On 9 Mar, Greg KH wrote:
> I ran the program on two devices tonight, using a OHCI host controller
> I didn't patch the kernel to modify HZ like you mentioned, so I don't
> know if that makes a difference with the numbers or not:
>
> Inside Out Networks Edgeport/4
> port = /devfs/usb/tts/0
> baud = 38400
> packet size = 1
> packet count = 10
> serial-test.c:utimer_init() - checking clock speed... 453MHz
> packetSize : 1
> 1 6.1/ 6.3/ 8.0
>
> Keyspan PDA:
> port = /devfs/usb/tts/0
> baud = 38400
> packet size = 1
> packet count = 10
> serial-test.c:utimer_init() - checking clock speed... 452MHz
> packetSize : 1
> 1 10.0/ 11.0/ 19.8
>
>
> And here is the serial port connected to my motherboard:
> port = /devfs/tts/0
> baud = 38400
> packet size = 1
> packet count = 10
> serial-test.c:utimer_init() - checking clock speed... 452MHz
> packetSize : 1
> 1 9.9/ 10.0/ 10.0
>
> Hm, the Edgeport driver is faster than the serial port on my
> motherboard? Does that make any sense, or is it an indication that I
> really need to apply the HZ patch?
I once discovered that the standard UART drivers will only deliver data
to user level processes on "kernel ticks". Since you are running a
100Hz scheduler, that would limit the native UARTS to 10ms minimum
latency and it looks like you got 10ms very consistently.
I don't know where this limitation is so I don't know of the USB serial
ports are subject to it but it looks like maybe the Edgeport isn't but
the keyspan is. ...is the Edgeport driver you used based on the
"standard USB serial support"? I think I recently saw a patch to
change it so it would be but recall that it previously stood alone.
I suspect that you would get different results if you patched the
kernel for 1024Hz scheduling. ...would you mind doing that experiment?
Also, try 115,200 bps if you can?
> Hope this helps,
Yup. At leat the Edgeport data suggests that the latency I am seeing
isn't inherent to USB. I'm still wondering if it is somehow tunable
though.
> greg k-h
Thanks!
Ty
--
Tyson D Sawyer iRobot Corporation
Senior Systems Engineer Real World Interface Div.
[EMAIL PROTECTED] Robots for the Real World
603-532-6900 ext 206 http://www.irobot.com
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
http://lists.sourceforge.net/lists/listinfo/linux-usb-devel
