The CDNS will report a M0 error if the data length is not what
it expects, but certain CCCs can have a variable length such as
GETMXDS and GETCAPS. This sets it up to ignore the M0 error if
it sees that ccc was GETMXDS or GETCAPS.
Signed-off-by: Ryan McClelland <ryanmcclelland@meta.com>
GETSTATUS and GETMRL where not checking the right argc length. This
corrects it to check for the right count.
Signed-off-by: Ryan McClelland <ryanmcclelland@meta.com>
For many of the state checks with timeout, the check
should be very fast (sub 1 microseconds) and the
previous implemention involving busy waits between
checks could add unneeded latency. Change the
timeout checking method to use WAIT_FOR(), with
no extra delays between checks.
Signed-off-by: Mike J. Chen <mjchen@google.com>
This implements basic driver to utilize the I3C IP block
on NPCX.
1. I3C mode: Main controller mode only.
2. Transfer: Support SDR only.
3. IBI: Support Hot-Join, IBI(MDB).
Controller request is not supported.
4. Support 3 I3C modules:
I3C1(3.3V), I3C2(1.8V, espi mode), (I3C3 1.8V or 3.3V)
Signed-off-by: Alvis Sun <yfsun@nuvoton.com>
During DAA, the responding device might not be in the device list.
This CL adds target device descriptor's pointer checking to prevent
getting unexpected results.
Signed-off-by: Alvis Sun <yfsun@nuvoton.com>
Add an I3C shell. This includes support of all I3C CCC commands that
currently have helper functions implemented. This also includes all
the read/write shell commands that the i2c shell supported. An Info
command is also provided which will print out all i3c and i2c info
of an i3c bus. Only SDR read/writes are currently implemented.
Signed-off-by: Ryan McClelland <ryanmcclelland@meta.com>
Have seen timeout error occur when doing an emit stop at
end of a multiple register write to a mmc5633 mag sensor IC.
The i3c_mcux driver would return on such an error without
doing the k_condvar_broadcast(), since the stop wasn't
technically done (or unclear if it was), and then future
transfers requests waiting on the condvar could be blocked
forever. Add a limited retry when a timeout occurs to
avoid such a stall condition from happening.
Signed-off-by: Mike J. Chen <mjchen@google.com>
Error check was incorrect, causing wait for complete
not to be done. This can result in emit stop not
working on writes because controller won't do the
stop if it is still busy processing transmit.
Signed-off-by: Mike J. Chen <mjchen@google.com>
The main improvement is that mcux_i3c_transfer() and
mcux_i3c_i2c_api_transfer() will try much harder to not return an error
that could be caused by the bus being busy. The bus could be busy because
of IBI handling, especially if there are multiple I3C devices all raising
IBI that need to be processed, which can involve a number of context
switches and delays and take considerable time such that an application
initiated I3C transfer request might have returned busy. Replaced the code
that polled for idle state with a timeout with an infinite loop on a
condvar. The condvar is broadcast to at the end of every stop, which should
be when the bus goes idle.
Details of other changes:
* Remove ibi_lock, which seemed not useful. Use the single lock (switched
from semaphore to mutex so it can be released by condvar) for both IBI
handling and application requests.
* Remove code that disables i3c controller interrupts during transfers.
Since the only interrupt is SLVSTART, and that just posts a work, it
didn't seem necessary to disable that interrupt.
* Don't clear SLVSTART interrupt in mcux_i3c_status_clear_all(), to prevent
any application transfers from accidentally clearing the SLVSTART
interrupt or the handling of one IBI clearing the START initiated by
another I3C device immeidately as the other one finishes.
The only clearing of SLVSTART is in the isr.
* Add back a wait in mcux_i3c_request_auto_ibi(), otherwise the ibitype
could be 0 if the processor is faster than the auto ibi handling.
An earlier change removed a wait for MCTRLDONE, which isn't
always set when AUTO_IBI is requested, but that could mean we try to
read the ibitype before it's ready. Waiting for IBIWON instead should be
correct and better, since the AUTO_IBI should result in IBIWON status bit
being set (and MCTRLDONE being set would not guarantee that IBIWON was
set).
* Change mcux_i3c_request_emit_stop() to still wait for idle if requested,
even if the STOP wasn't actually issued, and add the release of the new
condvar
* Change mcux_i3c_do_one_xfer_read() to handle the IBI use case differently
than the regular application requested transfer case. In the application
requested transfer case, the rx_len is known and set in RDTERM, so we
could check for the COMPLETE status bit, but for IBI transfers, we
just do a read to the payload buffer without knowing how many bytes
the target may send us so never get a COMPLETE. Rather than check for a
COMPLETE bit that might never occur, just have both cases read until we
either read all requested bytes or we get a timeout error. But for the
timeout error, if it's IBI and non-zero bytes were received, return
the bytes received instead of an error.
* Change mcux_i3c_do_one_xfer() to return the error returned by
mcux_i3c_do_one_xfer_read/write().
* Remove spurious return -EIO from end of mcux_i3c_do_daa()
* Change mcux_i3c_ibi_enable() to restore SLVSTART on error, and
add some LOG_ERR() messages for error cases. Also add check to
make sure idx is valid since there is a limit to how many IBI
this controller can support.
* Change mcux_i3c_ibi_disable() to always reenable SLVSTART interrupt,
even if the CCC to tell the target to disable IBI events fails.
* Change mcux_i3c_isr() to reenable SLVSTART interrupt if the
work_enqueue() fails.
Signed-off-by: Mike J. Chen <mjchen@google.com>
It might happens that DT(_INST)_PROP_OR is used with boolean properties.
For instance:
.single_wire = DT_INST_PROP_OR(index, single_wire, false), \
.tx_rx_swap = DT_INST_PROP_OR(index, tx_rx_swap, false), \
This is not required as boolean properties are generated with false
value when not present, so the _OR macro extension is superflous
and the above code can be replaced by:
.single_wire = DT_INST_PROP(index, single_wire), \
.tx_rx_swap = DT_INST_PROP(index, tx_rx_swap), \
Signed-off-by: Roman Studenikin <srv@meta.com>
Some targets do not give EoD at the end of a register read. They will
auto increment their address pointer on to the next address, but that
may not be of interest to the application where the buffer size will
only be set to the size of only that register. If the target, does
not give an EoD, then the Controller will give an Abort... but this
should not be treated as an error in this case.
There is still however a case where an abort Error shall still be
considered as an error. Athough the driver does not support it yet,
threshold interrupts are to be used if the length of the buffer
exceeds the size of the fifo. There could be the case where the
cpu can not get around fast enough to pop out data out of the rx
fifo and it will fill up. The controller will just give an abort
as it can not take any more data.
Signed-off-by: Ryan McClelland <ryanmcclelland@meta.com>
The command response buffer will return the total number of bytes
transfered. This will write back to the pointer which is to contain
the number of bytes sent or received.
Signed-off-by: Ryan McClelland <ryanmcclelland@meta.com>
This renames the I2C 'DCR' mode to 'LVR' as that is the variable it
should be looking at and not the dcr value. This also fixes the get
'lvr' mode argument.
Signed-off-by: Ryan McClelland <ryanmcclelland@meta.com>
Previously, the idle bit would be read for X amount of times. This
could vary alot depend on the CPU speed. Timeout is now to happen
from the cycle time.
Signed-off-by: Ryan McClelland <ryanmcclelland@meta.com>
It is just used to be able to DEVICE_DT_GET() bus devices from
tests/drivers/build_all in an upcoming commit.
Signed-off-by: Armando Visconti <armando.visconti@st.com>
For example, if a driver needed to reserve address before it does a
ENTDAA, it would need to get free address in a loop, but the get
free address func would return the same address everytime. It needs
the start address, which would be the last free address it go, to
be passed in to get the next free address.
Signed-off-by: Ryan McClelland <ryanmcclelland@meta.com>
Move the syscall_handler.h header, used internally only to a dedicated
internal folder that should not be used outside of Zephyr.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
The default is that the high time for open-drain clk is one
PPBAUD, which is typically very short. Some device require
a longer high time during the open-drain address phase so
add a property to allow device tree to override the default.
Signed-off-by: Mike J. Chen <mjchen@google.com>
Remove the MCTRLDONE wait in mcux_i3c_request_auto_ibi().
I've seen this code getting stuck where the MCTRLDONE
bit is never set in the MSTATUS register by the controller
and this function spins forever. Documentaiton of the
MCTRLDONE bit only mentions it being set for EmitStartAddr
and ProcessDAA, but not for AutoIBI requests.
All the calls to this function do completion checks
afterwards, and with a timeout, so I believe the MCTRLDONE
check is not needed (and may not even be correct).
Signed-off-by: Mike J. Chen <mjchen@google.com>
At high i3c rates, the mcux_i3c_do_one_xfer_read()
could get into an infinite loop where the rx_count
kept returning 0 but the complete status bit
was never set. I believe the problem was that
the function was not emptying the FIFO fast enough,
so tighten the loop that processes the FIFO.
Signed-off-by: Mike J. Chen <mjchen@google.com>
mcux_i3c_configure() was saving values to a ctrl_config_hal
struct, but config_get() was not returning the values in
that struct. Remove that struct from the static data of
the driver and instead just have it on the stack. We init
that struct as needed just before calling the SDK API
I3C_MasterInit(). There's no reason to keep it around.
Change mcux_i3c_configure() to save a copy of the configuration
in the static data common.ctrl_config, which is what is
returned by config_get().
Signed-off-by: Mike J. Chen <mjchen@google.com>
The Cadence I3C was not building with CONFIG_I3C_USE_IBI, this fixes
the build and will give a small code size reduction when enabled.
Signed-off-by: Ryan McClelland <ryanmcclelland@meta.com>
If a transfer happen in rapid sucession. It was possible for
the core to not be ready to accept another command. Poll on
the idle status bit until the core is ready to accept new data.
Signed-off-by: Ryan McClelland <ryanmcclelland@meta.com>
Due to a bug, after a completed transfer happen. Only the first
command response error was read. This fixes the issue so all
commands are read for if an error occurred.
Signed-off-by: Ryan McClelland <ryanmcclelland@meta.com>
This adds a few line use zephyr_syscall_header() to include
headers containing syscall function prototypes.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
Fixes for bug:
https://github.com/zephyrproject-rtos/zephyr/issues/57560
* don't do CCC if no i3c devices in device tree
* don't wait for MCTRLDONE status when issuing stop
* don't do data part of transfer if buf_sz is 0
* don't limit transfers to only i2c devices in the device tree
so "i2c scan" shell cmd works as expected
Signed-off-by: Mike J. Chen <mjchen@google.com>
The MCUX platform always uses pinctrl, there's no need to keep extra
macrology around pinctrl. Also updated driver's Kconfig options to
`select PINCTRL` (note that some already did).
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
The init infrastructure, found in `init.h`, is currently used by:
- `SYS_INIT`: to call functions before `main`
- `DEVICE_*`: to initialize devices
They are all sorted according to an initialization level + a priority.
`SYS_INIT` calls are really orthogonal to devices, however, the required
function signature requires a `const struct device *dev` as a first
argument. The only reason for that is because the same init machinery is
used by devices, so we have something like:
```c
struct init_entry {
int (*init)(const struct device *dev);
/* only set by DEVICE_*, otherwise NULL */
const struct device *dev;
}
```
As a result, we end up with such weird/ugly pattern:
```c
static int my_init(const struct device *dev)
{
/* always NULL! add ARG_UNUSED to avoid compiler warning */
ARG_UNUSED(dev);
...
}
```
This is really a result of poor internals isolation. This patch proposes
a to make init entries more flexible so that they can accept sytem
initialization calls like this:
```c
static int my_init(void)
{
...
}
```
This is achieved using a union:
```c
union init_function {
/* for SYS_INIT, used when init_entry.dev == NULL */
int (*sys)(void);
/* for DEVICE*, used when init_entry.dev != NULL */
int (*dev)(const struct device *dev);
};
struct init_entry {
/* stores init function (either for SYS_INIT or DEVICE*)
union init_function init_fn;
/* stores device pointer for DEVICE*, NULL for SYS_INIT. Allows
* to know which union entry to call.
*/
const struct device *dev;
}
```
This solution **does not increase ROM usage**, and allows to offer clean
public APIs for both SYS_INIT and DEVICE*. Note that however, init
machinery keeps a coupling with devices.
**NOTE**: This is a breaking change! All `SYS_INIT` functions will need
to be converted to the new signature. See the script offered in the
following commit.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
init: convert SYS_INIT functions to the new signature
Conversion scripted using scripts/utils/migrate_sys_init.py.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
manifest: update projects for SYS_INIT changes
Update modules with updated SYS_INIT calls:
- hal_ti
- lvgl
- sof
- TraceRecorderSource
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
tests: devicetree: devices: adjust test
Adjust test according to the recently introduced SYS_INIT
infrastructure.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
tests: kernel: threads: adjust SYS_INIT call
Adjust to the new signature: int (*init_fn)(void);
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
When a controller is running at full SDR speed at 12.5MHz, there needs
to be enough time for the processor get around to writing more data in
the fifo. Previously at -1 the size, this was enough for 1MHz with a
decent processor, but not enough at a 12.5MHz SCL.
Signed-off-by: Ryan McClelland <ryanmcclelland@meta.com>
The cadence i3c ip requires it's retaining registers to be updated
when a device is detached or attached.
Signed-off-by: Ryan McClelland <ryanmcclelland@meta.com>
There are some needs to attach and reattach i3c/i2c devices at runtime
Some I2C devices can have special registers where the address can be
changed at runtime. Also some I3C devices can be powered off at runtime
freeing up the address space they take up. These new APIs allow for these
to be changed at runtime. This also moves some config/data in to a common
i3c config/data structure which would allow the api to operate on to be
common for all I3C drivers.
Signed-off-by: Ryan McClelland <ryanmcclelland@meta.com>
This adds the reattach api necessary for writing the i3c retaining
registers within the cdns i3c when the dynamic address changes.
Signed-off-by: Ryan McClelland <ryanmcclelland@meta.com>
Some I3C controllers have retaining registers which are used to contain
the DA of the i3c device. This needs to be updated every time the DA is
updated with SETNEWDA or SETDASA
Signed-off-by: Ryan McClelland <ryanmcclelland@meta.com>
