Commit Graph
109 Commits
Author SHA1 Message Date
Mark Charlebois 62eb403e4d Linus: print format fixes to build with clang on IFC6410
Signed-off-by: Mark Charlebois <charlebm@gmail.com>
2015-04-20 11:15:12 -07:00
Mark Charlebois cd30b4d5ca Linux: added I2C class
Signed-off-by: Mark Charlebois <charlebm@gmail.com>
2015-04-20 11:13:24 -07:00
Mark Charlebois 2feeecdab1 Linux: Added config and stubs to compile I2C device for Linux
Not yet functional. Full implementation will provide an IOCTL interface to
do bi-directional transfer. will model the interface after Linux.

Signed-off-by: Mark Charlebois <charlebm@gmail.com>
2015-04-20 11:13:24 -07:00
Mark Charlebois 4a6bd4ddc0 Linux: replaced getopt in ms5611_linux.cpp
ms5611 uses getopt to parse args but the static variable
optind was not being properly updated.

Replaced use of external getopt call with simple parser;

Signed-off-by: Mark Charlebois <charlebm@gmail.com>
2015-04-20 11:07:00 -07:00
Mark Charlebois 99399115f2 Linux: Minor fixes of error caught by gcc-4.9.1
- Missing static declarations for functions not used outside a file.

Signed-off-by: Mark Charlebois <charlebm@gmail.com>
2015-04-20 11:07:00 -07:00
Mark Charlebois d013ac0927 Support for building more modules with Linux
Added more queue support to linux/px4_layer.

Use virt char devices for ms5611, and mavlink.

Added more HRT functionality. uORB latency test
now fails. Likely due to bad HRT impl for Linux.

Signed-off-by: Mark Charlebois <charlebm@gmail.com>
2015-04-20 11:07:00 -07:00
Mark Charlebois 9758112e31 Use px4_config.h instead of nuttx/config.h
Modified code to use OS independent header file for config settings.

Signed-off-by: Mark Charlebois <charlebm@gmail.com>
2015-04-20 10:55:41 -07:00
Daniel Agar c2abb0f82a fix code style if trivial one line difference 2015-03-19 23:49:36 +01:00
Lorenz Meier a27395a15d MS5611: Move to 0 based indices 2015-02-09 22:56:24 +01:00
Andrew Tridgell 52bff67076 ms5611: removed debug code 2015-02-01 22:13:09 +01:00
Andrew Tridgell 04a9050410 ms5611: allow for all 4 bus combinations in ms5611 driver
this uses the same table driven approach as the hmc5883 driver, and
allows for a ms5611 baro on any of the 4 bus combinations. A simple
"ms5611 start" will start all baros that are found.
2015-02-01 13:09:44 +01:00
Lorenz Meier d851a630d8 Move all drivers to multi pub/sub API 2015-01-29 16:33:53 +01:00
Lorenz Meier 6324513b91 MS5611: Comment-only fix 2015-01-17 18:48:41 +01:00
Lorenz Meier 08297db39a Support for 2nd baro 2015-01-02 12:28:22 +01:00
Lorenz Meier 4c3ebee15b Hackery on option parsing to make MS5611 comply 2015-01-02 12:28:22 +01:00
Lorenz Meier a77c9225df Allow access to both device handles, make external use a special handle 2015-01-02 12:28:21 +01:00
Lorenz Meier 2effc9a23d MS5611: Allow two instances 2015-01-02 12:28:21 +01:00
Thomas Gubler 1913ae19f1 add param for qnh 2014-08-17 10:45:30 +02:00
Lorenz Meier 7f0e675228 Pass over drivers to reduce build size 2014-07-23 08:24:56 +02:00
Lorenz Meier e696ed5509 Merged master 2014-07-15 22:07:03 +02:00
Lorenz Meier 67e3a904b6 fix ms5611 code style for usage call 2014-07-12 19:05:13 +02:00
Lorenz Meier a42ec7df1b MS5611: Deal with missing external bus 2014-07-08 13:50:24 +02:00
Andrew Tridgell ab90fe7832 ms5611: added -X option for external SPI bus 2014-07-08 11:44:32 +02:00
Lorenz Meier becfed9fbd MS5611: Add support for multi uORB devices 2014-06-10 15:08:47 +02:00
Lorenz Meier aa312f96f8 drivers: Fix compile warnings and non-standard performance counter names 2014-05-21 14:20:48 +02:00
Ash Charles e5508a1aa0 Add Gumstix AeroCore device
Based on the work of Andrew Smith [1], add board configuration and device
drivers to support the Gumstix AeroCore (previously Aerodroid) board [2]. The
AeroCore is an autopilot board based on a STM32F427 similar to the FMUv2.

[1] https://github.com/smithandrewc/Firmware
[2] https://store.gumstix.com/index.php/products/585/

Signed-off-by: Ash Charles <ashcharles@gmail.com>
2014-05-13 09:41:41 -07:00
Kynos 2de01964e2 Reset MS5611 baro sensor after an error
Reset MS5611 baro sensor after an error in order to avoid endless error
loops
2014-05-02 22:00:34 +02:00
Lorenz Meier d199710651 Fixed up init sequence of all sensors - we can publish in interrupt context, but not advertise! All advertisements now contain valid data 2014-01-14 15:40:46 +01:00
Lorenz Meier e6a67b1deb Support for publication blocking: MS5611, cleaned up device start 2014-01-12 11:54:55 +01:00
Andrew Tridgell d6088efd34 ms5611: report P and T in ms5611 info 2014-01-07 09:34:32 +01:00
Lorenz Meier f4ac204f46 Cranking up bus speeds for all sensors to achievable 10.4 MHz, will cut the bus lock time to half 2013-12-15 18:32:46 +01:00
Andrew Tridgell 8744aa7536 ms5611: check for all zero in the prom
when SPI CLK fails we get all zero data
2013-12-10 11:39:28 +01:00
Andrew Tridgell b0bb5a3450 ms5611: change bus speed to 5MHz
this gives 5MHz SPI bus speed (by asking for 6MHz due to timer
granularity).

Tests with a logic analyser show that the ms5611 is actually more
reliable at 5MHz than lower speeds
2013-12-10 11:01:53 +01:00
Andrew Tridgell a52e70ca93 ms5611: removed unused variable 2013-12-10 11:01:33 +01:00
Andrew Tridgell 3f0f34a4c7 ms5611: give cleaner SPI traces
this makes logic traces cleaner by zeroing extra bytes written
2013-12-10 11:01:20 +01:00
Lorenz Meier 012adc9e33 Minor fixes to bus reset 2013-12-04 09:25:07 +01:00
Lorenz Meier 90c4664dce Merged status changes 2013-10-06 14:17:37 +02:00
px4dev 19fdaf2009 Use the generic device::SPI locking strategy. 2013-09-12 23:56:53 -07:00
Andrew Tridgell 4893509344 drivers: report error_count in drivers where possible 2013-09-12 09:26:49 +02:00
px4dev cefc7ac00e Rework the ringbuffer class so that it's not templated, and refactor its clients so they aren't dancing around the linker anymore. 2013-09-12 00:53:08 +02:00
Andrew Tridgell a5821d2928 ms5611: converted to using RingBuffer 2013-09-12 00:53:01 +02:00
Lorenz Meier 5e6d3604a3 Made MS5611 startup exclusive as well 2013-09-11 22:46:00 +02:00
Andrew Tridgell fdbc09e2a5 avoid counters going above limit in INCREMENT()
when using INCREMENT() the counter would temporarily read equal to
limit, which could cause an issue if the task is preempted.

(this macro should be in a common header, though which header?)
2013-08-28 11:16:30 +02:00
Andrew Tridgell ad732ee3a1 free perf counters in driver destructor
this prevents drivers that probe on one bus then instantiate on
another from leaving behind stale/duplicate perf counters
2013-08-28 11:15:43 +02:00
Julian Oes 58a4c5d544 Added missing include for MS5611 2013-08-05 16:32:59 +02:00
Lorenz Meier 17da1e3f36 Fixed MS5611 startup on V1 boards 2013-08-04 19:19:48 +02:00
px4dev 33e33d71b8 Just the changes required to avoid SPI bus conflicts for MS5611 2013-07-24 22:53:10 -07:00
px4dev d57d12818a Revert "Disable interrupts during ms5611 SPI transactions to prevent interruption by sensor drivers polling other sensors from interrupt context."
This reverts commit 31ded04c80.
2013-07-24 22:51:27 -07:00
px4dev 31ded04c80 Disable interrupts during ms5611 SPI transactions to prevent interruption by sensor drivers polling other sensors from interrupt context. 2013-07-20 19:09:59 -07:00
px4dev f6daafba03 Revert "Changed the MS5611 from the workq to hrt_call_every implementation, this seems to solve the SPI chip select overlaps"
This reverts commit c93e743374.
2013-07-20 18:51:46 -07:00