jffs2 - mtd - nand flash works....except....
MQuicquaro at necii.com
MQuicquaro at necii.com
Wed Dec 15 10:19:17 EST 2004
Ok. Thanks. My problem must be elsewhere then.
>From what you say, jffs2 will never do subsector writes.
Is that correct? Is there a configuration option maybe that
controls this?
My problem is that, in my 2 MB kernel MTD partition, I place
my 750K kernel. There is like a 50-50% chance that when I
write this kernel, when I boot the system, it reports a CRC error.
If I diff the file with a known good one it shows no differences.
Hmm, maybe I have a problem with my U-boot reading NAND.
If anyone has any ideas please let me know.
Thanks.
- Mike
Here is the output:
U-Boot 1.0.2 (Dec 10 2004 - 21:31:52)
CPU: MPC870ZPnn at 50 MHz: 8 kB I-Cache 8 kB D-Cache FEC present
Board: NEC Infrontia, Inc. - DSX rev. 1.0 DRAM: (16 MB SDRAM) 16 MB
FLASH: 1 MB
AMD AM29LV800B (8 Mbit, bottom boot sect)
Size: 1 MB in 19 Sectors
Sector Start Addresses:
FE000000 (RO) FE004000 (RO) FE006000 (RO) FE008000 (RO) FE010000 (RO)
FE020000 (RO) FE030000 FE040000 FE050000 FE060000
FE070000 FE080000 FE090000 FE0A0000 FE0B0000
FE0C0000 FE0D0000 FE0E0000 FE0F0000
In: serial
Out: serial
Err: serial
NAND: 16 MB
Net: FEC ETHERNET
Hit any key to stop autoboot: 0
NAND read: device 0 offset 0, size 2097152 ... 2097152 bytes read: OK
### JFFS2 loading 'vmlinux.pkg' to 0x100000
Scanning JFFS2 FS: . done.
### JFFS2 load complete: 750562 bytes loaded to 0x100000
## Booting image at 00100000 ...
Image Name: Linux kernel for PQ2FADS
Image Type: PowerPC Linux Kernel Image (gzip compressed)
Data Size: 750498 Bytes = 732.9 kB
Load Address: 00000000
Entry Point: 00000000
Verifying Checksum ... Bad Data CRC
=>
Thomas Gleixner
<tglx at linutronix.
de> To
MQuicquaro at necii.com
12/14/2004 06:17 cc
PM linux-mtd at lists.infradead.org
Subject
Re: jffs2 - mtd - nand flash
Please respond to works....except....
tglx at linutronix.d
e
On Tue, 2004-12-14 at 18:09 -0500, MQuicquaro at necii.com wrote:
> Just about everything seems to be working great as far as I can see
> except subsector writing.
Subsector writes are not permitted in the nand driver.
I'm curious why you don not get an error on write.
Hmm, someone should fix nandtest and remove the subsector stuff.
tglx
More information about the linux-mtd
mailing list