Scubapro Aladin Square

Linus Torvalds torvalds at linux-foundation.org
Fri Nov 10 18:53:51 PST 2017


On Fri, Nov 10, 2017 at 4:33 PM, vavincavent <vavincavent at gmail.com> wrote:
> Linus, Jef,
>
> I've made dump file report with USBlyser.
> I've attach the file in this message.

>From a quick look, i tdoes look like the standard Uwatec thing
(without a handshake):

LogTrak sends: 01 10 to query the model.

Dive computer responds:

00000000  01 22 7F 7F BF C1 C4 B1 C1 FB 0E 03 00 00 00 00  ."..............
00000010  FE FE FE 01 00 00 00 00 C1 7D 7E 7C 04 7F 7C BF  .........}~|..|.
00000020  C1 BF C1 12 7E 7F B1 0F 62 C1 03 B1 7E 0B 75 04  ....~...b...~.u.
00000030  7D BF 7E B1 09 7D 7D BF 7D B1 C1 7E 01 01 BF 02  }.~..}}.}..~....

where only that "01 22" is actually valid (the rest is just garbage
from the dive computer memory - it seems to be quite common that the
USB HID stack uses a static buffer and sends garbage padding).

So the model is 0x22.

Then LogTrak does that 0x1b thing (that we use as a handshake, but it
clearly means something): 01 1B

And the dive computer responds with

00000000  01 01 7F 7F BF C1 C4 B1 C1 FB 0E 03 00 00 00 00  ................
00000010  FE FE FE 01 00 00 00 00 C1 7D 7E 7C 04 7F 7C BF  .........}~|..|.
00000020  C1 BF C1 12 7E 7F B1 0F 62 C1 03 B1 7E 0B 75 04  ....~...b...~.u.
00000030  7D BF 7E B1 09 7D 7D BF 7D B1 C1 7E 01 01 BF 02  }.~..}}.}..~....

which is just "01 01" and the same garbage padding.

Then it does that "05 1C 10 27 00 00" thing that we do as part of the
handshake, and we get 01 01 back (with garbage padding).

And then LogTrak asks for devtime, and gets "04 D6 FD 31 43" (and
garbage padding) back.

So everything looks very much like the G2 should work.

So I don't know why it subsurface gets that

  ERROR: Usb read interrupt transfer failed (LIBUSB_ERROR_TIMEOUT).
[in ../../src/usbhid.c:518 (dc_usbhid_read)]

error.  Maybe the Aladin Square has a fairly stupid firmware, and
really needs the commands to be in one particular order, and wants
that "model query" first.

Does anybody else see anything odd?

            Linus


More information about the subsurface mailing list