Merge with upstream.
* git://github.com/libdivecomputer/libdivecomputer:
Simplify the conditional compilation
Disable some Windows specific compiler warnings
Use SDP to auto-detect the rfcomm port number
Use the actual IrDA device names
Use the correct data type for the temperature
Disable O2 sensors with default calibration values
Add support for the Seac Jack
Check for memory allocation errors
Erase the buffer before calling the vtable function
Replace hardcoded size with the sizeof operator
Use the correct model number for the Shearwater Nerd 2
Rough merge of upstream libdivecomputer.
This is mainly about making the new iostream code upstream, although we
don't actually use it.
It abstracts out the the old serial and usbhid code, but we end up still
using our own 'custom_io' interface because the iostream code doesn't do
it right.
* jef/master:
Correctly determine git SHA if libdivecomputer is a git submodule
Don't accept a NULL pointer as parameter
Add support for semi-closed circuit diving
Detect dives with invalid profile data
Implement the serial communication functions as no-ops
Move the socket code to a common file
Add support for a custom I/O implementation
Port the USB HID code to the new I/O interface
Port the bluetooth code to the new I/O interface
Port the IrDA code to the new I/O interface
Port the serial code to the new I/O interface
Add a new abstract I/O interface
Post release version bump to 0.7.0
It works like the G2, but has a different model number, and different USB device ID.
Lots-of-testing-by: Vincent <vavincavent@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
It turns out that the Scubapro G2 itself does not care at all, but
apparently some of the other dive computers that use the same protocol
do).
In particular, Vincent reports that his Scubapro Aladin Square downloads
ok with the USB ID's switched to c251:2006, but only if we send 32-byte
USB HID payload packets, even though the command itself is much smaller.
Otherwise it will simply not reply at all.
To actually download correctly, we'll still need to do some model-
specific USB ID updates, and there will be some model changes for the
Aladin Square, but this at least fixes the transport side.
Jef had actually already done this because of the Windows behavior side,
I just hadn't thought it could possibly matter. Mea Culpa.
Reported-and-tested-by: Vincent <vavincavent@gmail.com>
Cc: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Cc: Jef Driesen <jef@libdivecomputer.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The model number was wrong. Mea culpa.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The protocol is identical to the G2 protocol, with the exception of a
missing handshake.
Signed-off-by: Berthold Stoeger <bstoeger@mail.tuwien.ac.at>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Merge with upstream libdivecomputer:
- workarounds for Windows libusb and hidapi issues
- misc random cleanups/noise
- rename DiveSystem to Ratio
- make Cochran with better with FTDI
- new support for: Suunto D4f, Ratio idive tank pressure, and Sherwood
Insight temperature and Oceanic ndl/deco sample.
* git://git.libdivecomputer.org/libdivecomputer:
Workaround for a Windows libusb issue
Use a fixed size packet for sending
Replace the size macro with the sizeof operator
Use the correct printf format for the size_t type
Move platform specific macros to a common header file
Use the correct data type for the return value
Rename the DiveSystem vendor to Ratio
Fix the Sherwood Insight temperature
Implement the ndl/deco sample
Change communication parameter to work better with FTDI
Retry read operations on failure
Add support for the Suunto D4f
Implement the tank pressure
Jef reports that this is required for HIDAPI, and while I would really
like to just make the report type part of the custom packet_io interface
instead of making it visible here, this is the minimal fix for now.
See commit d251b373becc ("Add a zero report ID to the commands"), which
does the same thing, except for the fact that we now need to treat BLE
and USB HID differently.
I may still end up teaching the USB HID custom-IO layer to add the
report ID byte, and just specify it at dc_usbhid_custom_io() time
instead. That would make the G2 code not have to care about the
transfer protocol again.
(But the other user of USB HID - the Suunto EON Steel - has much bigger
protocol differences between USB HID and BLE, so the whole "try to be
protocol-agnostic" hope may be just a pipe dream anyway, and it's just
the Scubapro G2 that _could_ work that way).
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Merge with upstream libdivecomputer from Jef:
- more Cochran work from John Van Ostrand
- new 'timesync' interface to synchronize the clock on a dive computer
- support for Aqualung i200
- misc updates (Cressi Leonardo fix, OSTC 3+ renaming, fix surface pressure on iX3M, idive salinity parsing)
- HIDAPI work.
It turns out that HIDAPI is not compatible with libusb in the actual
packet sending path, so this will need some more cleanups - Jef doesn't
see the issue because he doesn't have a generic packet IO layer and
doesn't share packets with the BLE code.
* git://git.libdivecomputer.org/libdivecomputer: (25 commits)
Add basic timezone support
Add time synchronization to the example application
Implement the new api for the HW devices
Add support for synchronizing the device clock
Use hidapi as the default USB HID library
Workaround for a Windows hidapi issue
Reset the number of bytes to zero on error
Add a zero report ID to the commands
Fix compatibility issue with hidapi
Implement the salinity field
Fix the atmospheric pressure for the iX3M
Rename the OSTC 3+ to OSTC Plus
Locate the most recent dive using the logbook pointers
Add support for the Aqualung i200
Add event handling to TM model
Fix profile buffer size and address size
Add three event codes
Add support for the Commander TM
Dump function no longer assumes reads begin at byte 0
Remove unneeded function
...
Merge with upstream libdivecomputer from Jef:
- Jef merged my Scubapro G2 work, but renamed everything, and didn't
get the newer IO model code. Very annoying.
I went along with changing the G2 model family name to
DC_FAMILY_UWATEC_G2 just to keep some of the basic infrastructure
more easily mergeable. But his uwatec_g2 version is not usable.
- Cochran updates from John Van Ostrand
- Misc improvements from Jef:
* divesystems idive improvements
* Oceanic OCS freedive mode
* ppO2 callback cleanup
- Some transport type work:
* changes to IRDA configuration
* basic bluetooth rfcomm transport mode
* 'master' of git://git.libdivecomputer.org/libdivecomputer: (35 commits)
Removed unused code
Fixed duplicate gasmix event reports
Added decompression event handling for the Commander
Fix bad profiles when profile ringbuffer wraps around
Changed cochran_comander_profile_size function parameters
Fixed location and encoding of Commander II pointers
Use a local variable for the layout pointer
Add new EMC device model string
Add support for Pre-21000 s/n Commander dive computers
Fix problems with wrapped logbook ringbuffer
Retry read operations on failure
Change profile download to be incremental
Fix the id string offset
Fix the progress events
Use the trimix data format
Use the correct model number
Enable more fine grained progress events
Abort with an error if the buffer is too small
Use the standard libdivecomputer error codes
Scubapro G2: add missed command packet logging
...
The code actually almost worked as-is, but for a tiny detail: the USBHID
packet reception code always receives a full 64-byte packet, while BLE
GATT will return how much it actually received.
The other difference is that USB HID is so fast that it didn't make any
difference where the progress was updated, it took about a second to
download everything.
BLE GATT is not fast to begin with, and the G2 may be particularly slow.
So with the BLE backend, you really do want progress updates for each
packet received, because the dump is going to take a while...
But with the trivial packet verification change, and with the progress
report updates, everything "JustWorks(tm)" over BLE.
Of course, I haven't committed the actual Subsurface BLE transfer parts
yet, because they are some incredibly ugly stuff with fragile bits and
pieces. But the fact that I can now download from two different dive
computers does mean that I think it's getting to the point where I will
just submit even my ugly code to Dirk.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This means that they still default to their respective USB devices, but
you can now set custom IO structure to pass in your own data.
Not only will we hopefully have some kind of BLE support, you could also
use this to simply emulate packets from a log-file by having a packet
replay (or by actually emulating a device). Of course, you can already
do this by actually emulating the device in a virtual environment, but
it might be useful for some kind of libdivecomputer testing environment.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
I did the packet logging for the received data side, but forgot to do
the same thing on the command transfer side, which makes the debug logs
a bit less useful.
Of course, it turns out that a bug in subsurface means that the logging
never gets enabled anyway even when you ask for it, so right now this
particular misfeature didn't matter. I'm hoping Tomaz/Dirk can fix the
UI widget problem.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
When doing the G2 downloader, I dropped the initial handshake as I tried
to keep the code minimal, and the handshake didn't seem to make any
difference what-so-ever to me.
And it probably doesn't matter for anybody else either. But the code
isn't working for some people, and maybe it does actually matter.
More importantly, Scubapro's own LogTRAK application does send those two
initial commands, and it's probably a good idea to minimize the
differences between the different downloaders anyway, so add the
handshake sequence back in.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Jef Driesen correctly points out that the 'address' field is just
leftover from the IrDA code, and is meaningless for the USB HID
transport version of the Scubapro G2.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The back-end parser seems to be the same as for the Uwatec Smart (aka
Galileo Sol). At least that's the assumption right now.
The downloader just uses USB HID (very similar to EON Steel) rather than
the horrible IrDA thing.
There's also eventually a BLE thing, but that's for the future.
This is an unholy mixture of the Uwatec Smart downloader logic and the
EON Steel usbhid transfer code. The back-end is pure Uwatec Smart
(model 0x11, same as Galileo Sol).
I'm not at all sure this gets everything right, but it downloads
*something*.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
