https://bugs.dragonflybsd.org/https://bugs.dragonflybsd.org/favicon.ico?16293952082008-12-15T13:52:02ZDragonFlyBSD bugtrackerDragonFlyBSD - Bug #1179: panic: tcp_setpersist: retransmit pendinghttps://bugs.dragonflybsd.org/issues/1179?journal_id=56322008-12-15T13:52:02Zsepherosa
<ul></ul><p>There is slight difference between current tcp timer and original code<br />(even with the priority message, I committed yesterday). The<br />difference lies within the callout_active() tests in !tcp_timer.c<br />files. I will work out the solution within next 2~3 days.</p>
<p>Best Regards,<br />sephe</p> DragonFlyBSD - Bug #1179: panic: tcp_setpersist: retransmit pendinghttps://bugs.dragonflybsd.org/issues/1179?journal_id=56332008-12-15T13:57:00Zrumcic
<ul></ul><p>Uploaded the core dump to leaf:~rumko/crash/{kernel,vmcore}.0</p> DragonFlyBSD - Bug #1179: panic: tcp_setpersist: retransmit pendinghttps://bugs.dragonflybsd.org/issues/1179?journal_id=56352008-12-15T23:08:03Zsepherosa
<ul></ul><p>Please test/review following patch:<br /><a class="external" href="http://leaf.dragonflybsd.org/~sephe/0001-Restore-the-semantic-of-callout_active-testing-on.patch">http://leaf.dragonflybsd.org/~sephe/0001-Restore-the-semantic-of-callout_active-testing-on.patch</a></p>
<p>Best Regards,<br />sephe</p> DragonFlyBSD - Bug #1179: panic: tcp_setpersist: retransmit pendinghttps://bugs.dragonflybsd.org/issues/1179?journal_id=56362008-12-16T01:10:03Zdillon
<ul></ul><p>:Please test/review following patch:<br />:http://leaf.dragonflybsd.org/~sephe/0001-Restore-the-semantic-of-callout_active-testing-on.patch<br />:<br />:Best Regards,<br />:sephe<br />:<br />:-- <br />:Live Free or Die</p>
<pre><code>It looks pretty good to me. I like the idea of just leaving the<br /> message queued and not trying to dequeue it, but cleaning up the<br /> tt_tasks bitmap.</code></pre>
<pre><code>Is it possible to get rid of all those tp->tt_msg != NULL tests<br /> by moving tt_tasks and tt_running_tasks out of the netmsg and<br /> into the tcpcb proper?</code></pre>
<pre><code>-Matt<br /> Matthew Dillon <br /> &lt;<a class="email" href="mailto:dillon@backplane.com">dillon@backplane.com</a>&gt;</code></pre> DragonFlyBSD - Bug #1179: panic: tcp_setpersist: retransmit pendinghttps://bugs.dragonflybsd.org/issues/1179?journal_id=56382008-12-16T10:04:13Zsepherosa
<ul></ul><p>On Tue, Dec 16, 2008 at 1:07 AM, Matthew Dillon<br /><<a class="email" href="mailto:dillon@apollo.backplane.com">dillon@apollo.backplane.com</a>> wrote:</p>
<blockquote>
<p>:Please test/review following patch:<br />:http://leaf.dragonflybsd.org/~sephe/0001-Restore-the-semantic-of-callout_active-testing-on.patch<br />:<br />:Best Regards,<br />:sephe<br />:<br />:--<br />:Live Free or Die</p>
<p>It looks pretty good to me. I like the idea of just leaving the<br />message queued and not trying to dequeue it, but cleaning up the<br />tt_tasks bitmap.</p>
<p>Is it possible to get rid of all those tp->tt_msg != NULL tests<br />by moving tt_tasks and tt_running_tasks out of the netmsg and<br />into the tcpcb proper?</p>
</blockquote>
<p>The callouts should not be used by listen TCP socket, whose tt_msg is<br />NULL (I cleared it intentionally). I think we could remove the<br />tp->tt_msg != NULL testing in tcp_callout_* functions; we probably<br />only need to test tp->tt_msg != NULL in tcp_close(). If it didn't<br />work out, I would put tt_tasks and tt_running_tasks into tcpcb.</p>
<p>Best Regards,<br />sephe</p> DragonFlyBSD - Bug #1179: panic: tcp_setpersist: retransmit pendinghttps://bugs.dragonflybsd.org/issues/1179?journal_id=58342009-01-20T15:58:38Zsepherosa
<ul></ul><p>Committed:<br />a48c5dd526<br />2d42d2b0d6</p>