Evaluating MCCI's driver for 2.6.23-rc3 and gadget ether
Aras Vaichas <[email protected]>
| Newsgroups | gmane.linux.usb.devel |
|---|---|
| Message-ID | <[email protected]> |
Hi, I'll make this a new thread and stop CC'ing everyone. Please join in if you have an opinion. Li Yang-r58472 wrote: >> Does this mean that USB gadget ether will now play nice with >> a Windows Host using RNDIS? > > IMO, not very nicely. We found loss of packets when using NFS. > >> We are in the process of evaluating the MCCI CDC-SAFE driver >> and it would be nice to know if this is not necessary. > > Please share your result of evaluating the MCCI driver. Thanks. We've never noticed any particular network transport problems as such, but we've never looked. I would appreciate some help/tips with the evaluation. What should I look for specifically? If anyone can suggest a specific test, then I am more than willing to try it. Our in-house experience with gadget ether and MS RNDIS has always been OK as long as we followed good practices like: 1) Plug the device in and wait for Windows to: enumerate, create the network connection and obtain an IP address (via DHCP running on the device) 2) don't rapidly unplug-plug the USB cable otherwise Windows will either: a) flag the USB controller as broken and it won't work for anything until reset. I've noticed that not all machines are susceptible to this problem. Some PCs will completely lose the USB host controller, some will recover over time. b) mark the connection as "Disabled" in the "Network Connections", it won't get removed when the device is disconnected and it will never work again until the OS is rebooted. If we tell our customer what *not* to do then it solves 99% of our problems. Some customers report "mysterious" problems but I've found that these are often machine specific and almost impossible to reproduce. A known potential problem is that our ethernet gadget is an RFID reader which works at 13.56MHz. This can be bad news for 12 Mbit/s USB if the RFID antennas are installed incorrectly as they can cause noisy current loops that travel back through the USB cable. The symptoms look like rapid unplug-plug events and this can have the same side-effects. i.e. Windows marks the connection as bad and you need to reboot. We figure that if the MCCI driver is more robust than RNDIS, then it will help alleviate our main gadget ether problems. I am open to suggestions. regards, Aras Vaichas ______________________________________________________________________ This email has been scanned by the MessageLabs Email Security System. For more information please visit http://www.messagelabs.com/email ______________________________________________________________________ ------------------------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/ _______________________________________________ [email protected] To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel