JFFS2 warning message

JFFS2 warning message

[Kevin Liao]

Dear all,

I had a mtd partition on NAND flash mounted as JFFS2. If the power was shutdown abnormally, after that each time when the system
rebooted and tried to mount it again, I always saw a warning message "Empty flash at 0x00000a14 ends at 0x00000c00". Although I
finally found that message was printed by jffs2_scan_eraseblock function in jffs2/scan.c, I still had no idea about what's really
happened. Does anyone could explain the reason about that warning message or tell me what should I do to avoid the kernel's
complaint? Thanks a lot.

Regards,
Kevin

Re: JFFS2 warning message

[Kevin Liao]

Does anyone have any idea about this problem? Thanks a lot.

Regards,
Kevin

----- Original Message ----- 
From: "Kevin Liao" <kevinliao@iei.com.tw>
To: <linux-mtd@lists.infradead.org>
Sent: Wednesday, June 25, 2003 6:50 PM
Subject: JFFS2 warning message


> Dear all,
> 
> I had a mtd partition on NAND flash mounted as JFFS2. If the power was shutdown abnormally, after that each time when the system
> rebooted and tried to mount it again, I always saw a warning message "Empty flash at 0x00000a14 ends at 0x00000c00". Although I
> finally found that message was printed by jffs2_scan_eraseblock function in jffs2/scan.c, I still had no idea about what's really
> happened. Does anyone could explain the reason about that warning message or tell me what should I do to avoid the kernel's
> complaint? Thanks a lot.
> 
> Regards,
> Kevin
> 
> 
> 
> ______________________________________________________
> Linux MTD discussion mailing list
> http://lists.infradead.org/mailman/listinfo/linux-mtd/
> 

Re: JFFS2 warning message

[crowe]

Kevin Liao wrote:
> Does anyone have any idea about this problem? Thanks a lot.
> 
> Regards,
> Kevin
> 
> ----- Original Message ----- 
> From: "Kevin Liao" <kevinliao@iei.com.tw>
> To: <linux-mtd@lists.infradead.org>
> Sent: Wednesday, June 25, 2003 6:50 PM
> Subject: JFFS2 warning message
> 
> 
> 
>>Dear all,
>>
>>I had a mtd partition on NAND flash mounted as JFFS2. If the power was shutdown abnormally, after that each time when the system
>>rebooted and tried to mount it again, I always saw a warning message "Empty flash at 0x00000a14 ends at 0x00000c00". Although I
>>finally found that message was printed by jffs2_scan_eraseblock function in jffs2/scan.c, I still had no idea about what's really
>>happened. Does anyone could explain the reason about that warning message or tell me what should I do to avoid the kernel's
>>complaint? Thanks a lot.
I see this warning message too, but the conditions are
different (Type I flash, normal boot, ...).
I haven't really dug into it, but I know that I've been creating
the jffs2 file system image on a different machine (creating a target filesystem)
on a i386 Linux box, then moving it over to my embedded arm target.  I just
assumed that it was generated by having slightly different versions jffs2, or
that I have debugging turned on.  Haven't had time to really look at it yet.
So what I guess I'm saying is I see this too, and would also like to know how to
get rid of it.

Thanks
-- 
Mike Crowe
crowe@psilongbeach.com

Re: JFFS2 warning message

[Thomas Gleixner]

On Thursday 26 June 2003 20:23, crowe@psilongbeach.com wrote:
> >>I had a mtd partition on NAND flash mounted as JFFS2. If the power was
> >> shutdown abnormally, after that each time when the system rebooted and
> >> tried to mount it again, I always saw a warning message "Empty flash at
> >> 0x00000a14 ends at 0x00000c00". Although I finally found that message
> >> was printed by jffs2_scan_eraseblock function in jffs2/scan.c, I still
> >> had no idea about what's really happened. Does anyone could explain the
> >> reason about that warning message or tell me what should I do to avoid
> >> the kernel's complaint? Thanks a lot.
>
> I see this warning message too, but the conditions are
> different (Type I flash, normal boot, ...).
> I haven't really dug into it, but I know that I've been creating
> the jffs2 file system image on a different machine (creating a target
> filesystem) on a i386 Linux box, then moving it over to my embedded arm
> target.  I just assumed that it was generated by having slightly different
> versions jffs2, or that I have debugging turned on.  Haven't had time to
> really look at it yet. So what I guess I'm saying is I see this too, and
> would also like to know how to get rid of it.
>
Nothing to be worried about. The page was not complete written to FLASH before 
power down, so the fs check on mount finds empty FLASH inside this page 
(0xff) and thinks the rest of the eraseblock is empty too. Then it finds at 
the beginning of the next page some new data and gives a message. That's just 
for information, as something is not as exspected. The fs can handle this and 
this is not a corruption or something like that, just a hint, which keeps you 
informed, that something happend.
If this bothers you, change the message level to KERN_DEBUG instead of 
KERN_WARNING. It will finally go away, when the offending eraseblock will be 
recycled sometimes.

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