Hi, David Brownell a écrit : >>> A USB-level trace log (via CATC, Beagle, or whatever) might be >>> more informative about what's really going on. Getting useful >>> diagnostics out of MS-Windows tends to be hopeless to anyone >>> not doing driver development on that platform. >> Do i need some special hardware to provide USB-level trace log? > >>>> USB-level trace log (via CATC, Beagle, or whatever) > ^^^^ ^^^^^^ ^^^^^^^^ > .... yes. The www.totalphase.com Beagle (full/low speed) is > quite affordable. > > >>> a simpler >>> hypothesis is that the S3C UDC code has some problem which has >>> recently been exposed by the gadget stack updates. >> So the problem is specific to openmoko hardware drivers? > > That's the hypothesis I would investigate. > I don't know if it could help but the s3c udc code has got some problem : - the controller is based on the same IP than lh7a40x, and some version if s3c controller (on 2412 at least) got a double fifo that need special care [1]. - the s3c controller seems to loose some interrupt, and the only reliable way we found is to poll interrupt register at regular interval. Matthieu [1] http://article.gmane.org/gmane.comp.handhelds.openmoko.kernel/3619/match= PS : the s3c UDC driver doesn't pass usbtest (you need to use pattern=1 module parameter to expose some bugs). -- 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
