Calls to flash write or erase operations can get preempted
in the middle of the API call if a concurrent call (from
other fiber or task) to the same API is made. The patch ensures
concurrent API calls from other fiber/task is put on hold until
the current operation is completed.
The mechanism is by default not activated. To enable
it, the following flag needs to be defined:
CONFIG_SOC_FLASH_QMSI_API_REENTRANCY
Jira: ZEP-414
Change-Id: I39429e40cb6ed446123dd1a1d7c7acc1b12417aa
Signed-off-by: Kuo-Lang Tseng <kuo-lang.tseng@intel.com>
pcal9535a driver is now using new system log macros and updated the
Kconfig variable to be a level rather than a bool.
JIRA: ZEP-311
Change-Id: Ic3d6c72630e66ac116c1e8954c32254677c51566
Signed-off-by: Genaro Saucedo Tejada <genaro.saucedo.tejada@intel.com>
GPIO drivers are now using new system log macros also updated the
Kconfig variable to be a level rather than a bool.
JIRA: ZEP-311
Change-Id: I9a49626d816080cb8081c2dd445bae31f5dbf409
Signed-off-by: Genaro Saucedo Tejada <genaro.saucedo.tejada@intel.com>
Drivers that implement power management should use the preferred
device_pm_ops method instead.
Change-Id: I9ae06e26b77325265bbe46bdee82ba39dedb6b79
Signed-off-by: Iván Briano <ivan.briano@intel.com>
Drivers that implement power management should use the preferred
device_pm_ops method instead.
Change-Id: I337722b1e06afe8508b5c84c00c3542571232e07
Signed-off-by: Iván Briano <ivan.briano@intel.com>
The suspend and resume hooks in the spi_driver_api struct are relics
from before the current power management infrastructure was in place.
The correct way to implement this now is through the device_pm_ops
struct, by way of the DEFINE_DEVICE_PM_OPS and DEVICE_AND_API_INIT_PM
macros, which make the hooks available through a generic mechanism for
all devices, rather than using per-type APIs.
Since the existing spi_suspend() and spi_resume() functions don't check
if the driver_api hooks are NULL, there's now a place holder function
to prevent breaking functionality until the hooks are removed.
Change-Id: I48287c58e9a8649d3e1be7547e3d0d293c84327a
Signed-off-by: Iván Briano <ivan.briano@intel.com>
The suspend and resume hooks in the spi_driver_api struct are relics
from before the current power management infrastructure was in place.
The correct way to implement this now is through the device_pm_ops
struct, by way of the DEFINE_DEVICE_PM_OPS and DEVICE_AND_API_INIT_PM
macros, which make the hooks available through a generic mechanism for
all devices, rather than using per-type APIs.
Since the existing spi_suspend() and spi_resume() functions don't check
if the driver_api hooks are NULL, there's now a place holder function
to prevent breaking functionality until the hooks are removed.
Change-Id: I6a3e3db370860ad46f428d287943b1ca58a80ae1
Signed-off-by: Iván Briano <ivan.briano@intel.com>
Drivers that implement power management should use the preferred
device_pm_ops method instead.
Change-Id: Ice9e0469a1fcb50eb64dcb240dddea56755b6e84
Signed-off-by: Iván Briano <ivan.briano@intel.com>
The suspend and resume hooks in the pwm_driver_api struct are relics
from before the current power management infrastructure was in place.
The correct way to implement this now is through the device_pm_ops
struct, by way of the DEFINE_DEVICE_PM_OPS and DEVICE_AND_API_INIT_PM
macros, which make the hooks available through a generic mechanism for
all devices, rather than using per-type APIs.
Since the existing pwm_suspend() and pwm_resume() functions don't check
if the driver_api hooks are NULL, there's now a place holder function
to prevent breaking functionality until the hooks are removed.
Change-Id: Ib6b03053b483f97a1f7441af5ba51503270b8674
Signed-off-by: Iván Briano <ivan.briano@intel.com>
Added CONFIG_SYS_LOG_I2C_LEVEL define, so that DW I2C driver will
not print traces unless requested with CONFIG_SYS_LOG_I2C_LEVEL.
Currently it doesn't obey it, so DW I2C spams DBG traces when
logging is enabled.
Change-Id: I87dc1d922ba61d39f26945627602864c75d9bf33
Signed-off-by: Jani Pajarinen <jani.pajarinen@intel.com>
Fixed undefined CONFIG_ISL29035_LUX_RANGE_4 in #elif
Change-Id: I2a7a6e46ffa0236da126d4e2c321e59302e54527
Signed-off-by: Javier B Perez <javier.b.perez.hernandez@intel.com>
Guard critical section of the driver API, so that multiple
simultaneous calls from tasks/fibers won't corrupt the driver
state.
The locking mechanism is by default disabled. To enable it,
the following flag needs to be defined:
CONFIG_PWM_QMSI_API_REENTRANCY
Jira: ZEP-430
Change-Id: Ia3eb2a962f4176a3ac94163b9843ab068abe4b4e
Signed-off-by: Kuo-Lang Tseng <kuo-lang.tseng@intel.com>
Signed-off-by: Inaky Perez-Gonzalez <inaky.perez-gonzalez@intel.com>
In order to restore GPIO context after entering
SYS_PM_DEEP_SLEEP, the suspend and resume
functions for GPIO and GPOIO_AON are called. The
following parameters are restored:
- All non-sticky RW registers for the GPIO device.
- The MASK register (interrupt routing register).
Note: No need to sve/resume AON_GPIO registers as
they all are sticky registers.
The suspend/resume functionality is implemented in
the QMSI shim layer as a fast and temporary solution,
it will be removed and migrated to QMSI later.
Change-Id: I68e25c9da4bb8ea65f312e3686d9ef090cb775ff
Signed-off-by: Dragan Cvetic <dragan.cvetic@intel.com>
Code was using CONFIG_MAX44009_I2C_MASTER_DEV_NAME which doesn't exist.
Use CONFIG_MAX44009_I2C_DEV_NAME instead.
Change-Id: I6bf6eedd2229bb91f64264dc5b32e4a3c00921b6
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
CONFIG_UART_INTERRUPT_DRIVE should be CONFIG_UART_INTERRUPT_DRIVEN
CONFIG_UART_LINE_CTR should be CONFIG_UART_LINE_CTRL
Change-Id: I11fba5b387e3d36711cd3c813fc8a297ae23af05
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
We don't currently have a SOC_NRF5 Kconfig symbol. Utilize the
SOC_FAMILY_NRF5 Kconfig symbol instead.
Change-Id: I062ecab230e9e7814fad19517d28ddbbae66bccb
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
__ASSERT was being used when __ASSERT_NO_MSG was appropriate.
This problem was previously only visible if CONFIG_ASSERT=y,
but a recent change to the assert macros fixed them so they
can't be improperly used with assertions disabled either.
Change-Id: I1e13c796873d79826d177a01246d2ca0e8f0a107
Signed-off-by: Andrew Boie <andrew.p.boie@intel.com>
Guard critical section of the driver API, so that
multiple simultaneous calls from tasks/fibers wont
corrupt the driver state.
The locking mechanism is by default disabled in the
build. To enable it, the following flag needs to be defined:
CONFIG_RTC_QMSI_API_REENTRANCY
Jira: ZEP-412
Change-Id: I430ebc07bf3277e5a5961cb6278c5500cd1e139b
Signed-off-by: Jithu Joseph <jithu.joseph@intel.com>
Signed-off-by: Inaky Perez-Gonzalez <inaky.perez-gonzalez@intel.com>
Signed-off-by: Kuo-Lang Tseng <kuo-lang.tseng@intel.com>
Uses locking in an API and removes a global variable
for improved safety/correctness in the face of concurrent
threads and fibers.
The locking mechanism is by default disabled in the
build. To enable it, the following flag needs to be defined:
CONFIG_WDT_QMSI_API_REENTRANCY
Jira: ZEP-440
Change-Id: Ibcd1501a4af628017b20d7e7ce20b988c8e4e4e2
Signed-off-by: Jithu Joseph <jithu.joseph@intel.com>
Signed-off-by: Inaky Perez-Gonzalez <inaky.perez-gonzalez@intel.com>
Signed-off-by: Kuo-Lang Tseng <kuo-lang.tseng@intel.com>
The macro defining the device_pm_ops structs is located inside an
ifdef for CONFIG_UART_QMSI_0, so when the UART_0 is disabled at build
time, but UART_1 is still used, it references a struct not defined.
Fix it by placing the DEFINE_DEVICE_PM_OPS line outside the ifdef block.
Change-Id: I7dafda162741201ee47b5b480c07ec8c1f373d64
Signed-off-by: Iván Briano <ivan.briano@intel.com>
The slip driver revert (3e63a74) was incomplete, missing the Kconfig and
Makefile changes, causing 'make clean' to fail.
Change-Id: I9d944148e6be3756b62d2371a5ab5528365e1ec1
Signed-off-by: Ricardo Salveti <ricardo.salveti@linaro.org>
The idle status check before a i2c transfer is not a reliable
way to check whether i2c master is ready to transfer data. If
the data transfer for the previous API call is done, but, a
stop condition is not sent by the previous API, the i2c master
is not idle, but, it can start new data transfer.
Change-Id: I93dbf46cef6272f3da69a0d67fb737c9b480791f
Signed-off-by: Baohong Liu <baohong.liu@intel.com>
We have already done this on x86 and ARM. The policy is as follows:
* IRQ priority levels starting at 0 all have the same semantics and
do not have special properties. The priority level is either ignored
on arches which do not support programmable priority levels, or lower
priority levels take precedence over higher ones.
* Special-case priorty levels are specified via flags, in which case
the supplied priority level is ignored.
Issue: ZEP-60
Change-Id: Ic603f49299ee1426fb9350ca29d0b8ef96a1d53a
Signed-off-by: Andrew Boie <andrew.p.boie@intel.com>
This will guard the driver API to be safe for concurrent
invocation from fibers or tasks.
The machanism is by default disabled. To enable it, the
following flag needs to be defined:
CONFIG_GPIO_QMSI_API_REENTRANCY
Jira: ZEP-441
Change-Id: Ia3ee738aff2f8e70e4f9a7ec76346138ff6f5031
Signed-off-by: Sergio Rodriguez <sergio.sf.rodriguez@intel.com>
Signed-off-by: Inaky Perez-Gonzalez <inaky.perez-gonzalez@intel.com>
Signed-off-by: Kuo-Lang Tseng <kuo-lang.tseng@intel.com>
This will allow the driver API to be safe for concurrent invocation
from fibers and task.
The machanism is by default disabled. To enable it, the following
flag needs to be defined:
CONFIG_GPIO_QMSI_SS_API_REENTRANCY
Jira: ZEP-411
Change-Id: I5b738532f2eefec6d5d78ab6a668790dfe9faaff
Signed-off-by: Sergio Rodriguez <sergio.sf.rodriguez@intel.com>
Signed-off-by: Kuo-Lang Tseng <kuo-lang.tseng@intel.com>
Signed-off-by: Inaky Perez-Gonzalez <inaky.perez-gonzalez@intel.com>
At the moment firmware reads everything we send to it over UART and
when an application tries to send too much notifications we are stuck.
The proper approach would be forcing firmware to read from UART only
when there is available memory. In this case we would block on UART
FIFO and everything handles gracefully.
This workaround mimics solution used by firmware project by allowing
to send next notification only after previous was sent correctly.
Change-Id: I56dd5abc4d13735d285415da04db75e7bb8bc459
Signed-off-by: Andrei Emeltchenko <andrei.emeltchenko@intel.com>
Depending on advertising options this can be non-connectable address.
NFC pairing application document also allows broadcaster/observer
roles.
Change-Id: I9b104ac8fb9752a083a7a31fc20598c66f23f608
Signed-off-by: Szymon Janc <ext.szymon.janc@tieto.com>
Follow up to TSC decission for further discussion in the networking
WIG.
Change-Id: I148b484dfe308661573e47ed3e60cceed673bddf
Signed-off-by: Inaky Perez-Gonzalez <inaky.perez-gonzalez@intel.com>
This adds some basic suspend/resume support in the uart_qmsi
driver. This is part of a change to enable device suspend/resume
and deep sleep support in sample PM application.
Jira: ZEP-512
Change-Id: I38cfbfb352d0027b7beaadebedbc10002dade50a
Signed-off-by: Ramesh Thomas <ramesh.thomas@intel.com>
This adds some basic suspend/resume support in the rtc qmsi
driver. This is part of a change to enable device suspend/resume
and Deep Sleep support in sample application.
Jira: ZEP-512
Change-Id: I79ccfb56de5b3cbfb5bc91e9cacca68554aa46ee
Signed-off-by: Ramesh Thomas <ramesh.thomas@intel.com>
This is part of a change to enable device suspend/resume and
Deep Sleep support in applications. Adds suspend/resume handling
in loapic timer.
Jira: ZEP-512
Change-Id: I9da2c8419bd9109fb71ef5a6caf736de7c7de9e1
Signed-off-by: Ramesh Thomas <ramesh.thomas@intel.com>
This will guard critical regions in the driver API so that the API
operation called from one thread (task or fiber) can complete
without the context being over-written by any simultaneous call
of the API from other thread (task or fiber).
The locking mechanism is by default disabled in the build. To enable it,
the following flag needs to be defined:
CONFIG_AON_API_REENTRANCY
Jira: ZEP-424
Change-Id: I1b0eae70a4ab97229931033153662874d7c4a68e
Signed-off-by: Kuo-Lang Tseng <kuo-lang.tseng@intel.com>
Signed-off-by: Inaky Perez-Gonzalez <inaky.perez-gonzalez@intel.com>
This allows to query information needed for Out Of Band pairing
or connection creation. Currently supports only BT addresses.
Change-Id: I60bf9344baee552e7743fa8fc1b3cfb3a4765334
Signed-off-by: Szymon Janc <ext.szymon.janc@tieto.com>
This lets the callback take advantage of the powerful net_buf API for
parsing the advertising data content.
Change-Id: Id65e6e83efd60c0f36c47bc5446a2e8ec2833d7c
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
This makes bt_gatt_write consistent with the rest of the API where the
parameters are passed back to its callback.
Change-Id: Ie94208aa661d3620d0cbc5be4a4fb5b3c3ef061c
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Fixes SEC/AUT/BV-11 PTS test case at least in GUI mode.
Change-Id: I60e7013f1c66027dabc1b94df87f2da991a5718c
Signed-off-by: Andrei Emeltchenko <andrei.emeltchenko@intel.com>
QMSI provides more than just peripheral drivers, and it may be used
from the SOC or by applications that need finer grained control of the
platform.
This also makes the QMSI HAL consistent with the others.
Change-Id: I27490c98ccf1c1afcc090ce1e65f0e066eddc777
Signed-off-by: Iván Briano <ivan.briano@intel.com>
Correct that handle is attribute one, print flag to notice cancel
prepared write requests.
Change-Id: I46d747f0466a44460e92fcbaef4e436066c09f4a
Signed-off-by: Andrei Emeltchenko <andrei.emeltchenko@intel.com>
The latest NBLE firmware includes a new version
exchange mechanism that is independent from the
RPC function definition. It also allows using
firmwares that are automatically generated by
the official builders without needing a local
build.
The upgrade of the softdevice fixes several
issues in PTS tests. The major change in
term of API introduced by this upgrade is that
upon write requests, the offset and data of the
write request must be sent back to the softdevice.
Change-Id: Ie574047a09a4e6fdcbf775bc98d8a61f981cee0c
Signed-off-by: Louis Caron <louis.caron@intel.com>
The comments state that higher-numbered vectors have priority,
but the function was doing the opposite. The code now agrees
with the comments.
Change-Id: I1a689fc22c8355458bdd734d95ff637da0655533
Signed-off-by: Andrew Boie <andrew.p.boie@intel.com>
Created menu for Ethernet driver and moved driver + debugging setting
under it. Changed prompt from "Grove log level" to more describing and
fixed typos.
Change-Id: If2f79fee1a002a42e89e2493a2294332abec05ee
Signed-off-by: Jani Pajarinen <jani.pajarinen@intel.com>
As there is a L2 in the middle, there is no need for a net_if to provide
such information.
Change-Id: I58a35d4d124cd8a67026f32e71713ef06f43b7c5
Signed-off-by: Tomasz Bursztyka <tomasz.bursztyka@linux.intel.com>
Net core then does not know anything about l2 related logic.
For instance ARP is used in ethernet l2 API and nowhere else.
This will be helpful when adding different technologies altogether.
Currently, only SLIP driver is enabled to use relevant l2 layer.
Change-Id: I03c93326321028d04222733ca4083e3c6b785202
Signed-off-by: Tomasz Bursztyka <tomasz.bursztyka@linux.intel.com>
