LTIB and M5329x Framebuffer - how to get it to work

[aLUNZ]

G'day,

We have a M5329EVB with the LTIB supplied from Freescale.  Most everything works out of the box but we are just starting to use the LCD controller.  We have enabled compile time config options as specified in the support documentation and when building the image it looks like m532xfb.o is built and linked into the kernel.

However, when booting the /var/log/messages does not show a line along the lines of 'Initing M532x Framebuffer' and using a scope to look at the control lines to the LCD show nothing happening. 

It looks like there is something simple that we are missing to actually tell the kernel to use the fb.  Is there a kernel command line (or some config file) that we need to set?

Any help would be appreciated,
Cheers,
aLUNZ

[obidon]

I have used the 5329EVB and and LogicPD 6.4" LCD with the current LTIB 20061117. The framebuffer does work but there are a some issues. There is an appnote on the LogicPD site that you may have to add a jumper to the LCD connecter.  Also I experienced a lot of streaks on the display. The solution to this was give the LCDC the highest priority in the cross-bar module. The ADS7843 touchscreen driver also did not work out of the box. I made some changes to get it to work somewhat, but I'm also anxious for a new BSP release. I was at FTF Orlando and attended a session on building a nano-X application with the 5329.

Don

[aLUNZ]

After my original post we found the embarrassing solution to our problem[1] and we have since been able to get the  M5329EVB to work with the Sharp LQ057Q3DC12 display (required some tweaking of parameters in the FB source).  Like you, we initially had some streaking problems but that turned out to be the fact that the pixel clock was over fast for the LCD panel.  Simply slowing the pixel clock (to ~6MHz) fixed that and the display now works brilliantly.

We still have some issues, one current perplexing problem (but low priority for us) is that it appears that the LTIB make always turns the Microwindows mouse config option off (i.e. we set "SERMOUSE  = Y" in the config file but the mou_ser.o is not linked in.

But apart from that, things are now working brilliantly.  Like you I am eagerly anticipating the next LTIB. 

aLUNZ

[1] <Ahem> It turns out that we (well, OK _I_) had stuffed up the make file resulting in the wrong version (older, without the FB driver) of the kernel being copied to the target.

[obidon]

I had some trouble with the nano-X compile originally also. When I changed to the SERMOUSE and rebuilt the project I still did not have the mouse pointer. Eventually I deleted the .a library files and rebuilt and then it worked. Since then I have been able to make changes to the config and rebuild with no problems.

[JWW]

aLUNZ,

We've got several kernel patches that need to be done to the current release.  The team just finished putting together a new set of patches and we tested/demo'd those at FTF last week in Orlando and they are working quite well.  We are going to try and get the patches posted ASAP.  It will probably be after the 4th of July holiday...

Just wanted to drop you a note and let you know that we have some brand new code, that needs to be posted.  Didn't want you to waste a lot of time poking at it.

-JWW

[jkaufmann]

Any update on when these changes will be posted?  Also, where would they be available?

[aLUNZ]

The last I heard was that message earlier in the thread where it was speculated that there would be a new LTIB release around about 4 July.  It might be worthwhile to post a new question onto this Forum to see if a new release is imminent.

aLUNZ