{"id":6505,"date":"2017-06-10T20:35:37","date_gmt":"2017-06-10T11:35:37","guid":{"rendered":"http:\/\/umezawa.dyndns.info\/wordpress\/?p=6505"},"modified":"2017-06-11T17:08:52","modified_gmt":"2017-06-11T08:08:52","slug":"%e8%87%aa%e5%8b%95%e3%83%99%e3%82%af%e3%83%88%e3%83%ab%e5%8c%96%e3%81%af%e3%81%a9%e3%82%8c%e3%81%8f%e3%82%89%e3%81%84%e4%bd%bf%e3%81%84%e7%89%a9%e3%81%ab%e3%81%aa%e3%82%8b%e3%81%8b%ef%bc%88%e3%81%9d-2","status":"publish","type":"post","link":"http:\/\/umezawa.dyndns.info\/wordpress\/?p=6505","title":{"rendered":"\u81ea\u52d5\u30d9\u30af\u30c8\u30eb\u5316\u306f\u3069\u308c\u304f\u3089\u3044\u4f7f\u3044\u7269\u306b\u306a\u308b\u304b\uff08\u305d\u306e\uff12\uff09"},"content":{"rendered":"<p><a href=\"http:\/\/umezawa.dyndns.info\/wordpress\/?p=6497\">\u305d\u306e\uff11<\/a>\u306e\u7d9a\u304d<\/p>\n<p>\u8981\u7d20\u306e\u4e26\u3079\u66ff\u3048\uff08\u6b63\u78ba\u306b\u306f\u30a2\u30f3\u30d1\u30c3\u30af\uff09\u3092\u884c\u3046\u5834\u5408\u306f\u3069\u3046\u3067\u3057\u3087\u3046\u304b\u3002 UtVideo \u3067\u306f packed \u30d5\u30a9\u30fc\u30de\u30c3\u30c8\u3068 planar \u30d5\u30a9\u30fc\u30de\u30c3\u30c8\u306e\u5909\u63db\u3067\u3053\u306e\u51e6\u7406\u304c\u51fa\u3066\u304d\u307e\u3059\u3002<\/p>\n<p><!--more--><\/p>\n<p>\u3068\u308a\u3042\u3048\u305a\u7c21\u5358\u306a\u4f8b\u3068\u3057\u3066\u4ee5\u4e0b\u306e\u3088\u3046\u306a\u30b3\u30fc\u30c9\u3067\u8a66\u3057\u307e\u3059\u3002<\/p>\n<blockquote>\n<pre>\r\ntemplate&lt;typename T&gt;\r\nvoid foo(T* __restrict dst, const T* __restrict a, const T* __restrict b, size_t n)\r\n{\r\n\tfor (size_t i = 0; i &lt; n; ++i)\r\n\t{\r\n\t\tdst[0] = *a++;\r\n\t\tdst[1] = *b++;\r\n\t\tdst += 2;\r\n\t}\r\n}\r\n\r\ntemplate void foo&lt;float&gt;(float* __restrict, const float* __restrict, const float* __restrict, size_t);\r\ntemplate void foo&lt;int&gt;(int* __restrict, const int* __restrict, const int* __restrict, size_t);\r\ntemplate void foo&lt;short&gt;(short* __restrict, const short* __restrict, const short* __restrict, size_t);\r\ntemplate void foo&lt;char&gt;(char* __restrict, const char* __restrict, const char* __restrict, size_t);\r\n<\/pre>\n<\/blockquote>\n<p>char \u306e\u5834\u5408\u306e\u6a21\u7bc4\u89e3\u7b54\u306f\u3053\u3093\u306a\u611f\u3058\u3002<\/p>\n<blockquote>\n<pre>\r\n    movdqu  xmm0, [rdx]\r\n    movdqu  xmm1, [r8]\r\n    movdqa  xmm2, xmm0\r\n    puhpcklbw xmm0, xmm1\r\n    punpckhbw xmm2, xmm1\r\n    movdqu  [rcx], xmm0\r\n    movdqu  [rcx+16], xmm2\r\n<\/pre>\n<\/blockquote>\n<p>\u3067\u3001\u5b9f\u969b\u306b\u3084\u3063\u3066\u307f\u308b\u3068\u3001<\/p>\n<blockquote>\n<pre>\r\nD:\\&gt;cl \/c \/O2 \/Qvec-report:2 autovec2.cpp\r\nMicrosoft(R) C\/C++ Optimizing Compiler Version 19.10.25019 for x64\r\nCopyright (C) Microsoft Corporation.  All rights reserved.\r\n\r\nautovec2.cpp\r\n\r\n--- \u5206\u6790\u95a2\u6570: void __cdecl foo&lt;float&gt;(float * __ptr64 __restrict,float const * __ptr64 __restrict,float const * __ptr64 __restrict,unsigned __int64)\r\nd:\\autovec2.cpp(4) : info C5002: \u30eb\u30fc\u30d7\u306f\u30d9\u30af\u30bf\u30fc\u5316\u3055\u308c\u307e\u305b\u3093\u3002\u7406\u7531: '1300'\r\n\r\n--- \u5206\u6790\u95a2\u6570: void __cdecl foo&lt;char&gt;(char * __ptr64 __restrict,char const * __ptr64 __restrict,char const * __ptr64 __restrict,unsigned __int64)\r\nd:\\autovec2.cpp(4) : info C5002: \u30eb\u30fc\u30d7\u306f\u30d9\u30af\u30bf\u30fc\u5316\u3055\u308c\u307e\u305b\u3093\u3002\u7406\u7531: '1300'\r\n\r\n--- \u5206\u6790\u95a2\u6570: void __cdecl foo&lt;int&gt;(int * __ptr64 __restrict,int const * __ptr64 __restrict,int const * __ptr64 __restrict,unsigned __int64)\r\nd:\\autovec2.cpp(4) : info C5002: \u30eb\u30fc\u30d7\u306f\u30d9\u30af\u30bf\u30fc\u5316\u3055\u308c\u307e\u305b\u3093\u3002\u7406\u7531: '1300'\r\n\r\n--- \u5206\u6790\u95a2\u6570: void __cdecl foo&lt;short&gt;(short * __ptr64 __restrict,short const * __ptr64 __restrict,short const * __ptr64 __restrict,unsigned __int64)\r\nd:\\autovec2.cpp(4) : info C5002: \u30eb\u30fc\u30d7\u306f\u30d9\u30af\u30bf\u30fc\u5316\u3055\u308c\u307e\u305b\u3093\u3002\u7406\u7531: '1300'\r\n<\/pre>\n<\/blockquote>\n<p><strong>\u3048\u3063<\/strong><\/p>\n<p>\u3061\u306a\u307f\u306b\u7406\u7531\u30b3\u30fc\u30c9\u306e\u4e00\u89a7\u306f<a href=\"https:\/\/msdn.microsoft.com\/ja-jp\/library\/jj658585.aspx\">\u3053\u3053<\/a>\u306b\u3042\u3063\u3066\u30011300 \u306f\u300c\u30eb\u30fc\u30d7\u672c\u4f53\u306b\u8a08\u7b97\u304c\u307e\u3063\u305f\u304f\u304b\u307b\u3068\u3093\u3069\u542b\u307e\u308c\u3066\u3044\u307e\u305b\u3093\u3002\u300d\u3060\u305d\u3046\u3067\u3059\u3002\u78ba\u304b\u306b\u8a08\u7b97\u306f\u3057\u3066\u306a\u3044\u3051\u3069\u3055\u3002<\/p>\n<p>\u3055\u3066\u3001\u6700\u8fd1\u306e Visual C++ \u306b\u306f Clang\/C2 \u3068\u3044\u3046\u300c\u5225\u306e\u30b3\u30f3\u30d1\u30a4\u30e9\u300d\u304c\u542b\u307e\u308c\u3066\u3044\u308b\u306e\u3067\u3001\u3053\u3063\u3061\u3082\u8a66\u3057\u3066\u307f\u307e\u3057\u3087\u3046\u3002<\/p>\n<blockquote>\n<pre>\r\nD:\\&gt;\"C:\\Program Files (x86)\\Microsoft Visual Studio\\2017\\Community\\VC\\Tools\\ClangC2\\14.10.25903\\bin\\HostX64\\clang.exe\" --version\r\nclang with Microsoft CodeGen version 3.8.0\r\nProvided as - is without support\r\nC:\\Program Files (x86)\\Microsoft Visual Studio\\2017\\Community\\VC\\Tools\\ClangC2\\14.10.25903\\bin\\HostX64\\x64\\c2.dll version 19.10.25903.0\r\nTarget: x86_64-pc-windows-msvc\r\nThread model: posix\r\nInstalledDir: C:\\Program Files (x86)\\Microsoft Visual Studio\\2017\\Community\\VC\\Tools\\ClangC2\\14.10.25903\\bin\\HostX64\r\n\r\nD:\\&gt;\"C:\\Program Files (x86)\\Microsoft Visual Studio\\2017\\Community\\VC\\Tools\\ClangC2\\14.10.25903\\bin\\HostX64\\clang.exe\" -c -O2 -msse4.1 autovec2.cpp\r\n<\/pre>\n<\/blockquote>\n<blockquote>\n<pre>\r\numezawa@selene \/cygdrive\/d\r\n$ objdump -d -M intel autovec2.o\r\n\r\nautovec2.o:     \u30d5\u30a1\u30a4\u30eb\u5f62\u5f0f pe-bigobj-x86-64\r\n\r\n\r\n\u30bb\u30af\u30b7\u30e7\u30f3 .text$mn \u306e\u9006\u30a2\u30bb\u30f3\u30d6\u30eb:\r\n\r\n0000000000000000 &lt;??$foo@D@@YAXPEIADPEIBD1_K@Z&gt;:\r\n   0:   4d 85 c9                test   r9,r9\r\n   3:   74 29                   je     2e &lt;??$foo@D@@YAXPEIADPEIBD1_K@Z+0x2e&gt;\r\n   5:   66 66 66 0f 1f 84 00    data16 data16 nop WORD PTR [rax+rax*1+0x0]\r\n   c:   00 00 00 00\r\n  10:   0f b6 02                movzx  eax,BYTE PTR [rdx]\r\n  13:   48 8d 52 01             lea    rdx,[rdx+0x1]\r\n  17:   88 01                   mov    BYTE PTR [rcx],al\r\n  19:   48 8d 49 02             lea    rcx,[rcx+0x2]\r\n  1d:   41 0f b6 00             movzx  eax,BYTE PTR [r8]\r\n  21:   4d 8d 40 01             lea    r8,[r8+0x1]\r\n  25:   88 41 ff                mov    BYTE PTR [rcx-0x1],al\r\n  28:   49 83 e9 01             sub    r9,0x1\r\n  2c:   75 e2                   jne    10 &lt;??$foo@D@@YAXPEIADPEIBD1_K@Z+0x10&gt;\r\n  2e:   c3                      ret\r\n<\/pre>\n<\/blockquote>\n<p><strong>\u30c0\u30e1\u3058\u3083\u306d\u3047\u304b<\/strong><\/p>\n<p><a href=\"https:\/\/wandbox.org\/\">wandbox<\/a> \u4e0a\u3067\u3044\u308d\u3044\u308d\u8a66\u3057\u305f\u7d50\u679c\u3001 <a href=\"https:\/\/wandbox.org\/permlink\/BsBS7JLSnKl5lqdI\" rel=\"nofollow\">GCC \u306e\u5834\u5408\u306f\u5c11\u306a\u304f\u3068\u3082\u3060\u3044\u3076\u524d\u306e\u30d0\u30fc\u30b8\u30e7\u30f3\u3067\u3042\u308b 4.4.7 \u306e\u6642\u70b9\u3067 -O3 \u30aa\u30d7\u30b7\u30e7\u30f3\uff08\u3088\u308a\u6b63\u78ba\u306b\u306f -ftree-vectorize \u30aa\u30d7\u30b7\u30e7\u30f3\u3002-O2 \u3067\u306f\u30c0\u30e1\uff09\u3092\u4ed8\u3051\u305f\u6642\u306b\u30d9\u30af\u30c8\u30eb\u5316\u3057\u3066\u304f\u308c\u307e\u3059<\/a>\u304c\u3001 Clang \u3060\u3068\u6700\u65b0\u7248\u3067\u3042\u308b 4.0.0 \u3067\u3088\u3046\u3084\u304f\u30d9\u30af\u30c8\u30eb\u5316\u3057\u3066\u304f\u308c\u308b\u3088\u3046\u3067\u3059\u3002<\/p>\n<p>\u4e00\u756a\u7c21\u5358\u306a2\u8981\u7d20\u306e\u30b1\u30fc\u30b9\u3067\u3059\u3089\u3053\u308c\u3067\u306f\u5148\u304c\u601d\u3044\u3084\u3089\u308c\u307e\u3059\u3002<\/p>\n<p><a href=\"http:\/\/umezawa.dyndns.info\/wordpress\/?p=6521\">\u305d\u306e\uff13<\/a>\u306b\u7d9a\u304f<\/p>\n","protected":false},"excerpt":{"rendered":"<p>\u305d\u306e\uff11\u306e\u7d9a\u304d \u8981\u7d20\u306e\u4e26\u3079\u66ff\u3048\uff08\u6b63\u78ba\u306b\u306f\u30a2\u30f3\u30d1\u30c3\u30af\uff09\u3092\u884c\u3046\u5834\u5408\u306f\u3069\u3046\u3067\u3057\u3087\u3046\u304b\u3002 UtVideo \u3067\u306f packed \u30d5\u30a9\u30fc\u30de\u30c3\u30c8\u3068 planar \u30d5\u30a9\u30fc\u30de\u30c3\u30c8\u306e\u5909\u63db\u3067\u3053\u306e\u51e6\u7406\u304c\u51fa\u3066\u304d\u307e\u3059\u3002<\/p>\n","protected":false},"author":2,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[27],"tags":[],"class_list":["post-6505","post","type-post","status-publish","format-standard","hentry","category-technology"],"_links":{"self":[{"href":"http:\/\/umezawa.dyndns.info\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/6505","targetHints":{"allow":["GET"]}}],"collection":[{"href":"http:\/\/umezawa.dyndns.info\/wordpress\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/umezawa.dyndns.info\/wordpress\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/umezawa.dyndns.info\/wordpress\/index.php?rest_route=\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"http:\/\/umezawa.dyndns.info\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=6505"}],"version-history":[{"count":14,"href":"http:\/\/umezawa.dyndns.info\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/6505\/revisions"}],"predecessor-version":[{"id":6528,"href":"http:\/\/umezawa.dyndns.info\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/6505\/revisions\/6528"}],"wp:attachment":[{"href":"http:\/\/umezawa.dyndns.info\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=6505"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/umezawa.dyndns.info\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=6505"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/umezawa.dyndns.info\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=6505"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}