Alan, I contacted the manufacturer and they confirmed that there is a problem with their USB 2.0 PHY, which may cause problems with handshaking. Part of their email is appended below. >From the USB log, I infer that there is framing error before the reset. OHC did not send Start of Frame (SOF) after an polling interval of 32ms. URBs submitted by HID device resulted in protocol errors. USBHID device wanted to reset after this condition. USBHID driver (workqueue routine usb_reset) saw this and reset the device. Thank you, Toan Email from manufacturer: ------------------------------------------------- Dear Toan, Regarding your query, indeed, there are some our clients reported that their USB devices sometimes could not work on EBOX-3300, but when they connected their devices to EBOX-3300 via an USB hub, the fault disappeared. We have found that the fault is caused by the USB 2.0 PHY. It will lead to USB 2.0 PHY and some devices sometimes miss commands when they are handshaking. I think that you can clarify the situation by conneting your USB CDC device via an external usb hub and checking if the error condition exist. Furthermore, the left usb port and onboard usb audio chip are connected to the same internal usb hub. Perhaps, there are some interference in their software interface. ----------------------------------------------------- On Tue, Sep 15, 2009 at 11:01 PM, Alan Stern <stern@xxxxxxxxxxxxxxxxxxx> wrote: > On Tue, 15 Sep 2009, Toan Pham wrote: > >> Sorry, >> >> The file I just sent was a Ms. Excel file. >> Here's a pdf version of it. >> >> And if you really want to analyze, i also attached usb_log.dtc file, >> which can be analyzed with Total Phase Data Center. It can be >> downloaded for free at the link below. >> No installation is needed for the software. >> >> The beagle log file shows USB activities for the first 1m 42sec after boot up. >> Dev #03 is the CDC device, 04 is the HID device. >> CDC devices starts communication at frame number 549 to EOF (Basic >> heart-beating between PC & CDC Device) >> Note, HID device resets rapidly at frame number: 685, 839, 1028 and 1183 >> >> If you filter out device 03 (CDC device), transactions look normal and >> without errors; >> However, OHCD reports ETIME (-62) and EPROTOCOL (-71) errors. > > Okay, I'll look at these files a few days from now. > > Alan Stern > > -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
