mtd erase error

mtd erase error

[Talat Ozyagcilar]

Hi All,

I am trying to write a rootfs on flash.  I have made an jffs2 image of
the rootfs and I am using "cat rootfs.jffs2 > /dev/mtdblock/3", but I
am getting the following output:
----------------------------------------------------------
# cat rootfs.jffs2 > /dev/mtdblock/3
mtdblock_open
ok
mtdblock: read on "RootFS" at 0x282000, size 0x1000
mtdblock: write on "RootFS" at 0x0, size 0x1000
mtdblock: write on "RootFS" at 0x1000, size 0x1000
mtdblock: write on "RootFS" at 0x2000, size 0x1000
mtdblock: write on "RootFS" at 0x3000, size 0x1000
mtdblock: write on "RootFS" at 0x4000, size 0x1000
mtdblock: write on "RootFS" at 0x5000, size 0x1000
mtdblock: write on "RootFS" at 0x6000, size 0x1000
mtdblock: write on "RootFS" at 0x7000, size 0x1000
mtdblock: write on "RootFS" at 0x8000, size 0x1000
mtdblock: write on "RootFS" at 0x9000, size 0x1000
mtdblock: write on "RootFS" at 0xa000, size 0x1000
mtdblock: write on "RootFS" at 0xb000, size 0x1000
mtdblock: write on "RootFS" at 0xc000, size 0x1000
mtdblock: write on "RootFS" at 0xd000, size 0x1000
mtdblock: write on "RootFS" at 0xe000, size 0x1000
mtdblock: write on "RootFS" at 0xf000, size 0x1000
mtdblock: write on "RootFS" at 0x10000, size 0x1000
mtdblock: write on "RootFS" at 0x11000, size 0x1000
mtdblock: write on "RootFS" at 0x12000, size 0x1000
mtdblock: write on "RootFS" at 0x13000, size 0x1000
mtdblock: write on "RootFS" at 0x14000, size 0x1000
mtdblock: write on "RootFS" at 0x15000, size 0x1000
mtdblock: write on "RootFS" at 0x16000, size 0x1000
mtdblock: write on "RootFS" at 0x17000, size 0x1000
mtdblock: write on "RootFS" at 0x18000, size 0x1000
mtdblock: write on "RootFS" at 0x19000, size 0x1000
mtdblock: write on "RootFS" at 0x1a000, size 0x1000
mtdblock: write on "RootFS" at 0x1b000, size 0x1000
mtdblock: write on "RootFS" at 0x1c000, size 0x1000
mtdblock: write on "RootFS" at 0x1d000, size 0x1000
mtdblock: write on "RootFS" at 0x1e000, size 0x1000
mtdblock: write on "RootFS" at 0x1f000, size 0x1000
mtdblock: write on "RootFS" at 0x20000, size 0x1000
mtdblock: writing cached data for "RootFS" at 0x0, size 0x20000
Chip reports voltage low on erase: status 0xa8
mtdblock: erase of region [0x0, 0x20000] on "RootFS" failed
end_request: I/O error, dev 1f:03 (mtdblock), sector 256
mtdblock: write on "RootFS" at 0x21000, size 0x1000
mtdblock: writing cached data for "RootFS" at 0x0, size 0x20000
Chip reports voltage low on erase: status 0xa8
mtdblock: erase of region [0x0, 0x20000] on "RootFS" failed
end_request: I/O error, dev 1f:03 (mtdblock), sector 264
mtdblock: write on "RootFS" at 0x22000, size 0x1000
mtdblock: writing cached data for "RootFS" at 0x0, size 0x20000
Chip reports voltage low on erase: status 0xa8
mtdblock: erase of region [0x0, 0x20000] on "RootFS" failed
end_request: I/O error, dev 1f:03 (mtdblock), sector 272
mtdblock: write on "RootFS" at 0x23000, size 0x1000
mtdblock: writing cached data for "RootFS" at 0x0, size 0x20000
Chip reports voltage low on erase: status 0xa8
mtdblock: erase of region [0x0, 0x20000] on "RootFS" failed
end_request: I/O error, dev 1f:03 (mtdblock), sector 280
mtdblock: write on "RootFS" at 0x24000, size 0x1000
mtdblock: writing cached data for "RootFS" at 0x0, size 0x20000
Chip reports voltage low on erase: status 0xa8
mtdblock: erase of region [0x0, 0x20000] on "RootFS" failed
end_request: I/O error, dev 1f:03 (mtdblock), sector 288
mtdblock: write on "RootFS" at 0x25000, size 0x1000
mtdblock: writing cached data for "RootFS" at 0x0, size 0x20000
Chip reports voltage low on erase: status 0xa8
mtdblock: erase of region [0x0, 0x20000] on "RootFS" failed
end_request: I/O error, dev 1f:03 (mtdblock), sector 296
mtdblock: write on "RootFS" at 0x26000, size 0x1000
------------------------------------------------------

Is this really a hardware problem, or has anybody come across this? 
Any help is appreciated...

Regards

Re: mtd erase error

[Todd Poynor]

Talat Ozyagcilar wrote:
...
> mtdblock: writing cached data for "RootFS" at 0x0, size 0x20000
> Chip reports voltage low on erase: status 0xa8
...
> Is this really a hardware problem, or has anybody come across this? 

It means the driver for the chip's connection to your board (probably) 
isn't raising the voltage to the proper level (called VPP in the mtd 
code) required by the chip for an erase operation.  Various code paths 
can nest VPP enables/disables in unbalanced ways, often causing 
problems.  If this is the case in your situation, info in
http://lists.infradead.org/pipermail/linux-mtd/2005-January/011517.html
may help.

-- 
Todd