[J-core] Weird console output

[Daniel Lezcano]

On 05/16/2016 04:57 AM, Rich Felker wrote:
> On Sun, May 15, 2016 at 05:41:32PM -0400, Christopher Friedt wrote:
>> On May 15, 2016 4:49 PM, "Daniel Lezcano" <daniel.lezcano at free.fr> wrote:
>>>
>>> On 05/15/2016 09:12 PM, Christopher Friedt wrote:
>>>>
>>>> The baud rate is 19200, iirc, not 115200
>>>
>>>
>>>
>>> IIUC, 19200 is for the binary upload. The console itself is 115200 if I
>> refer to the web site.
>>>
>>> Anyway, I tried 19200 and I have the same issue.
>>
>> Weird. I don't recall changing my baud rate at all between uploading a new
>> bitstream and running vmlinux.
>
> As I understand it, baud rate only makes sense for the physical serial
> line, which is between the FPGA uartlite core and whatever the chip is
> provides the USB ACM device. This is what is operating at 19200. At
> least on Linux, the configured baud rate for /dev/ttyACM* seems to be
> completely ignored; the transfer rate is controlled by the USB ACM
> device. Perhaps this is different on other operating systems and you
> have to do something to tell the OS the rate to expect...?

Ok, so I did the simplest test:

cat /dev/ttyACM0

on x86:
-------

CPU tests passed
GDB Stub for HS-2J0 SH2 ROM
revision: open
build: Mon Apr 18 22:36:49 UTC 2016
Booting
SD:start
SD:init good
Storage device initialized
ELF not found
Probe storage device failed

Run elf FAILED.
....

on Raspberry:
-------------

1st attempt:

eison eJ 2RO
uidMn p12::9 T26
Dstrt
S:n oo
Sta ec iaid
EL ntond
o tae ecfid
R lfFL.

2nd attempt:

CP ts se
DBSbo -JS O
vsn e
ud Mnp1 :6:9T21
Dsr
S:i o
E tfnevcitazd
restreei aie
Rnl FID

Sounds like the console is coughing when running on raspberry pi.

I tried with different powered USB hubs (low speed and high speed) and a 
direct usb connection on the raspberry.

The only difference I can see between the x86 and Raspberry for the 
ttyACM0 usb device is the "Self Powered" and "Device Status" set to 
0x0001 on x86.

When looking at dmesg:

on x86:
-------

new full-speed USB device number 20 using xhci_hcd

on Raspberry:
-------------

new full-speed USB device number 96 using dwc_otg

Apparently there are a lot of issues with USB serial on Raspberry [1]

   -- Daniel

[1] 
http://raspberrypi.stackexchange.com/questions/1886/what-kernel-parameters-are-available-for-fixing-usb-problems