<html>
<body>
<div style="font-family: Verdana, Arial, Helvetica, Sans-Serif;">
<table bgcolor="#f9f3c9" width="100%" cellpadding="8" style="border: 1px #c9c399 solid;">
<tr>
<td>
This is an automatically generated e-mail. To reply, visit:
<a href="https://reviewboard.asterisk.org/r/3175/">https://reviewboard.asterisk.org/r/3175/</a>
</td>
</tr>
</table>
<br />
<blockquote style="margin-left: 1em; border-left: 2px solid #d0d0d0; padding-left: 10px;">
<p style="margin-top: 0;">On February 4th, 2014, 4:53 p.m. UTC, <b>Corey Farrell</b> wrote:</p>
<blockquote style="margin-left: 1em; border-left: 2px solid #d0d0d0; padding-left: 10px;">
<pre style="white-space: pre-wrap; white-space: -moz-pre-wrap; white-space: -pre-wrap; white-space: -o-pre-wrap; word-wrap: break-word;">I like what you've done in general, I think the use of ao2 objects in timing implementations can be possibly stopped.
Most (if not all) users of struct ast_timer are already protected by a lock or are owned exclusively by a single thread. If I'm right the locking within timing implementations are of no use, except maybe in res_timing_pthread.
As for refcounting, most of it has been removed since you no longer ao2_find the timer data. Since timer lifetime is controlled by ast_timer_open / ast_timer_close, refcounting is not useful.</pre>
</blockquote>
<p>On February 4th, 2014, 5:02 p.m. UTC, <b>Joshua Colp</b> wrote:</p>
<blockquote style="margin-left: 1em; border-left: 2px solid #d0d0d0; padding-left: 10px;">
<pre style="white-space: pre-wrap; white-space: -moz-pre-wrap; white-space: -pre-wrap; white-space: -o-pre-wrap; word-wrap: break-word;">I thought so as well but after looking back at why locking exists in timerfd for example it was added because in certain cases it wasn't protected by a lock. I wanted to keep my changes as minimal as possible so as to not cause regressions in that area.</pre>
</blockquote>
</blockquote>
<pre style="white-space: pre-wrap; white-space: -moz-pre-wrap; white-space: -pre-wrap; white-space: -o-pre-wrap; word-wrap: break-word;">I'm in agreement with both of you with regards to the ao2 ref thing. Since keeping the diff minimal is a good idea, I think that removing refcounting from the kqueue, pthread, and timerfd implementations can wait.
However, I think that adding refcounting to the dahdi implementation is unnecessary and can be managed with a single malloc/free pairing.</pre>
<br />
<p>- Mark</p>
<br />
<p>On February 4th, 2014, 3:16 p.m. UTC, Joshua Colp wrote:</p>
<table bgcolor="#fefadf" width="100%" cellspacing="0" cellpadding="8" style="background-image: url('https://reviewboard.asterisk.org/static/rb/images/review_request_box_top_bg.ab6f3b1072c9.png'); background-position: left top; background-repeat: repeat-x; border: 1px black solid;">
<tr>
<td>
<div>Review request for Asterisk Developers.</div>
<div>By Joshua Colp.</div>
<p style="color: grey;"><i>Updated Feb. 4, 2014, 3:16 p.m.</i></p>
<div style="margin-top: 1.5em;">
<b style="color: #575012; font-size: 10pt;">Repository: </b>
Asterisk
</div>
<h1 style="color: #575012; font-size: 10pt; margin-top: 1.5em;">Description </h1>
<table width="100%" bgcolor="#ffffff" cellspacing="0" cellpadding="10" style="border: 1px solid #b8b5a0">
<tr>
<td>
<pre style="margin: 0; padding: 0; white-space: pre-wrap; white-space: -moz-pre-wrap; white-space: -pre-wrap; white-space: -o-pre-wrap; word-wrap: break-word;">The timing API is currently optimized for the case where timing implementations provide a file descriptor and any API invocations directly act on the file descriptor. In practice this turns out to not be the most common case. Timing implementations commonly store a structure with additional information and then have to use a container lookup on any API invocation to get to it.
The attached change removes this container lookup requirement by allowing implementations to store a pointer to their data directly. This removes the container lookup and the world is happier.</pre>
</td>
</tr>
</table>
<h1 style="color: #575012; font-size: 10pt; margin-top: 1.5em;">Testing </h1>
<table width="100%" bgcolor="#ffffff" cellspacing="0" cellpadding="10" style="border: 1px solid #b8b5a0">
<tr>
<td>
<pre style="margin: 0; padding: 0; white-space: pre-wrap; white-space: -moz-pre-wrap; white-space: -pre-wrap; white-space: -o-pre-wrap; word-wrap: break-word;">Ran the timing test using the various timing implementations on Linux and confirmed they still work. I also placed some calls to a Playback and confirmed it works.</pre>
</td>
</tr>
</table>
<h1 style="color: #575012; font-size: 10pt; margin-top: 1.5em;">Diffs</b> </h1>
<ul style="margin-left: 3em; padding-left: 0;">
<li>/trunk/res/res_timing_timerfd.c <span style="color: grey">(407196)</span></li>
<li>/trunk/res/res_timing_pthread.c <span style="color: grey">(407196)</span></li>
<li>/trunk/res/res_timing_kqueue.c <span style="color: grey">(407196)</span></li>
<li>/trunk/res/res_timing_dahdi.c <span style="color: grey">(407196)</span></li>
<li>/trunk/main/timing.c <span style="color: grey">(407196)</span></li>
<li>/trunk/include/asterisk/timing.h <span style="color: grey">(407196)</span></li>
</ul>
<p><a href="https://reviewboard.asterisk.org/r/3175/diff/" style="margin-left: 3em;">View Diff</a></p>
</td>
</tr>
</table>
</div>
</body>
</html>