Hi,
It looks like project has slowed down a bit now. Many of Foundation employees had to move on to new jobs so probably it's busy time too. Also one can get the feeling that Symbian becomes less attractive these days..
However - I still believe that there must be some passionate people out there who'd fancy bring up a new platform with Symbian.
How about pandaboard? It might attract more of you -> it is so much more stuffed with peripherals..
I am quite impressed - it's like a small computer: HDMI, DVI, LCD (and old LCDs should work - the interface is the same), USB, Ethernet, WiFi / bluetooth / FM.. Not to mention about Dual Cortex A9 at 1GHz, DSP, dual Cortex M3 subsystem, Image and video accelerator, 1GB SDRAM and many others.. It's more powerful than my mom's laptop ;) Is like HD set-top-box/multimedia computer - that could fit into cookie-box… It's hard to imagine what these 'phones' are now to be - it has less and less to do with 'embedded'
Obviously - you probably know all that.. a lot of info can be found at pandaboard.org and elswhere.
Few other OS distributions already support it. I know it's a hard work.. but maybe we could add Symbian to the list too?
I had a chance to play with it this weekend and since un-boxing and playing with other OSes - I have managed to get the boot environment that should be sufficient for start.
Later - hacking some beagle code - I managed to get some 'Symbian' code executed!
Not much of it (debug output of the bootstrap) - but at least something! And it is a very hard work to bring such a complicated platform. I don't know if it's doable by one person in reasonable time..
It is, however very similar to omap3530 which we pretty-much have working - many peripherals are similar or the same and even are mapped to similar or even the same addresses. Many drivers should be easy to move with small changes.
These are:
- bootloader (for now taken from ubuntu distribution as I did not have with few different branches of uboot/x-loader that would run more stable with some recent changes ..(or I didn't spend enough time on that yet;) )
- panda.rom.img -> hacked code of the beagleboard that successfully runs the bootstrap up to the point, where MMU is initialized.
To run it - all files have to be extracted and copied to the SD card. Only requirements are that it has to be bootable partition on that SD card and MLO should be copied first after that partition is formated (as first file).They you can attach the serial and power (either from usb or supply), stick the card it - and it should start, although it's really nothing yet.
I've managed to finish a bit more the bootstrap with proper mapping - and I'm trying to clean the code up a bit now. As soon as I get something more or less 'submittable' I shall push it to the repo along with instruction how to build etc.
Hopefully there'll be more people interested :)
Cheers,
Lukasz.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Getting something work on the Pandaboard is definitely a big plus and would be much more interesting as well … , the possibilities then are infinite …
May be we can see if we can do a "boot-camp" or something , to sit together and hack it down …. :-) , just a thought ..
Anyways , as you correctly said that many of us had disappeared (including me) , but I think now that people are getting settled , we should be able to put in some more time and passion into the project and keep it alive :-) .. at least I am committed to put my two cents to the project :-) …
In next few weeks , I will be focusing more to get the wiki in a better shape and support Antrax in shipping the product ..
Arunabh
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
So I sat down and hacked a bit myself.. which resulted in an early drop of a well structured (I believe so) baseport for pandaboard.
I have copied / refactored relevant drivers of the beagleboard, however many of them need work to actually port / update them.
So far mapping and some addresses for timers were updated. Bootstrap still needs some debugging which I didn't have time to bring further. The implementation of serial debug port is fine for the bootstrap and it might need not many changes (if any) in uart / euart. Although there's still lots of work to do and if it happens that I'll have another 'sit-down' session I shall share it / commit it :)
It should properly export headers and other stuff and nicely compile. Rom can also be built - see 'submission_log.txt' for more info.
I think that maybe beagleboard structure could be re-arranged a little bit too in a similar way at some point (I might try to do that on a branch or something)
Any ideas /thoughts / feedback - please let me know. Regards, L.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Small update - next good news: after last changes boot gets further->some of the log:
Mmu::Init1
MmuBase::Init1
A::Init1()
Panda::Init1()
Omap4430Assp::Init1()
Omap4430Interrupt::Init1()
Omap4430Interrupt::DisableAndClearAll
It obviously crashes at the last one, as interrupts need re-implementation (GIC has to be configured for single core option).
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi,
It looks like project has slowed down a bit now. Many of Foundation employees had to move on to new jobs so probably it's busy time too. Also one can get the feeling that Symbian becomes less attractive these days..
However - I still believe that there must be some passionate people out there who'd fancy bring up a new platform with Symbian.
How about pandaboard? It might attract more of you -> it is so much more stuffed with peripherals..
I am quite impressed - it's like a small computer: HDMI, DVI, LCD (and old LCDs should work - the interface is the same), USB, Ethernet, WiFi / bluetooth / FM.. Not to mention about Dual Cortex A9 at 1GHz, DSP, dual Cortex M3 subsystem, Image and video accelerator, 1GB SDRAM and many others.. It's more powerful than my mom's laptop ;) Is like HD set-top-box/multimedia computer - that could fit into cookie-box… It's hard to imagine what these 'phones' are now to be - it has less and less to do with 'embedded'
Obviously - you probably know all that.. a lot of info can be found at pandaboard.org and elswhere.
Few other OS distributions already support it. I know it's a hard work.. but maybe we could add Symbian to the list too?
I had a chance to play with it this weekend and since un-boxing and playing with other OSes - I have managed to get the boot environment that should be sufficient for start.
Later - hacking some beagle code - I managed to get some 'Symbian' code executed!
Not much of it (debug output of the bootstrap) - but at least something! And it is a very hard work to bring such a complicated platform. I don't know if it's doable by one person in reasonable time..
It is, however very similar to omap3530 which we pretty-much have working - many peripherals are similar or the same and even are mapped to similar or even the same addresses. Many drivers should be easy to move with small changes.
Anyway - I can't still edit the wiki page.. nor attach files to project so I have pushed rom files to the repo under:
http://wildducks.hg.sourceforge.net/hgweb/wildducks/wildducks/rev/76393e12d751
These are:
- bootloader (for now taken from ubuntu distribution as I did not have with few different branches of uboot/x-loader that would run more stable with some recent changes ..(or I didn't spend enough time on that yet;) )
- panda.rom.img -> hacked code of the beagleboard that successfully runs the bootstrap up to the point, where MMU is initialized.
To run it - all files have to be extracted and copied to the SD card. Only requirements are that it has to be bootable partition on that SD card and MLO should be copied first after that partition is formated (as first file).They you can attach the serial and power (either from usb or supply), stick the card it - and it should start, although it's really nothing yet.
I've managed to finish a bit more the bootstrap with proper mapping - and I'm trying to clean the code up a bit now. As soon as I get something more or less 'submittable' I shall push it to the repo along with instruction how to build etc.
Hopefully there'll be more people interested :)
Cheers,
Lukasz.
That's great news !!!
Getting something work on the Pandaboard is definitely a big plus and would be much more interesting as well … , the possibilities then are infinite …
May be we can see if we can do a "boot-camp" or something , to sit together and hack it down …. :-) , just a thought ..
Anyways , as you correctly said that many of us had disappeared (including me) , but I think now that people are getting settled , we should be able to put in some more time and passion into the project and keep it alive :-) .. at least I am committed to put my two cents to the project :-) …
In next few weeks , I will be focusing more to get the wiki in a better shape and support Antrax in shipping the product ..
Arunabh
Hi,
So I sat down and hacked a bit myself.. which resulted in an early drop of a well structured (I believe so) baseport for pandaboard.
I have copied / refactored relevant drivers of the beagleboard, however many of them need work to actually port / update them.
So far mapping and some addresses for timers were updated. Bootstrap still needs some debugging which I didn't have time to bring further. The implementation of serial debug port is fine for the bootstrap and it might need not many changes (if any) in uart / euart. Although there's still lots of work to do and if it happens that I'll have another 'sit-down' session I shall share it / commit it :)
Anyway - code is here:
http://wildducks.hg.sourceforge.net/hgweb/wildducks/pandaboard/rev/c1e3c0808699
Simply clone it to \sf\adaptation\pandaboard.
It should properly export headers and other stuff and nicely compile. Rom can also be built - see 'submission_log.txt' for more info.
I think that maybe beagleboard structure could be re-arranged a little bit too in a similar way at some point (I might try to do that on a branch or something)
Any ideas /thoughts / feedback - please let me know. Regards, L.
Small update - next good news: after last changes boot gets further->some of the log:
Mmu::Init1
MmuBase::Init1
A::Init1()
Panda::Init1()
Omap4430Assp::Init1()
Omap4430Interrupt::Init1()
Omap4430Interrupt::DisableAndClearAll
It obviously crashes at the last one, as interrupts need re-implementation (GIC has to be configured for single core option).