[PATCH] fix type mismatch in jffs.

[PATCH] fix type mismatch in jffs.

[Stephen Hemminger]

On 2.6.0-test5 jffs generates a warning about type mismatch because it casting a short
to a pointer.  Look like an obvious typo.

Builds clean, not tested on real hardware.

diff -Nru a/fs/jffs/inode-v23.c b/fs/jffs/inode-v23.c
--- a/fs/jffs/inode-v23.c	Tue Sep  9 14:41:53 2003
+++ b/fs/jffs/inode-v23.c	Tue Sep  9 14:41:53 2003
@@ -1734,7 +1734,7 @@
 		   the device should be read from the flash memory and then
 		   added to the inode's i_rdev member.  */
 		u16 val;
-		jffs_read_data(f, (char *)val, 0, 2);
+		jffs_read_data(f, (char *)&val, 0, 2);
 		init_special_inode(inode, inode->i_mode,
 			old_decode_dev(val));
 	}

Re: [PATCH] fix type mismatch in jffs.

[viro]

On Tue, Sep 09, 2003 at 02:44:20PM -0700, Stephen Hemminger wrote:
> On 2.6.0-test5 jffs generates a warning about type mismatch because it casting a short
> to a pointer.  Look like an obvious typo.

Which it is.  Thanks for spotting.  Linux, please apply.
 
> Builds clean, not tested on real hardware.
> 
> diff -Nru a/fs/jffs/inode-v23.c b/fs/jffs/inode-v23.c
> --- a/fs/jffs/inode-v23.c	Tue Sep  9 14:41:53 2003
> +++ b/fs/jffs/inode-v23.c	Tue Sep  9 14:41:53 2003
> @@ -1734,7 +1734,7 @@
>  		   the device should be read from the flash memory and then
>  		   added to the inode's i_rdev member.  */
>  		u16 val;
> -		jffs_read_data(f, (char *)val, 0, 2);
> +		jffs_read_data(f, (char *)&val, 0, 2);
>  		init_special_inode(inode, inode->i_mode,
>  			old_decode_dev(val));
>  	}

Re: [PATCH] fix type mismatch in jffs.

[Geert Uytterhoeven]

On Wed, 10 Sep 2003 viro@parcelfarce.linux.theplanet.co.uk wrote:
> On Tue, Sep 09, 2003 at 02:44:20PM -0700, Stephen Hemminger wrote:
> > On 2.6.0-test5 jffs generates a warning about type mismatch because it casting a short
> > to a pointer.  Look like an obvious typo.
> 
> Which it is.  Thanks for spotting.  Linux, please apply.
>  
> > Builds clean, not tested on real hardware.
> > 
> > diff -Nru a/fs/jffs/inode-v23.c b/fs/jffs/inode-v23.c
> > --- a/fs/jffs/inode-v23.c	Tue Sep  9 14:41:53 2003
> > +++ b/fs/jffs/inode-v23.c	Tue Sep  9 14:41:53 2003
> > @@ -1734,7 +1734,7 @@
> >  		   the device should be read from the flash memory and then
> >  		   added to the inode's i_rdev member.  */
> >  		u16 val;
> > -		jffs_read_data(f, (char *)val, 0, 2);
> > +		jffs_read_data(f, (char *)&val, 0, 2);
> >  		init_special_inode(inode, inode->i_mode,
> >  			old_decode_dev(val));
> >  	}

Is this endian-safe?

Gr{oetje,eeting}s,

						Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
							    -- Linus Torvalds

Re: [PATCH] fix type mismatch in jffs.

[viro]

On Wed, Sep 10, 2003 at 07:14:37PM +0200, Geert Uytterhoeven wrote:
 
> Is this endian-safe?

JFFS is host-endian.  If you want to make it swing both ways - feel free,
but AFAICS it's not worth the trouble.

Re: [PATCH] fix type mismatch in jffs.

[Linus Torvalds]

On Wed, 10 Sep 2003 viro@parcelfarce.linux.theplanet.co.uk wrote:
> 
> JFFS is host-endian.  If you want to make it swing both ways - feel free,

Please don't.

Dual-endianness is _evil_.

Admittedly host-endian is stupid too, but it's less stupid than being 
dual.

The only sane thing to do is fixed-endianness. I'm sure the m68k people 
remember being forced to fix their ext2 partitions back in the bad old 
days. It's painful once, but after that, fixed-endian is a lot more 
efficient and much simpler to handle.

		Linus

Re: [PATCH] fix type mismatch in jffs.

[viro]

On Wed, Sep 10, 2003 at 11:54:13AM -0700, Linus Torvalds wrote:
> 
> On Wed, 10 Sep 2003 viro@parcelfarce.linux.theplanet.co.uk wrote:
> > 
> > JFFS is host-endian.  If you want to make it swing both ways - feel free,
> 
> Please don't.
> 
> Dual-endianness is _evil_.
> 
> Admittedly host-endian is stupid too, but it's less stupid than being 
> dual.
> 
> The only sane thing to do is fixed-endianness. I'm sure the m68k people 
> remember being forced to fix their ext2 partitions back in the bad old 
> days. It's painful once, but after that, fixed-endian is a lot more 
> efficient and much simpler to handle.

	a) you've snipped the critical part ;-)
	b) nobody sane uses that beast these days
	c) if somebody wants to grow a private patch - it's their time,
after all...

Seriously, though, by now fs/jffs/* has only one real use - extracting
data from old filesystem.  IIRC, there was even a talk about having it go
the way of ext and xiafs.  He's dead, Jim...

Re: [PATCH] fix type mismatch in jffs.

[Russell King]

On Wed, Sep 10, 2003 at 08:03:04PM +0100, viro@parcelfarce.linux.theplanet.co.uk wrote:
> 
> 	a) you've snipped the critical part ;-)
> 	b) nobody sane uses that beast these days
> 	c) if somebody wants to grow a private patch - it's their time,
> after all...
> 
> Seriously, though, by now fs/jffs/* has only one real use - extracting
> data from old filesystem.  IIRC, there was even a talk about having it go
> the way of ext and xiafs.  He's dead, Jim...

It isn't that dead - I get the occasional patch from people wanting to keep
it working, although I really wish people would send them to dwmw2 rather
than myself.

-- 
Russell King (rmk@arm.linux.org.uk)	http://www.arm.linux.org.uk/personal/
Linux kernel maintainer of:
  2.6 ARM Linux   - http://www.arm.linux.org.uk/
  2.6 PCMCIA      - http://pcmcia.arm.linux.org.uk/
  2.6 Serial core

Re: [PATCH] fix type mismatch in jffs.

[Jörn Engel]

On Wed, 10 September 2003 20:16:07 +0100, Russell King wrote:
> On Wed, Sep 10, 2003 at 08:03:04PM +0100, viro@parcelfarce.linux.theplanet.co.uk wrote:
> > 
> > Seriously, though, by now fs/jffs/* has only one real use - extracting
> > data from old filesystem.  IIRC, there was even a talk about having it go
> > the way of ext and xiafs.  He's dead, Jim...
> 
> It isn't that dead - I get the occasional patch from people wanting to keep
> it working, although I really wish people would send them to dwmw2 rather
> than myself.

Yes, it still beats jffs2 in cases where people have <=5 flash blocks
and want a r/w filesystem on them.  Until David, myself or someone
else finds the time to improve jffs2 for corner cases like this, jffs
has a use - sadly.

Jörn

-- 
When you close your hand, you own nothing. When you open it up, you
own the whole world.
-- Li Mu Bai in Tiger & Dragon