SMI712 driver 16bpp color problem (maybe endian related?)

SMI712 driver 16bpp color problem (maybe endian related?)

[Rolf Offermanns]

Hi All,
I am trying to port a fb driver for the SMI712 chip from kernel 2.4 to 2.6.

I use the 2.4 driver from SMIs homepage as a template. Having tried for
several days now, I could need some help in understanding a few things.

I know, that some of the below writings may look confused, because that's
the state my mind is in right now. :)

8bpp pseudo-color mode seems to work OK right now, but I am struggeling with
16bpp mode. 

According the the manual the chips support RGB565 16bpp. I changed the rgb
offset/length fields to:

red: 11/5
green: 5/6
blue: 0/5

I test the driver with the XF86FBDev server from x.org-6.8.1. Using the
xsetroot command for setting the root window color to red/green/blue, I
noticed, that the color were mixed up like this:

red -> blue
green -> red
blue -> green

So I changed the offsets to:

red: 5
green: 0
blue: 11

Now I get the expected result for xsetroot. However the color of my window
decorations look wrong. I use wm2 as window manager. I then tried fbtest
from CVS (w/o the image test, as I do not have netpbm on my target) and
there the color look wrong, too. I can post a screenshot (digicam) if that
is of any help (only tomorrow, because I have the setup in my office).

Does anyone has got an idea of what might be wrong here?

I read about endian problems in the archive.  My target is a ppc440ep (big
endian) and the chip is on a PCI card. Could this be the problem?

Any hint is greatly appreciated.

Thanks,
Rolf



-------------------------------------------------------
This SF.Net email is sponsored by the JBoss Inc.
Get Certified Today * Register for a JBoss Training Course
Free Certification Exam for All Training Attendees Through End of 2005
Visit http://www.jboss.com/services/certification for more information

SMI712 driver 16bpp color problem (maybe endian related?)

[Rolf Offermanns]

Hi All,
I am trying to port a fb driver for the SMI712 chip from kernel 2.4 to 2.6.

I use the 2.4 driver from SMIs homepage as a template. Having tried for
several days now, I could need some help in understanding a few things.

I know, that some of the below writings may look confused, because that's
the state my mind is in right now. :)

8bpp pseudo-color mode seems to work OK right now, but I am struggeling with
16bpp mode. 

According the the manual the chips support RGB565 16bpp. I changed the rgb
offset/length fields to:

red: 11/5
green: 5/6
blue: 0/5

I test the driver with the XF86FBDev server from x.org-6.8.1. Using the
xsetroot command for setting the root window color to red/green/blue, I
noticed, that the color were mixed up like this:

red -> blue
green -> red
blue -> green

So I changed the offsets to:

red: 5
green: 0
blue: 11

Now I get the expected result for xsetroot. However the color of my window
decorations look wrong. I use wm2 as window manager. I then tried fbtest
from CVS (w/o the image test, as I do not have netpbm on my target) and
there the color look wrong, too. I can post a screenshot (digicam) if that
is of any help (only tomorrow, because I have the setup in my office).

Does anyone has got an idea of what might be wrong here?

I read about endian problems in the archive.  My target is a ppc440ep (big
endian) and the chip is on a PCI card. Could this be the problem?

Any hint is greatly appreciated.

Thanks,
Rolf



-------------------------------------------------------
This SF.Net email is sponsored by the JBoss Inc.
Get Certified Today * Register for a JBoss Training Course
Free Certification Exam for All Training Attendees Through End of 2005
Visit http://www.jboss.com/services/certification for more information