Cleaning JFFS2 errors

Cleaning JFFS2 errors

[Bernd Büttner]

Hi,
after working some time with the JFFS2 filesystem (on a uclinux-system with a non-mmu freescale processor)
I'm getting errors on every mount like this:

   jffs2_scan_inode(): CRC failed on node at 0x00c999f0: Read 0xffffffff, calculated 0x7621c0c4
   Empty flash at 0x00c99a3c ends at 0x00c99c00
   Empty flash at 0x0134201c ends at 0x01342200
   jffs2_scan_eraseblock(): Node at 0x01d88df8 {0x1985, 0xe001, 0x0000002f) has invalid
     CRC 0x1985e001 (calculated 0x392fe03e)
   JFFS2 notice: (13) read_dnode: node CRC failed on dnode at 0x1341fd4:
     read 0xffffffff, calculated 0xce9c70e9

The kernel is 2.6.19 from the latest uclinux distribution 20070130.

I think these errors are the result of power-loss, misbehaving new drivers and system crashes,
and I suppose they are not critical and will be corrected by jffs2 after some time.

But delivering such systems to customers doesn't look good.

Is there a possibility to force jffs2 to correct these errors at once?

I hope so.

Bernd

Re: Cleaning JFFS2 errors

[Bernd Büttner]

No answer jet.
Does this mean it is a silly question or is there no solution for it?

Bernd


Bernd Büttner schrieb:
> Hi,
> after working some time with the JFFS2 filesystem (on a uclinux-system with a non-mmu freescale processor)
> I'm getting errors on every mount like this:
> 
>    jffs2_scan_inode(): CRC failed on node at 0x00c999f0: Read 0xffffffff, calculated 0x7621c0c4
>    Empty flash at 0x00c99a3c ends at 0x00c99c00
>    Empty flash at 0x0134201c ends at 0x01342200
>    jffs2_scan_eraseblock(): Node at 0x01d88df8 {0x1985, 0xe001, 0x0000002f) has invalid
>      CRC 0x1985e001 (calculated 0x392fe03e)
>    JFFS2 notice: (13) read_dnode: node CRC failed on dnode at 0x1341fd4:
>      read 0xffffffff, calculated 0xce9c70e9
> 
> The kernel is 2.6.19 from the latest uclinux distribution 20070130.
> 
> I think these errors are the result of power-loss, misbehaving new drivers and system crashes,
> and I suppose they are not critical and will be corrected by jffs2 after some time.
> 
> But delivering such systems to customers doesn't look good.
> 
> Is there a possibility to force jffs2 to correct these errors at once?
> 
> I hope so.
> 
> Bernd
> 
> ______________________________________________________
> Linux MTD discussion mailing list
> http://lists.infradead.org/mailman/listinfo/linux-mtd/
> 

Re: Cleaning JFFS2 errors

[Jörn Engel]

On Wed, 30 January 2008 14:23:00 +0100, Bernd Büttner wrote:
> 
> No answer jet.
> Does this mean it is a silly question or is there no solution for it?


It's an FAQ.  Most of these are harmless, they simply indicate a power
loss.  They will get cleaned eventually, you cannot force them to get
cleaned, you cannot easily distinguish power loss from data corruption
and you should add this to the FAQ if it's not part of it yet.

Jörn

-- 
Why do musicians compose symphonies and poets write poems?
They do it because life wouldn't have any meaning for them if they didn't.
That's why I draw cartoons.  It's my life.
-- Charles Shultz