src/devices/low-ram.jd

page.title=Running Android with low RAM
@jd:body

<!--
    Copyright 2013 The Android Open Source Project

    Licensed under the Apache License, Version 2.0 (the "License");
    you may not use this file except in compliance with the License.
    You may obtain a copy of the License at

        http://www.apache.org/licenses/LICENSE-2.0

    Unless required by applicable law or agreed to in writing, software
    distributed under the License is distributed on an "AS IS" BASIS,
    WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
    See the License for the specific language governing permissions and
    limitations under the License.
-->
<div id="qv-wrapper">
  <div id="qv">
    <h2>In this document</h2>
    <ol id="auto-toc">
    </ol>
  </div>
</div>

<h2 id="intro">Introduction</h2>

<p>Android now supports devices with 512MB of RAM. This documentation is intended 
to help OEMs optimize and configure Android 4.4 for low-memory devices. Several 
of these optimizations are generic enough that they can be applied to previous 
releases as well.</p>

<h2 id="optimizations">Android 4.4 platform optimizations</h2>

<h3 id="opt-mgmt">Improved memory management</h3>
<ul>
<li>Validated memory-saving kernel configurations: Kernel Same-page Merging
(KSM), and Swap to ZRAM.</li>
<li>Kill cached processes if about to be uncached and too large.</li>
<li>Don’t allow large services to put themselves back into A Services (so they
can’t cause the launcher to be killed).</li>
<li>Kill processes (even ordinarily unkillable ones such as the current IME)
that get too large in idle maintenance.</li>
<li>Serialize the launch of background services.</li>
<li>Tuned memory use of low-RAM devices: tighter out-of-memory (OOM) adjustment
levels, smaller graphics caches, etc.</li>
</ul>

<h3 id="opt-mem">Reduced system memory</h3>
<ul>
<li>Trimmed system_server and SystemUI processes (saved several MBs).</li>
<li>Preload dex caches in Dalvik (saved several MBs).</li>
<li>Validated JIT-off option (saves up to 1.5MB per process).</li>
<li>Reduced per-process font cache overhead.</li>
<li>Introduced ArrayMap/ArraySet and used extensively in framework as a
lighter-footprint replacement for HashMap/HashSet.</li>
</ul>

<h3 id="opt-proc">Procstats</h3>
<p>
Added a new Developer Option to show memory state and application memory usage
ranked by how often they run and amount of memory consumed.
</p>

<h3 id="opt-api">API</h3>
<p>
Added a new ActivityManager.isLowRamDevice() to allow applications to detect
when running on low memory devices and choose to disable large-RAM features.
</p>

<h3 id="opt-track">Memory tracking</h3>
<p>
New memtrack HAL to track graphics memory allocations, additional information
in dumpsys meminfo, clarified summaries in meminfo (for example reported free
RAM includes RAM of cached processes, so that OEMs don’t try to optimize the
wrong thing).
</p>

<h2 id="build-time">Build-time configuration</h2>
<h3 id="flag">Enable Low Ram Device flag</h3>
<p>We are introducing a new API called <code>ActivityManager.isLowRamDevice()</code> for applications to  determine if they should turn off specific memory-intensive 
  features that work poorly on low-memory devices.</p>
<p>For 512MB devices, this API is expected to return: "true" It can be enabled by 
  the following system property in the device makefile.<br/>
<code>PRODUCT_PROPERTY_OVERRIDES += ro.config.low_ram=true</code></p>

<h3 id="jit">Disable JIT</h3>

  <p>System-wide JIT memory usage is dependent on the number of applications 
  running and the code footprint of those applications. The JIT establishes a 
  maximum translated code cache size and touches the pages within it as needed. 
  JIT costs somewhere between 3M and 6M across a typical running system.<br/>
  <br/>
  The large apps tend to max out the code cache fairly quickly (which by default 
  has been 1M). On average, JIT cache usage runs somewhere between 100K and 200K 
  bytes per app. Reducing the max size of the cache can help somewhat with 
  memory usage, but if set too low will send the JIT into a thrashing mode.  For 
the really low-memory devices, we recommend the JIT be disabled entirely.<code>
</code></p>

<p>This can be achieved by adding the following line to the product makefile:<br/>
<code>PRODUCT_PROPERTY_OVERRIDES += dalvik.vm.jit.codecachesize=0</code></p>
<h3 id="launcher">Launcher Configs</h3>


  <p>Ensure the default wallpaper setup on launcher is <strong>not</strong>
using live-wallpaper. Low-memory devices should not pre-install any live wallpapers. </p>


<h2 id="kernel">Kernel configuration</h2>
<h3 id="kernel-tuning">Tuning kernel/ActivityManager to reduce direct reclaim </h3>


  <p>Direct reclaim happens when a process or the kernel tries to allocate a page 
  of memory (either directly or due to faulting in a new page) and the kernel 
  has used all available free memory. This requires the kernel to block the 
  allocation while it frees up a page. This in turn often requires disk I/O to 
  flush out a dirty file-backed page or waiting for <code>lowmemorykiller</code> to kill a 
  process. This can result in extra I/O in any thread, including a UI thread.</p>
  
  <p>To avoid direct reclaim, the kernel has watermarks that trigger <code>kswapd</code> or 
  background reclaim.  This is a thread that tries to free up pages so the next 
  time a real thread allocates it can succeed quickly.</p>
  
  <p>The default threshold to trigger background reclaim is fairly low, around 2MB 
  on a 2GB device and 636KB on a 512MB device. And the kernel reclaims only a 
  few MB of memory in background reclaim. This means any process that quickly 
  allocates more than a few megabytes is going to quickly hit direct reclaim.</p>
  
<p>Support for a new kernel tunable is added in the android-3.4 kernel branch as 
  patch 92189d47f66c67e5fd92eafaa287e153197a454f ("add extra free kbytes 
  tunable").  Cherry-picking this patch to a device's kernel will allow 
  ActivityManager to tell the kernel to try to keep 3 full-screen 32 bpp buffers 
  of memory free.</p>
  
<p>These thresholds can be configured via the framework config.xml</p>
<p><code> &lt;!-- Device configuration setting the /proc/sys/vm/extra_free_kbytes tunable in the kernel (if it exists).  A high value will increase the amount of memory that the kernel tries to keep free, reducing allocation time and causing the lowmemorykiller to kill earlier.  A low value allows more memory to be used by processes but may cause more allocations to block waiting on disk I/O or lowmemorykiller.  Overrides the default value chosen by ActivityManager based on screen size.  0 prevents keeping any extra memory over what the kernel keeps by default.  -1 keeps the default. --&gt;<br />
&lt;integer name=&quot;config_extraFreeKbytesAbsolute&quot;&gt;-1&lt;/integer&gt;</code></p>

<code>
<p> &lt;!-- Device configuration adjusting the /proc/sys/vm/extra_free_kbytes tunable in the kernel (if it exists).  0 uses the default value chosen by ActivityManager.  A positive value  will increase the amount of memory that the kernel tries to keep free, reducing allocation time and causing the lowmemorykiller to kill earlier.  A negative value allows more memory to be used by processes but may cause more allocations to block waiting on disk I/O or lowmemorykiller.  Directly added to the default value chosen by  ActivityManager based on screen size. --&gt;<br />
  &lt;integer name=&quot;config_extraFreeKbytesAdjust&quot;&gt;0&lt;/integer&gt;</code>

<h3 id="lowmem">Tuning LowMemoryKiller</h3>


  <p>ActivityManager configures the thresholds of the LowMemoryKiller to match its 
  expectation of the working set of file-backed pages (cached pages) required to 
  run the processes in each priority level bucket.  If a device has high 
  requirements for the working set, for example if the vendor UI requires more 
memory or if more services have been added, the thresholds can be increased. </p>
<p>The thresholds can be reduced if too much memory is being reserved for file 
  backed pages, so that background processes are being killed long before disk 
thrashing would occur due to the cache getting too small.</p>
<p> <code>&lt;!-- Device configuration setting the minfree tunable in the lowmemorykiller in the kernel.  A high value will cause the lowmemorykiller to fire earlier, keeping more memory in the file cache and preventing I/O thrashing, but allowing fewer processes to stay in memory.  A low value will keep more processes in memory but may cause thrashing if set too low.  Overrides the default value chosen by ActivityManager based on screen size and total memory for the largest lowmemorykiller bucket, and scaled proportionally to the smaller buckets.  -1 keeps the default. --&gt;<br />
  &lt;integer name=&quot;config_lowMemoryKillerMinFreeKbytesAbsolute&quot;&gt;-1&lt;/integer&gt;</code></p>
<p> <code>&lt;!-- Device configuration adjusting the minfree tunable in the lowmemorykiller in the kernel.  A high value will cause the lowmemorykiller to fire earlier, keeping more memory in the file cache and preventing I/O thrashing, but allowing fewer processes to stay in memory.  A low value will keep more processes in memory but may cause thrashing if set too low.  Directly added to the default value chosen by          ActivityManager based on screen size and total memory for the largest lowmemorykiller bucket, and scaled proportionally to the smaller buckets. 0 keeps the default. --&gt;<br />
  &lt;integer name=&quot;config_lowMemoryKillerMinFreeKbytesAdjust&quot;&gt;0&lt;/integer&gt;</code></p>
<h3 id="ksm">KSM (Kernel samepage merging)</h3>


  <p>KSM is a kernel thread that runs in the background and compares pages in 
  memory that have been marked <code>MADV_MERGEABLE</code> by user-space. If two pages are 
  found to be the same, the KSM thread merges them back as a single 
  copy-on-write page of memory.</p>
  
  <p>KSM will save memory over time on a running system, gaining memory duplication 
  at a cost of CPU power, which could have an impact on battery life. You should 
  measure whether the power tradeoff is worth the memory savings you get by 
  enabling KSM.</p>
  
  <p>To test KSM, we recommend looking at long running devices (several hours) and 
  seeing whether KSM makes any noticeable improvement on launch times and 
  rendering times.</p>
  
<p>To enable KSM, enable <code>CONFIG_KSM</code> in the kernel and then add the following lines to your` <code>init.&lt;device&gt;.rc</code> file:<br>
  <code>write /sys/kernel/mm/ksm/pages_to_scan 100<br>
  write /sys/kernel/mm/ksm/sleep_millisecs 500<br>
write /sys/kernel/mm/ksm/run 1</code></p>
<p>Once enabled, there are few utilities that will help in the debugging namely : 
  procrank, librank, &amp; ksminfo. These utilities allow you to see which KSM 
  memory is mapped to what process, which processes use the most KSM memory. 
  Once you have found a chunk of memory that looks worth exploring you can use 
  either the hat utility if it's a duplicate object on the dalvik heap. </p>
<h3 id="zram">Swap to zRAM</h3>


  <p>zRAM swap can increase the amount of memory available in the system by 
  compressing memory pages and putting them in a dynamically allocated swap area 
  of memory.</p>
  
  <p>Again, since this is trading off CPU time for a small increase in memory, you 
  should be careful about measuring the performance impact zRAM swap has on your 
  system.</p>


<p>Android handles swap to zRAM at several levels:</p>

<ul>
  <li>First, the following kernel options must be enabled to use zRAM swap 
    effectively:
    <ul>
      <li><code>CONFIG_SWAP</code></li>
      <li><code>CONFIG_CGROUP_MEM_RES_CTLR</code></li>
      <li><code>CONFIG_CGROUP_MEM_RES_CTLR_SWAP</code></li>
      <li><code>CONFIG_ZRAM</code></li>
    </ul>
  </li>
  <li>Then, you should add a line that looks like this to your fstab:<br />
    <code>/dev/block/zram0 none swap defaults zramsize=&lt;size in bytes&gt;,swapprio=&lt;swap partition priority&gt;</code><br />
  <code><br />
  zramsize</code> is mandatory and indicates how much uncompressed memory you want 
    the zram area to hold. Compression ratios in the 30-50% range are usually 
  observed.<br />
  <br />
  <code>swapprio</code> is optional and not needed if you don't have more than one swap 
  area.<br />
  <br />
  </li>
  <li>By default, the Linux kernel swaps in 8 pages of memory at a time. When 
    using ZRAM, the incremental cost of reading 1 page at a time is negligible 
    and may help in case the device is under extreme memory pressure. To read 
    only 1 page at a time, add the following to your init.rc:<br />
  `write /proc/sys/vm/page-cluster 0`</li>
  <li>In your init.rc, after the `mount_all /fstab.X` line, add:<br />
  `swapon_all /fstab.X`</li>
  <li>The memory cgroups are automatically configured at boot time if the 
    feature is enabled in kernel.</li>
  <li>If memory cgroups are available, the ActivityManager will mark lower 
    priority threads as being more swappable than other threads. If memory is 
    needed, the Android kernel will start migrating memory pages to zRAM swap, 
    giving a higher priority to those memory pages that have been marked by 
    ActivityManager. </li>
</ul>
<h3 id="carveouts">Carveouts, Ion and Contiguous Memory Allocation (CMA)</h3>

  <p>It is especially important on low memory devices to be mindful about 
  carveouts, especially those that will not always be fully utilized -- for 
  example a carveout for secure video playback. There are several solutions to 
  minimizing the impact of your carveout regions that depend on the exact 
  requirements of your hardware.</p>
  <p>If hardware permits discontiguous memory 
  allocations, the ion system heap allows memory allocations from system memory, 
  eliminating the need for a carveout. It also attempts to make large 
  allocations to eliminate TLB pressure on peripherals. If memory regions must 
  be contiguous or confined to a specific address range, the contiguous memory 
  allocator (CMA) can be used.</p>
<p>This creates a carveout that the system can also 
  use of for movable pages. When the region is needed, movable pages will be 
  migrated out of it, allowing the system to use a large carveout for other 
  purposes when it is free. CMA can be used directly or more simply via ion by 
  using the ion cma heap.</p>

<h2 id="app-opts">Application optimization tips</h2>
<ul>
   <li>Review <a 
href="http://developer.android.com/training/articles/memory.html">Managing your
App's Memory</a> and these past blog posts on the same topic:
  <ul>
    <li><a
href="http://android-developers.blogspot.com/2009/01/avoiding-memory-leaks.html">http://android-developers.blogspot.com/2009/01/avoiding-memory-leaks.html</a></li>
    <li><a
href="http://android-developers.blogspot.com/2011/03/memory-analysis-for-android.html">http://android-developers.blogspot.com/2011/03/memory-analysis-for-android.html</a></li>
    <li><a
href="http://android-developers.blogspot.com/2009/02/track-memory-allocations.html">http://android-developers.blogspot.com/2009/02/track-memory-allocations.html</a></li>
    <li> <a
href="http://tools.android.com/recent/lintperformancechecks">http://tools.android.com/recent/lintperformancechecks</a></li>
    </ul>
</li>
  <li>Check/remove any unused assets from preinstalled apps - 
development/tools/findunused (should help make the app smaller).</li>
<li>Use PNG format for assets, especially when they have transparent areas</li>
<li>If writing native code, use calloc() rather than malloc/memset</li>
<li>Don't enable code that is writing Parcel data to disk and reading it later.</li>
<li>Don't subscribe to every package installed, instead use ssp filtering. Add
filtering like below:
<br />
  <code>&lt;data android:scheme=&quot;package&quot; android:ssp=&quot;com.android.pkg1&quot; /&gt;<br />
  &lt;data android:scheme=&quot;package&quot; android:ssp=&quot;com.myapp.act1&quot; /&gt;</code></li>
</ul>

<h3 id="process-states">Understand the various process states in Android</h3>

  <ul>
  <li><p>SERVICE - SERVICE_RESTARTING<br/>
  Applications that are making themselves run in the background for their own 
  reason.  Most common problem apps have when they run in the background too 
  much.  %duration * pss is probably a good "badness" metric, although this set 
  is so focused that just doing %duration is probably better to focus on the 
  fact that we just don't want them running at all.</p></li>
  <li><p>IMPORTANT_FOREGROUND - RECEIVER<br/>
  Applications running in the background (not directly interacting with the 
  user) for any reason.  These all add memory load to the system.  In this case 
  the (%duration * pss) badness value is probably the best ordering of such 
  processes, because many of these will be always running for good reason, and 
  their pss size then is very important as part of their memory load.</p></li>
  <li><p>PERSISTENT<br/>
  Persistent system processes.  Track pss to watch for these processes getting 
  too large.</p></li>
  <li><p>TOP<br/>
  Process the user is currently interacting with.  Again, pss is the important 
  metric here, showing how much memory load the app is creating while in use.</p></li>
  <li><p>HOME - CACHED_EMPTY<br/>
  All of these processes at the bottom are ones that the system is keeping 
  around in case they are needed again; but they can be freely killed at any 
  time and re-created if needed.  These are the basis for how we compute the 
  memory state -- normal, moderate, low, critical is based on how many of these 
  processes the system can keep around.  Again the key thing for these processes 
  is the pss; these processes should try to get their memory footprint down as 
  much as possible when they are in this state, to allow for the maximum total 
  number of processes to be kept around.  Generally a well behaved app will have 
  a pss footprint that is significantly smaller when in this state than when 
  TOP.</p></li>
  <li>
    <p>TOP vs. CACHED_ACTIVITY-CACHED_ACTIVITY_CLIENT<em><br/>
  </em>The difference in pss between when a process is TOP vs. when it is in either 
  of these specific cached states is the best data for seeing how well it is 
  releasing memory when going into the background.  Excluding CACHED_EMPTY state 
  makes this data better, since it removes situations when the process has 
  started for some reasons besides doing UI and so will not have to deal with 
  all of the UI overhead it gets when interacting with the user.</p></li>
  </ul>


<h2 id="analysis">Analysis</h2>
<h3 id="app-startup">Analyzing app startup time</h3>


  <p>Use "<code>adb shell am start</code>" with the <code>-P</code> or <code>--start-profiler</code> option to run 
  the profiler when your app starts. This will start the profiler almost 
  immediately after your process is forked from zygote, before any of your code 
is loaded into it.</p>
<h3 id="bug-reports">Analyze using bugreports </h3>


  <p>Now contains various information that can be used for debugging. The services 
  include <code>batterystats</code>, <code>netstats</code>, <code>procstats</code>, and <code>usagestats</code>.  You can 
  find them with lines like this:</p>


<pre>------ CHECKIN BATTERYSTATS (dumpsys batterystats --checkin) ------
7,0,h,-2558644,97,1946288161,3,2,0,340,4183
7,0,h,-2553041,97,1946288161,3,2,0,340,4183
</pre>
<h3 id="persistent">Check for any persistent processes</h3>


  <p>Reboot the device and check the processes.<br/>
  Run for a few hours and check the processes again. There should not be any 
long running processes.</p>
<h3 id="longevity">Run longevity tests</h3>


  <p>Run for longer durations and track the memory of the process. Does it 
  increase? Does it stay constant? Create Canonical use cases and run longevity tests on these scenarios</p>