On 14/07/17 13:40, Petr Mladek wrote:

On Wed 2017-07-12 13:11:17, Petr Mladek wrote:

quoted

On Tue 2017-07-11 15:41:50, Matt Redfearn wrote:

quoted

On 11/07/17 13:43, Petr Mladek wrote:

quoted

IMHO, the reasonable solution is to move early console code and data
out of the init sections. We should do this for the early consoles
where the corresponding real console is registered using a deferred
probe. Others should be already replaced by the real console when
printk_late_init() is called. At least this is how I understand it.

This seems like the most reasonable way forward to me as well,
though sadly will lead to some post-init kernel bloat.

I still think, however, that this patch is a reasonable change to
make.

The thing is that this patch "silently" makes the keep_bootcon
option almost unusable.

I was wrong here. I thought that most early consoles used the init
section. It was mentioned somewhere and I looked a wrong way.
But this is not true. In fact, it seems that there are
only few of them. Most early consoles have struct console
and the write() callback in the normal section that is preserved.

Matt's patch and the keep_bootcon option makes sense to me
after all. Let me to resend Matt's patch with some small
improvements and one more patch that improves the check
of early consoles that use init section. I'll keep Matt
as the author of the first patch.

Thanks for looking into this and your fix Petr!

Matt

Best Regards,
Petr