Pete Zaitcev wrote: > Here's my proposal: forget the clever-by-half tricks and simply use > transfer_buffer. This should not only fix Opteron, but also future IOMMUs > that Intel and AMD promise. It may also make usbmon useable on PPC at last. > > I think downsides are negligible. The ones I see are: > - A driver may pass an address of one buffer down as transfer_buffer, > and entirely different entity mapped for DMA, resulting in misleading > output of usbmon. > - Out of tree drivers may crash usbmon if they store garbage in > transfer_buffer. I went over the tree with a comb and fixed obvious > bugs, and clarified the documentation in comments. > - Drivers that use get_user_pages will not be possible to monitor > - Similar deal is with usb_storage transferring from highmem, but > it works fine on 64-bit systems, so I think it's not a concern. > > I'm adding a sign-off line in case, but in general this is a patch > for testing. In particular, sisusb is a concern. > > David & John, I am quite certain that your Opterons will not crash now, > but please give it a try, and also you can use this code to produce usbmon > traces you needed for debugging of other things. > > Signed-off-by: Pete Zaitcev <zaitcev@xxxxxxxxxx> > > Works fine on my system! Cheers David -- 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
