- From: Bernd Porr <Bernd.Porr_at_cn.stir.ac.uk>
- Date: Wed, 24 Sep 2003 21:56:31 +0100
Hi! This might be of general interest: yesterday our data aquisition board for USB (for LINUX/comedi) passed the EMC tests. In particular it was interesting how the board reacted to the 8kV sparks to its connectors and 500V bursts to its inputs/outputs. It was connected to a SONY VAIO. Module "usb-uhci". VIA USB controller. Kernel 2.4.21. The driver registers itself to the usb stack and to COMEDI. Both the 8kV sparks on the board and the 500V bursts basically completely confused the USB controller in the VAIO (not the FX2 in my board :-). However, the kernel did a really good job: Sep 23 14:00:01 localhost kernel: comedi0: successfully attached to usbdux. ##### SPARK ##### Sep 23 14:00:03 localhost kernel: hub.c: port 1, portstatus 100, change 0, 12 Mb /s Sep 23 14:00:03 localhost kernel: hub.c: port 2, portstatus 101, change 3, 12 Mb /s Sep 23 14:00:03 localhost kernel: hub.c: port 2 connection change Sep 23 14:00:03 localhost kernel: hub.c: port 2, portstatus 101, change 3, 12 Mb /s Sep 23 14:00:03 localhost kernel: usb.c: USB disconnect on device 00:07.2-2 addr ess 11 Sep 23 14:00:03 localhost kernel: usb.c: kusbd: /sbin/hotplug remove 11 Sep 23 14:00:03 localhost /sbin/hotplug: arguments (usb) env (PWD=/etc/hotplug D EVICE=/proc/bus/usb/001/011 ACTION=remove DEBUG=kernel OLDPWD=/ DEVFS=/proc/bus/ usb TYPE=255/255/255 SHLVL=1 HOME=/ PATH=/bin:/sbin:/usr/sbin:/usr/bin PRODUCT=4 b4/8613/4 _=/usr/bin/env) Sep 23 14:00:03 localhost /sbin/hotplug: invoke /etc/hotplug/usb.agent () Sep 23 14:00:03 localhost kernel: hub.c: port 2, portstatus 101, change 2, 12 Mb /s Sep 23 14:00:03 localhost kernel: hub.c: port 2, portstatus 101, change 2, 12 Mb /s Sep 23 14:00:03 localhost kernel: hub.c: port 2, portstatus 101, change 3, 12 Mb /s Sep 23 14:00:03 localhost kernel: hub.c: port 2, portstatus 101, change 2, 12 Mb /s Sep 23 14:00:04 localhost last message repeated 3 times Sep 23 14:00:04 localhost kernel: hub.c: port 2, portstatus 103, change 0, 12 Mb /s Sep 23 14:00:04 localhost kernel: hub.c: new USB device 00:07.2-2, assigned addr ess 12 Sep 23 14:00:04 localhost kernel: usb.c: kmalloc IF ce537d40, numif 1 Sep 23 14:00:04 localhost kernel: usb.c: new device strings: Mfr=0, Product=0, S erialNumber=0 Sep 23 14:00:04 localhost kernel: comedi_: usbdux0 has been successfully initial ized. Sep 23 14:00:04 localhost kernel: usb.c: usbdux driver claimed interface ce537d4 0 Sep 23 14:00:04 localhost kernel: usb.c: kusbd: /sbin/hotplug add 12 Sep 23 14:00:04 localhost kernel: hub.c: port 1, portstatus 100, change 0, 12 Mb /s Sep 23 14:00:04 localhost kernel: hub.c: port 2, portstatus 103, change 0, 12 Mb /s Sep 23 14:00:04 localhost /sbin/hotplug: arguments (usb) env (PWD=/etc/hotplug D EVICE=/proc/bus/usb/001/012 ACTION=add DEBUG=kernel OLDPWD=/ DEVFS=/proc/bus/usb TYPE=255/255/255 SHLVL=1 HOME=/ PATH=/bin:/sbin:/usr/sbin:/usr/bin PRODUCT=4b4/ 8613/4 _=/usr/bin/env) Sep 23 14:00:04 localhost /sbin/hotplug: invoke /etc/hotplug/usb.agent () Sep 23 14:00:04 localhost /etc/hotplug/usb.agent: Setup usbdux for USB product 4 b4/8613/4 Sep 23 14:00:04 localhost /etc/hotplug/usb.agent: Module setup usbdux for USB pr oduct 4b4/8613/4 Sep 23 14:00:04 localhost modprobe: modprobe: Can't locate module char-major-98- 0 Sep 23 14:00:04 localhost kernel: comedi0: usbdux: usb-device 0 is attached to c omedi. Sep 23 14:00:04 localhost kernel: comedi0: successfully attached to usbdux. I'm not sure how the disconnect is "triggered". However, the probe function is called again and also the hotplug script is called again. The system recovers itself which is needed for the test. Praise to the authors of the low level parts of the USB stack. Great work. Also comedi itself worked nicely. (I used ao_waveform for most of the tests). /Bernd -- http://www.cn.stir.ac.uk/~bp1/ mailto:bp1_at_cn.stir.ac.uk
Received on 2003-09-24Z19:56:31