Tom Horsley writes:
> The most reproducible one is claws-mail which always
> takes 25 seconds to start, and the delay seems to be
> something to do with dbus messages. Here's some
> debug-monitor output that shows the delay:
30s is a common timeout for Internet functions, such as DNS. I
reme
On Sun, 8 Dec 2019 12:40:50 -0700
stan via users wrote:
> The fact that the delay is exactly 30 seconds supports this, as that
> sounds like a timeout value that is set.
It is definitely a timeout. An strace showed it sitting
in a poll() and timing out talking to someone somewhere
possibly on dbu
On Sun, 8 Dec 2019 14:25:30 -0500
Tom Horsley wrote:
> If I use virt-manager to tell it to emulate IvyBridge
> instead of native cpu, there is no hang!
And I've now submitted this bug:
https://bugzilla.redhat.com/show_bug.cgi?id=1780965
___
users maili
On Sun, 8 Dec 2019 14:25:30 -0500
Tom Horsley wrote:
> On Sun, 08 Dec 2019 19:12:02 +
> Patrick O'Callaghan wrote:
>
> > Maybe do an 'strace' on the process.
>
> I've done that, but nothing obvious is going on. However
> I now have a definite clue:
>
> In my fedora 31 virtual machine, th
On Sun, 08 Dec 2019 19:12:02 +
Patrick O'Callaghan wrote:
> Maybe do an 'strace' on the process.
I've done that, but nothing obvious is going on. However
I now have a definite clue:
In my fedora 31 virtual machine, the hang happens the
same way.
If I use virt-manager to tell it to emulate I
On Sun, 2019-12-08 at 14:03 -0500, Tom Horsley wrote:
> On Sun, 8 Dec 2019 11:49:41 -0700
> stan via users wrote:
>
> > What is
> > happening on top when the delay happens?
>
> Nothing I can see. I even created a new virtual machine from scratch,
> and it hangs there - I posted another message wi
On Sun, 8 Dec 2019 11:49:41 -0700
stan via users wrote:
> What is
> happening on top when the delay happens?
Nothing I can see. I even created a new virtual machine from scratch,
and it hangs there - I posted another message with details, but
it seems to be the combination of gtk2 and an Intel I9
On Sun, 8 Dec 2019 09:39:52 -0500
Tom Horsley wrote:
> The most reproducible one is claws-mail which always
> takes 25 seconds to start, and the delay seems to be
> something to do with dbus messages. Here's some
> debug-monitor output that shows the delay:
>
> sig 1575815215.835943 42
The most reproducible one is claws-mail which always
takes 25 seconds to start, and the delay seems to be
something to do with dbus messages. Here's some
debug-monitor output that shows the delay:
sig 1575815215.835943 4294967295 org.freedesktop.DBus
/org/freedesktop/DBusor
