SCC2 and SCC3 serial ports on RPX Classic (MPC860)

SCC2 and SCC3 serial ports on RPX Classic (MPC860)

[Dmitriy.Zavin]

Hey all,

I'm using the Embedded Planet RPX Classic LF SBC (MPC860 onboard).
I compiled the 2.4.7 kernel (from TimeSYS) with SCC2 and SCC3 enabled for IO. I have a break-out board where I manually hook into the TXD/RXD/GND pins (Port A, pins 12/13). Now, I'm sending a pattern (FOOOOOO) from the SBC to /dev/tts/1 (SCC2) which is connected to a windows box on COM1. All I get on COM1 is garbage. But its not random. I made sure to set all the baud/stop bits/parity manually through IOCTL calls to make sure its right, but still no go. And SCC3 (/dev/tts/2) is completely unresponsive. Writing data to it always blocks the program. I have to ^C, because nothing shows up on the other side of the serial link. Its very puzzling.

Do I have to hook up all the CTS/RTS lines in order to get it to work? I thought that TX/RX/GND was enough?

Any ideas??

Thanks a lot in advance.

--Dima

Some info:

CPM UART driver version 0.03
ttyS00 at 0x0280 is a SMC
ttyS01 at 0x0100 is a SCC
ttyS02 at 0x0200 is a SCC

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

Re: SCC2 and SCC3 serial ports on RPX Classic (MPC860)

[Dan Malek]

Dmitriy.Zavin@infineon.com wrote:

> Any ideas??

My first guess would be some clock routing for the baud rate
generators isn't configured correctly.  There are some driver
assumptions about which clocks/paths are associated with
each port.  They could be used by something else, or the
driver configuration may not have selected the proper ones.

My second guess would be to check the logic external to the
860.  Ensure the external drivers are enabled and that any
data paths through asics are properly routed.  IMHO, this
shouldn't be part of the uart driver due to the number
of possible configurations.

good luck.

	-- Dan


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

Re: SCC2 and SCC3 serial ports on RPX Classic (MPC860)

[Wolfgang Denk]

In message <FD90897E35ABD311AE0E0008C7E6404A07192EDE@ca02exch4.sjc.infineon.com> you wrote:
>
> I compiled the 2.4.7 kernel (from TimeSYS) with SCC2 and SCC3 enabled for IO. I have a break-out board where I manually hook into the TXD/RXD/GND pins (Port A, pins 12/13). Now, I'm sending a pattern (FOOOOOO) from the SBC to /dev/tts/1 (SCC2) which is
> connected to a windows box on COM1. All I get on COM1 is garbage. But its not random. I made sure to set all the baud/stop bits/parity manually through IOCTL calls to make sure its right, but still no go. And SCC3 (/dev/tts/2) is completely unresponsive

What exactly do you mean with "garbage, but not copmpletely random"?

Is it the ordinary type of "line noise" as you get when the baud rate
is wrong, or do you get  valid  characters,  but  only  every  second
character  gets  output?  If  you  get  every  second character, this
usually means that your console port is misconfigured.

Best regards,

Wolfgang Denk

--
Software Engineering:  Embedded and Realtime Systems,  Embedded Linux
Phone: (+49)-8142-4596-87  Fax: (+49)-8142-4596-88  Email: wd@denx.de
Always try to do things in chronological order; it's  less  confusing
that way.

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

RE: SCC2 and SCC3 serial ports on RPX Classic (MPC860)

[Chuck Partridge]

Is this string sent by a program loaded via insmod and busybox?

I had a problem with drivers loaded as modules when using busybox.  The wrong strings would be sent when I did printk.  It drove me nuts for days.  I installed the true version of insmod and all worked fine.

Maybe it helps,

Chuck

>>> <Dmitriy.Zavin@infineon.com> 03/26/03 04:11PM >>>

Wolfgang,

What I mean by "not random" is this:
I write a string to /dev/tts/1, which will cause another string to be printed on the remote console. If I repeat this 50 times, the same messed up string will appear.

Thanks.

I hope that answers the question.

--Dima

>In message
><FD90897E35ABD311AE0E0008C7E6404A07192EDE@ca02exch4.sjc.infineo
>n.com> you wrote:
>>
>> I compiled the 2.4.7 kernel (from TimeSYS) with SCC2 and
>SCC3 enabled for IO. I have a break-out board where I manually
>hook into the TXD/RXD/GND pins (Port A, pins 12/13). Now, I'm
>sending a pattern (FOOOOOO) from the SBC to /dev/tts/1 (SCC2) which is
>> connected to a windows box on COM1. All I get on COM1 is
>garbage. But its not random. I made sure to set all the
>baud/stop bits/parity manually through IOCTL calls to make
>sure its right, but still no go. And SCC3 (/dev/tts/2) is
>completely unresponsive
>
>What exactly do you mean with "garbage, but not completely random"?
>
>Is it the ordinary type of "line noise" as you get when the baud rate
>is wrong, or do you get  valid  characters,  but  only  every  second
>character  gets  output?  If  you  get  every  second character, this
>usually means that your console port is misconfigured.
>
>Best regards,
>
>Wolfgang Denk
>
>--
>Software Engineering:  Embedded and Realtime Systems,  Embedded Linux
>Phone: (+49)-8142-4596-87  Fax: (+49)-8142-4596-88  Email: wd@denx.de
>Always try to do things in chronological order; it's  less  confusing
>that way.
>


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

RE: SCC2 and SCC3 serial ports on RPX Classic (MPC860)

[Dmitriy.Zavin]

>Is this string sent by a program loaded via insmod and busybox?
>
>I had a problem with drivers loaded as modules when using
>busybox.  The wrong strings would be sent when I did printk.
>It drove me nuts for days.  I installed the true version of
>insmod and all worked fine.

This is using busybox, but no modules have been loaded. SCC Uart support was built-in to the kernel.

I also took a look at the hex values that I actually receive and it does NOT look like the bit-string is offset by any bits. There's no direct mapping that I can see..

Thanks.

--Dima

>
>Maybe it helps,
>
>Chuck
>
>>>> <Dmitriy.Zavin@infineon.com> 03/26/03 04:11PM >>>
>
>Wolfgang,
>
>What I mean by "not random" is this:
>I write a string to /dev/tts/1, which will cause another
>string to be printed on the remote console. If I repeat this
>50 times, the same messed up string will appear.
>
>Thanks.
>
>I hope that answers the question.
>
>--Dima
>
>>In message
>><FD90897E35ABD311AE0E0008C7E6404A07192EDE@ca02exch4.sjc.infineo
>>n.com> you wrote:
>>>
>>> I compiled the 2.4.7 kernel (from TimeSYS) with SCC2 and
>>SCC3 enabled for IO. I have a break-out board where I manually
>>hook into the TXD/RXD/GND pins (Port A, pins 12/13). Now, I'm
>>sending a pattern (FOOOOOO) from the SBC to /dev/tts/1 (SCC2) which is
>>> connected to a windows box on COM1. All I get on COM1 is
>>garbage. But its not random. I made sure to set all the
>>baud/stop bits/parity manually through IOCTL calls to make
>>sure its right, but still no go. And SCC3 (/dev/tts/2) is
>>completely unresponsive
>>
>>What exactly do you mean with "garbage, but not completely random"?
>>
>>Is it the ordinary type of "line noise" as you get when the baud rate
>>is wrong, or do you get  valid  characters,  but  only  every  second
>>character  gets  output?  If  you  get  every  second character, this
>>usually means that your console port is misconfigured.
>>
>>Best regards,
>>
>>Wolfgang Denk
>>
>>--
>>Software Engineering:  Embedded and Realtime Systems,  Embedded Linux
>>Phone: (+49)-8142-4596-87  Fax: (+49)-8142-4596-88  Email: wd@denx.de
>>Always try to do things in chronological order; it's  less  confusing
>>that way.
>>
>


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

Re: SCC2 and SCC3 serial ports on RPX Classic (MPC860)

[Wolfgang Denk]

In message <FD90897E35ABD311AE0E0008C7E6404A07192EE0@ca02exch4.sjc.infineon.com> you wrote:
>
> I also took a look at the hex values that I actually receive and it does NOT look like the bit-string is offset by any bits. There's no direct mapping that I can see..

Well, can you post an example, i. e. which output  do  you  get  when
sending the string "01234567890123456789abcdefghabcdefgh" ?


Wolfgang Denk

--
Software Engineering:  Embedded and Realtime Systems,  Embedded Linux
Phone: (+49)-8142-4596-87  Fax: (+49)-8142-4596-88  Email: wd@denx.de
I usually tell my classes "if you are using @ and [] together in this
class, you will almost certainly NOT get what you want. That's  going
down  the wrong tunnel. There's no cheese at the end of that tunnel."
         -- Randal L. Schwartz in <8czptuomey.fsf@gadget.cscaper.com>

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/