Author Topic: usb woes  (Read 14893 times)

andy_code

  • Member
  • ***
  • Posts: 2
usb woes
« on: September 22, 2011, 04:38:11 am »
we have been struggling to develop a project based around the mcf 52772.

we have written our own OS and usb stack but are have some difficulty with a large number of sticks
a handful of sticks work perfectly whilst others are either not detected at all or fall over whilst trying to get the descriptors.

some just sit there naking away.

on some sticks i have noticed they recive a big packet from the host which contain everything we have done in the enumeration process.


help

regards

Jan Axelson

  • Administrator
  • Frequent Contributor
  • *****
  • Posts: 3033
    • Lakeview Research
Re: usb woes
« Reply #1 on: September 22, 2011, 09:16:04 am »
If by "sticks", you mean flash drives, see the Debugging section here:

http://www.lvr.com/mass_storage.htm#Hosts

Jan

andy_code

  • Member
  • ***
  • Posts: 2
Re: usb woes
« Reply #2 on: September 23, 2011, 04:42:04 am »
thanks,

I have found that that a lot of devices stall on a request for get_capacity. at which point we send a bot_msd reset then set clear feature for the endpoints

however our reset doesnt get sent. nothing shows up in the analyser. if the device has stalled i should still be able to send data and see it in the traffic

many thanks

Jan Axelson

  • Administrator
  • Frequent Contributor
  • *****
  • Posts: 3033
    • Lakeview Research
Re: usb woes
« Reply #3 on: September 23, 2011, 07:28:00 pm »
If you're attempting to send a bulk only mass storage reset request and it's not appearing on the bus, there is a problem with your host firmware. Maybe the endpoint is set to receive data from the device?

Jan