If CONFIG_SPI_STATS is enabled, the device state for all SPI controller
drivers must contain the SPI stats. This space is allocated by calling
Z_SPI_INIT_FN as part of the device definition; this is done automatically
when using SPI_DEVICE_DT_DEFINE instead of DEVICE_DT_DEFINE. If space for
statistics is not properly allocated but CONFIG_SPI_STATS is enabled, an
unexpected write to memory outside of the stats region may occur on a SPI
transfer. This commit uses SPI_DEVICE_DT_DEFINE or
SPI_DEVICE_DT_INST_DEFINE for all in-tree SPI controller drivers.
Signed-off-by: Dane Wagner <dane.wagner@gmail.com>
- Following similar approach followed on spi_mcux_lpspi driver.
- Enabling DMA by default when SPI RTIO is selected to favor
non-blocking transfers.
Signed-off-by: Luis Ubieda <luisf@croxel.com>
Transmit/write buffers are expected to be constant values given to the
operation to transmit but not mutate. Seperate the OP_TX and OP_RX
operation description unions.
Signed-off-by: Tom Burdick <thomas.burdick@intel.com>
I/O Devices were meant to be handles of sorts and had a built in mpsc queue
as this made sense initially. As time has gone on it turned out that often
we wanted the mpsc queue to be an implementation detail hidden in a driver.
In fact pretty much all drivers work this way now.
Keeping the struct mpsc queue as a member of rtio_iodev meant wasted memory
in cases where it wasn't used. It also meant a bit of confusion as the
queue might be accidently used in places where it shouldn't be.
Remove the mpsc queue member from struct rtio_iodev and the last remaining
usages of it. Will ensure RTIO for 3.7 LTS avoids causing unneeded churn
for future users.
Signed-off-by: Tom Burdick <thomas.burdick@intel.com>
Moves the rtio_ prefixed lockfree queues to sys alongside existing
mpsc/spsc pbuf, ringbuf, and similar queue-like data structures.
Signed-off-by: Tom Burdick <thomas.burdick@intel.com>
Previously the logic was inverted for error_callback_en where 0 was
enablement and 1 was disable. This was likely done so that the default,
sensibly so, was to enable the error callback if possible. A variety of
in tree users had confused the enable/disable value.
Change the name of the flag to error_callback_dis where the default
remains 0 (do not disable the callback!) and correct in tree uses of the
flag where it seemed incorrect.
Signed-off-by: Tom Burdick <thomas.burdick@intel.com>
Rather than the rings, which weren't shared between userspace and kernel
space in Zephyr like they are in Linux with io_uring, use atomic mpsc
queues for submission and completion queues.
Most importantly this removes a potential head of line blocker in the
submission queue as the sqe would be held until a task is completed.
As additional bonuses this avoids some additional locks and restrictions
about what can be submitted and where. It also removes the need for
two executors as all chains/transactions are done concurrently.
Lastly this opens up the possibility for a common pool of sqe's to
allocate from potentially saving lots of memory.
Signed-off-by: Tom Burdick <thomas.burdick@intel.com>
When using gpio chip select the clock line seems to get stuck low after
some transactions. When attempting to use other SPI_CSR registers
the peripheral fails to work as expected.
Always using SPI_CSR[0] when using gpio chip selects resolves the issue.
Signed-off-by: Tom Burdick <thomas.burdick@intel.com>
Implements the SPI RTIO API. Refactors many of the internal transfer
functions to work on simple types to allow for both the RTIO and existing
SPI API functions to co-exist.
Removes the unneeded shift_master specialization. Either a polling or DMA
transfer flow is used when used with RTIO as all operations are normalized.
When SPI_RTIO is enabled the spi_transceive blocking call translates
the request into an RTIO transaction placed in the queue of transactions
the device is meant to do.
Signed-off-by: Tom Burdick <thomas.burdick@intel.com>
A MACRO argument naming mismatch causes a bug when trying to use DMA.
Fix the MACRO argument and conditional DMA configuration.
Signed-off-by: Pieter De Gendt <pieter.degendt@basalte.be>
Uninitialized sam spinlocks were causing an error if the stack
happens to have 'thread_cpu' memory set to something other than 0.
Signed-off-by: Yuval Peress <peress@google.com>
The fast tx/rx/txrx functions will leave the SPI bus in an
inoperable state if interrupted, potentially spinning forever waiting on
some data. Wrapping these operations in what amounts to a critical section
using spin locks to avoid the issue.
Signed-off-by: Tom Burdick <thomas.burdick@intel.com>
For larger transfers DMA can be used enabling other tasks
to continue running. A threshold of 32 byte transfers
is about right and is defined threshold value for using DMA.
This does not currently support multiple SPI transactions changing
chip select with DMA (though the hardware supports this) currently.
Instead opting for the simpler first change of enabling one shot
DMA SPI transfers for those where the size warrants it.
Adds the loopback binding option to enable the spi_loopback test.
Signed-off-by: Tom Burdick <thomas.burdick@intel.com>
Adds a new spi_transcieve_cb API which enables asynchronous
SPI transactions with callback notification.
The exist spi_transcieve_async API remains and uses the new
spi_transcieve_cb API to provide a k_poll_signal notifier.
The driver API changes to provide a callback and userdata
parameter to async transcieve. All drivers in the tree
have been updated to the change.
Signed-off-by: Tom Burdick <thomas.burdick@intel.com>
In order to bring consistency in-tree, migrate all drivers to the new
prefix <zephyr/...>. Note that the conversion has been scripted, refer
to #45388 for more details.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
This update Atmel sam spi driver to use pinctrl driver and API. It
updates all boards with new pinctrl groups format.
Signed-off-by: Gerson Fernando Budke <nandojve@gmail.com>
In case when we have multiple devices connected to the
one SPI interface the initial state of CS gpios after
MCU reset is floating and it might be low that prevents us from
communicating between particular devices. Fix that by
initializing all provided cs gpios and setting them as inactive.
Signed-off-by: Bartosz Bilas <bartosz.bilas@hotmail.com>
The macros are used to get the pin(s) of a given driver instance. Add
_INST prefix to match convention used by the devicetree.h. The original
macros can now be used to obtain pin(s) of an arbitrary device instance
identified by the nodelabel.
Signed-off-by: Piotr Mienkowski <piotr.mienkowski@gmail.com>
Keep locking for SPI_LOCK_ON from the first call of transceive until
spi_release release the lock. Use owner parameter to in the spi_context
to store the owner of the lock.
The locking is in line with the SPI_HOLD_ON_CS
Signed-off-by: Stefan Bigler <stefan@bigler.io>
Replaces all existing variants of value clamping with the MIN and MAX
macros with the CLAMP macro.
Signed-off-by: Trond Einar Snekvik <Trond.Einar.Snekvik@nordicsemi.no>
Now that device_api attribute is unmodified at runtime, as well as all
the other attributes, it is possible to switch all device driver
instance to be constant.
A coccinelle rule is used for this:
@r_const_dev_1
disable optional_qualifier
@
@@
-struct device *
+const struct device *
@r_const_dev_2
disable optional_qualifier
@
@@
-struct device * const
+const struct device *
Fixes#27399
Signed-off-by: Tomasz Bursztyka <tomasz.bursztyka@linux.intel.com>
Usually, we want to operate only on "available" device
nodes ("available" means "status is okay and a matching binding is
found"), but that's not true in all cases.
Sometimes we want to operate on special nodes without matching
bindings, such as those describing memory.
To handle the distinction, change various additional devicetree APIs
making it clear that they operate only on available device nodes,
adjusting gen_defines and devicetree.h implementation details
accordingly:
- emit macros for all existing nodes in gen_defines.py, regardless
of status or matching binding
- rename DT_NUM_INST to DT_NUM_INST_STATUS_OKAY
- rename DT_NODE_HAS_COMPAT to DT_NODE_HAS_COMPAT_STATUS_OKAY
- rename DT_INST_FOREACH to DT_INST_FOREACH_STATUS_OKAY
- rename DT_ANY_INST_ON_BUS to DT_ANY_INST_ON_BUS_STATUS_OKAY
- rewrite DT_HAS_NODE_STATUS_OKAY in terms of a new DT_NODE_HAS_STATUS
- resurrect DT_HAS_NODE in the form of DT_NODE_EXISTS
- remove DT_COMPAT_ON_BUS as a public API
- use the new default_prop_types edtlib parameter
Signed-off-by: Martí Bolívar <marti.bolivar@nordicsemi.no>
Remove semicolon between instance invocations of DT_FOREACH_IMPL_ and
thus DT_INST_FOREACH. This provides more flexibility to the user. This
requires we fixup in tree users to add semicolon where needed.
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
Reworked spi_sam driver to utilize new DT_INST macros as part of
this rework we also now get pin ctrl/mux configuration information
from the device tree instead of via Kconfig and defines in soc_pinmap.h
We remove defines from dts_fixup.h and soc_pinmap.h and associated
Kconfig symbols that are no longer needed due to getting all that
information from devicetree.
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
move spi.h to drivers/spi.h and
create a shim for backward-compatibility.
No functional changes to the headers.
A warning in the shim can be controlled with CONFIG_COMPAT_INCLUDES.
Related to #16539
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
Found a few annoying typos and figured I better run script and
fix anything it can find, here are the results...
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
Asyc transceive is currently not implemented for this driver,
so don't claim support.
spi_context_lock() is already being called in spi_sam_transceive()
so calling it in the wrapper function will cause a deadlock.
This reverts eae05d928e
Signed-off-by: Benjamin Valentin <benjamin.valentin@ml-pa.com>
There are issues using lowercase min and max macros when compiling a C++
application with a third-party toolchain such as GNU ARM Embedded when
using some STL headers i.e. <chrono>.
This is because there are actual C++ functions called min and max
defined in some of the STL headers and these macros interfere with them.
By changing the macros to UPPERCASE, which is consistent with almost all
other pre-processor macros this naming conflict is avoided.
All files that use these macros have been updated.
Signed-off-by: Carlos Stuart <carlosstuart1970@gmail.com>
initialize both tx and rx in the spi_sam0 driver. Make the spi_sam
driver look the same by splitting the declaration into 2 lines.
Discovered with gcc 8.2
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
If the configuration is already installed, there will no need to
reconfigure the controller all over again.
This was missing for mcux_dspi, mcux_lspi, sam and sam0.
Signed-off-by: Tomasz Bursztyka <tomasz.bursztyka@linux.intel.com>
Move to using Kconfig (like other Atmel SAM drivers do) to specify the
pinmux setting for SAMe70 SoCs. Updated the sam_e70_xplained board to
set the default in Kconfig.defconfig instead of via board.h
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>