[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: jffs_erasable_size: offset = 0x0034b000



On Fri, 7 Jul 2000, Sébastien Côté wrote:

> Alexander Larsson wrote:
> > 
> > On Thu, 6 Jul 2000, David Woodhouse wrote:
> > 
> > >
> > > bjorn.wesen@xxxxxxx.com said:
> > > >  We'll probably leave the 2.0 as it is, and when we switch to 2.4
> > > > we'll either use mtd or just keep going with our own little flash.c
> > > > (which works fine :)
> > >
> > > It looks like 2.4 will have the MTD code in it already - Linus included
> > > Tuesday's snapshot in 2.4.0-test3-pre4.
> > 
> > Ouch, it's got jffs too. The invalidate_inode_pages(inode) patch should
> > really be applied to it. I haven't had time to test it, but it's almost
> > certainly correct, and it is broken without it.
> > 
> 
> I don't think JFFS is ready to be included in the kernel.  It hasn't
> been tested a lot (the 2.2 and 2.4 ports) and I'm sure we'll find many
> bugs (the invalidate_inode_pages bug is a good example of a huge bug
> that we almost missed).  Just today, I untarred the source code of a
> program on a JFFS partition, tried to compile it and got a read error
> from jffs_readpage().  

Doing write()'s with the O_APPEND flag is not working correctly in jffs. Some
strange read errors might occur because of that.

// Simon Kågström