Quote:
Originally Posted by loviedovie
@DaltonST
That did not help at all ;( Somehow some hours later I loose puldown menus. I do not attach or detach devices or do any other thing between a fresh restart and loosing the menus. The menu icons are there but clicking on any of the menus with a pulldown window does not do anything. The weird thing is that the debug does not create any debug event either.
|
I'm not surprised, since calibre's builtin code will not randomly mess up like that just because a user plugin fails to load with a Qt4 error.
But to double-check, you can start calibre with
Code:
calibre --ignore-plugins
to make sure that your problems aren't plugin-related in any way.
Afraid I don't know why the menus would break for you though.
That X11 error still looks weird to me.
Quote:
Originally Posted by gbm
Code:
calibre Debug log
calibre 2.47 isfrozen: True is64bit: True
Linux-4.1.3-customkernel-x86_64-with-debian-stretch-sid Linux ('64bit', 'ELF')
('Linux', '4.1.3-customkernel', '#1 SMP Wed Jul 22 12:55:12 CDT 2015')
Have you tried a 4.1.3 generic kernel?
I would go back to and try an older kernel also.
bernie |
I highly doubt that this is a kernel error. But FWIW I don't use ancient kernels either, and I haven't gotten this error.
As for custom kernels vs. generic kernels, WTF is a generic kernel anyway. Every kernel out there makes a choice about how to build, and the kernel developers don't bless any one choice.
The specific kernel config blessed by Debian sid is no more likely to fix this than the config options turned on/off by the OP's custom changes.
The kernel is not supposed to be configured to break userspace
I customize the kernel by using the -ck patchset/bfq/bfs scheduler plus cutting down on installed modules that I will never load plus a couple other random tweaks (don't bother checking for microcode, I already know my laptop doesn't have any available, GCC optimize for my CPU, etc.).
I might expect the boot process to error out, but not calibre...
(Actually my boot did recently break, but that was a kernel bug not a -ck bug.)