Commit Graph

12 Commits

Author SHA1 Message Date
Tomasz Bursztyka bf40601c2a cc2520: Fix how FREGs and SREGs are accessed
- SREGs should use MEMRG/MEMWR instructions, FREGs use REGRD/REGWR
- REGRD/REGWR get the register address in the first byte along the
  instruction.

Reported-by: Igor Modino <igor@emutex.com>

Change-Id: I0da9de5ecd4cab914c8711bd4276e7a2bc6147a5
Signed-off-by: Tomasz Bursztyka <tomasz.bursztyka@linux.intel.com>
2016-06-22 04:47:43 +00:00
Kumar Gala 9ec2f3be80 Cleanup whitespace in Kconfig files
Convert leading whitespace into tabs in Kconfig files.  Also replaced
double spaces between config and <prompt>.

Change-Id: I341c718ecf4143529b477c239bbde88e18f37062
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
2016-05-25 13:28:07 -05:00
Tomasz Bursztyka 7cd63abccb ieee802154: cc2520: SPI workaround when QMSI drivers are in use
Current QMSI SPI API does not handle asymmetric tx/rx buffers, and thus
it's up to cc2520 driver to manage the tx dummy bytes.

Change-Id: I97900946bcd8d96e9f039646bccd765c574c363e
Signed-off-by: Tomasz Bursztyka <tomasz.bursztyka@linux.intel.com>
2016-05-04 13:47:17 +00:00
Tomasz Bursztyka 6fc7566cdd ieee802154: cc2520: Invalid argument should return -EINVAL
-EIO is an error code for operationnal error, not about invalid
argument.

Change-Id: I7064eaf01c6040d2e93ed43df3932d8baed1256e
Signed-off-by: Tomasz Bursztyka <tomasz.bursztyka@linux.intel.com>
2016-04-27 14:49:41 +02:00
Andre Guedes 04cdf8b6e8 ieee802154: Fix returning codes
DEV_* error codes are deprecated and should not be used anymore. This
patch replaces DEV_* occurrences by the corresponding errno.h code,
according to the error code conventions.

Change-Id: I7b098a6107c390b46ed15aaeb9141627e9502727
Signed-off-by: Andre Guedes <andre.guedes@intel.com>
2016-04-26 22:49:48 +00:00
Tomasz Bursztyka 51503a2388 cc2520: Using new GPIO API callbacks
Adapting CC2520 driver to use the new callback format. That way, cc2520
will work on boards where FIFOP and SFD are hooked to a different GPIO
controller.

Change-Id: Ia40b17867000de26f332f36422f84bb3f20be2aa
Signed-off-by: Tomasz Bursztyka <tomasz.bursztyka@linux.intel.com>
2016-04-22 22:17:53 -04:00
Tomasz Bursztyka c239a8cb7b cc2520: Properly start and stop the device
Disable reception and the oscillator as well when stopping. Revert
when starting.

Change-Id: I2a61066602267ac61e691ec9c20eb243de6fa076
Signed-off-by: Tomasz Bursztyka <tomasz.bursztyka@linux.intel.com>
2016-04-14 07:49:23 +00:00
Tomasz Bursztyka d8dcf56334 cc2520: Enable hardware filtering all the time
This prevents false packets to get accepted/generated into RX fifo.
Also, this will make AUTOACK working properly, when enabled. For some
reasons FRMFILT1 and SRCMATCH need to get their reset values written at
initialization, or then hw filtering won't work. This behavior is not
documented as it seems.

Change-Id: Ic0fe664dbc3b17d85d794c12b77bdbaafeb601f5
Signed-off-by: Tomasz Bursztyka <tomasz.bursztyka@linux.intel.com>
2016-04-14 07:49:23 +00:00
Tomasz Bursztyka a9af3c0512 cc2520: Set short address and ieee address from the driver
For hw filtering and autoack to work, all the information should be
properly set in the chip.

That's a fix for the legacy radio/net API. From original code, these
were set from the application which is bad. But setting it from the
driver is not any better. ieee802154 and net stack should know what to
do, that will be fixed in the future.

Change-Id: I1688223e9488d10a423e788eb88ba1e251cb3f88
Signed-off-by: Tomasz Bursztyka <tomasz.bursztyka@linux.intel.com>
2016-04-14 07:49:23 +00:00
Jukka Rissanen 12485eb8c9 cc2520: Make the generated MAC address to look more random
Because the random number generator is very dumb at the moment,
the generated MAC address has always last byte set to 0.
The MAC address is used when generating IPv6 address.
In order to avoid IPv6 address collision, swap the last four
bytes of the MAC address so that the most random data is
at the end of the MAC.

Change-Id: I96c03654359e32f407bab3b29be0e3b08ee91bd7
Signed-off-by: Jukka Rissanen <jukka.rissanen@linux.intel.com>
2016-04-12 16:27:21 +03:00
Tomasz Bursztyka a1adbcb125 ieee802154: Replace the CC2520 driver with a new implementation
This is a complete new cc2520 driver for zephyr. Intention is to fit
better within Zephyr device driver model.

- It's (almost*) ready to be instanciated as many times as necessary
- It's fully interrupt based on SFD and FIFOP (no pin polling)
- It's nicer to other sub-systems (it sleeps, no busy-wait loop)
- It still loosely complies to old legacy radio device driver model

*: GPIO API needs to be fixed in order to accept multiple callbacks, as
well as enabling callbacks to retrieve private data.

Notes:
- Hardware filtering does not work yet as the net stack, above, needs to
  provide the relevant information for it (src/dst ieee802154 extended
  addresses, short addresses...)
- A embryo of generic functions (txpower, channel, addresses...)
  have been implemented but don't belong yet to any radio device driver
  model. Such new driver model will come afterwards (soon?)
- SPI API would need to be improved to avoid as much as possible memcpy
  as well as spi_slave_select() call.

Change-Id: I1fd6dfff28fba3984f6006d394ea12f1e763ac18
Signed-off-by: Tomasz Bursztyka <tomasz.bursztyka@linux.intel.com>
2016-04-07 17:25:02 +00:00
Tomasz Bursztyka f7d5d2c8f5 drivers: Renaming directory "802.15.4" into "ieee802154"
ieee802154 is a more relevant name. Applying the change in
include/drivers as well.

Change-Id: I7f7188ed0421045d7667303c375eeb8af1298b97
Signed-off-by: Tomasz Bursztyka <tomasz.bursztyka@linux.intel.com>
2016-04-07 17:24:55 +00:00