[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: CFI Devices in x8 mode



Folks,
 
I'm trying to bring the sbc_mediagx and elan-104nc drivers into line with the latest and greatest
JFFS/MTD. Unfortunately something seems to have broken along the way...
 
Running cfi_probe() from the mediagx driver now results in no CFI chip being found. A brief look
through the detection code shows that we are writing 0x98 (Read Query Command) to offset 0x55.
Next we read back and expect to see 'Q', 'R' 'Y' starting at 0x10 (given a device type of x8 (equ 1) and an
interleave of 1).
 
Now, with an Intel Strataflash part in x8 mode the Intel spec. gives the 0x10, 0x11, 0x12 addresses as
N/A. Rather it points us toward addresses 0x20... where we should see 'Q', 'Q', 'R', 'R', 'Y', 'Y'.
 
A quick session with DOS DEBUG gives:
 
-e dc00:0 98
 
-
d dc00:0
 
DC00:0000  89 89 15 15 00 00 00 00-00 00 00 00 00 00 00 00   ................
DC00:0010  00 00 00 00 00 00 00 00-00 00 00 00 00 00 00 00   ................
DC00:0020  51 51 52 52 59 59 01 01-00 00 31 31 00 00 00 00   QQRRYY....11....
DC00:0030  00 00 00 00 00 00 45 45-55 55 00 00 00 00 07 07   ......EEUU......
DC00:0040  07 07 0A 0A 00 00 04 04-04 04 04 04 00 00 17 17   ................
DC00:0050  02 02 00 00 05 05 00 00-01 01 3F 3F 00 00 00 00   ..........??....
DC00:0060  02 02 50 50 52 52 49 49-31 31 31 31 0A 0A 00 00   ..PPRRII1111....
DC00:0070  00 00 00 00 01 01 01 01-00 00 50 50 00 00 00 00   ..........PP....
-
q
 
This all seems to make sense to me, but before I delve any further I'd be grateful if someone
could confirm - or tell me I'm talking out of my hat :-)
 
Thanks,
 
    Alex