd + 2];</span>
<span class="lineNum"> 56 </span> :
<span class="lineNum"> 57 </span> : raid_sse_begin();
<span class="lineNum"> 58 </span> :
<span class="lineNum"> 59 </span><span class="lineCov"> 1176 : asm volatile (&quot;movdqa %0,%%xmm7&quot; : : &quot;m&quot; (gfzconst16.poly[0]));</span>
<span class="lineNum"> 60 </span><span class="lineCov"> 1176 : asm volatile (&quot;movdqa %0,%%xmm3&quot; : : &quot;m&quot; (gfzconst16.half[0]));</span>
<span class="lineNum"> 61 </span><span class="lineCov"> 1176 : asm volatile (&quot;movdqa %0,%%xmm6&quot; : : &quot;m&quot; (gfzconst16.low7[0]));</span>
<span class="lineNum"> 62 </span> :
<span class="lineNum"> 63 </span><span class="lineCov"> 1378536 : for (i = 0; i &lt; size; i += 16) {</span>
<span class="lineNum"> 64 </span><span class="lineCov"> 1377360 : asm volatile (&quot;movdqa %0,%%xmm0&quot; : : &quot;m&quot; (v[l][i]));</span>
<span class="lineNum"> 65 </span><span class="lineCov"> 1377360 : asm volatile (&quot;movdqa %xmm0,%xmm1&quot;);</span>
<span class="lineNum"> 66 </span><span class="lineCov"> 1377360 : asm volatile (&quot;movdqa %xmm0,%xmm2&quot;);</span>
<span class="lineNum"> 67 </span><span class="lineCov"> 11089152 : for (d = l - 1; d &gt;= 0; --d) {</span>
<span class="lineNum"> 68 </span><span class="lineCov"> 9711792 : asm volatile (&quot;pxor %xmm4,%xmm4&quot;);</span>
<span class="lineNum"> 69 </span><span class="lineCov"> 9711792 : asm volatile (&quot;pcmpgtb %xmm1,%xmm4&quot;);</span>
<span class="lineNum"> 70 </span><span class="lineCov"> 9711792 : asm volatile (&quot;paddb %xmm1,%xmm1&quot;);</span>
<span class="lineNum"> 71 </span><span class="lineCov"> 9711792 : asm volatile (&quot;pand %xmm7,%xmm4&quot;);</span>
<span class="lineNum"> 72 </span><span class="lineCov"> 9711792 : asm volatile (&quot;pxor %xmm4,%xmm1&quot;);</span>
<span class="lineNum"> 73 </span> :
<span class="lineNum"> 74 </span><span class="lineCov"> 9711792 : asm volatile (&quot;movdqa %xmm2,%xmm4&quot;);</span>
<span class="lineNum"> 75 </span><span class="lineCov"> 9711792 : asm volatile (&quot;pxor %xmm5,%xmm5&quot;);</span>
<span class="lineNum"> 76 </span><span class="lineCov"> 9711792 : asm volatile (&quot;psllw $7,%xmm4&quot;);</span>
<span class="lineNum"> 77 </span><span class="lineCov"> 9711792 : asm volatile (&quot;psrlw $1,%xmm2&quot;);</span>
<span class="lineNum"> 78 </span><span class="lineCov"> 9711792 : asm volatile (&quot;pcmpgtb %xmm4,%xmm5&quot;);</span>
<span class="lineNum"> 79 </span><span class="lineCov"> 9711792 : asm volatile (&quot;pand %xmm6,%xmm2&quot;);</span>
<span class="lineNum"> 80 </span><span class="lineCov"> 9711792 : asm volatile (&quot;pand %xmm3,%xmm5&quot;);</span>
<span class="lineNum"> 81 </span><span class="lineCov"> 9711792 : asm volatile (&quot;pxor %xmm5,%xmm2&quot;);</span>
<span class="lineNum"> 82 </span> :
<span class="lineNum"> 83 </span><span class="lineCov"> 9711792 : asm volatile (&quot;movdqa %0,%%xmm4&quot; : : &quot;m&quot; (v[d][i]));</span>
<span class="lineNum"> 84 </span><span class="lineCov"> 9711792 : asm volatile (&quot;pxor %xmm4,%xmm0&quot;);</span>
<span class="lineNum"> 85 </span><span class="lineCov"> 9711792 : asm volatile (&quot;pxor %xmm4,%xmm1&quot;);</span>
<span class="lineNum"> 86 </span><span class="lineCov"> 9711792 : asm volatile (&quot;pxor %xmm4,%xmm2&quot;);</span>
<span class="lineNum"> 87 </span> : }
<span class="lineNum"> 88 </span><span class="lineCov"> 1377360 : asm volatile (&quot;movntdq %%xmm0,%0&quot; : &quot;=m&quot; (p[i]));</span>
<span class="lineNum"> 89 </span><span class="lineCov"> 1377360 : asm volatile (&quot;movntdq %%xmm1,%0&quot; : &quot;=m&quot; (q[i]));</span>
<span class="lineNum"> 90 </span><span class="lineCov"> 1377360 : asm volatile (&quot;movntdq %%xmm2,%0&quot; : &quot;=m&quot; (r[i]));</span>
<span class="lineNum"> 91 </span> : }
<span class="lineNum"> 92 </span> :
<span class="lineNum"> 93 </span> : raid_sse_end();
<span class="lineNum"> 94 </span><span class="lineCov"> 1176 : }</span>
<span class="lineNum"> 95 </span> : #endif
<span class="lineNum"> 96 </span> :
<span class="lineNum"> 97 </span> : #if defined(CONFIG_X86_64) &amp;&amp; defined(CONFIG_SSE2)
<span class="lineNum"> 98 </span> : /*
<span class="lineNum"> 99 </span> : * GENz (triple parity with powers of 2^-1) SSE2 implementation
<span class="lineNum"> 100 </span> : *
<span class="lineNum"> 101 </span> : * Note that it uses 16 registers, meaning that x64 is required.
<span class="lineNum"> 102 </span> : */
<span class="lineNum"> 103 </span> : void raid_genz_sse2ext(int nd, size_t size, void **vv)
<span class="lineNum"> 104 </span> : {
<span class="lineNum"> 105 </span> : uint8_t **v = (uint8_t**)vv;
<span class="lineNum"> 106 </span> : uint8_t *p;
<span class="lineNum"> 107 </span> : uint8_t *q;
<span class="lineNum"> 108 </span> : uint8_t *r;
<span class="lineNum"> 109 </span> : int d, l;
<span class="lineNum"> 110 </span> : size_t i;
<span class="lineNum"> 111 </span> :
<span class="lineNum"> 112 </span> : l = nd - 1;
<span class="lineNum"> 113 </span> : p = v[nd];
<span class="lineNum"> 114 </span> : q = v[nd + 1];
<span class="lineNum"> 115 </span> : r = v[nd + 2];
<span class="lineNum"> 116 </span> :
<span class="lineNum"> 117 </span> : raid_sse_begin();
<span class="lineNum"> 118 </span> :
<span class="lineNum"> 119 </span> : asm volatile (&quot;movdqa %0,%%xmm7&quot; : : &quot;m&quot; (gfzconst16.poly[0]));
<span class="lineNum"> 120 </span> : asm volatile (&quot;movdqa %0,%%xmm3&quot; : : &quot;m&quot; (gfzconst16.half[0]));
<span class="lineNum"> 121 </span> : asm volatile (&quot;movdqa %0,%%xmm11&quot; : : &quot;m&quot; (gfzconst16.low7[0]));
<span class="lineNum"> 122 </span> :
<span class="lineNum"> 123 </span> : for (i = 0; i &lt; size; i += 32) {
<span class="lineNum"> 124 </span> : asm volatile (&quot;movdqa %0,%%xmm0&quot; : : &quot;m&quot; (v[l][i]));
<span class="lineNum"> 125 </span> : asm volatile (&quot;movdqa %0,%%xmm8&quot; : : &quot;m&quot; (v[l][i + 16]));
<span class="lineNum"> 126 </span> : asm volatile (&quot;movdqa %xmm0,%xmm1&quot;);
<span class="lineNum"> 127 </span> : asm volatile (&quot;movdqa %xmm8,%xmm9&quot;);
<span class="lineNum"> 128 </span> : asm volatile (&quot;movdqa %xmm0,%xmm2&quot;);
<span class="lineNum"> 129 </span> : asm volatile (&quot;movdqa %xmm8,%xmm10&quot;);
<span class="lineNum"> 130 </span> : for (d = l - 1; d &gt;= 0; --d) {
<span class="lineNum"> 131 </span> : asm volatile (&quot;movdqa %xmm2,%xmm6&quot;);
<span class="lineNum"> 132 </span> : asm volatile (&quot;movdqa %xmm10,%xmm14&quot;);
<span class="lineNum"> 133 </span> : asm volatile (&quot;pxor %xmm4,%xmm4&quot;);
<span class="lineNum"> 134 </span> : asm volatile (&quot;pxor %xmm12,%xmm12&quot;);
<span class="lineNum"> 135 </span> : asm volatile (&quot;pxor %xmm5,%xmm5&quot;);
<span class="lineNum"> 136 </span> : asm volatile (&quot;pxor %xmm13,%xmm13&quot;);
<span class="lineNum"> 137 </span> : asm volatile (&quot;psllw $7,%xmm6&quot;);
<span class="lineNum"> 138 </span> : asm volatile (&quot;psllw $7,%xmm14&quot;);
<span class="lineNum"> 139 </span> : asm volatile (&quot;psrlw $1,%xmm2&quot;);
<span class="lineNum"> 140 </span> : asm volatile (&quot;psrlw $1,%xmm10&quot;);
<span class="lineNum"> 141 </span> : asm volatile (&quot;pcmpgtb %xmm1,%xmm4&quot;);
<span class="lineNum"> 142 </span> : asm volatile (&quot;pcmpgtb %xmm9,%xmm12&quot;);
<span class="lineNum"> 143 </span> : asm volatile (&quot;pcmpgtb %xmm6,%xmm5&quot;);
<span class="lineNum"> 144 </span> : asm volatile (&quot;pcmpgtb %xmm14,%xmm13&quot;);
<span class="lineNum"> 145 </span> : asm volatile (&quot;paddb %xmm1,%xmm1&quot;);
<span class="lineNum"> 146 </span> : asm volatile (&quot;paddb %xmm9,%xmm9&quot;);
<span class="lineNum"> 147 </span> : asm volatile (&quot;pand %xmm11,%xmm2&quot;);
<span class="lineNum"> 148 </span> : asm volatile (&quot;pand %xmm11,%xmm10&quot;);
<span class="lineNum"> 149 </span> : asm volatile (&quot;pand %xmm7,%xmm4&quot;);
<span class="lineNum"> 150 </span> : asm volatile (&quot;pand %xmm7,%xmm12&quot;);
<span class="lineNum"> 151 </span> : asm volatile (&quot;pand %xmm3,%xmm5&quot;);
<span class="lineNum"> 152 </span> : asm volatile (&quot;pand %xmm3,%xmm13&quot;);
<span class="lineNum"> 153 </span> : asm volatile (&quot;pxor %xmm4,%xmm1&quot;);
<span class="lineNum"> 154 </span> : asm volatile (&quot;pxor %xmm12,%xmm9&quot;);
<span class="lineNum"> 155 </span> : asm volatile (&quot;pxor %xmm5,%xmm2&quot;);
<span class="lineNum"> 156 </span> : asm volatile (&quot;pxor %xmm13,%xmm10&quot;);
<span class="lineNum"> 157 </span> :
<span class="lineNum"> 158 </span> : asm volatile (&quot;movdqa %0,%%xmm4&quot; : : &quot;m&quot; (v[d][i]));
<span class="lineNum"> 159 </span> : asm volatile (&quot;movdqa %0,%%xmm12&quot; : : &quot;m&quot; (v[d][i + 16]));
<span class="lineNum"> 160 </span> : asm volatile (&quot;pxor %xmm4,%xmm0&quot;);
<span class="lineNum"> 161 </span> : asm volatile (&quot;pxor %xmm4,%xmm1&quot;);
<span class="lineNum"> 162 </span> : asm volatile (&quot;pxor %xmm4,%xmm2&quot;);
<span class="lineNum"> 163 </span> : asm volatile (&quot;pxor %xmm12,%xmm8&quot;);
<span class="lineNum"> 164 </span> : asm volatile (&quot;pxor %xmm12,%xmm9&quot;);
<span class="lineNum"> 165 </span> : asm volatile (&quot;pxor %xmm12,%xmm10&quot;);
<span class="lineNum"> 166 </span> : }
<span class="lineNum"> 167 </span> : asm volatile (&quot;movntdq %%xmm0,%0&quot; : &quot;=m&quot; (p[i]));
<span class="lineNum"> 168 </span> : asm volatile (&quot;movntdq %%xmm8,%0&quot; : &quot;=m&quot; (p[i + 16]));
<span class="lineNum"> 169 </span> : asm volatile (&quot;movntdq %%xmm1,%0&quot; : &quot;=m&quot; (q[i]));
<span class="lineNum"> 170 </span> : asm volatile (&quot;movntdq %%xmm9,%0&quot; : &quot;=m&quot; (q[i + 16]));
<span class="lineNum"> 171 </span> : asm volatile (&quot;movntdq %%xmm2,%0&quot; : &quot;=m&quot; (r[i]));
<span class="lineNum"> 172 </span> : asm volatile (&quot;movntdq %%xmm10,%0&quot; : &quot;=m&quot; (r[i + 16]));
<span class="lineNum"> 173 </span> : }
<span class="lineNum"> 174 </span> :
<span class="lineNum"> 175 </span> : raid_sse_end();
<span class="lineNum"> 176 </span> : }
<span class="lineNum"> 177 </span> : #endif
<span class="lineNum"> 178 </span> :
<span class="lineNum"> 179 </span> : #if defined(CONFIG_X86_64) &amp;&amp; defined(CONFIG_AVX2)
<span class="lineNum"> 180 </span> : /*
<span class="lineNum"> 181 </span> : * GENz (triple parity with powers of 2^-1) AVX2 implementation
<span class="lineNum"> 182 </span> : *
<span class="lineNum"> 183 </span> : * Note that it uses 16 registers, meaning that x64 is required.
<span class="lineNum"> 184 </span> : */
<span class="lineNum"> 185 </span> : void raid_genz_avx2ext(int nd, size_t size, void **vv)
<span class="lineNum"> 186 </span> : {
<span class="lineNum"> 187 </span> : uint8_t **v = (uint8_t**)vv;
<span class="lineNum"> 188 </span> : uint8_t *p;
<span class="lineNum"> 189 </span> : uint8_t *q;
<span class="lineNum"> 190 </span> : uint8_t *r;
<span class="lineNum"> 191 </span> : int d, l;
<span class="lineNum"> 192 </span> : size_t i;
<span class="lineNum"> 193 </span> :
<span class="lineNum"> 194 </span> : l = nd - 1;
<span class="lineNum"> 195 </span> : p = v[nd];
<span class="lineNum"> 196 </span> : q = v[nd + 1];
<span class="lineNum"> 197 </span> : r = v[nd + 2];
<span class="lineNum"> 198 </span> :
<span class="lineNum"> 199 </span> : raid_avx_begin();
<span class="lineNum"> 200 </span> :
<span class="lineNum"> 201 </span> : asm volatile (&quot;vbroadcasti128 %0,%%ymm7&quot; : : &quot;m&quot; (gfzconst16.poly[0]));
<span class="lineNum"> 202 </span> : asm volatile (&quot;vbroadcasti128 %0,%%ymm3&quot; : : &quot;m&quot; (gfzconst16.half[0]));
<span class="lineNum"> 203 </span> : asm volatile (&quot;vbroadcasti128 %0,%%ymm11&quot; : : &quot;m&quot; (gfzconst16.low7[0]));
<span class="lineNum"> 204 </span> : asm volatile (&quot;vpxor %ymm15,%ymm15,%ymm15&quot;);
<span class="lineNum"> 205 </span> :
<span class="lineNum"> 206 </span> : for (i = 0; i &lt; size; i += 64) {
<span class="lineNum"> 207 </span> : asm volatile (&quot;vmovdqa %0,%%ymm0&quot; : : &quot;m&quot; (v[l][i]));
<span class="lineNum"> 208 </span> : asm volatile (&quot;vmovdqa %0,%%ymm8&quot; : : &quot;m&quot; (v[l][i + 32]));
<span class="lineNum"> 209 </span> : asm volatile (&quot;vmovdqa %ymm0,%ymm1&quot;);
<span class="lineNum"> 210 </span> : asm volatile (&quot;vmovdqa %ymm8,%ymm9&quot;);
<span class="lineNum"> 211 </span> : asm volatile (&quot;vmovdqa %ymm0,%ymm2&quot;);
<span class="lineNum"> 212 </span> : asm volatile (&quot;vmovdqa %ymm8,%ymm10&quot;);
<span class="lineNum"> 213 </span> : for (d = l - 1; d &gt;= 0; --d) {
<span class="lineNum"> 214 </span> : asm volatile (&quot;vpsllw $7,%ymm2,%ymm6&quot;);
<span class="lineNum"> 215 </span> : asm volatile (&quot;vpsllw $7,%ymm10,%ymm14&quot;);
<span class="lineNum"> 216 </span> : asm volatile (&quot;vpsrlw $1,%ymm2,%ymm2&quot;);
<span class="lineNum"> 217 </span> : asm volatile (&quot;vpsrlw $1,%ymm10,%ymm10&quot;);
<span class="lineNum"> 218 </span> : asm volatile (&quot;vpcmpgtb %ymm1,%ymm15,%ymm4&quot;);
<span class="lineNum"> 219 </span> : asm volatile (&quot;vpcmpgtb %ymm9,%ymm15,%ymm12&quot;);
<span class="lineNum"> 220 </span> : asm volatile (&quot;vpcmpgtb %ymm6,%ymm15,%ymm5&quot;);
<span class="lineNum"> 221 </span> : asm volatile (&quot;vpcmpgtb %ymm14,%ymm15,%ymm13&quot;);
<span class="lineNum"> 222 </span> : asm volatile (&quot;vpaddb %ymm1,%ymm1,%ymm1&quot;);
<span class="lineNum"> 223 </span> : asm volatile (&quot;vpaddb %ymm9,%ymm9,%ymm9&quot;);
<span class="lineNum"> 224 </span> : asm volatile (&quot;vpand %ymm11,%ymm2,%ymm2&quot;);
<span class="lineNum"> 225 </span> : asm volatile (&quot;vpand %ymm11,%ymm10,%ymm10&quot;);
<span class="lineNum"> 226 </span> : asm volatile (&quot;vpand %ymm7,%ymm4,%ymm4&quot;);
<span class="lineNum"> 227 </span> : asm volatile (&quot;vpand %ymm7,%ymm12,%ymm12&quot;);
<span class="lineNum"> 228 </span> : asm volatile (&quot;vpand %ymm3,%ymm5,%ymm5&quot;);
<span class="lineNum"> 229 </span> : asm volatile (&quot;vpand %ymm3,%ymm13,%ymm13&quot;);
<span class="lineNum"> 230 </span> : asm volatile (&quot;vpxor %ymm4,%ymm1,%ymm1&quot;);
<span class="lineNum"> 231 </span> : asm volatile (&quot;vpxor %ymm12,%ymm9,%ymm9&quot;);
<span class="lineNum"> 232 </span> : asm volatile (&quot;vpxor %ymm5,%ymm2,%ymm2&quot;);
<span class="lineNum"> 233 </span> : asm volatile (&quot;vpxor %ymm13,%ymm10,%ymm10&quot;);
<span class="lineNum"> 234 </span> :
<span class="lineNum"> 235 </span> : asm volatile (&quot;vmovdqa %0,%%ymm4&quot; : : &quot;m&quot; (v[d][i]));
<span class="lineNum"> 236 </span> : asm volatile (&quot;vmovdqa %0,%%ymm12&quot; : : &quot;m&quot; (v[d][i + 32]));
<span class="lineNum"> 237 </span> : asm volatile (&quot;vpxor %ymm4,%ymm0,%ymm0&quot;);
<span class="lineNum"> 238 </span> : asm volatile (&quot;vpxor %ymm4,%ymm1,%ymm1&quot;);
<span class="lineNum"> 239 </span> : asm volatile (&quot;vpxor %ymm4,%ymm2,%ymm2&quot;);
<span class="lineNum"> 240 </span> : asm volatile (&quot;vpxor %ymm12,%ymm8,%ymm8&quot;);
<span class="lineNum"> 241 </span> : asm volatile (&quot;vpxor %ymm12,%ymm9,%ymm9&quot;);
<span class="lineNum"> 242 </span> : asm volatile (&quot;vpxor %ymm12,%ymm10,%ymm10&quot;);
<span class="lineNum"> 243 </span> : }
<span class="lineNum"> 244 </span> : asm volatile (&quot;vmovntdq %%ymm0,%0&quot; : &quot;=m&quot; (p[i]));
<span class="lineNum"> 245 </span> : asm volatile (&quot;vmovntdq %%ymm8,%0&quot; : &quot;=m&quot; (p[i + 32]));
<span class="lineNum"> 246 </span> : asm volatile (&quot;vmovntdq %%ymm1,%0&quot; : &quot;=m&quot; (q[i]));
<span class="lineNum"> 247 </span> : asm volatile (&quot;vmovntdq %%ymm9,%0&quot; : &quot;=m&quot; (q[i + 32]));
<span class="lineNum"> 248 </span> : asm volatile (&quot;vmovntdq %%ymm2,%0&quot; : &quot;=m&quot; (r[i]));
<span class="lineNum"> 249 </span> : asm volatile (&quot;vmovntdq %%ymm10,%0&quot; : &quot;=m&quot; (r[i + 32]));
<span class="lineNum"> 250 </span> : }
<span class="lineNum"> 251 </span> :
<span class="lineNum"> 252 </span> : raid_avx_end();
<span class="lineNum"> 253 </span> : }
<span class="lineNum"> 254 </span> : #endif
<span class="lineNum"> 255 </span> :
</pre>
</td>
</tr>
</table>
<br>
<table width="100%" border=0 cellspacing=0 cellpadding=0>
<tr><td class="ruler"><img src="../glass.png" width=3 height=3 alt=""></td></tr>
<tr><td class="versionInfo">Generated by: <a href="http://ltp.sourceforge.net/coverage/lcov.php" target="_parent">LCOV version 1.10</a></td></tr>
</table>
<br>
</body>
</html>