> From: Alan Stern [mailto:stern@xxxxxxxxxxxxxxxxxxx] > Sent: Monday, October 17, 2011 9:23 AM > > On Mon, 17 Oct 2011, Felipe Balbi wrote: > > > > > The analyzer screenshot shows that the device did send back only one > > > > byte of data. So this appears to be a bug in the CV program itself, > > > > not a violation of any sort. In fact, the error message indicates that > > > > the mass-storage gadgets' current behavior is correct. (But maybe the > > > > CV program prints out the wrong error message -- this would still be a > > > > bug, of course.) > > > > > > I just took a look at the USB20CV documentation (release 1.4.7.0). It > > > says that receiving a single byte from Get Max LUN with wLength > 1 is > > > a valid response. > > > > > > On the other hand, it also says that the class-specific Reset request > > > must have wLength = 0, and this requirement is more logical. We > > > currently do not check for this; adding such a check would be a > > > reasonable thing to do. > > > > meaning Paul's patch should be applied ?? > > Meaning that Paul should rewrite his patch so that it changes the > USB_BULK_RESET_REQUEST case rather than (or in addition to) the > USB_BULK_GET_MAX_LUN_REQUEST base. I will send a separate patch to add the check for the USB_BULK_RESET_REQUEST case. Then you guys can decide whether to apply the first patch or not. -- Paul ��.n��������+%������w��{.n�����{���)�� jg��� �����ݢj����G���� ���j:+v���w�m������w�������h�����٥
