Re: [ECOS] YAFFS2 software license and eCOS software license compatiable?

[Rutger Hofman <rutger at cs dot vu dot nl>]

Jürgen Lambrecht wrote:
> Rutger Hofman wrote:
> > Jürgen Lambrecht wrote:
> >  
> > > Andrew Lunn wrote:
> > >    
> > > > On Thu, Oct 23, 2008 at 07:23:24AM -0700, Tim michals wrote:
> > > >
> > > >      
> > > > > Is the YAFFS2 software license compatible with eCOS license?
> > > > >
> > > > >         
> > > > It depends on how you define compatible.
> > > >
> > > > YAFFS2 CVS is as far as i understand GPL.
> > > >       
> > [snip]
> >  
> > > > For many commercial application of eCos this is not wanted. YAFFS has
> > > > a solution to this problem. You can pay some money and get a copy of
> > > > YAFFS which is not GPL, it has some other commercial license. See the
> > > > YAFFS website.
> > > >
> > > >       
> > > Indeed. We will probably pay for YAFFS.
> > > But is it not possible to put the YAFFS "interface code" into ecos? :
> > > - HW interface code in /devs/flash/nand and /devs/flash/[arm]/[eb9260]
> > > - file system "interface" functions
> > >  a sort of yaffs_ecos.c , just like fs/jffs2/../fs-ecos.c
> > >     
> >
> > I am working on this (announced on the list ca. a month ago). Work on
> > NAND Flash stuff is in the testing phase, with my BlackFin NAND Flash
> > Controller and an ST-Micro chip. Currently there is support for multiple
> > controllers, multiple chips, code for erase block, read/write page, bad
> > block query/mark, and under the hood a Bad Block Table that aims to be
> > compatible with MTD's, and ECC that also aims to be compatible.
> >
> > This week, I have been typing at yaffs/2 support in the manner you
> > suggest. My assumption is that the yaffs *API* alone shouldn't be GPL.
> >   
> indeed, if we are carefull, like Andrew also mailed.
> > When we are satisfied with stress tests, we will release both packages
> > (NAND Flash and yaffs wrapper) to eCos.
> >
> > That said, I have only one (soon 2) platforms, both BlackFin, and only
> > large-page nand chips. I would really like to support and test more
> > controllers and small-page nand chips (their command set is partly
> > different). But testing is impossible for me because I don't have these
> > other platforms. So I would welcome cooperation.
> >
> > Rutger
> >
> >   
> Hello Rutger,
>
> we want to cooperate; how do you see it done?

Do you agree that your fixes are fed back to eCos with its LGPL?

> I suppose you mail your latest code to us, and that we integrate it into 
> our ecos (we have a copy of a recent ecos in our CVS, 20080704), and

OK, I'll polish up, bring the documentation up-to-date, and publish this.

> then test it on our board. Our board has this nand flash: 
> "MT29F2G08AADWP-ET:D 256MB 3V3 NAND FLASH", and an Atmel ARM9: 
> AT91SAM9260 and 16MB SDRAM.

I have difficulty finding a datasheet for your flash chip. MT29F2G08AAC* 
is easily found, though. Is it similar?

Your Atmel Arm9 has an on-board NFC (NAND Flash Controller). You will 
have to make a controller driver for it. You can use my BlackFin NFC or 
my (sketchy) GPIO NFC as an example. My guess is that is won't be very 
difficult, although using the ECC hardware might involve some work. You 
might first ignore it, and use my bundled ECC software.

> Then we need your YAFFS "interface code" as mentioned above, and also 
> the YAFFS2 code itself. (We have already the latest YAFFS2 code from 
> yaffs-cvs.)

Well, this hasn't progressed much. There must be something that I don't 
understand about the first steps. I will have to ask on the YAFFS 
mailing list.

Rutger

--
Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos
and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss