<p dir="ltr">Throwing in a "crazy" idea: Replace the default malloc with Google's performance tools tcmalloc. Not the tcmalloc_minimal that anyone can preload, but the full implementation. </p>
<p dir="ltr">-Alex</p>
<div class="gmail_quote">On Nov 22, 2013 3:09 AM, "John Marino" <<a href="mailto:dragonflybsd@marino.st">dragonflybsd@marino.st</a>> wrote:<br type="attribution"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
On 11/21/2013 20:00, Justin Sherrill wrote:<br>
> On Thu, Nov 21, 2013 at 11:03 AM, John Marino <<a href="mailto:dragonflybsd@marino.st">dragonflybsd@marino.st</a><br>
> <mailto:<a href="mailto:dragonflybsd@marino.st">dragonflybsd@marino.st</a>>> wrote:<br>
><br>
><br>
><br>
> This is a good list to have. It's all maintenance material, though.<br>
> Updates from other BSDs. That's valuable, because it saves us a lot of<br>
> labor, but doesn't anybody have some crazy ideas?<br>
<br>
<br>
I would not classify much of this work as maintenance.<br>
<br>
The semaphores is pthreads work -- nobody has touched threads in years<br>
from what I can see. Not trivial.<br>
<br>
Bringing in llvm/clang/libc++ is definitely not trivial. It a huge,<br>
complex task.<br>
<br>
the new USB has never reached production so that's not maintenance either.<br>
<br>
OSS/ALSA - When was the last time somebody in dragonfly touched audio?<br>
<br>
<br>
Sure there's room for "crazy" but this list is plenty challenging.<br>
</blockquote></div>