[asterisk-dev] 1.6 vs. 1.8

[Alec Davis]

> If a patch for multiple simultaneous call pickups to a local 
> channel is included in 1.8.7.0, and the storesipcause 
> configuration change solves the "deadlock-like" problems (we 
> haven't had an opportunity to test this under heavy load 
> yet), then all the 1.8.X issues we're aware of at this time 
> will be resolved.
> 

There seems to be a simpler problem than with multiple localchan calls, and
I would image very common.
A simple dialplan line of DIAL(sip/phone1&sip/phone2&sip/phone3) and then
attempt mulitple simutaneous pickups with *8, causes NULL object messages,
and will leave orpanded channels.
Doing the same with DAHDI to some fxs ports as in
DIAL(DAHDI/1&DAHDI/2&DAHDI/3), results in a segfault when multiple
simultaneous *8 pickups are attempted.

I've attempted a couple of approaches, but both were too focused on the
wrong point of view, they were reviewed at
https://reviewboard.asterisk.org/r/1353/

Richard Mudgett has an idea as how these are happening;
[
https://issues.asterisk.org/jira/browse/ASTERISK-18222?page=com.atlassian.ji
ra.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=181511#com
ment-181511 ] hopefully he can put something together.

Alistair: When the simple mulitple dial issue is fixed, hopefully it will
also covers dialling multiple localchans.
Then comes the question does it work with multiple simultaneous pickups
using multiple localchans to attempt the pickups.

Alec Davis