[PATCH 0/3] Maverick ti-omap4, Modularize syslink drivers

Bryan Wu bryan.wu at canonical.com
Wed Jun 30 12:12:31 UTC 2010


On 06/30/2010 05:38 PM, Sebastien Jan wrote:
> Looping Mathieu and Nicolas
> 
> + See my answers inlined
> 
> On 06/30/2010 11:21 AM, Bryan Wu wrote:
>> On 06/30/2010 03:25 AM, Tim Gardner wrote:
>>> Bryan,
>>>
>>> TI is interested in modularizing some of the syslink drivers. The first 2
>>> patches are required before that can be done. Please have a look to make
>>> sure its the right thing to do. I assume the right way to get everything loaded is to add
>>> notify_ducatidriver to /etc/modules.
>>>
>>
>> Sebjan, will you confirm this request? IIRC, you mentioned we need built-in 
>> syslink driver before. And if we modularize it, does TI notice this failure 
>> which Tim fixed in this patchset?
> 
> Yes: up to now, our syslink driver used to be statically linked mainly because it did not support beeing built as a module. After working with our dev team, this is now fixed. I have just push some patches for that in our dev.omapzoom tree.
> 
> Now we were wondering which was the best way to auto install the syslink driver modules in the system, hence the discussion with Mathieu and Tim.
> 
> I did not try yet any auto install strategy, and have not seen the patches from Tim, so I cannot comment further.
> 
>>
>>> There is also a wicked symbol reference dependency between 2 modules that you ought to have a look at:
>>>
>>> (maverick-armel)root at tyler:/home/rtg/kern/maverick/armel/junk# depmod -b `pwd` 2.6.34-901-omap4
>>> WARNING: Module /lib/modules/2.6.34-901-omap4/kernel/drivers/dsp/syslink/multicore_ipc/syslink_ipc.ko ignored, due to loop
>>> WARNING: Loop detected: /home/rtg/kern/maverick/armel/junk/lib/modules/2.6.34-901-omap4/kernel/drivers/dsp/syslink/ipu_pm/syslink_ipu_pm.ko needs syslink_ipc.ko which needs syslink_ipu_pm.ko again!
>>> WARNING: Module /lib/modules/2.6.34-901-omap4/kernel/drivers/dsp/syslink/ipu_pm/syslink_ipu_pm.ko ignored, due to loop
>>>
>>

Tim,

I think the patch in Sebjan's tree will fix this.
http://dev.omapzoom.org/?p=integration/kernel-ubuntu.git;a=commitdiff;h=277e48777de3471b6d1f85a5c4e6762661af9c4f

And it also fixed the symbol export error.
I think we can wait for Sebjan to provide a kernel branch for testing.

Sebjan, I saw many important fix in you tree. is that ready for our Maverick kernel?

-- 
Bryan Wu <bryan.wu at canonical.com>
Kernel Developer    +86.138-1617-6545 Mobile
Ubuntu Kernel Team | Hardware Enablement Team
Canonical Ltd.      www.canonical.com
Ubuntu - Linux for human beings | www.ubuntu.com




More information about the kernel-team mailing list