https://bobcopeland.com/blog/tags/memory/ Recent content in Memory on Insignificant Bits Hugo en-US Tue, 15 Mar 2016 12:27:35 +0000 https://bobcopeland.com/blog/2016/03/memory-%C2%B5-optimizing/ Tue, 15 Mar 2016 12:27:35 +0000 http://bobcopeland.com/blog/?p=2279 <p>While looking at some kmalloc statistics for my little corner of wireless, I found a new tool for the toolbox: <code>perf kmem</code>. To test it out, I ran a simulated connected 25-node mesh in mac80211_hwsim.</p> <p>Here&rsquo;s what <code>perf kmem stat --caller</code> had to say:</p> <div class="highlight"><pre tabindex="0" style="color:#f8f8f2;background-color:#272822;-moz-tab-size:4;-o-tab-size:4;tab-size:4;-webkit-text-size-adjust:none;"><code class="language-gdscript3" data-lang="gdscript3"><span style="display:flex;"><span><span style="color:#f92672">---------------------------------------------------------------------</span> </span></span><span style="display:flex;"><span> Callsite <span style="color:#f92672">|</span> Total_alloc<span style="color:#f92672">/</span>Per <span style="color:#f92672">|</span> Total_req<span style="color:#f92672">/</span>Per <span style="color:#f92672">|</span> Hit <span style="color:#f92672">|</span> <span style="color:#f92672">...</span> </span></span><span style="display:flex;"><span><span style="color:#f92672">---------------------------------------------------------------------</span> </span></span><span style="display:flex;"><span> mesh_table_alloc<span style="color:#f92672">+</span><span style="color:#ae81ff">3</span>e <span style="color:#f92672">|</span> <span style="color:#ae81ff">1600</span><span style="color:#f92672">/</span><span style="color:#ae81ff">32</span> <span style="color:#f92672">|</span> <span style="color:#ae81ff">400</span><span style="color:#f92672">/</span><span style="color:#ae81ff">8</span> <span style="color:#f92672">|</span> <span style="color:#ae81ff">50</span> <span style="color:#f92672">|</span> </span></span><span style="display:flex;"><span> mesh_rmc_init<span style="color:#f92672">+</span><span style="color:#ae81ff">23</span> <span style="color:#f92672">|</span> <span style="color:#ae81ff">204800</span><span style="color:#f92672">/</span><span style="color:#ae81ff">8192</span> <span style="color:#f92672">|</span> <span style="color:#ae81ff">102600</span><span style="color:#f92672">/</span><span style="color:#ae81ff">4104</span> <span style="color:#f92672">|</span> <span style="color:#ae81ff">25</span> <span style="color:#f92672">|</span> </span></span><span style="display:flex;"><span> mesh_path_new<span style="color:#f92672">.</span><span style="color:#66d9ef">const</span> <span style="color:#f92672">|</span> <span style="color:#ae81ff">315904</span><span style="color:#f92672">/</span><span style="color:#ae81ff">512</span> <span style="color:#f92672">|</span> <span style="color:#ae81ff">241864</span><span style="color:#f92672">/</span><span style="color:#ae81ff">392</span> <span style="color:#f92672">|</span> <span style="color:#ae81ff">617</span> <span style="color:#f92672">|</span> </span></span><span style="display:flex;"><span>[<span style="color:#f92672">...</span>] </span></span></code></pre></div><p>The number of mesh nodes shows up in the &lsquo;Hit&rsquo; column in various ways: we can pretty easily see that mesh_table_alloc is called twice per mesh device, mesh_rmc_init is called once, and mesh_path_new is called at least once for every peer (617 ~= 25 * 24). We might scan the hit column to see if there are any cases in our code that perform unexpectedly high numbers of allocations, signifying a bug or poor design.</p>