Re: jffs_erasable_size: offset = 0x0034b000

Re: jffs_erasable_size: offset = 0x0034b000

[David Woodhouse]

bjorn.wesen@axis.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.

/me digs into a trench and waits for the bug reports and support questions 
to roll in :)



--
dwmw2




To unsubscribe, send "unsubscribe mtd" to majordomo@infradead.org

Re: jffs_erasable_size: offset = 0x0034b000

[Alexander Larsson]

On Thu, 6 Jul 2000, David Woodhouse wrote:

> 
> bjorn.wesen@axis.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.

/ Alex




To unsubscribe, send "unsubscribe mtd" to majordomo@infradead.org

Re: jffs_erasable_size: offset = 0x0034b000

[Sébastien Côté]

Alexander Larsson wrote:
> 
> On Thu, 6 Jul 2000, David Woodhouse wrote:
> 
> >
> > bjorn.wesen@axis.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().  

I'm not sure if Linus plans on releasing 2.4 soon but I wouldn't want
JFFS to be released in the kernel is it's full of bugs.

-- 
Sébastien Côté


To unsubscribe, send "unsubscribe mtd" to majordomo@infradead.org

Re: jffs_erasable_size: offset = 0x0034b000

[Simon Kagstrom]

On Fri, 7 Jul 2000, Sébastien Côté wrote:

> Alexander Larsson wrote:
> > 
> > On Thu, 6 Jul 2000, David Woodhouse wrote:
> > 
> > >
> > > bjorn.wesen@axis.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



To unsubscribe, send "unsubscribe mtd" to majordomo@infradead.org

Re: jffs_erasable_size: offset = 0x0034b000

[David Woodhouse]

scote1@Matrox.COM said:
> 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().  

It wasn't really my intention to send it in - it was in the tarball I sent 
to Linus out of habit because I've been doing it every time I've released 
one for some time - I'm amazed that he actually took it.

But now it's in, it does seem to have attracted some attention, and with it 
at least one bug fix so far - which can only really be a good thing. 

It's marked with CONfIG_EXPERIMENTAL, and as soon as I submit the Documentation
patch which I've always promised "after the code gets merged" it'll be even 
more clearly documented as such.

I'm of the opinion that it might as well stay there now that it's in -
although it's really up to Axis to decide as it's their baby.



--
dwmw2




To unsubscribe, send "unsubscribe mtd" to majordomo@infradead.org

Re: jffs_erasable_size: offset = 0x0034b000

[Simon Kagstrom]

On Sun, 9 Jul 2000, David Woodhouse wrote:

> 
> scote1@Matrox.COM said:
> > 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().  
> 
> It wasn't really my intention to send it in - it was in the tarball I sent 
> to Linus out of habit because I've been doing it every time I've released 
> one for some time - I'm amazed that he actually took it.
> 
> But now it's in, it does seem to have attracted some attention, and with it 
> at least one bug fix so far - which can only really be a good thing. 

What bug fix is that?
 
// Simon Kågström



To unsubscribe, send "unsubscribe mtd" to majordomo@infradead.org

Re: jffs_erasable_size: offset = 0x0034b000

[David Woodhouse]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: TEXT/PLAIN; charset=UTF8, Size: 467 bytes --]

On Wed, 5 Jul 2000, [iso-8859-1] Sébastien Côté wrote:

> Are you going to port MTD to 2.0 or will you switch your kernel to
> 2.2/2.4 ?

There has already been some work on porting MTD to 2.0. I cruelly reverted
a few of the changes which were made - but they were mostly fine in
principle, it's just that the result didn't actually compile on 2.[24] any
more for me.

-- 
dwmw2




To unsubscribe, send "unsubscribe mtd" to majordomo@infradead.org