Comment 16 for bug 1458014

Revision history for this message
John Johansen (jjohansen) wrote :

Alessio,

so from the boot chart I am not able to say what is causing the delay. What I do see is a large gap in activity for both the cpu and i/o.
That gap lines up roughly with the start of pulse audio, but that doesn't necessarily make it the culprit. We then get a large gap of little to no activity and then what looks like the user sessions starting up, another pulse daemon, notify osd, compiz, nautilus, ...
So it looks to me like there is something waiting, and then timing out, the question is what.

The apparmor logging you are seeing is largely an artifact of policy being loaded piece meal. Apparmor actually does at least some of its load during second 4 of your boot (look for apparmor_parser), the load around 34s appears to be a secondary load initiated by cups, which is being started, and it is ensuring its policy is loaded.