[Home]

Summary:ASTERISK-01121: TDM PCI Master abort
Reporter:ktsaou (ktsaou)Labels:
Date Opened:2004-02-28 18:10:52.000-0600Date Closed:2011-06-07 14:00:24
Priority:BlockerRegression?No
Status:Closed/CompleteComponents:Core/General
Versions:Frequency of
Occurrence
Related
Issues:
Environment:Attachments:
Description:I am having problems with a TDM40B card. After some time of operation the driver logs "TDM PCI Master abort" a few thousand times per second and asterisk becomes useless.

All Feb the card worked perfectly, but on Feb 27, using the latest cvs tree the card showed the problem:

Feb 27 23:03:01 [kernel] TDM PCI Master abort
               - Last output repeated 712919 times -

I have a blue card (REV E/F). I got this card in Jan and had this problem then too, but after a few days I updated the source tree and everything was working properly. Since then I updated the source tree once a week, without any further problems.

Now with the latest source tree, the problem is back.

To fix the problem I have to rmmod wcfxs and zaptel and modprobe them again.


****** ADDITIONAL INFORMATION ******

The card is installed on a GA-6VXDC7 motherboard with two PIII @933MHz, running Gentoo Linux with kernel 2.4.24 vanilla (compiled SMP).
Comments:By: ktsaou (ktsaou) 2004-02-28 18:17:59.000-0600

The dates are GMT+2.

I have not noticed anything external to trigger the problem. No calls, no Zap phones picked up or changed state, nothing. It happens while the system is left alone.

Also, I found no pattern on how much time it needs to start this. It might need a few seconds, a few minutes or several hours.
By: Mark Spencer (markster) 2004-02-28 19:06:00.000-0600

Does going back fix it?
By: ktsaou (ktsaou) 2004-02-28 19:54:28.000-0600

Already reverted to "1 week back" for the source tree of zaptel.
I'll let it run for 24 hours before responding that it is ok with that version.
By: ktsaou (ktsaou) 2004-02-28 21:03:14.000-0600

"1 week ago" still gives the problem.
Going "2 weeks ago"...
By: ktsaou (ktsaou) 2004-02-28 21:30:25.000-0600

No, this does not work either (the error produced from the older drivers is 'FXS PCI Master abort').

I have logs from Feb 19 showing that until Feb 27 there were no problems.

I can't figure why, but I cannot go back to where I was. Only the stable updates from Gentoo and asterisk CVS have been applied.

Any help?
By: Mark Spencer (markster) 2004-02-29 13:41:49.000-0600

If going back doesn't fix it, it's possible that you have some bad hardware, although as far as I know this problem only has been known to occur on the older Rev C boards.  I'm going to close this out since in any case it's clearly not an asterisk issue, but I would (if you haven't already) open a ticket with support@digium.com.