Please sync 1:1.4.10~dfsg-1 from Debian Unstable (Main)

Bug #203948 reported by Daniel Swarbrick
18
Affects Status Importance Assigned to Milestone
zaptel (Debian)
Fix Released
Unknown
zaptel (Ubuntu)
Fix Released
Medium
Unassigned
Hardy
Fix Released
Medium
Unassigned

Bug Description

Binary package hint: zaptel

Related to Digium bug http://bugs.digium.com/view.php?id=11141 but even after applying those, compilation fails further down.

Please update 1.4.9.2 (current upstream release as of writing) and ensure it builds with m-a.

Changed in zaptel:
status: Unknown → Fix Committed
Revision history for this message
Daniel Swarbrick (pressureman) wrote :

Package has been updated to 1.4.9.2 in Debian Sid:

   * New upstream release (Closes: #464900).
   - All kernel modules moved to under kernel/
   * Converting patches to quilt (because bristuff uses them).
   * Include the separate xpp changelog file.
   * Fix a possible crash with oslec when a fax tone is detected:
     http://sourceforge.net/mailarchive/message.php?msg_name=20080217212421.GT15415%40xorcom.com
     (Closes: #447245).
   * Adjusted lintian overrides: mknod is now called from a function.
   * Adjust vzaphfc to netdevice API changes in kernel 2.6.24.
   * Once again ignoring 'make distclean' errors: it will fail if we don't
     have kernel sources / headers for current kernel :-( .
   * Remove some unnecessary changes from the florz zaphfc patch - fixes zaphfc
     warning.

I've tested compiling from this package on both Debian Lenny and Ubuntu Hardy. Both compile, load, look fine.

Please sync Ubuntu package with Debian Sid.

Changed in zaptel:
status: New → Confirmed
Revision history for this message
Test-tools (roland-verifysoft) wrote :

Hello,

I can confirm this issue, an update to the mentioned "sid" packages solves this problem.
 libtonezone1 1:1.4.9.2~dfsg
 zaptel 1:1.4.9.2~dfsg
 zaptel-source 1:1.4.9.2~dfsg
This is a blocker for hardy release...

Roland "Test-tools" Bär

Revision history for this message
Shake (shake234) wrote :

This needs to be fixed asap. I upgraded to Hardy today --> now Asterisk doesn't work anymore...
I think this is a blocker for others too.

Steve Langasek (vorlon)
Changed in zaptel:
importance: Undecided → Medium
milestone: none → ubuntu-8.04
Revision history for this message
James Westby (james-w) wrote :

Hi,

Release team, could you approve this exception request please?

Note that there has since been another upload in Debian, hence
it should be synced from *testing*, not unstable, as that is the
version that has been checked on Ubuntu.

Thanks,

James

Revision history for this message
James Westby (james-w) wrote :

Hi,

The diffstat of this change is interesting, I'll just post the summary:

   699 files changed, 201198 insertions(+), 198750 deletions(-)

It's far larger than I expected.

However, this is a milestoned bug and it doesn't compile with the
hardy kernel. If we don't take the new version we will have to pick
through the changes to pick the ones that make the current
version compile.

Thanks,

James

Revision history for this message
StefanPotyra (sistpoty) wrote :

will asterisk still build with the new package? Also, can you attach a diff of the symbols for libtonezone1?
Thanks.

Revision history for this message
Jamie Strandboge (jdstrand) wrote :

FYI, I ran into this too and rebuilt the latest sid zaptel (1.4.10~dfsg), and it compiles and works fine. Note I only use the zaptel, zttranscode and ztdummy modules.

Revision history for this message
Scott Kitterman (kitterman) wrote : Re: [Bug 203948] Re: zaptel 1.4.8~dfsg-1 fails to compile under kernel 2.6.24

Ack for upload.

Revision history for this message
Jamie Strandboge (jdstrand) wrote : Re: zaptel 1.4.8~dfsg-1 fails to compile under kernel 2.6.24

Seems I spoke too soon. While 1.4.10~dfsg compiles and loads fine, it prevents sound from working. See http://forums.digium.com/viewtopic.php?p=66616&sid=659fd5b339c9017810ddef37bf8123c0.

I will be trying the version in Debian testing and report back.

Revision history for this message
Jamie Strandboge (jdstrand) wrote :

Version 1:1.4.9.2~dfsg-1 is same as 1.4.10~dfsg-- compiles and loads, but no sound when ztdummy is loaded.

Revision history for this message
Daniel Swarbrick (pressureman) wrote :

@ Jamie Strandboge:

Using what type of channel? A colleague of mine just upgraded a Feisty box running Asterisk 1.2 to Hardy, running Asterisk 1.4 and reported no problems. He is using ztdummy from the zaptel 1.4.9.2 (Debian) package. The box is just handling SIP and IAX channels.

Revision history for this message
Scott Kitterman (kitterman) wrote :

OK. Un-Ack then. Please let us know when you get it sorted out.

Revision history for this message
Jamie Strandboge (jdstrand) wrote :

Daniel: that is exactly my configuration: Feisty to Hardy upgrade (via gutsy), asterisk 1.2 to 1.4, ztdummy from 1:1.4.9.2~dfsg-1. Handling SIP and IAX. I found if I changed kernels from -server to -generic or -386, I could get sound, but it is *very* choppy. Seems likely to just be a kernel+ztdummy+hardware issue. Bottom line, 1:1.4.9.2~dfsg-1 or 1.4.10 actually compiles and will be more usable than what is in the archive.

Revision history for this message
Scott Kitterman (kitterman) wrote : Re: [Bug 203948] Re: zaptel 1.4.8~dfsg-1 fails to compile under kernel 2.6.24

Then ack to go ahead as you see best.

Revision history for this message
Jamie Strandboge (jdstrand) wrote : Re: zaptel 1.4.8~dfsg-1 fails to compile under kernel 2.6.24

Last point of interest-- I started looking into the whole ztdummy issue and found that it typically works best when CONFIG_HZ_1000=y is in the kernel compile. -386, -generic and -server do not have this, but -rt does. So I tried -rt with the 1:1.4.9.2~dfsg-1 zaptel-source and ztdummy works great now. However, it is odd that I was running the -server kernel on feisty and did not have this problem.

Revision history for this message
Daniel Swarbrick (pressureman) wrote :

The configuration I am referring to was a fresh install of Hardy Heron (i386, server) to replace a Feisty box that had been running for about the last year. As I said, the zaptel-source (v 1.4.9.2) package was taken from the Debian repos, and seems to be working fine.

I'm not at all surprised that CONFIG_HZ=1000 makes things run better. I think the Asterisk crowd recommends this, somewhere in their documentation. However, we're running the default Ubuntu -server kernel, and not experiencing any problems (fingers crossed - the new box has only been in place a few days).

Revision history for this message
Jamie Strandboge (jdstrand) wrote :

It is likely a combination of hardware/kernel/zaptel as I mentioned previously. While the -server kernel worked the worse (no sound at all), when I recompiled -server with CONFIG_HZ=1000, it works beautifully. Note, I did not have to do this on Feisty.

Reference:
http://archives.free.net.ph/message/20070612.171818.cfeb10c0.en.html

And for posterity, seems for ztdummy to reliably work well, seems you need the following configured in your kernel:
CONFIG_RTC=y
# CONFIG_HZ_100 is not set
# CONFIG_HZ_250 is not set
# CONFIG_HZ_300 is not set
CONFIG_HZ_1000=y
CONFIG_HZ=1000
CONFIG_HPET_EMULATE_RTC=y

(note in the -server install, only need to change the CONFIG_HZ lines).

Revision history for this message
Jamie Strandboge (jdstrand) wrote :

Just tried 1.4.10~dfsg and it works well for me with CONFIG_HZ=1000 in my kernel.

Revision history for this message
Jamie Strandboge (jdstrand) wrote :

Please sync so that zaptel is usable again.

Revision history for this message
Scott Kitterman (kitterman) wrote :

Ack from motu-release.

Changed in zaptel:
status: Confirmed → Fix Released
Changed in zaptel:
status: Fix Committed → Fix Released
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.