<div dir="ltr"><div>There are numerous possibilities. For example, tmpfs use. You could check if the wired page count has become excessive (could be an indication of a leak). There was a bug fix made in master related to a memory leak from locked memory that was fixed on July 12 (a51ba7a69d2c5084f2 in master), you could try cherry-picking that one into your local tree and see if it helps.<br><br>You'll need to do some more investigation. The most likely possibility is tmpfs use. The wired memory leak is a possibility too but depends on the system workload.<br><br></div>-Matt<br></div><div class="gmail_extra"><br><div class="gmail_quote">On Sun, Aug 16, 2015 at 2:25 AM, YONETANI Tomokazu <span dir="ltr"><<a href="mailto:y0n3t4n1@gmail.com" target="_blank">y0n3t4n1@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Hi.<br>
<br>
One of my DragonFly box keeps eating memory and eventually it starts<br>
killing some random processes, and I need to reboot it every two<br>
weeks because of this. This is a swap-less system running AMD C60<br>
and 8G bytes of RAM, which is much more than the sum of the virtual<br>
memory of all processes, which is less than 1G bytes. I had no problem<br>
before upgrading from 4.0-RELEASE to 4.2. The kernel has been built<br>
from source as of 5f93900a, and the kernel config file X86_64_GENERIC.<br>
Do I have anywhere else to look at?<br>
<br>
Best Regards,<br>
YONETANI Tomokazu.<br>
</blockquote></div><br></div>