On Tue 2021-01-05 17:17:08, Petr Mladek wrote:
> On Tue 2020-12-22 09:43:48, David Gow wrote:
> > On Tue, Dec 22, 2020 at 4:02 AM Andy Shevchenko
> > <andriy.shevchenko@xxxxxxxxxxxxxxx> wrote:
> > >
> > > On Mon, Dec 21, 2020 at 09:40:08PM +0200, Andy Shevchenko wrote:
> > > > +Cc people from culprit commit
> > >
> > > Guys, revert helps. I am open to test any solution you may propose, thanks!
> > >
> > > ...
> > >
> > > > # first bad commit: [757055ae8dedf5333af17b3b5b4b70ba9bc9da4e] init/console: Use ttynull as a fallback when there is no console
> > >
> > > --
> >
> > +CC linux-um
> >
> > There appear to be two problems here:
> > 1. UML now no longer has console output by default (which KUnit needs
> > to get results)
>
> > This can be worked around for KUnit by passing console=tty to the
> > kernel. I don't think this is a "correct" fix
>
> It is rather a workaround. ttynull was supposed to be an ultimate
> fallback to provide a "valid" stdin, stdout, and stderr for
> the init process. ttyX still should be used by default when
> there is no console defined on the command line.
>
> So the question is why ttyX was not registered with this patch.
>
> I see the problem even when I revert the commit
> 757055ae8dedf5333af ("init/console: Use ttynull as a fallback when
> there is no console") and enable the ttynull driver as built in:
>
> CONFIG_NULL_TTY=y
>
> By other words, the problem existed even before. The commit only
> made it visible by default.
>
> I am still trying to understand arch/um and kunit code. I wonder
> if it is somehow related to stdiocons implemented in
> arch/um/drivers/stdio_console.c.
The following change solved the problem for me as well. It causes
that ttynull is initialized after stdiocons console.
diff --git a/drivers/tty/ttynull.c b/drivers/tty/ttynull.c
index eced70ec54e1..602af4d30bd4 100644
--- a/drivers/tty/ttynull.c
+++ b/drivers/tty/ttynull.c
@@ -121,7 +121,6 @@ static void __exit ttynull_exit(void)
tty_port_destroy(&ttynull_port);
}
-module_init(ttynull_init);
-module_exit(ttynull_exit);
+late_initcall_sync(ttynull_init);
MODULE_LICENSE("GPL v2");
But I am not completely sure that it is the right solution.
It is strange. Console should get registered only when
it was added by add_preferred_console(). It means that
ttynull_init() should not register by default.
Some clue might be in stderr_console. It has
to be explicitly unregistered to avoid staying as
the default console, see unregister_stderr() in
arch/um/drivers/stderr_console.c
I am going to dig more into it.
Best Regards,
Petr
