Scubapro Aladin Square
vavincavent
vavincavent at gmail.com
Tue Nov 14 13:08:22 PST 2017
Linus,
i change back
if (1) {
to
if (io->packet_size <64) {
here is the usb.pcap
in windows, i see idle and busy on the divecomputer, but not in debian.
Vincent
Le mardi 14 novembre 2017 à 12:33 -0800, Linus Torvalds a écrit :
> On Tue, Nov 14, 2017 at 12:15 PM, Jef Driesen <jef at libdivecomputer.or
> g> wrote:
> >
> > In of you previous emails, you asked Vincent to patch the
> >
> > if (io->packet_size <64) {
> >
> > to a
> >
> > if (1) {
>
> Oh, my bad.
>
> That was actually an earlier "let's try to avoid the report byte",
> and
> I'd forgotten entirely about that (and it was pointless, since the
> hidusb.c code actually removes the report byte anyway for libusb).
>
> But yes, if Vincent has that code, then that explains why he might
> get
> the smaller size.
>
> Vincent, please remove that particular broken test patch if you have
> it.
>
> > Ah that explains the difference. In upstream libdivecomputer the
> > buffer is
> > one byte larger:
> >
> > unsigned char buf[TX_PACKET_SIZE + 1];
>
> And that may actually end up being relevant.
>
> It makes no sense that the receiver cares about the number of padding
> bytes, but the code originally always padded out to the full USB HID
> packet size of 64 bytes.
>
> The whole confusion between 32/33 is all kinds of crazy, but yes,
> maybe some broken other end really wants to see 32 bytes plus report
> byte.
>
> So Vincent, also do change that
>
> unsigned char buf[TX_PACKET_SIZE];
>
> in src/scubapro_g2.c to have the "+1" while you're at it, and see if
> that matters..
>
> Linus
-------------- next part --------------
A non-text attachment was scrubbed...
Name: usb.pcap
Type: application/vnd.tcpdump.pcap
Size: 186444 bytes
Desc: not available
URL: <http://lists.subsurface-divelog.org/pipermail/subsurface/attachments/20171114/6bd9d83e/attachment-0001.pcap>
More information about the subsurface
mailing list