Skip to contentSkip to conversation

OMAP FB drivers muck up the FB menu?

DISCUSSION Dirk Behme 2 participants 2 messages Oct 22, 2007 – Oct 31, 2007 Framebuffer layer

lore
Share & tools
lore

Dirk Behme

author
#0 lore raw

Please see below answer from Randy. Seems that it didn't find its way to the list because Randy isn't subscribed.

Randy: Thanks for explanation, investigation and patch!

Dirk

-------- Original Message --------
Subject: Re: OMAP FB drivers muck up the FB menu?
Date: Sun, 21 Oct 2007 21:55:06 -0700
From: Randy Dunlap <[email protected]>
Organization: YPO4
To: [email protected]
CC: [email protected]

[I'm having problems sending email thru oracle.com, so I'm replying from my personal account instead...]

Quoted 17 lines · Dirk Behme

In

http://www.kernel.org/pub/linux/kernel/v2.6/snapshots/patch-2.6.23-git15.log

I just found this:

-- cut --
commit 0058f479e52d0c0718c843cb34223bc1bfce36e1

Author: Randy Dunlap <[email protected]>
Date:   Tue Oct 16 01:29:35 2007 -0700

video gfx: fix menu ordering

Move video graphics driver configs to fix menus:

Quoted 8 lines

...
The OMAP FB drivers still muck up the FB menu a bit, so I put
OMAP drivers at the end of the FB menu.

...
-- cut --

Anybody with an idea what is meant there?

All of the frame buffer drivers should be listed (indented) under the menu heading "Support for frame buffer devices". If I TI_OMAP board type and then view the Graphics Support menu, it looks like this:

x x [ ] Backlight & LCD device support --->
x x
x x Display device support --->
x x
x x <M> Lowlevel video output switch controls
x x
x x <*> Support for frame buffer devices
x x
x x [ ] Enable firmware EDID (NEW)
x x
x x [ ] Enable Video Mode Handling Helpers (NEW)
x x
x x [ ] Enable Tile Blitting Support (NEW)
x x
x x --- Frame buffer hardware drivers
x x
x x < > Epson S1D13XXX framebuffer support (NEW)
x x
x x < > OMAP frame buffer support (EXPERIMENTAL) (NEW)
x x
x x < > Virtual Frame Buffer support (ONLY FOR TESTING!) (NEW)
x x
x x Console display driver support --->
x x
x x [ ] Bootup logo (NEW) --->

This is usually caused by some missing config symbol(s) on the "depends on" line, but I was not able to determine the exact cause of it in this case.

Notice how OMAP is not indented like the fb drivers above it. Neither is the Virtual FB, but that's expected once a config symbol above causes a problem.

In this menu, there were also problems with PMAGB_B and with IBM_GXT4500. After those were fixed, all FB drivers except for OMAP are indented as expected. If anyone has ideas or suggestions for how to make the OMAP FB entry better, please send it.

[/me does more investigation and testing ...]

OK, I know what's happening. In drivers/video/Kconfig, these lines:

if ARCH_OMAP
source "drivers/video/omap/Kconfig"
endif

introduce a dependency on ARCH_OMAP for all config symbols in the drivers/video/omap/Kconfig file. NOTE: The construct above does NOT prevent kconfig from reading that file for all non-ARCH_OMAP architectures or platforms. It justs add a dependency. In this case that added dependency mucks up the FB menu (in not-so-technical terms).

This is easy enough to fix. Patch is attached. I'll submit it to the [email protected] list now.

[Comment from Dirk: Attachment renamed to .txt to get it properly through my mailer and into list archive]

Quoted 16 lines

In same changelog there is from Randy as well:

-- cut --
commit e82ce352335b643755a7fa4130b11d6d7adcf10c

Author: Randy Dunlap <[email protected]>
Date:   Tue Oct 16 23:29:44 2007 -0700

maintainers: linux-omap list is subscribers only

You are not allowed to post to this mailing list, and your message
has been
automatically rejected. If you think that your messages are being
rejected
in error, contact the mailing list owner at [email protected].

Quoted 7 lines

-- cut --

Randy: Did you like to contact linux-omap list regarding OMAP FB issue and got a reject?

Yes, as quoted above.

Thanks,
---
~Randy

From: Randy Dunlap <[email protected]>

Fix FB_OMAP dependencies so that the OMAP FB driver options are presented correctly.

---
 drivers/video/Kconfig      |    4 +---
 drivers/video/omap/Kconfig |    2 +-
 2 files changed, 2 insertions(+), 4 deletions(-)
drivers/video/Kconfig +1 -3
@@ -1892,9 +1892,7 @@ config FB_VIRTUAL
 
 	  If unsure, say N.
 
-if ARCH_OMAP
-	source "drivers/video/omap/Kconfig"
-endif
+source "drivers/video/omap/Kconfig"
 
 source "drivers/video/backlight/Kconfig"
 source "drivers/video/display/Kconfig"
drivers/video/omap/Kconfig +1 -1
@@ -1,6 +1,6 @@
 config FB_OMAP
 	tristate "OMAP frame buffer support (EXPERIMENTAL)"
-	depends on FB
+	depends on FB && ARCH_OMAP
 	select FB_CFB_FILLRECT
 	select FB_CFB_COPYAREA
 	select FB_CFB_IMAGEBLIT
#1 lore raw
Quoted 5 lines · Dirk Behme

Please see below answer from Randy. Seems that it didn't find its way to the list because Randy isn't subscribed.

Randy: Thanks for explanation, investigation and patch!

Great, looks like this is alreay applied today in Linus' tree.

Regards,

Tony