Jaya,

The docs are amazingly confusing on the way these switches map to the
polarity of signals. They use all the terms on/off, up/down, "0/1" instead
of making things easy and clear... I spent a day messing around with this
before finally getting it to work.

If you look at the EVM technical ref page 3-29, you'll see the table of
settings for SW3.  You'll notice the polarity of the "switch" bits and the
"pins" are reversed; this means that TO MAKE A BIT A LOGICAL 1, THE SWITCH
NEEDS TO BE IN THE DOWN POSITION. The "down" position means the little white
switch is on the side where the little numbers (1-8) are labeled on the
switch casing, the LEFT if you have the board oriented so that you can read
the "SW3" label underneath the switch (long way up with the RJ45 on the
right side).

So, for example, to do PCI with no auto init, switches 1-4 (using the little
numbers on the side of the switch casing to index switches) corresponding to
the BM bits should look like this (where XXXXX is the white switch, and
_____ is blank space

1. XXXXX  _____  (LEFT  = DOWN = PIN VALUE "1")
2. _____  XXXXX  (RIGHT = UP   = PIN VALUE "0")
3. XXXXX  _____  (LEFT  = DOWN = PIN VALUE "1")
4. XXXXX  _____  (LEFT  = DOWN = PIN VALUE "1")

By the way, we have the other switches as follows:

5. LEFT
6. LEFT
7. LEFT
8. RIGHT

With the switches set this way, and external 5V power disconnected, and
board in PCI slot, I was able to bring up my Fedora box and then do "lspci"
and see the TI board enumerated as a PCI device.

I am still working on getting the actual booting working, but am almost
there...

Steve



On 8/6/09 4:07 AM, "Jaya krishnan" <jaya.krish...@samsung.com> wrote:

> 
> Hemant,
> Thanks for the  reply. My  SW3  settings  are now SW3[1:8]=0100 0100, here
> '1' means  ON (UP)  and  '0' means OFF(DOWN)
> and BM[3:0] means positions 4:1
> 
> But  still  the  problem  is  not solved. I  am keeping the 6467 EVM in the
> PCI  slot  of a  PowerPc  (460EX) based  system  running  in Linux kernel
> 2.6.30.
> 
> We modified the  PCI boot driver  supplied  with the dvsdk_2_00_00_22  to
> work  with  this  Kernel and  are  able  to insmod  also.
> 
> Error is posted  below
> 
> thanks
> Jayakrishnan
>  
> $ insmod pcidrv_org.ko
> pcidrv: module license 'unspecified' taints kernel.
> Disabling lock debugging due to kernel taint
> Found DM646x PCI device at 0xdf828800
> PCIDRV - Found PCI Devices
> PCIDRV - BAR Configuration -
>            Start        |       Length  |       Flags
>         0x81420000      |       32768   |       0x00020200
>         0x81428000      |       32768   |       0x00020200
>         0x81000000      |       4194304 |       0x00020200
>         0x81400000      |       131072  |       0x00021208
> ARM TCM RAM memory mapped to 0xe9070000
> DDR2 Memory Controller registers mapped to 0xe9080000
> MMR registers mapped to 0xe9100000
> Machine check in kernel mode.
> Data Read PLB Error
> Machine Check exception is imprecise
> Oops: Machine check, sig: 7 [#1]
> PowerPC 44x Platform
> Modules linked in: pcidrv(P+) atp8624 stwdrv solo6010
> NIP: e9063400 LR: e9068264 CTR: e90633ec
> REGS: dfff7f10 TRAP: 0214   Tainted: P            (2.6.30)
> MSR: 00029000 <EE,ME,CE>  CR: 24002028  XER: 20000000
> TASK = df854410[874] 'insmod' THREAD: dfbd6000
> GPR00: 00690000 dfbd7e20 df854410 00000004 00000003 ffffffff c01773e8 00000031
> GPR08: 00000000 7d000000 00002fd9 e90633ec 44002022 100d1258 1ffb1600 00000000
> GPR16: 10005748 00000000 1007b8cc 10086b3e 100cd254 bff2e008 e90658cc e9064b54
> GPR24: dfbd7e6c dfbd7e54 dfbd7e3c e9064b38 e9060000 e9060000 82000000 e911a000
> NIP [e9063400] dm646x_pci_remap_BAR+0x3f0/0x48c [pcidrv]
> LR [e9068264] init_module+0x264/0x9a0 [pcidrv]
> Call Trace:
> [dfbd7e20] [00000401] 0x401 (unreliable)
> [dfbd7e30] [e9068264] init_module+0x264/0x9a0 [pcidrv]
> [dfbd7eb0] [c00013d8] do_one_initcall+0x34/0x1a0
> [dfbd7f20] [c004e05c] sys_init_module+0xb0/0x1b8
> [dfbd7f40] [c000defc] ret_from_syscall+0x0/0x3c
> Instruction dump:
> 480013c9 80010014 38600000 bbc10008 7c0803a6 38210010 4e800020 893f01d0
> 881f01d1 889f01d2 5529c00e 5400801e <7c004b78> 5484402e 893f01d3 7c840378
> ---[ end trace 4e666c929b59e7cd ]---
> Bus error
> 
>  
> 
> ------- Original Message -------
> Sender : Pedanekar, Hemant<hema...@ti.com>
> Date   : Aug 06, 2009 16:15 (GMT+09:00)
> Title  : RE: DM6467 EVM  SW3  settings for  PCI boot
> 
> Jayakrishnan,
> 
> Make sure that you set PCI to &#39;1&#39; (SW3 position 6) with BM[3:0] to
> either &#39;0010&#39; (w/o auto init) or &#39;0011&#39; (w/ auto init).
> 
> -
> Hemant
>> -----Original Message-----
>> From: davinci-linux-open-source-boun...@linux.davincidsp.com
>> [mailto:davinci-linux-open-source-boun...@linux.davincidsp.com] On Behalf
>> Of Jaya krishnan
>> Sent: Thursday, August 06, 2009 12:20 PM
>> To: davinci-linux-open-source@linux.davincidsp.com
>> Subject: DM6467 EVM SW3 settings for PCI boot
>> 
>> Can  somebody tell  me  the SW3  switch  settings in  the DM6467 EVM  to
>> boot it in PCI  mode?
>> SW3[1:8]=0100 0000  where 1 is ON and  0 is OFF  is  not  working  for
>> me. I referred  the technical reference manual also, but bit confusing.
>> 
>> --Jayakrishnan M M
> _______________________________________________
> Davinci-linux-open-source mailing list
> Davinci-linux-open-source@linux.davincidsp.com
> http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source

------ End of Forwarded Message


_______________________________________________
Davinci-linux-open-source mailing list
Davinci-linux-open-source@linux.davincidsp.com
http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source

Reply via email to