<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/">
  <channel>
    <title>Ext4 on Taras&#39; Blog on AI, Perf, Hacks</title>
    <link>/tags/ext4/</link>
    <description>Recent content in Ext4 on Taras&#39; Blog on AI, Perf, Hacks</description>
    <image>
      <title>Taras&#39; Blog on AI, Perf, Hacks</title>
      <url>/images/papermod-cover.png</url>
      <link>/images/papermod-cover.png</link>
    </image>
    <generator>Hugo -- 0.147.0</generator>
    <language>en</language>
    <copyright>Taras Glek</copyright>
    <lastBuildDate>Tue, 15 Mar 2022 13:26:36 +0100</lastBuildDate>
    <atom:link href="/tags/ext4/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>mmap Page Fault Tracing with bpftrace and ext4</title>
      <link>/posts/mmap-page-fault-tracing-via-ext4/</link>
      <pubDate>Tue, 15 Mar 2022 13:26:36 +0100</pubDate>
      <guid>/posts/mmap-page-fault-tracing-via-ext4/</guid>
      <description>&lt;p&gt;Previous blog post on how to &lt;a href=&#34;../ebpf-mmap-page-fault-tracing/&#34;&gt;trace Firefox IO using bpftrace&lt;/a&gt; via official page_fault_user tracepoint left me a bit unsatisfied with how complicated it turned out. Complexity has potential to be error-prone and the syscall-tracing dependency makes it impossible to trace IO within the main executable.&lt;/p&gt;
&lt;p&gt;I decided to try reimplement the trace using my old approach of tracing ext4 functions that handle page-faults. This turned out to be much more robust. This is now documented on my &lt;a href=&#34;https://github.com/tarasglek/bpftrace_pagefaults/tree/main/ext4&#34;&gt;github&lt;/a&gt;. It&amp;rsquo;s ugly in that it&amp;rsquo;s dependent on internal kernel structures, but it catches 100% of the IO and requires no post-trace syscall-correlation/fudging.&lt;/p&gt;</description>
    </item>
  </channel>
</rss>
