opreport ã®XML出力機能。
2007å¹´12月12æ—¥ (水曜日) 01:36:31 # Life opreport ã®XML出力機能。 oprofile ã¯ãƒ—ãƒãƒ•ã‚¡ã‚¤ãƒªãƒ³ã‚°ã®ãŸã‚ã®ãƒ„ールã§ã€ãã®å‡ºåŠ›ã‚’ã™ã‚‹ãŸã‚ã®ãƒ„ール㌠opreport ã§ã™ã€‚ opreport ã¯é€šå¸¸ã¯äººé–“ãŒèªã‚€ãŸã‚ã®å‡ºåŠ›ã‚’出ã—ã¾ã™ã€‚ãã®å‡ºåŠ›ã¯ã‚³ãƒ³ãƒ”ュータ処ç†ã‚’ã™ã‚‹ã«ã¯ã¡ã‚‡ã£ã¨é¢å€’ãªå½¢å¼ã§ã™ã€‚ ã—ã‹ã—最近ã®ãƒãƒ¼ã‚¸ãƒ§ãƒ³ã§ã¯ã€-X オプションを指定ã—ãŸã‚‰XMLã®å‡ºåŠ›ã‚’ã—ã¦ãれるよã†ã«ãªã£ã¦ã„ã¾ã™ã€‚ 出力形å¼ã¯ã‚·ãƒ³ãƒ—ルã§ã€XSLTを作æˆã—ã¦XML出力をHTMLファイルã«å¤‰æ›ã—ã¦ã¿ã‚‹ã®ã‚‚ç°¡å˜ã§ã™ã€‚ sudo opcontrol --shutdown sudo opcontrol --reset sudo opcontrol --setup \ --vmlinux=/lib/modules/$(uname -r)/build/vmlinux \ --separ
CPU内ã‚ャッシュ効率ã¨ãƒ™ã‚¯ãƒˆãƒ«æ¼”算（SIMD）を利用ã—ãŸè¨ˆç®—密度å‘上ã®ãŸã‚ã®2分木構築 - maropuã®ãƒ¡ãƒ¢å¢“å ´
最近ã€ç¾å®Ÿé€ƒé¿ã§ã‚„ã£ã¦ã„ãŸä½œæ¥çµæžœã®ãƒ¡ãƒ¢ï¼Ž SIGMOD'10ã®intel lab.è«–æ–‡ã®å†…容を実際ã«è‡ªåˆ†ã§ãƒ—ãƒã‚°ãƒ©ãƒ ã—ã¦ã€è©•ä¾¡ã—ã¦ã¿ãŸï¼Ž 2000年以é™DBç ”ç©¶ç•Œéšˆã§ã¯ã€ãƒ‡ãƒ¼ã‚¿é ˜åŸŸãŒå…¨ã¦ãƒ¡ãƒ¢ãƒªä¸Šã«è¼‰ã‚‹ç’°å¢ƒã‚’å‰æã¨ã—ãŸãƒœãƒˆãƒ«ãƒãƒƒã‚¯åˆ†æž/改善æ案をç©æ¥µçš„ã«è¡Œã£ã¦ã„ã¦ã€æœ€è¿‘ã¯ç‰¹ã«ãã®å‚¾å‘ãŒå¼·ã„． 今後ã®ç ”究作æ¥ã®å½¹ã«ç«‹ã¤ã ã‚ã†ã¨æ€ã„ã€ä¸Šã®è«–文を題æã«oprofileを使ã£ã¦CPU内処ç†åˆ†æžã‚’ã‚„ã£ã¦ã¿ãŸï¼ˆãƒŽã‚¦ãƒã‚¦ç²å¾—ã®ãŸã‚ã«ï¼‰ï¼Žoprofileを使ã†ã¨ã€xxstatç³»ã§ã¯åˆ†ã‹ã‚‰ãªã„ã€CPU使用率ã®å†…訳をå–å¾—ã™ã‚‹ã“ã¨ãŒã§ãる． å‚考論文ã¯ä»¥ä¸‹ï¼Ž http://portal.acm.org/citation.cfm?id=1807167.1807206 論文内ã§ã¯ã€ç°¡å˜ã«ã„ãˆã°ä»¥ä¸‹3点を行ã£ã¦ã„る． 1. ãƒã‚¤ãƒ³ã‚¿ã‚’利用ã—ãŸ2分木実装ã§ã¯ãƒŽãƒ¼ãƒ‰ç§»å‹•æ¯Žã«Cache/TLBミスãŒç™ºç”Ÿã—ã€CPU内処ç†ãŒã‚¹
White Paper: Piranha - Load-balanced Web and FTP Clusters
Our teams collaborate with you to ensure you accomplish your goals with Red Hat solutions. The relationship we build with you is designed to provide you with the tools and resources you need to find success on your IT journey. We develop a holistic understanding of your experience as a customer by ensuring our support and engineering teams work together. Our support team works hand in hand with th
Linux カーãƒãƒ«ãƒ—ãƒãƒ•ã‚¡ã‚¤ãƒªãƒ³ã‚° - Bishopドã‚ュメント
ã“ã®ãƒ‰ã‚ュメントã§ã¯çµ„込㿠Linux 開発用 CPU ボード Bishop ã§ã‚«ãƒ¼ãƒãƒ«ãƒ¬ãƒ™ãƒ«ã®ãƒ—ãƒãƒ•ã‚¡ã‚¤ãƒªãƒ³ã‚°ã‚’è¡Œã†æ‰‹é †ã¨ã—ã¦ã€ 最もシンプル㪠readprofile ã¨ä»Šå¾Œ Linux カーãƒãƒ«ã®æ¨™æº–çš„ãªãƒ—ãƒãƒ•ã‚¡ã‚¤ãƒ©ã¨ãªã‚‹ã¨æ€ã‚れる OProfile ã® ç°¡å˜ãªä½¿ç”¨æ–¹æ³•ã‚’紹介ã—ã¾ã™ã€‚ 1. readprofile1.1. 概è¦Linuxカーãƒãƒ«ã«æ¨™æº–ã§å«ã¾ã‚Œã‚‹ã‚·ãƒ³ãƒ—ルãªãƒ—ãƒãƒ•ã‚¡ã‚¤ãƒ© (kernel/profile.c) 㯠tick 毎ã«å®Ÿè¡Œã‚¢ãƒ‰ãƒ¬ã‚¹ã‚’ カーãƒãƒ«å†…部ã®ãƒ—ãƒãƒ•ã‚¡ã‚¤ãƒªãƒ³ã‚°ãƒãƒƒãƒ•ã‚¡ã«ä¿æŒã—ã¾ã™ã€‚ ã“ã®ãƒãƒƒãƒ•ã‚¡ã¯ä»®æƒ³ãƒ•ã‚¡ã‚¤ãƒ« /proc/profile を通ã˜ã¦ãƒ¦ãƒ¼ã‚¶ç©ºé–“ã‹ã‚‰ã‚¢ã‚¯ã‚»ã‚¹ã§ãã¾ã™ãŒã€ /proc/profile ã¯ãƒã‚¤ãƒŠãƒªãƒ•ã‚¡ã‚¤ãƒ«ã§ã‚ã‚‹ãŸã‚ãã®ã¾ã¾ã§ã¯èªã‚ã¾ã›ã‚“。 readprofile 㯠/proc/profile ã‚’å¯èªåŒ–ã™ã‚‹ãŸã‚ã«ä½¿ç”¨ã•ã‚Œã‚‹ãƒ„ールã§ã™ã€‚1.
Performance Monitoring with OProfile | Beyond Syntax | Michael J. Schultz
oprofile is a low overhead, open-source tool that hooks into Linux and can keep track of CPU event monitoring information. This is a fairly general statement and for this post I'll be using the Intel Penryn microarchitecture, which should have similar event counters to most recent Intel processors. You can get the canonical list of event counters from Intel's own documentation in Chapter 30, Perfo
OS層ã®è©•ä¾¡
2005 年度上期 ã‚ªãƒ¼ãƒ—ãƒ³ã‚½ãƒ¼ã‚¹ã‚½ãƒ•ãƒˆã‚¦ã‚§ã‚¢æ´»ç”¨åŸºç›¤æ•´å‚™äº‹æ¥ ã€ŒOSS æ€§èƒ½ãƒ»ä¿¡é ¼æ€§è©•ä¾¡ / 障害解æžãƒ„ール開発〠OS 層 ～CPU スケーラビリティ評価編～ ä½œæˆ OSS 技術開発・評価コンソーシアム商標表記 ・ Alicia ã¯ã€ãƒ¦ãƒ‹ã‚¢ãƒ‡ãƒƒã‚¯ã‚¹æ ªå¼ä¼šç¤¾ã®ç™»éŒ²å•†æ¨™ã§ã™ã€‚ ・ Asianux ã¯ã€ãƒŸãƒ©ã‚¯ãƒ«ãƒ»ãƒªãƒŠãƒƒã‚¯ã‚¹æ ªå¼ä¼šç¤¾ã®æ—¥æœ¬ã«ãŠã‘る登録商標ã§ã™ã€‚ ・ Intelã€Itanium ãŠã‚ˆã³ Intel Xeon ã¯ã€ã‚¢ãƒ¡ãƒªã‚«åˆè¡†å›½ãŠã‚ˆã³ãã®ä»–ã®å›½ã«ãŠã‘るイ ンテルコーãƒãƒ¬ãƒ¼ã‚·ãƒ§ãƒ³ã¾ãŸã¯ãã®å会社ã®å•†æ¨™ã¾ãŸã¯ç™»éŒ²å•†æ¨™ã§ã™ã€‚ ・ Intel ã¯ã€Intel Corporation ã®ä¼šç¤¾åã§ã™ã€‚ ・ Linux ã¯ã€Linus Torvalds ã®ç±³å›½ãŠã‚ˆã³ãã®ä»–ã®å›½ã«ãŠã‘る登録商標ã‚ã‚‹ã„ã¯å•†æ¨™ã§ ã™ã€‚ ・ MIRACLE LINUX ã¯ã€ãƒŸãƒ©ã‚¯ãƒ«ãƒ»ãƒªãƒŠãƒƒã‚¯ã‚¹æ ªå¼ä¼šç¤¾ãŒä½¿ç”¨è¨±è«¾
Memory part 7: Memory performance tools [LWN.net]
This article brought to you by LWN subscribersSubscribers to LWN.net made this article — and everything that surrounds it — possible. If you appreciate our content, please buy a subscription and make the next set of articles possible. [Editor's note: welcome to part 7 of Ulrich Drepper's "What every programmer should know about memory"; this section is concerned with tools which can help programme
mmi2010040065 65..79
.......................................................................................................................................................................................................................... GOOGLE-WIDE PROFILING: A CONTINUOUS PROFILING INFRASTRUCTURE FOR DATA CENTERS .......................................................................................................
Oprofile @ RHEL5.5 on KVM : Mazn.net
2010å¹´09月24æ—¥12:40 Oprofile @ RHEL5.5 on KVM カテゴリIT技術kvm システムã®ãƒ—ãƒãƒ•ã‚¡ã‚¤ãƒ«ã«å½¹ã«ç«‹ã¤Opfofileã§ã™ãŒã€KVM上ã§Oprofileを実行ã—ã¦ã‚‚ã€CPUã®ã‚«ã‚¦ãƒ³ã‚¿ãŒåˆ©ç”¨ã§ããªã„ãŸã‚ã€é€šå¸¸ä½•ã‚‚å–å¾—ã§ãã¾ã›ã‚“。 ã“ã†ã„ã†ã¨ãã¯ã€opfofile 使用å‰ã«ãƒ¢ã‚¸ãƒ¥ãƒ¼ãƒ«ã®ã‚ªãƒ—ションã§ã€ # modprobe oprofile timer=1 ã¨ã—ã¦ã‚ã’ã‚‹ã¨ã€å–å¾—ã§ãるよã†ã§ã™ã€‚ ã™ã§ã«ãƒ¢ã‚¸ãƒ¥ãƒ¼ãƒ«ãŒå‹•ã„ã¦ã„ã‚‹å ´åˆã¯ã€ # opcontrol --deinit ã§åœæ¢ã•ã›ã‚‹ã“ã¨ãŒã§ãã¾ã™ã€‚ 「IT技術ã€ã‚«ãƒ†ã‚´ãƒªã®æœ€æ–°è¨˜äº‹
AndroidNDKã«ã‚ˆã‚‹JNI開発手法第4ç« AndroidNDKã«ã‚ˆã‚‹ JNI開発手法 ç·’æ–¹ è¡ O g a t a S a t o s h i æ ªå¼ä¼šç¤¾ã‚¤ãƒ¼ãƒ•ãƒãƒ¼ 事æ¥çµ±æ‹¬æœ¬éƒ¨ 海外開発部 第 4 ç« ã‚‹æœ€å¤§ã®ãƒ¡ãƒªãƒƒãƒˆã¯ã€Œå®Ÿè¡Œå‡¦ç†ã®å‘上ã€ã¨ã€Œæ—¢å˜ï¿½ï¿½
AndroidNDKã«ã‚ˆã‚‹JNI開発手法第4ç« AndroidNDKã«ã‚ˆã‚‹ JNI開発手法 ç·’æ–¹ è¡ O g a t a S a t o s h i æ ªå¼ä¼šç¤¾ã‚¤ãƒ¼ãƒ•ãƒãƒ¼ 事æ¥çµ±æ‹¬æœ¬éƒ¨ 海外開発部 第 4 ç« ã‚‹æœ€å¤§ã®ãƒ¡ãƒªãƒƒãƒˆã¯ã€Œå®Ÿè¡Œå‡¦ç†ã®å‘上ã€ã¨ã€Œæ—¢å˜ã®ãƒ©ã‚¤ ブラリã®æœ‰åŠ¹æ´»ç”¨ã€ã§ã™ã€‚第2ç« ã®ã€ŒAndroid高速化 テクニックã€ã§ç´¹ä»‹ã•ã‚Œã¦ã„る高速化テクニックã®æœ€å¾Œ ã®æ‰‹æ®µã¨ã—ã¦Javaコードをãƒã‚¤ãƒ†ã‚£ãƒ–コードã«ç½®ãæ› ãˆã‚‹ã®ã¯ã€å®Ÿè¡Œå‡¦ç†ã®å‘上ã¨ã„ã†å´é¢ã§ã¯æœ‰åŠ¹ã§ã™ã€‚ ãŸã ã—ã€ãƒã‚¤ãƒ†ã‚£ãƒ–コードã«ã—ã¦ã—ã¾ã†ã¨ç‰¹å®šã®ãƒ—ラッ トフォームã«ä¾å˜ã—ã¦ã—ã¾ã†ãŸã‚ã€ãŸã¨ãˆã°ARMå‘ã‘ ã«ä½œæˆã—ãŸãƒã‚¤ãƒ†ã‚£ãƒ–コードã¯x86ã§ã¯å‹•ä½œã—ãªã„ã® ã§ã€çµæžœã¨ã—ã¦ä½œæˆã—ãŸã‚¢ãƒ—リã¯ARM専用ã¨ã„ã†ã“㨠ã«ãªã£ã¦ã—ã¾ã„ã¾ã™ã€‚ 　メリットã¨ãƒ‡ãƒ¡ãƒªãƒƒãƒˆã‚’比較ã—ã€JNIを使ã†ã‹ä½¿ã‚ ãªã„ã‹ã‚’よã考ãˆã¦ãã ã•ã„。もã—JNIを使ã†
未æ¥ã®ã„ã¤ã‹/hyoshiokã®æ—¥è¨˜ | Rubyã®ã‚ャッシュミスを測定ã™ã‚‹ã€‚
カーãƒãƒ«èªæ›¸ä¼šã®è°è«–ã‹ã‚‰æ´¾ç”Ÿã—ãŸã€mallocã®alignmentã®ä»¶ã«ã¤ã„ã¦noocyteã•ã‚“ãŒè©³ç´°ãªã‚³ãƒ¡ãƒ³ãƒˆã‚’ã¤ã‘ã¦ã„ãŸã ã„ã¦ã„る。ã¡ã‚ƒã‚“ã¨è¿½ã„ãã‚Œã¦ã„ãªã„。ã™ã„ã¾ã›ã‚“。 ã¾ã‚ã€ã‚ãŸã—ã®ã‚¢ãƒã‚¦ãƒˆãªç†è§£ã§ã¯mallocを８ãƒã‚¤ãƒˆã‚¢ãƒ©ã‚¤ãƒ³ã—ãªã‘ã‚Œã°ã„ã‘ãªã„ã¨ã„ã†POSIXçš„ãªæ¨™æº–ã¯ãªã„ã‘ã©ã€glibcã®å®Ÿè£…ã§ã¯ãã†ãªã£ã¦ã„ã‚‹ã¨ã„ã†ã‚ˆã†ãªæ„Ÿã˜ã§ã¯ãªã„ã‹ã¨ã€‚IA32(x86)ã§ã¯ï¼˜ãƒã‚¤ãƒˆã‚¢ãƒ©ã‚¤ãƒ³ã§ãªã‘ã‚Œã°ã„ã‘ãªã„ã¨ã„ã†ã“ã¨ã¯ãªã•ãã†ãªã®ã§ã€‚ã‚‚ã¡ã‚ã‚“ã€ï¼˜ãƒã‚¤ãƒˆã‚¢ãƒ©ã‚¤ãƒ³ã®æ–¹ãŒå¥½ã¾ã—ã„ã‘ã©ã€‚ #ifndef INTERNAL_SIZE_T #define INTERNAL_SIZE_T size_t #endif /* The corresponding word size */ #define SIZE_SZ (sizeof(INTERNAL_SIZE_T)) #define MALLOC_ALI
4. OProfile results with JIT samples
After profiling a Java (or other supported VM) application, the OProfile JIT support creates ELF binaries from the intermediate files that were written by the agent library. The ELF binaries are named <tgid>.jo. With the symbol information stored in these ELF files, it is possible to map samples to the appropriate symbols. The usual analysis tools (opreport and/or opannotate) can now be used to ge
Profiling Java Code - Fiji
If your Java code seems to have bottlenecks that make it run slow, you need to profile the code. There are a couple of options you have for that. JProfiler JProfiler is a commercial program, but you can evaluate a trial version. HPROF Use HPROF, a profiler included in Sun's Java since version 1.5. VisualVM Use VisualVM (a tutorial can be found here), which is a graphical alternative to HPROF, a
リリースã€éšœå®³æƒ…å ±ãªã©ã®ã‚µãƒ¼ãƒ“スã®ãŠçŸ¥ã‚‰ã›
最新ã®äººæ°—エントリーã®é…ä¿¡
j次ã®ãƒ–ックマーク
kå‰ã®ãƒ–ックマーク
lã‚ã¨ã§èªã‚€
eコメント一覧を開ã
oページを開ã
Intel Westmere microarchitecture performance counter events
Intel Nehalem microarchitecture performance counter events
FAST: fast architecture sensitive tree search on modern CPUs and GPUs
IBM Developer
Dynamic Tracing and Profiling
Google Code Archive - Long-term storage for Google Code Project Hosting.
OProfile JIT agent developer guide
{{#tags}}- {{label}}
{{/tags}}