why can't kernel read from NAND/SMC device?

[Aubrey]

Did you erase the whole root file system partition before you write
the filesystem image to it?

Regards,
-Aubrey

On 3/16/06, Selmeci, Tamas <tamas.selmeci at siemens.com> wrote:
> Hello!
>
> After spending a few hours examining S3C2410 NAND-code, I could successfully bring to life the NAND subsystem of the S3C2410 processor. (I'm working with 2.6.15-5 kernel). The following appears:
>
> ------
> s3c2410-nand: Tacls=3, 30ns Twrph0=7 70ns, Twrph1=7 70ns
> NAND device: Manufacturer ID: 0xec, Chip ID: 0x79 (Samsung NAND 128MiB 3,3V 8-bit)
> Scanning device for bad blocks
> nand_read_oob: from = 0x00000000, len = 16
> ...
> (media check, not too interesting)
> ------
>
> After the check has finished, it assigns every partition to an MTD:
>
> ------
> Creating 4 MTD partitions on "NAND 128MiB 3,3V 8-bit":
> 0x00000000-0x00020000 : "vivi"
> mtd: Giving out device 0 to vivi
> 0x00020000-0x00030000 : "param"
> mtd: Giving out device 1 to param
> 0x00030000-0x000f0000 : "kernel"
> mtd: Giving out device 2 to kernel
> 0x00100000-0x00240000 : "root"
> mtd: Giving out device 3 to root
> ------
>
> (this partition structure is made by vivi to the SMC and compiled in the board's machine file (in s3c2410_nand_set structure) )
>
> cramfs, romfs and jffs2 support is compiled in the kernel, FTL and NFTL not.
>
> If I give the kernel the root=/dev/mtdblock3 parameter, then this can be seen:
>
> ------
> mtdblock_open
> ok
> mtdblock: read on "root" at 0x0, size 0x200
> nand_read_ecc: from = 0x00100000, len = 512
> nand_read_ecc: Failed ECC read, page 0x00000800
> nand_read_ecc: Failed ECC read, page 0x00000800
> end_request: I/O error, dev mtdblock3, sector 0
> Buffer I/O error on device mtdblock3, logical block 0
> mtdblock: read on "root" at 0x0, size 0x200
> nand_read_ecc: from = 0x00100000, len = 512
> nand_read_ecc: Failed ECC read, page 0x00000800
> nand_read_ecc: Failed ECC read, page 0x00000800
> end_request: I/O error, dev mtdblock3, sector 0
> Buffer I/O error on device mtdblock3, logical block 0
> mtdblock: read on "root" at 0x1000, size 0x200
> nand_read_ecc: from = 0x00101000, len = 512
> mtdblock: read on "root" at 0x1200, size 0x200
> nand_read_ecc: from = 0x00101200, len = 512
> nand_read_ecc: Failed ECC read, page 0x00000809
> end_request: I/O error, dev mtdblock3, sector 8
> Buffer I/O error on device mtdblock3, logical block 1
> mtdblock: read on "root" at 0x1000, size 0x200
> nand_read_ecc: from = 0x00101000, len = 512
> mtdblock: read on "root" at 0x1200, size 0x200
> nand_read_ecc: from = 0x00101200, len = 512
> nand_read_ecc: Failed ECC read, page 0x00000809
> end_request: I/O error, dev mtdblock3, sector 8
> Buffer I/O error on device mtdblock3, logical block 1
> mtdblock: read on "root" at 0x2000, size 0x200
> nand_read_ecc: from = 0x00102000, len = 512
> nand_read_ecc: Failed ECC read, page 0x00000810
> end_request: I/O error, dev mtdblock3, sector 16
> Buffer I/O error on device mtdblock3, logical block 2
> mtdblock: read on "root" at 0x2000, size 0x200
> nand_read_ecc: from = 0x00102000, len = 512
> nand_read_ecc: Failed ECC read, page 0x00000810
> end_request: I/O error, dev mtdblock3, sector 16
> Buffer I/O error on device mtdblock3, logical block 2
> mtdblock: read on "root" at 0x3000, size 0x200
> nand_read_ecc: from = 0x00103000, len = 512
> nand_read_ecc: Failed ECC read, page 0x00000818
> nand_read_ecc: Failed ECC read, page 0x00000818
> end_request: I/O error, dev mtdblock3, sector 24
> Buffer I/O error on device mtdblock3, logical block 3
> mtdblock: read on "root" at 0x3000, size 0x200
> nand_read_ecc: from = 0x00103000, len = 512
> nand_read_ecc: Failed ECC read, page 0x00000818
> nand_read_ecc: Failed ECC read, page 0x00000818
> end_request: I/O error, dev mtdblock3, sector 24
> Buffer I/O error on device mtdblock3, logical block 3
> mtdblock_release
> nand_sync: called
> ok
> VFS: tried fs_name = <cramfs> err= -22
> mtdblock_open
> ok
> mtdblock: read on "root" at 0x0, size 0x200
> nand_read_ecc: from = 0x00100000, len = 512
> nand_read_ecc: Failed ECC read, page 0x00000800
> nand_read_ecc: Failed ECC read, page 0x00000800
> end_request: I/O error, dev mtdblock3, sector 0
> romfs: unable to read superblock
> mtdblock_release
> nand_sync: called
> ok
> VFS: tried fs_name = <romfs> err= -22
> Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(31,3)
> ------
> There are ECC error messages and I/O errors when reading the SmartMedia card, but in the last line (unknown-block(31,3) ) the device ID seems to be OK. The NAND device is recognized but works improperly. Why can't be the partition read?
>
> With root=/dev/mtd3 setting:
> ------
> VFS: test name = </dev/root>
> VFS: fs_name = <cramfs>
> VFS: fs_name = <romfs>
> VFS: tried fs_name = <cramfs> err= -6
> VFS: Cannot open root device "mtd3" or unknown-block(0,0)
> Please append a correct "root=" boot option
> Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(0,0)
> ------
>
> There are absolutely no ECC error messages - because /dev/mtd3 refers to no device (unknown-block(0,0) ), I suppose.
>
> I don't know what's the problem with the NAND card (it's a rather new one, with only one bad eraseblock) or driver... those ECC errors are not too nice (in the kernel I switched off S3C2410 NAND hardware ECC option). Does anybody know what to do now, or any good advices?
>
> Many thanks
>
> Bye
> --
> Tamas Selmeci
>
> ______________________________________________________
> Linux MTD discussion mailing list
> http://lists.infradead.org/mailman/listinfo/linux-mtd/
>