On Sat, Aug 21, 2010 at 10:26:26PM -0000, Lennie wrote:
> I think maybe the debug I enabled made startup of consolekit slow enough
> that it doesn't happen for me anymore. Because I can't reproduce. :-)
>
> Can I suggest adding some debug to consolekit in the if "if (errno ==
> EINVAL && data->retries++ < 10) {", so we can see if we trigger it ?
You should still get "Error waiting for native console 1 activation:
Invalid argument" etc. with my patch. Furthermore, if you arrange for
console-kit-daemon to be run with --debug, then you should get
"VT_WAITACTIVE for vt 1" and "VT_WAITACTIVE for vt 1 returned
<something>" around any calls to that ioctl, whether successful or not,
which should be enough to see what's going on.
Still, I've added a few more bits of debug to the version of my patch
that I'll send upstream if this works out.
On Sat, Aug 21, 2010 at 10:26:26PM -0000, Lennie wrote:
> I think maybe the debug I enabled made startup of consolekit slow enough
> that it doesn't happen for me anymore. Because I can't reproduce. :-)
>
> Can I suggest adding some debug to consolekit in the if "if (errno ==
> EINVAL && data->retries++ < 10) {", so we can see if we trigger it ?
You should still get "Error waiting for native console 1 activation:
Invalid argument" etc. with my patch. Furthermore, if you arrange for
console-kit-daemon to be run with --debug, then you should get
"VT_WAITACTIVE for vt 1" and "VT_WAITACTIVE for vt 1 returned
<something>" around any calls to that ioctl, whether successful or not,
which should be enough to see what's going on.
Still, I've added a few more bits of debug to the version of my patch
that I'll send upstream if this works out.