[Trac #425] cafe_ccic sporadic oops when running gstreamer

Zarro Boogs per Child bugtracker at laptop.org
Mon Nov 13 20:54:38 EST 2006


#425: cafe_ccic sporadic oops when running gstreamer
---------------------+------------------------------------------------------
 Reporter:  cjb      |        Owner:  corbet  
     Type:  defect   |       Status:  assigned
 Priority:  blocker  |    Milestone:  BTest-2 
Component:  kernel   |   Resolution:          
 Keywords:           |  
---------------------+------------------------------------------------------
Comment (by dwmw2):

 Oh, actually I think your stress test was trying to read from bad blocks.
 The ECC syndrome registers (0x50-0x58) are precisely what I'd expect from
 a read of all zeroes. Don't Do That Then.

 You can confirm that it was a bad block by looking in /var/log/dmesg for
 the output when the NAND driver starts up. And I can reproduce the problem
 by doing the same...

 {{{
 -bash-3.1# dmesg -c | grep Bad
 Bad block table found at page 262080, version 0x01
 Bad block table found at page 262016, version 0x01
 nand_read_bbt: Bad block at 0x06400000
 nand_read_bbt: Bad block at 0x0a1a0000
 nand_read_bbt: Bad block at 0x0c200000
 -bash-3.1# modprobe mtdchar
 -bash-3.1# dd if=/dev/mtd0 bs=2048 skip=$((0xa1a0000 / 2048)) count=1
 of=/dev/null
 1+0 records in
 1+0 records out
 2048 bytes (2.0 kB) copied, 0.214659 seconds, 9.5 kB/s
 -bash-3.1# dmesg
 CAFÉ NAND 0000:00:0c.0: Failed to correct ECC at 0a1a0000
 Register 0: 64000000
 Register 4: 28000000
 Register 8: 40000000
 Register c: 40000000
 Register 10: 00000000
 Register 14: ffffffff
 Register 18: 00000840
 Register 1c: 00000000
 Register 20: 00014340
 Register 24: 01010a0a
 Register 28: 24121212
 Register 2c: 11000000
 Register 30: 00000000
 Register 34: 00000000
 Register 38: 00000000
 Register 3c: 00040000
 Register 40: a0000840
 Register 44: 01316000
 Register 48: 00000000
 Register 4c: 00005fbe
 Register 50: 0e9200ff
 Register 54: 0f57058c
 Register 58: 082301b2
 }}}

-- 
Ticket URL: <http://dev.laptop.org/ticket/425#comment:7>
One Laptop Per Child <http://laptop.org/>



More information about the Bugs mailing list