Tested new package from proposed. Run Jackdbus, Run zita-j2a with USB
device plugged in. That device shows up on the jack graph and will take
connections. Start htop and show zita processes. Remove USB device while
watching htop. Cpu use remains stable at .5 to 2% (same as running
normally). Device remains on the graph (normal) which allows studio-
controls time to switch connections to another device before kill SIGINT
zita process.

So the bug fix does correct the problem of high cpu
The bug fix does not cause any other regressions in the software that relies on 
zita-ajbridge

-- 
You received this bug notification because you are a member of Ubuntu
Studio Bugs, which is subscribed to zita-ajbridge in Ubuntu.
Matching subscriptions: Ubuntu Studio Bugs
https://bugs.launchpad.net/bugs/1889146

Title:
  [SRU] Loss of USB device causes zita-ajbridge to run cpu at 100% and
  hang

Status in zita-ajbridge package in Ubuntu:
  Fix Released
Status in zita-ajbridge source package in Focal:
  Fix Committed
Status in zita-ajbridge source package in Groovy:
  Fix Released

Bug description:
  [Impact]

   * zita-ajbridge allows bridging extra alsa devices to jackd as a client. 
     If that extra device is a USB device and is unplugged, the bridge uses 
     100% cpu and slows the rest of the system to a crawl. The only way to 
     stop this process is with a kill -9.

  [Test Case]

   * Using (Ubuntu) Studio Controls, start Jack with built-in sound card and 
     a USB device.

   * Unplug USB device. Bridge will now use 100% CPU process and stall the 
     system.

  [Regression Potential]

   * The upstream version has one single change consisting of the bugfix to 
     prevent the 100% CPU usage on disconnect. While it is possible that 
     this fix doesn't do what it advertises, the regression potential here 
     is slim to none.

  [Additional Info]

   * As already noted, this fix is already in Groovy.

  
  Original bug report follows:

  zita-ajbridge allows bridging extra alsa devices to jackd as a client.
  If that extra device is a USB device and is unplugged, the bridge uses
  100% cpu and slows the rest of the system to a crawl. The only way to
  stop this process is with a kill -9.

  However, there is already a bugfix release upstream. The release notes for 
version 0.8.4 (which is already in Groovy) are only one line:
  "zita-ajbridge 0.8.4     (05/04/2020)

  * Fixed 100% CPU on sound card disconnect.

  ProblemType: Bug
  DistroRelease: Ubuntu 20.04
  Package: zita-ajbridge 0.8.2-1
  ProcVersionSignature: Ubuntu 5.4.0-42.46-lowlatency 5.4.44
  Uname: Linux 5.4.0-42-lowlatency x86_64
  ApportVersion: 2.20.11-0ubuntu27.4
  Architecture: amd64
  CasperMD5CheckResult: skip
  CurrentDesktop: KDE
  Date: Mon Jul 27 20:28:00 2020
  InstallationDate: Installed on 2020-04-23 (96 days ago)
  InstallationMedia: Kubuntu 20.04 LTS "Focal Fossa" - Release amd64 (20200422)
  SourcePackage: zita-ajbridge
  UpgradeStatus: No upgrade log present (probably fresh install)

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/zita-ajbridge/+bug/1889146/+subscriptions

-- 
Mailing list: https://launchpad.net/~ubuntustudio-bugs
Post to     : ubuntustudio-bugs@lists.launchpad.net
Unsubscribe : https://launchpad.net/~ubuntustudio-bugs
More help   : https://help.launchpad.net/ListHelp

Reply via email to