JFFS2 basic question

JFFS2 basic question

[xuli]

Hello all,
i want to implement the jffs2 file system under kernel2.4.4,
but when i compile the kernel with the MTD and Jffs2 support
i got a lot of errors: multiple definition. Because there is 
a file zlib.c both located in fs/jffs2/ and drivers/net/.
Does anybody meet the same problem with me?
Li

Re: JFFS2 basic question

[Wolfgang Denk]

In message <E18243X-0002eo-00@pentafluge.infradead.org> you wrote:
>
> i want to implement the jffs2 file system under kernel2.4.4,
> but when i compile the kernel with the MTD and Jffs2 support
> i got a lot of errors: multiple definition. Because there is 

You could be a bit more specific _which_ errors you get.

> a file zlib.c both located in fs/jffs2/ and drivers/net/.
> Does anybody meet the same problem with me?

It worked fine for me, in a couple of configurations,  running  2.4.4
on misc PowerPC systems.

Best regards,

Wolfgang Denk

-- 
Software Engineering:  Embedded and Realtime Systems,  Embedded Linux
Phone: (+49)-8142-4596-87  Fax: (+49)-8142-4596-88  Email: wd@denx.de
To program is to be.

Re: JFFS2 basic question

[Xiaogeng (Shawn) Jin]

Denk,

> It worked fine for me, in a couple of configurations,  running  2.4.4
> on misc PowerPC systems.

Do you mean that the latest mtd/jffs2 code from CVS can be compiled with 
2.4.4 distribution from your website? I did succeed in doing this.

Shawn.

Re: JFFS2 basic question

[Xiaogeng (Shawn) Jin]

> Denk,
> 
>> It worked fine for me, in a couple of configurations,  running  2.4.4
>> on misc PowerPC systems.
> 
> Do you mean that the latest mtd/jffs2 code from CVS can be compiled with 
> 2.4.4 distribution from your website? I did succeed in doing this.
                                           ^^^^^^
Sorry. Made a mistake. I did NOT succeed in doing this.

Shawn.

JFFS2 Config / FAQ

[Herman Oosthuysen]

Stupid question of the month:
I need to know how many blocks to leave for the GC at a minimum.

One?

Could someone point me to a FAQ that discusses this issue?

Thanks,
-- 

------------------------------------------------------------------------
Herman Oosthuysen
B.Eng.(E), Member of IEEE
Wireless Networks Inc.
http://www.WirelessNetworksInc.com
E-mail: Herman@WirelessNetworksInc.com
Phone: 1.403.569-5687, Fax: 1.403.235-3965
------------------------------------------------------------------------

Re: JFFS2 Config / FAQ

[David Woodhouse]

Herman@WirelessNetworksInc.com said:
>  Stupid question of the month: I need to know how many blocks to leave
> for the GC at a minimum.

> One?

There's a reason it's not configurable. Reducing it from its current value 
would require significant effort in fixing the corner cases where the file 
system expands, and then _proving_ it to be correct.


--
dwmw2

RE: JFFS2 Config / FAQ

[Joakim Tjernlund]

> 
> Stupid question of the month:
> I need to know how many blocks to leave for the GC at a minimum.
> 
> One?

I think it's 5 blocks currently. It should be possible to lower that count,
but that needs testing.

    Jocke 

Re: JFFS2 Config / FAQ

[David Woodhouse]

joakim.tjernlund@lumentis.se said:
>  I think it's 5 blocks currently. It should be possible to lower that
> count, but that needs testing. 

It was tested. It broke. http://mhonarc.axis.se/jffs-dev/msg01495.html


--
dwmw2

Re: JFFS2 Config / FAQ

[Jörn Engel]

On Mon, 25 November 2002 16:19:53 +0000, David Woodhouse wrote:
> joakim.tjernlund@lumentis.se said:
> >  I think it's 5 blocks currently. It should be possible to lower that
> > count, but that needs testing. 
> 
> It was tested. It broke. http://mhonarc.axis.se/jffs-dev/msg01495.html

Yes, it broke with a smaller number. But wether that smaller number
was 4, 1 or 0 is unknown.

I would expect 3 or 4 to be fine, but not bet more than a beer on it.

Jörn

Re: JFFS2 Config / FAQ

[Thomas Gleixner]

On Monday 25 November 2002 19:38, Jörn Engel wrote:
> On Mon, 25 November 2002 16:19:53 +0000, David Woodhouse wrote:
> > joakim.tjernlund@lumentis.se said:
> > >  I think it's 5 blocks currently. It should be possible to lower that
> > > count, but that needs testing.
> >
> > It was tested. It broke. http://mhonarc.axis.se/jffs-dev/msg01495.html
>
> Yes, it broke with a smaller number. But wether that smaller number
> was 4, 1 or 0 is unknown.
>
> I would expect 3 or 4 to be fine, but not bet more than a beer on it.

In August this year I made some cleanup in the GC algorithms, as they went 
into an endless GC loop, when the fs was filling up to 100%. I tried also to 
reduce blocks, 4 worked sometimes, 3 broke always. 

It depends on the filesizes you have on your fs (deleted or overwritten). But 
with 5 I was not able to break it again.

-- 
Thomas
____________________________________________________
linutronix - competence in embedded & realtime linux
http://www.linutronix.de
mail: tglx@linutronix.de

Re: JFFS2 Config / FAQ

[johan.adolfsson]

----- Original Message -----
From: "Jörn Engel" <joern@wohnheim.fh-wedel.de>
To: "David Woodhouse" <dwmw2@infradead.org>
Cc: <joakim.tjernlund@lumentis.se>; "Herman Oosthuysen"
<Herman@wirelessnetworksinc.com>; <linux-mtd@lists.infradead.org>
Sent: den 25 november 2002 19:38
Subject: Re: JFFS2 Config / FAQ


> On Mon, 25 November 2002 16:19:53 +0000, David Woodhouse wrote:
> > joakim.tjernlund@lumentis.se said:
> > >  I think it's 5 blocks currently. It should be possible to lower that
> > > count, but that needs testing.
> >
> > It was tested. It broke. http://mhonarc.axis.se/jffs-dev/msg01495.html
>
> Yes, it broke with a smaller number. But wether that smaller number
> was 4, 1 or 0 is unknown.
>
> I would expect 3 or 4 to be fine, but not bet more than a beer on it.
>
> Jörn

We applied the following patch to test:
--- os/linux/fs/jffs2/nodelist.h        23 Apr 2002 13:18:37 -0000
1.1.1.3
+++ os/linux/fs/jffs2/nodelist.h        12 Jul 2002 10:11:57 -0000
@@ -223,10 +223,10 @@
 #define ALLOC_DELETION 1       /* Deletion node. Best to allow it */
 #define ALLOC_GC       2       /* Space requested for GC. Give it or die */

-#define JFFS2_RESERVED_BLOCKS_BASE 3
/* Number
 of free blocks there must be before we... */
-#define JFFS2_RESERVED_BLOCKS_WRITE (JFFS2_RESERVED_BLOCKS_BASE + 2)
/* ... al
low a normal filesystem write */
+#define JFFS2_RESERVED_BLOCKS_BASE 0
/* Number
 of free blocks there must be before we... */
+#define JFFS2_RESERVED_BLOCKS_WRITE (JFFS2_RESERVED_BLOCKS_BASE + 1)
/* ... al
low a normal filesystem write */
 #define JFFS2_RESERVED_BLOCKS_DELETION (JFFS2_RESERVED_BLOCKS_BASE + 1)
/
* ... allow a normal filesystem deletion */
-#define JFFS2_RESERVED_BLOCKS_GCTRIGGER (JFFS2_RESERVED_BLOCKS_BASE + 3)
/* ... wa
ke up the GC thread */
+#define JFFS2_RESERVED_BLOCKS_GCTRIGGER (JFFS2_RESERVED_BLOCKS_BASE + 1)
/* ... wa
ke up the GC thread */
 #define JFFS2_RESERVED_BLOCKS_GCBAD (JFFS2_RESERVED_BLOCKS_BASE + 1)
/* ... pi
ck a block from the bad_list to GC */
 #define JFFS2_RESERVED_BLOCKS_GCMERGE (JFFS2_RESERVED_BLOCKS_BASE)
/* ... me
rge pages when garbage collecting */

Our JFFS2 partition was only 5x64kB (2MB flash total, kernel+cramfs the
rest)

/Johan