RE: JFFS patch

RE: JFFS patch

[Mikael Starvik]

Even at Axis (the original copyright holder of JFFS1) we are now using 
JFFS2. I'll check if someone inhouse is interrested in keeping JFFS1 
maintained 2.6.

PS. I tested JFFS1 in 2.5 a while ago and it seamed to kind of work after 
one small patch. But I guess that there is much more to be done to make
it reliable. DS

/Mikael

-----Original Message-----
From: owner-jffs-dev@axis.com [mailto:owner-jffs-dev@axis.com] On Behalf Of
David Woodhouse
Sent: Tuesday, June 24, 2003 12:42 PM
To: Alexander Viro
Cc: Allen Curtis; Infradead Org; jffs-dev@axis.com
Subject: Re: JFFS patch


On Tue, 2003-06-24 at 11:35, Alexander Viro wrote:
> It would be nice.  There are several ugly craps in there and I'd 
> rather see it gone than try and fix them...

Me too -- I started cleaning up at one point and the result was JFFS2 :)

So unless I hear anguished screams from a huge user base who are using it
and not JFFS2 for reasons which they haven't yet reported as JFFS2 bugs, and
someone also steps forward to clean it up for 2.6, I'll submit a patch to
Linus to remove it in a couple of weeks time.

-- 
dwmw2


To unsubscribe from this list: send the line "unsubscribe jffs-dev" in the
body of a message to majordomo@axis.com

RE: JFFS patch

[Allen Curtis]

I would be screaming a lot if we were not moving to JFFS2 ourselves. 
However this will take us a few months to accomplish. If there is any 
interest in JFFS and if someone is kind enough to point out the work TBD, I 
will take a shot.

At 12:48 PM 6/24/03 +0200, Mikael Starvik wrote:
>Even at Axis (the original copyright holder of JFFS1) we are now using
>JFFS2. I'll check if someone inhouse is interrested in keeping JFFS1
>maintained 2.6.
>
>PS. I tested JFFS1 in 2.5 a while ago and it seamed to kind of work after
>one small patch. But I guess that there is much more to be done to make
>it reliable. DS
>
>/Mikael
>
>-----Original Message-----
>From: owner-jffs-dev@axis.com [mailto:owner-jffs-dev@axis.com] On Behalf Of
>David Woodhouse
>Sent: Tuesday, June 24, 2003 12:42 PM
>To: Alexander Viro
>Cc: Allen Curtis; Infradead Org; jffs-dev@axis.com
>Subject: Re: JFFS patch
>
>
>On Tue, 2003-06-24 at 11:35, Alexander Viro wrote:
> > It would be nice.  There are several ugly craps in there and I'd
> > rather see it gone than try and fix them...
>
>Me too -- I started cleaning up at one point and the result was JFFS2 :)
>
>So unless I hear anguished screams from a huge user base who are using it
>and not JFFS2 for reasons which they haven't yet reported as JFFS2 bugs, and
>someone also steps forward to clean it up for 2.6, I'll submit a patch to
>Linus to remove it in a couple of weeks time.
>
>--
>dwmw2
>
>
>To unsubscribe from this list: send the line "unsubscribe jffs-dev" in the
>body of a message to majordomo@axis.com

Re: JFFS patch

[Andrew Goodney]

Allen Curtis wrote:

> I would be screaming a lot if we were not moving to JFFS2 ourselves. 
> However this will take us a few months to accomplish. If there is any 
> interest in JFFS and if someone is kind enough to point out the work 
> TBD, I will take a shot.


My $0.02;

It would be nice to keep JFFS around. We are using it because on our 
fairly slow platform, JFFS seems to be faster and gives us more usable 
space for a given partition size. So thanks for stepping up to the plate.

-Andrew Goodney
andrew@blueskylabs.com
---------------------------
www.blueskylabs.com

JFFS patch

[Allen Curtis]

[-- Attachment #1: Type: text/plain, Size: 40 bytes --]

Patch for updating JFFS in 2.4.20 kernel

[-- Attachment #2: jffs.patch --]
[-- Type: application/octet-stream, Size: 1513 bytes --]

diff -ru fs/jffs/inode-v23.c ../thales/linuxppc/fs/jffs/inode-v23.c
--- fs/jffs/inode-v23.c	Thu Mar 13 02:04:52 2003
+++ ../thales/linuxppc/fs/jffs/inode-v23.c	Sun Jun 22 22:31:37 2003
@@ -59,7 +59,7 @@
 #include "jffs_proc.h"
 #endif
 
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,5,2)
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,4,20)
 #define minor(x) MINOR(x)
 #define major(x) MAJOR(x)
 #endif
Only in ../thales/linuxppc/fs/jffs: inode-v23.c~
diff -ru fs/jffs/intrep.c ../thales/linuxppc/fs/jffs/intrep.c
--- fs/jffs/intrep.c	Tue Mar  5 06:07:39 2002
+++ ../thales/linuxppc/fs/jffs/intrep.c	Sun Jun 22 22:31:37 2003
@@ -71,6 +71,8 @@
 #include <linux/sched.h>
 #include <linux/ctype.h>
 
+#include <linux/mtd/compatmac.h>
+
 #include "intrep.h"
 #include "jffs_fm.h"
 
@@ -3427,7 +3429,8 @@
 			unsigned long signr;
 
 			spin_lock_irq(&current->sigmask_lock);
-			signr = dequeue_signal(&current->blocked, &info);
+			signr = dequeue_signal(current, &current->blocked, &info);
+/*			signr = dequeue_signal(&current->blocked, &info);*/
 			spin_unlock_irq(&current->sigmask_lock);
 
 			switch(signr) {
diff -ru fs/jffs/jffs_fm.c ../thales/linuxppc/fs/jffs/jffs_fm.c
--- fs/jffs/jffs_fm.c	Tue Jan 22 01:48:16 2002
+++ ../thales/linuxppc/fs/jffs/jffs_fm.c	Sun Jun 22 22:31:37 2003
@@ -23,7 +23,7 @@
 #include <linux/compatmac.h>
 #include "jffs_fm.h"
 
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,5,2)
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,4,20)
 #define minor(x) MINOR(x)
 #define major(x) MAJOR(x)
 #endif

[-- Attachment #3: Type: text/plain, Size: 0 bytes --]

Re: JFFS patch

[David Woodhouse]

On Mon, 2003-06-23 at 06:58, Allen Curtis wrote:
> Patch for updating JFFS in 2.4.20 kernel

Looks good. Want to take over maintenance for JFFS? I'm getting very
tempted to remove it from the 2.5 tree unless someone steps forward to
look after it.

-- 
dwmw2

Re: JFFS patch

[Alexander Viro]

On Tue, Jun 24, 2003 at 10:17:42AM +0100, David Woodhouse wrote:
> On Mon, 2003-06-23 at 06:58, Allen Curtis wrote:
> > Patch for updating JFFS in 2.4.20 kernel
> 
> Looks good. Want to take over maintenance for JFFS? I'm getting very
> tempted to remove it from the 2.5 tree unless someone steps forward to
> look after it.

It would be nice.  There are several ugly craps in there and I'd
rather see it gone than try and fix them...

Re: JFFS patch

[David Woodhouse]

On Tue, 2003-06-24 at 11:35, Alexander Viro wrote:
> It would be nice.  There are several ugly craps in there and I'd
> rather see it gone than try and fix them...

Me too -- I started cleaning up at one point and the result was JFFS2 :)

So unless I hear anguished screams from a huge user base who are using
it and not JFFS2 for reasons which they haven't yet reported as JFFS2
bugs, and someone also steps forward to clean it up for 2.6, I'll submit
a patch to Linus to remove it in a couple of weeks time.

-- 
dwmw2

Re: JFFS patch

[Allen Curtis]

Sure... I had CVS commit access at one time. It has been so long I am not 
sure if it still works.

At 10:17 AM 6/24/03 +0100, David Woodhouse wrote:
>On Mon, 2003-06-23 at 06:58, Allen Curtis wrote:
> > Patch for updating JFFS in 2.4.20 kernel
>
>Looks good. Want to take over maintenance for JFFS? I'm getting very
>tempted to remove it from the 2.5 tree unless someone steps forward to
>look after it.
>
>--
>dwmw2

Re: JFFS patch

[David Woodhouse]

On Tue, 2003-06-24 at 15:34, Allen Curtis wrote:
> Sure... I had CVS commit access at one time. It has been so long I am not 
> sure if it still works.

It should do, and you have SSHv2 keys installed so you're not likely to
fall into the trap that everyone else does -- trying to log in with
SSHv2 when you only have v1 keys.

Off the top of my head, I know that the kdev_t abuse (i.e. writing it
directly to the medium) needs to be fixed, and ideally you could port
the mount code from JFFS2 which doesn't need the mtdblock device to be
present. I'm sure Al will have plenty more 'suggestions'... ;)

-- 
dwmw2

Re: JFFS patch

[David Woodhouse]

On Tue, 2003-06-24 at 07:34 -0700, Allen Curtis wrote:
> Sure... I had CVS commit access at one time. It has been so long I am not 
> sure if it still works.

Dud it? Are you still interested in maintaining JFFS 1? It's getting on
for time for it to be dropped from 2.6 if nobody wants to look after it.

-- 
dwmw2

Re: JFFS patch

[Allen Curtis]

>
> Dud it? Are you still interested in maintaining JFFS 1? It's getting on
> for time for it to be dropped from 2.6 if nobody wants to look after 
> it.

I am sorry. I have been busy teaching an embedded systems course. This 
is now completed. Please send me the current issues and will address 
them over break.

Thanks