Buildroot Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
To: buildroot@busybox.net
Subject: [Buildroot] [RFC v2] linux: add fbtft kernel extension
Date: Wed, 31 Dec 2014 09:20:41 +0100	[thread overview]
Message-ID: <20141231092041.2b7cfa41@free-electrons.com> (raw)
In-Reply-To: <20141230232807.GD3928@free.fr>

Dear Yann E. MORIN,

On Wed, 31 Dec 2014 00:28:07 +0100, Yann E. MORIN wrote:

> OK, so it seems it will not be possible to biuld it as an out-of-tree
> module, after all. DAmn, that's really, really bad... :-(
> 
> So, here's why: the code uses FB-related functions that are built from
> separate source files, most of them conditionalised to a Kconfig symbol.
> But those symbols are promptless, and can only be selected.
> 
> Thus, unless another frambuffer implementation is enabled, that selects
> them, they wonlt be enabled, and thus the corresponding source files
> will not be built, and the needed functions will not be in the Kernel
> (neither built-in, nor in a module).
> 
> Moreover, one of the symbol, CONFIG_FB_BACKLIGHT, can only be selected
> by PCI-dependent symbols (DRM stuff, like radeon or nouvea), or SuperH
> dependent symbols (for some SuperH framebuffers).
> 
> Almost all needed symbols can be selected by cheating, and selecting the
> virtual framebuffer, but that's not a real sane option. And we'd still
> miss two symbols anyway.

Hum, indeed, that is a very good reason why external modules cannot
work. Thanks for the analysis.

> The proper solution would be to patch the kernel to add a kind of
> "framebuffer library for external FB drivers" (a bit like we have in the
> "Library routines" sub-menu for CRC and a few other functions, that
> would select all the hidden knobs for helper functions, and build them
> (built-in or module), and send that patch upstream.

No the proper solution would be for those fbtft people to submit their
drivers to the upstream kernel. But of course, since those drivers are
developed by RasberryPi folks, they cannot play with the normal
open-source rules of contributing things upstream: they *have* to do
some utter crap and do everything in their own fork.

Sigh. RasberryPi is definitely the worst platform ever when it comes to
learning embedded Linux.

Best regards,

Thomas
-- 
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com

  reply	other threads:[~2014-12-31  8:20 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-12-30 19:58 [Buildroot] [RFC v2] linux: add fbtft kernel extension Peter Seiderer
2014-12-30 19:58 ` Peter Seiderer
2014-12-30 22:36 ` Yann E. MORIN
2014-12-30 23:00   ` Peter Seiderer
2014-12-30 23:28   ` Yann E. MORIN
2014-12-31  8:20     ` Thomas Petazzoni [this message]
2014-12-31  9:14       ` Thomas Petazzoni
2014-12-31  9:23         ` Angelo Compagnucci

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20141231092041.2b7cfa41@free-electrons.com \
    --to=thomas.petazzoni@free-electrons.com \
    --cc=buildroot@busybox.net \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox