Added BUILD_ASSERT check for Tx and Rx thread priorities.
The Tx thread priority shall be higher than Rx thread
priority in order to correctly detect transaction violations
in ATT and SMP protocols. The Number of Completed Packets
for a connection shall be processed before any new data is
received and processed for that connection.
The Controller's priority receive thread priority shall be
higher than the Host's Tx and the Controller's Rx thread
priority.
Signed-off-by: Vinayak Kariappa Chettimada <vich@nordicsemi.no>
Upcoming memory protection features will be placing some additional
constraints on kernel objects:
- They need to reside in memory owned by the kernel and not the
application
- Certain kernel object validation schemes will require some run-time
initialization of all kernel objects before they can be used.
Per Ben these initializer macros were never intended to be public. It is
not forbidden to use them, but doing so requires care: the memory being
initialized must reside in kernel space, and extra runtime
initialization steps may need to be peformed before they are fully
usable as kernel objects. In particular, kernel subsystems or drivers
whose objects are already in kernel memory may still need to use these
macros if they define kernel objects as members of a larger data
structure.
It is intended that application developers instead use the
K_<object>_DEFINE macros, which will automatically put the object in the
right memory and add them to a section which can be iterated over at
boot to complete initiailization.
There was no K_WORK_DEFINE() macro for creating struct k_work objects,
this is now added.
k_poll_event and k_poll_signal are intended to be instatiated from
application memory and have not been changed.
Signed-off-by: Andrew Boie <andrew.p.boie@intel.com>
The controller and host-side RX threads recently had their priorities
lowered to 8. Make the driver RX threads consistent with this.
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Since more and more code is going to be reused by both the Host and the
Controller, this commit introduces a common/ folder that will contain
everything that is not tied to one of the two components but shared by
them.
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
The switch from C99 integer types to u16_t, etc. caused misalignment
in structs and function definitions with multi-line parameter lists.
Change-Id: Ic0e33dc199f834ad7772417bca4c0b2d2f779d15
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Convert code to use u{8,16,32,64}_t and s{8,16,32,64}_t instead of C99
integer types.
Jira: ZEP-2051
Change-Id: I8f57a17f78e674aca5400f005db8975c9f9e150e
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
Declaring these as const lets the linker generate more optimal code.
Some extra care is needed with hci_ecc.c since it was overwriting the
send callback. Now the choice of send() call is done directly in the
bt_send() function
Change-Id: Iac74f5ee9bee097bbb34c11bd13d1d886700f5cc
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Reduce the pressure on the common RX buffer pool by reusing HCI
command buffers also for the Command Status or Command Complete
response to them. This also implies removing the existing Kconfig
variable for the command buffer sizes since the size is also dependent
on maximum Command Complete event sizes. Instead, reuse the RX buffer
size also for HCI Command buffers.
Change-Id: I006b287d64a0c9ca40de741aa9a424a49a927385
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
If we are low on RX buffers and receive an LE Advertising Report or a
BR/EDR Inquiry response we should just discard this. If we don't
discard the event we increase the risk of deadlock where the RX
interrupt is enabled but bt_recv() is doing a synchronous HCI command
sending, i.e. waiting for a cmd_status/cmd_complete (which will never
come since the RX interrupt is disabled).
Change-Id: I6266625c9790d68bcf8e8718c8c36f127946c4c6
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Replace the existing Apache 2.0 boilerplate header with an SPDX tag
throughout the zephyr code tree. This patch was generated via a
script run over the master branch.
Also updated doc/porting/application.rst that had a dependency on
line numbers in a literal include.
Manually updated subsys/logging/sys_log.c that had a malformed
header in the original file. Also cleanup several cases that already
had a SPDX tag and we either got a duplicate or missed updating.
Jira: ZEP-1457
Change-Id: I6131a1d4ee0e58f5b938300c2d2fc77d2e69572c
Signed-off-by: David B. Kinder <david.b.kinder@intel.com>
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
The new IS_ENABLED macro allows exposing conditionally enabled code
always to the compiler, even though it may not ultimately end up being
built. This is in particular useful for letting the compiler catch any
logging format string errors. Introduce a new BT_DBG_ENABLED macro
that c-files need to define before including <bluetooth/log.h> in
order to choose whether BT_DBG() logs are enabled or not.
When no Bluetooth logs are enabled the patch also modifies the log
macros to have the format strings checked with the help of the
__printf_like annotation and empty static inline functions.
Change-Id: Ie6bc8e10727b5b306f3ed0f94089a07a22583d9b
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Since callers of bt_recv() have so far anyway been required to know in
which context to call it (based on e.g. bt_hci_evt_is_prio) it's
cleaner to have two separate APIs: bt_recv and bt_recv_prio.
Change-Id: Icd0d9aed9c51ffd2def31432c4ffcc16a9f13ccd
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Use interrupt based transmission instead of polling.
Change-Id: Iebfd67372044d3e6de9bdbdd6f0c9c6d01d46cb9
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
So far the use of k_sem meant that there was no major benefit of
having more than 2 or 3 RX buffers since there was no queuing
mechanism. Instead of using k_sem, introduce a k_fifo and use that to
queue up incoming buffers. This way the RX buffer count can be
increased with measurable effects on throughput.
Change-Id: I8122b233aeee7c8e145de3fff5f10bcfe348efaa
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Completely redesign the H:4 driver to utilize its own thread and
select the new RECV_IS_RX_THREAD Kconfig option.
Jira: ZEP-1483
Change-Id: I0ca0661b66d564e5edc1b8505706d6bb76632e79
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
This helps simplify code that was previously combining net_buf_add()
with memcpy().
Change-Id: If44cf9cd651aba5e544e36567869f147468663eb
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Not all users are in an ISR context where we can't block, so give the
callers the freedom to choose if they want to block or not.
Jira: ZEP-1481
Change-Id: I19bd7e2df94c4eeb60886a17a78f872bd7bea887
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Fix compilation issues that show up if SYS_LOG is mapped to printk
instead of printf. Unlike printf, printk is annotated so that the
compiler catches incorrect format specifiers passed to it.
Change-Id: I4d6f635a0ed61de698727028ea8767dc0ef28bb1
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Don't use the deprecated NANOKERNEL init level. The only requirement
for Bluetooth drivers is for them to be registered before the
application main() runs, so POST_KERNEL should be good enough.
Change-Id: I02a8609bf63e9d608b802576214a2e76211b3965
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
In 3 of our UART driver implementations the ISR is exclusively
used and enabled for the RX path. The existing logic was
susceptible to a stall situation where a polled out transmission
would be interrupted by a reception and then the ISR code
would loop forever due to the TX event being signalled (although
the interrupt itself was disabled) causing the ISR to keep
looping for an RX interrupt.
Change-Id: Ic379e58b1c974aca3cee37d2d81f12c3726fb160
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
The bt_driver API was created when Zephyr only had a Bluetooth host
stack, but no controller-side functionality. The only "driver" that
was needed for the host was the HCI driver, and hence "HCI" was
omitted from the name.
With support both for host and controller Zephyr will be getting more
Bluetooth driver types, in particular radio drivers. To prepare for
this, move all HCI drivers to drivers/bluetooth/hci/ and rename the
bt_driver API bt_hci_driver.
Change-Id: I82829da80aa61f26c2bb2005380f1e88d069ac7d
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
