unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
/*
|
|
|
|
|
* Copyright (c) 2016 Wind River Systems, Inc.
|
|
|
|
|
*
|
2017-01-18 17:01:01 -08:00
|
|
|
* SPDX-License-Identifier: Apache-2.0
|
unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
*/
|
|
|
|
|
|
|
|
|
|
/**
|
|
|
|
|
* @file
|
|
|
|
|
* @brief Message queues.
|
|
|
|
|
*/
|
|
|
|
|
|
|
|
|
|
|
2022-05-06 11:04:23 +02:00
|
|
|
#include <zephyr/kernel.h>
|
2021-04-18 23:24:40 -04:00
|
|
|
|
2022-05-06 11:04:23 +02:00
|
|
|
#include <zephyr/toolchain.h>
|
|
|
|
|
#include <zephyr/linker/sections.h>
|
unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
#include <string.h>
|
2019-10-25 00:08:21 +09:00
|
|
|
#include <ksched.h>
|
kernel: wait_q: add z_unpend_first_thread_locked() and migrate callers
Replace z_unpend_first_thread() with z_unpend_first_thread_locked() and
migrate every caller across the kernel. The old function dropped the
scheduler spinlock before returning, exposing a race window between
its caller's "arch_thread_return_value_set + z_ready_thread" pair and
a still-in-flight timeout handler that could ready the thread first --
the woken thread might then run on another CPU and see an uninitialized
swap_retval. Pre-1b8c7a3 the dticks-cancel check made the handler bail;
here we fix it cleanly by requiring the caller to hold _sched_spinlock
across the entire wake, so the handler is blocked for the duration and
runs as a no-op afterwards.
z_unpend_first_thread_locked() requires the caller to be inside a
locked region and must be paired with z_sched_ready_locked() (and
whatever return-value setup is needed) under the same lock acquisition.
Sites migrated:
Simple "set retval [+ swap_data] and ready" callers use the existing
z_sched_wake() convenience wrapper, refactored to use the new
helper internally:
sem (give, reset), mem_slab (free), stack (push),
condvar (signal, broadcast), msgq (purge),
queue (cancel_wait, queue_insert, append_list),
futex (wake).
Sites that need additional setup on the woken thread use
LOCK_SCHED_SPINLOCK + z_unpend_first_thread_locked() + custom wake:
mutex (unlock -- needs the thread reference to track new owner),
msgq put / get (needs memcpy into the receiver's swap_data
buffer before the return value is set).
The dticks-cancel check in z_thread_timeout() is left in place; it is
no longer load-bearing once z_abort_thread_timeout() has no callers,
and is removed in the next commit.
Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
2026-05-27 18:28:32 -04:00
|
|
|
#include <scheduler.h>
|
2023-08-29 17:03:12 +00:00
|
|
|
#include <wait_q.h>
|
2022-05-06 11:04:23 +02:00
|
|
|
#include <zephyr/sys/dlist.h>
|
|
|
|
|
#include <zephyr/sys/math_extras.h>
|
|
|
|
|
#include <zephyr/init.h>
|
2023-09-26 22:46:01 +00:00
|
|
|
#include <zephyr/internal/syscall_handler.h>
|
2018-09-05 10:13:38 -07:00
|
|
|
#include <kernel_internal.h>
|
2022-05-06 11:04:23 +02:00
|
|
|
#include <zephyr/sys/check.h>
|
2016-10-19 16:10:46 -05:00
|
|
|
|
kernel: Integrate object cores into kernel
Integrates object cores into the following kernel structures
sys_mem_blocks, k_mem_slab
_cpu, z_kernel
k_thread, k_timer
k_condvar, k_event, k_mutex, k_sem
k_mbox, k_msgq, k_pipe, k_fifo, k_lifo, k_stack
Signed-off-by: Peter Mitsis <peter.mitsis@intel.com>
2023-05-11 14:06:46 -04:00
|
|
|
#ifdef CONFIG_OBJ_CORE_MSGQ
|
|
|
|
|
static struct k_obj_type obj_type_msgq;
|
2024-03-08 12:00:10 +01:00
|
|
|
#endif /* CONFIG_OBJ_CORE_MSGQ */
|
kernel: Integrate object cores into kernel
Integrates object cores into the following kernel structures
sys_mem_blocks, k_mem_slab
_cpu, z_kernel
k_thread, k_timer
k_condvar, k_event, k_mutex, k_sem
k_mbox, k_msgq, k_pipe, k_fifo, k_lifo, k_stack
Signed-off-by: Peter Mitsis <peter.mitsis@intel.com>
2023-05-11 14:06:46 -04:00
|
|
|
|
2026-06-20 06:50:05 -05:00
|
|
|
static inline bool msgq_handle_poll_events(struct k_msgq *msgq)
|
2021-04-12 12:35:18 -07:00
|
|
|
{
|
2024-12-16 14:48:13 -08:00
|
|
|
#ifdef CONFIG_POLL
|
|
|
|
|
return z_handle_obj_poll_events(&msgq->poll_events,
|
|
|
|
|
K_POLL_STATE_MSGQ_DATA_AVAILABLE);
|
|
|
|
|
#else
|
|
|
|
|
ARG_UNUSED(msgq);
|
|
|
|
|
return false;
|
2021-04-12 12:35:18 -07:00
|
|
|
#endif /* CONFIG_POLL */
|
2024-12-16 14:48:13 -08:00
|
|
|
}
|
2021-04-12 12:35:18 -07:00
|
|
|
|
2019-06-19 07:30:50 -04:00
|
|
|
void k_msgq_init(struct k_msgq *msgq, char *buffer, size_t msg_size,
|
2020-05-27 11:26:57 -05:00
|
|
|
uint32_t max_msgs)
|
unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
{
|
2026-02-18 23:45:46 -08:00
|
|
|
__ASSERT_NO_MSG(!size_mul_overflow(max_msgs, msg_size, &(size_t){0}));
|
|
|
|
|
__ASSERT_NO_MSG(!size_add_overflow((size_t)(uintptr_t)buffer, max_msgs * msg_size,
|
|
|
|
|
&(size_t){0}));
|
|
|
|
|
|
2019-06-19 07:30:50 -04:00
|
|
|
msgq->msg_size = msg_size;
|
|
|
|
|
msgq->max_msgs = max_msgs;
|
|
|
|
|
msgq->buffer_start = buffer;
|
|
|
|
|
msgq->buffer_end = buffer + (max_msgs * msg_size);
|
|
|
|
|
msgq->read_ptr = buffer;
|
|
|
|
|
msgq->write_ptr = buffer;
|
|
|
|
|
msgq->used_msgs = 0;
|
|
|
|
|
msgq->flags = 0;
|
|
|
|
|
z_waitq_init(&msgq->wait_q);
|
|
|
|
|
msgq->lock = (struct k_spinlock) {};
|
2021-04-12 12:35:18 -07:00
|
|
|
#ifdef CONFIG_POLL
|
|
|
|
|
sys_dlist_init(&msgq->poll_events);
|
|
|
|
|
#endif /* CONFIG_POLL */
|
2021-03-26 12:39:53 +01:00
|
|
|
|
kernel: Integrate object cores into kernel
Integrates object cores into the following kernel structures
sys_mem_blocks, k_mem_slab
_cpu, z_kernel
k_thread, k_timer
k_condvar, k_event, k_mutex, k_sem
k_mbox, k_msgq, k_pipe, k_fifo, k_lifo, k_stack
Signed-off-by: Peter Mitsis <peter.mitsis@intel.com>
2023-05-11 14:06:46 -04:00
|
|
|
#ifdef CONFIG_OBJ_CORE_MSGQ
|
|
|
|
|
k_obj_core_init_and_link(K_OBJ_CORE(msgq), &obj_type_msgq);
|
2024-03-08 12:00:10 +01:00
|
|
|
#endif /* CONFIG_OBJ_CORE_MSGQ */
|
kernel: Integrate object cores into kernel
Integrates object cores into the following kernel structures
sys_mem_blocks, k_mem_slab
_cpu, z_kernel
k_thread, k_timer
k_condvar, k_event, k_mutex, k_sem
k_mbox, k_msgq, k_pipe, k_fifo, k_lifo, k_stack
Signed-off-by: Peter Mitsis <peter.mitsis@intel.com>
2023-05-11 14:06:46 -04:00
|
|
|
|
2021-03-26 12:39:53 +01:00
|
|
|
SYS_PORT_TRACING_OBJ_INIT(k_msgq, msgq);
|
|
|
|
|
|
2023-09-26 21:32:13 +00:00
|
|
|
k_object_init(msgq);
|
unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
}
|
|
|
|
|
|
2019-06-19 07:30:50 -04:00
|
|
|
int z_impl_k_msgq_alloc_init(struct k_msgq *msgq, size_t msg_size,
|
2020-05-27 11:26:57 -05:00
|
|
|
uint32_t max_msgs)
|
2018-04-12 18:35:56 -07:00
|
|
|
{
|
|
|
|
|
void *buffer;
|
|
|
|
|
int ret;
|
|
|
|
|
size_t total_size;
|
|
|
|
|
|
2021-03-26 12:39:53 +01:00
|
|
|
SYS_PORT_TRACING_OBJ_FUNC_ENTER(k_msgq, alloc_init, msgq);
|
|
|
|
|
|
2019-05-07 10:17:35 -07:00
|
|
|
if (size_mul_overflow(msg_size, max_msgs, &total_size)) {
|
2018-04-12 18:35:56 -07:00
|
|
|
ret = -EINVAL;
|
|
|
|
|
} else {
|
|
|
|
|
buffer = z_thread_malloc(total_size);
|
2018-09-20 16:30:45 -07:00
|
|
|
if (buffer != NULL) {
|
2019-06-19 07:30:50 -04:00
|
|
|
k_msgq_init(msgq, buffer, msg_size, max_msgs);
|
|
|
|
|
msgq->flags = K_MSGQ_FLAG_ALLOC;
|
2018-04-12 18:35:56 -07:00
|
|
|
ret = 0;
|
|
|
|
|
} else {
|
|
|
|
|
ret = -ENOMEM;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
2021-03-26 12:39:53 +01:00
|
|
|
SYS_PORT_TRACING_OBJ_FUNC_EXIT(k_msgq, alloc_init, msgq, ret);
|
2018-04-12 18:35:56 -07:00
|
|
|
return ret;
|
|
|
|
|
}
|
|
|
|
|
|
2017-10-02 10:53:06 -07:00
|
|
|
#ifdef CONFIG_USERSPACE
|
2021-03-29 10:54:23 -04:00
|
|
|
int z_vrfy_k_msgq_alloc_init(struct k_msgq *msgq, size_t msg_size,
|
2020-05-27 11:26:57 -05:00
|
|
|
uint32_t max_msgs)
|
2017-10-02 10:53:06 -07:00
|
|
|
{
|
2023-09-27 11:20:28 +00:00
|
|
|
K_OOPS(K_SYSCALL_OBJ_NEVER_INIT(msgq, K_OBJ_MSGQ));
|
2017-10-02 10:53:06 -07:00
|
|
|
|
2021-03-29 10:54:23 -04:00
|
|
|
return z_impl_k_msgq_alloc_init(msgq, msg_size, max_msgs);
|
2017-10-02 10:53:06 -07:00
|
|
|
}
|
2024-01-24 17:35:04 +08:00
|
|
|
#include <zephyr/syscalls/k_msgq_alloc_init_mrsh.c>
|
2024-03-08 12:00:10 +01:00
|
|
|
#endif /* CONFIG_USERSPACE */
|
2017-10-02 10:53:06 -07:00
|
|
|
|
2026-08-10 15:51:18 -07:00
|
|
|
int z_msgq_cleanup(struct k_msgq *msgq, __maybe_unused bool locked)
|
2018-04-12 18:35:56 -07:00
|
|
|
{
|
2026-08-10 15:51:18 -07:00
|
|
|
SYS_PORT_TRACING_OBJ_FUNC_ENTER(k_msgq, cleanup, msgq);
|
|
|
|
|
|
2025-07-30 19:55:05 -04:00
|
|
|
int ret = 0;
|
2026-08-10 14:12:36 -07:00
|
|
|
k_spinlock_key_t key = k_spin_lock(&msgq->lock);
|
|
|
|
|
|
2026-08-10 15:51:18 -07:00
|
|
|
CHECKIF(locked && (z_waitq_head_locked(&msgq->wait_q) != NULL)) {
|
|
|
|
|
ret = -EBUSY;
|
|
|
|
|
goto out;
|
|
|
|
|
}
|
2021-03-26 12:39:53 +01:00
|
|
|
|
2026-08-10 15:51:18 -07:00
|
|
|
CHECKIF(!locked && (z_waitq_head(&msgq->wait_q) != NULL)) {
|
2025-07-30 19:55:05 -04:00
|
|
|
ret = -EBUSY;
|
2026-08-09 11:06:05 -04:00
|
|
|
goto out;
|
2019-06-16 08:43:48 -04:00
|
|
|
}
|
2018-04-12 18:35:56 -07:00
|
|
|
|
2021-03-29 10:03:49 -04:00
|
|
|
if ((msgq->flags & K_MSGQ_FLAG_ALLOC) != 0U) {
|
2026-04-10 21:52:56 -04:00
|
|
|
k_free(msgq->buffer_start);
|
2019-06-19 07:30:50 -04:00
|
|
|
msgq->flags &= ~K_MSGQ_FLAG_ALLOC;
|
2018-04-12 18:35:56 -07:00
|
|
|
}
|
2021-03-26 12:39:53 +01:00
|
|
|
|
2026-08-09 11:06:05 -04:00
|
|
|
out:
|
2026-08-10 14:12:36 -07:00
|
|
|
k_spin_unlock(&msgq->lock, key);
|
2025-07-30 19:55:05 -04:00
|
|
|
SYS_PORT_TRACING_OBJ_FUNC_EXIT(k_msgq, cleanup, msgq, ret);
|
|
|
|
|
return ret;
|
2018-04-12 18:35:56 -07:00
|
|
|
}
|
|
|
|
|
|
2026-08-10 15:51:18 -07:00
|
|
|
int k_msgq_cleanup(struct k_msgq *msgq)
|
|
|
|
|
{
|
|
|
|
|
return z_msgq_cleanup(msgq, false);
|
|
|
|
|
}
|
|
|
|
|
|
2025-07-09 22:18:09 +01:00
|
|
|
static inline int put_msg_in_queue(struct k_msgq *msgq, const void *data,
|
|
|
|
|
k_timeout_t timeout, bool put_at_back)
|
unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
{
|
kernel/timeout: Make timeout arguments an opaque type
Add a k_timeout_t type, and use it everywhere that kernel API
functions were accepting a millisecond timeout argument. Instead of
forcing milliseconds everywhere (which are often not integrally
representable as system ticks), do the conversion to ticks at the
point where the timeout is created. This avoids an extra unit
conversion in some application code, and allows us to express the
timeout in units other than milliseconds to achieve greater precision.
The existing K_MSEC() et. al. macros now return initializers for a
k_timeout_t.
The K_NO_WAIT and K_FOREVER constants have now become k_timeout_t
values, which means they cannot be operated on as integers.
Applications which have their own APIs that need to inspect these
vs. user-provided timeouts can now use a K_TIMEOUT_EQ() predicate to
test for equality.
Timer drivers, which receive an integer tick count in ther
z_clock_set_timeout() functions, now use the integer-valued
K_TICKS_FOREVER constant instead of K_FOREVER.
For the initial release, to preserve source compatibility, a
CONFIG_LEGACY_TIMEOUT_API kconfig is provided. When true, the
k_timeout_t will remain a compatible 32 bit value that will work with
any legacy Zephyr application.
Some subsystems present timeout (or timeout-like) values to their own
users as APIs that would re-use the kernel's own constants and
conventions. These will require some minor design work to adapt to
the new scheme (in most cases just using k_timeout_t directly in their
own API), and they have not been changed in this patch, instead
selecting CONFIG_LEGACY_TIMEOUT_API via kconfig. These subsystems
include: CAN Bus, the Microbit display driver, I2S, LoRa modem
drivers, the UART Async API, Video hardware drivers, the console
subsystem, and the network buffer abstraction.
k_sleep() now takes a k_timeout_t argument, with a k_msleep() variant
provided that works identically to the original API.
Most of the changes here are just type/configuration management and
documentation, but there are logic changes in mempool, where a loop
that used a timeout numerically has been reworked using a new
z_timeout_end_calc() predicate. Also in queue.c, a (when POLL was
enabled) a similar loop was needlessly used to try to retry the
k_poll() call after a spurious failure. But k_poll() does not fail
spuriously, so the loop was removed.
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
2020-03-05 15:18:14 -08:00
|
|
|
__ASSERT(!arch_is_in_isr() || K_TIMEOUT_EQ(timeout, K_NO_WAIT), "");
|
2016-11-09 19:45:19 -05:00
|
|
|
|
kernel: wait_q: add z_unpend_first_thread_locked() and migrate callers
Replace z_unpend_first_thread() with z_unpend_first_thread_locked() and
migrate every caller across the kernel. The old function dropped the
scheduler spinlock before returning, exposing a race window between
its caller's "arch_thread_return_value_set + z_ready_thread" pair and
a still-in-flight timeout handler that could ready the thread first --
the woken thread might then run on another CPU and see an uninitialized
swap_retval. Pre-1b8c7a3 the dticks-cancel check made the handler bail;
here we fix it cleanly by requiring the caller to hold _sched_spinlock
across the entire wake, so the handler is blocked for the duration and
runs as a no-op afterwards.
z_unpend_first_thread_locked() requires the caller to be inside a
locked region and must be paired with z_sched_ready_locked() (and
whatever return-value setup is needed) under the same lock acquisition.
Sites migrated:
Simple "set retval [+ swap_data] and ready" callers use the existing
z_sched_wake() convenience wrapper, refactored to use the new
helper internally:
sem (give, reset), mem_slab (free), stack (push),
condvar (signal, broadcast), msgq (purge),
queue (cancel_wait, queue_insert, append_list),
futex (wake).
Sites that need additional setup on the woken thread use
LOCK_SCHED_SPINLOCK + z_unpend_first_thread_locked() + custom wake:
mutex (unlock -- needs the thread reference to track new owner),
msgq put / get (needs memcpy into the receiver's swap_data
buffer before the return value is set).
The dticks-cancel check in z_thread_timeout() is left in place; it is
no longer load-bearing once z_abort_thread_timeout() has no callers,
and is removed in the next commit.
Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
2026-05-27 18:28:32 -04:00
|
|
|
struct k_thread *pending_thread = NULL;
|
2019-06-19 07:30:50 -04:00
|
|
|
k_spinlock_key_t key;
|
unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
int result;
|
2024-12-16 17:12:30 -08:00
|
|
|
bool resched = false;
|
unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
|
2019-06-19 07:30:50 -04:00
|
|
|
key = k_spin_lock(&msgq->lock);
|
|
|
|
|
|
2025-07-16 15:23:00 -03:00
|
|
|
if (put_at_back) {
|
|
|
|
|
SYS_PORT_TRACING_OBJ_FUNC_ENTER(k_msgq, put, msgq, timeout);
|
|
|
|
|
} else {
|
|
|
|
|
SYS_PORT_TRACING_OBJ_FUNC_ENTER(k_msgq, put_front, msgq, timeout);
|
|
|
|
|
}
|
2021-03-26 12:39:53 +01:00
|
|
|
|
2019-06-19 07:30:50 -04:00
|
|
|
if (msgq->used_msgs < msgq->max_msgs) {
|
kernel: wait_q: add z_unpend_first_thread_locked() and migrate callers
Replace z_unpend_first_thread() with z_unpend_first_thread_locked() and
migrate every caller across the kernel. The old function dropped the
scheduler spinlock before returning, exposing a race window between
its caller's "arch_thread_return_value_set + z_ready_thread" pair and
a still-in-flight timeout handler that could ready the thread first --
the woken thread might then run on another CPU and see an uninitialized
swap_retval. Pre-1b8c7a3 the dticks-cancel check made the handler bail;
here we fix it cleanly by requiring the caller to hold _sched_spinlock
across the entire wake, so the handler is blocked for the duration and
runs as a no-op afterwards.
z_unpend_first_thread_locked() requires the caller to be inside a
locked region and must be paired with z_sched_ready_locked() (and
whatever return-value setup is needed) under the same lock acquisition.
Sites migrated:
Simple "set retval [+ swap_data] and ready" callers use the existing
z_sched_wake() convenience wrapper, refactored to use the new
helper internally:
sem (give, reset), mem_slab (free), stack (push),
condvar (signal, broadcast), msgq (purge),
queue (cancel_wait, queue_insert, append_list),
futex (wake).
Sites that need additional setup on the woken thread use
LOCK_SCHED_SPINLOCK + z_unpend_first_thread_locked() + custom wake:
mutex (unlock -- needs the thread reference to track new owner),
msgq put / get (needs memcpy into the receiver's swap_data
buffer before the return value is set).
The dticks-cancel check in z_thread_timeout() is left in place; it is
no longer load-bearing once z_abort_thread_timeout() has no callers,
and is removed in the next commit.
Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
2026-05-27 18:28:32 -04:00
|
|
|
/* message queue isn't full. Try to hand the message
|
|
|
|
|
* directly to the longest-waiting receiver, atomically
|
2026-08-10 12:21:47 -07:00
|
|
|
* under the scheduler's spinlock so a racing in-flight timeout
|
kernel: wait_q: add z_unpend_first_thread_locked() and migrate callers
Replace z_unpend_first_thread() with z_unpend_first_thread_locked() and
migrate every caller across the kernel. The old function dropped the
scheduler spinlock before returning, exposing a race window between
its caller's "arch_thread_return_value_set + z_ready_thread" pair and
a still-in-flight timeout handler that could ready the thread first --
the woken thread might then run on another CPU and see an uninitialized
swap_retval. Pre-1b8c7a3 the dticks-cancel check made the handler bail;
here we fix it cleanly by requiring the caller to hold _sched_spinlock
across the entire wake, so the handler is blocked for the duration and
runs as a no-op afterwards.
z_unpend_first_thread_locked() requires the caller to be inside a
locked region and must be paired with z_sched_ready_locked() (and
whatever return-value setup is needed) under the same lock acquisition.
Sites migrated:
Simple "set retval [+ swap_data] and ready" callers use the existing
z_sched_wake() convenience wrapper, refactored to use the new
helper internally:
sem (give, reset), mem_slab (free), stack (push),
condvar (signal, broadcast), msgq (purge),
queue (cancel_wait, queue_insert, append_list),
futex (wake).
Sites that need additional setup on the woken thread use
LOCK_SCHED_SPINLOCK + z_unpend_first_thread_locked() + custom wake:
mutex (unlock -- needs the thread reference to track new owner),
msgq put / get (needs memcpy into the receiver's swap_data
buffer before the return value is set).
The dticks-cancel check in z_thread_timeout() is left in place; it is
no longer load-bearing once z_abort_thread_timeout() has no callers,
and is removed in the next commit.
Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
2026-05-27 18:28:32 -04:00
|
|
|
* handler cannot wake the receiver before the message
|
|
|
|
|
* has been copied into its buffer.
|
|
|
|
|
*/
|
|
|
|
|
LOCK_SCHED_SPINLOCK {
|
|
|
|
|
pending_thread = z_unpend_first_thread_locked(&msgq->wait_q);
|
|
|
|
|
if (pending_thread != NULL) {
|
|
|
|
|
/* copy into the receiver's buffer */
|
|
|
|
|
(void)memcpy(pending_thread->base.swap_data, data,
|
|
|
|
|
msgq->msg_size);
|
|
|
|
|
arch_thread_return_value_set(pending_thread, 0);
|
|
|
|
|
z_sched_ready_locked(pending_thread);
|
|
|
|
|
resched = true;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
if (pending_thread == NULL) {
|
2026-02-17 13:34:09 -08:00
|
|
|
__ASSERT_NO_MSG((msgq->write_ptr >= msgq->buffer_start) &&
|
2026-02-19 09:55:27 -08:00
|
|
|
(msgq->write_ptr <= (msgq->buffer_end - 1)) &&
|
|
|
|
|
((size_t)(uintptr_t)(msgq->buffer_end - msgq->write_ptr) >=
|
|
|
|
|
msgq->msg_size));
|
2025-07-09 22:18:09 +01:00
|
|
|
if (put_at_back) {
|
|
|
|
|
/*
|
|
|
|
|
* to write a message to the back of the queue,
|
|
|
|
|
* copy the message and increment write_ptr
|
|
|
|
|
*/
|
|
|
|
|
(void)memcpy(msgq->write_ptr, (char *)data, msgq->msg_size);
|
|
|
|
|
msgq->write_ptr += msgq->msg_size;
|
|
|
|
|
if (msgq->write_ptr == msgq->buffer_end) {
|
|
|
|
|
msgq->write_ptr = msgq->buffer_start;
|
|
|
|
|
}
|
|
|
|
|
} else {
|
|
|
|
|
/*
|
|
|
|
|
* to write a message to the head of the queue,
|
|
|
|
|
* first decrement the read pointer (to open
|
|
|
|
|
* space at the front of the queue) then copy
|
|
|
|
|
* the message to the newly created space.
|
|
|
|
|
*/
|
|
|
|
|
if (msgq->read_ptr == msgq->buffer_start) {
|
|
|
|
|
msgq->read_ptr = msgq->buffer_end;
|
|
|
|
|
}
|
|
|
|
|
msgq->read_ptr -= msgq->msg_size;
|
|
|
|
|
(void)memcpy(msgq->read_ptr, (char *)data, msgq->msg_size);
|
unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
}
|
2019-06-19 07:30:50 -04:00
|
|
|
msgq->used_msgs++;
|
2026-06-20 06:50:05 -05:00
|
|
|
resched = msgq_handle_poll_events(msgq);
|
unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
}
|
|
|
|
|
result = 0;
|
kernel/timeout: Make timeout arguments an opaque type
Add a k_timeout_t type, and use it everywhere that kernel API
functions were accepting a millisecond timeout argument. Instead of
forcing milliseconds everywhere (which are often not integrally
representable as system ticks), do the conversion to ticks at the
point where the timeout is created. This avoids an extra unit
conversion in some application code, and allows us to express the
timeout in units other than milliseconds to achieve greater precision.
The existing K_MSEC() et. al. macros now return initializers for a
k_timeout_t.
The K_NO_WAIT and K_FOREVER constants have now become k_timeout_t
values, which means they cannot be operated on as integers.
Applications which have their own APIs that need to inspect these
vs. user-provided timeouts can now use a K_TIMEOUT_EQ() predicate to
test for equality.
Timer drivers, which receive an integer tick count in ther
z_clock_set_timeout() functions, now use the integer-valued
K_TICKS_FOREVER constant instead of K_FOREVER.
For the initial release, to preserve source compatibility, a
CONFIG_LEGACY_TIMEOUT_API kconfig is provided. When true, the
k_timeout_t will remain a compatible 32 bit value that will work with
any legacy Zephyr application.
Some subsystems present timeout (or timeout-like) values to their own
users as APIs that would re-use the kernel's own constants and
conventions. These will require some minor design work to adapt to
the new scheme (in most cases just using k_timeout_t directly in their
own API), and they have not been changed in this patch, instead
selecting CONFIG_LEGACY_TIMEOUT_API via kconfig. These subsystems
include: CAN Bus, the Microbit display driver, I2S, LoRa modem
drivers, the UART Async API, Video hardware drivers, the console
subsystem, and the network buffer abstraction.
k_sleep() now takes a k_timeout_t argument, with a k_msleep() variant
provided that works identically to the original API.
Most of the changes here are just type/configuration management and
documentation, but there are logic changes in mempool, where a loop
that used a timeout numerically has been reworked using a new
z_timeout_end_calc() predicate. Also in queue.c, a (when POLL was
enabled) a similar loop was needlessly used to try to retry the
k_poll() call after a spurious failure. But k_poll() does not fail
spuriously, so the loop was removed.
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
2020-03-05 15:18:14 -08:00
|
|
|
} else if (K_TIMEOUT_EQ(timeout, K_NO_WAIT)) {
|
unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
/* don't wait for message space to become available */
|
|
|
|
|
result = -ENOMSG;
|
|
|
|
|
} else {
|
2025-07-16 15:23:00 -03:00
|
|
|
if (put_at_back) {
|
|
|
|
|
SYS_PORT_TRACING_OBJ_FUNC_BLOCKING(k_msgq, put, msgq, timeout);
|
|
|
|
|
} else {
|
|
|
|
|
SYS_PORT_TRACING_OBJ_FUNC_BLOCKING(k_msgq, put_front, msgq, timeout);
|
|
|
|
|
}
|
2021-03-26 12:39:53 +01:00
|
|
|
|
unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
/* wait for put message success, failure, or timeout */
|
2025-01-07 12:00:43 -05:00
|
|
|
_current->base.swap_data = (void *) data;
|
2021-03-26 12:39:53 +01:00
|
|
|
|
|
|
|
|
result = z_pend_curr(&msgq->lock, key, &msgq->wait_q, timeout);
|
2025-07-16 15:23:00 -03:00
|
|
|
|
|
|
|
|
if (put_at_back) {
|
|
|
|
|
SYS_PORT_TRACING_OBJ_FUNC_EXIT(k_msgq, put, msgq, timeout, result);
|
|
|
|
|
} else {
|
|
|
|
|
SYS_PORT_TRACING_OBJ_FUNC_EXIT(k_msgq, put_front, msgq, timeout, result);
|
|
|
|
|
}
|
|
|
|
|
|
2021-03-26 12:39:53 +01:00
|
|
|
return result;
|
unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
}
|
|
|
|
|
|
2025-07-16 15:23:00 -03:00
|
|
|
if (put_at_back) {
|
|
|
|
|
SYS_PORT_TRACING_OBJ_FUNC_EXIT(k_msgq, put, msgq, timeout, result);
|
|
|
|
|
} else {
|
|
|
|
|
SYS_PORT_TRACING_OBJ_FUNC_EXIT(k_msgq, put_front, msgq, timeout, result);
|
|
|
|
|
}
|
2021-03-26 12:39:53 +01:00
|
|
|
|
2024-12-16 17:12:30 -08:00
|
|
|
if (resched) {
|
|
|
|
|
z_reschedule(&msgq->lock, key);
|
|
|
|
|
} else {
|
|
|
|
|
k_spin_unlock(&msgq->lock, key);
|
|
|
|
|
}
|
unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
|
|
|
|
|
return result;
|
|
|
|
|
}
|
|
|
|
|
|
2025-07-09 22:18:09 +01:00
|
|
|
|
|
|
|
|
int z_impl_k_msgq_put(struct k_msgq *msgq, const void *data, k_timeout_t timeout)
|
|
|
|
|
{
|
|
|
|
|
return put_msg_in_queue(msgq, data, timeout, true);
|
|
|
|
|
}
|
|
|
|
|
|
2025-07-30 13:43:18 +08:00
|
|
|
int z_impl_k_msgq_put_front(struct k_msgq *msgq, const void *data)
|
2025-07-09 22:18:09 +01:00
|
|
|
{
|
2025-07-30 13:43:18 +08:00
|
|
|
return put_msg_in_queue(msgq, data, K_NO_WAIT, false);
|
2025-07-09 22:18:09 +01:00
|
|
|
}
|
|
|
|
|
|
2017-10-02 10:53:06 -07:00
|
|
|
#ifdef CONFIG_USERSPACE
|
2021-03-29 10:54:23 -04:00
|
|
|
static inline int z_vrfy_k_msgq_put(struct k_msgq *msgq, const void *data,
|
kernel/timeout: Make timeout arguments an opaque type
Add a k_timeout_t type, and use it everywhere that kernel API
functions were accepting a millisecond timeout argument. Instead of
forcing milliseconds everywhere (which are often not integrally
representable as system ticks), do the conversion to ticks at the
point where the timeout is created. This avoids an extra unit
conversion in some application code, and allows us to express the
timeout in units other than milliseconds to achieve greater precision.
The existing K_MSEC() et. al. macros now return initializers for a
k_timeout_t.
The K_NO_WAIT and K_FOREVER constants have now become k_timeout_t
values, which means they cannot be operated on as integers.
Applications which have their own APIs that need to inspect these
vs. user-provided timeouts can now use a K_TIMEOUT_EQ() predicate to
test for equality.
Timer drivers, which receive an integer tick count in ther
z_clock_set_timeout() functions, now use the integer-valued
K_TICKS_FOREVER constant instead of K_FOREVER.
For the initial release, to preserve source compatibility, a
CONFIG_LEGACY_TIMEOUT_API kconfig is provided. When true, the
k_timeout_t will remain a compatible 32 bit value that will work with
any legacy Zephyr application.
Some subsystems present timeout (or timeout-like) values to their own
users as APIs that would re-use the kernel's own constants and
conventions. These will require some minor design work to adapt to
the new scheme (in most cases just using k_timeout_t directly in their
own API), and they have not been changed in this patch, instead
selecting CONFIG_LEGACY_TIMEOUT_API via kconfig. These subsystems
include: CAN Bus, the Microbit display driver, I2S, LoRa modem
drivers, the UART Async API, Video hardware drivers, the console
subsystem, and the network buffer abstraction.
k_sleep() now takes a k_timeout_t argument, with a k_msleep() variant
provided that works identically to the original API.
Most of the changes here are just type/configuration management and
documentation, but there are logic changes in mempool, where a loop
that used a timeout numerically has been reworked using a new
z_timeout_end_calc() predicate. Also in queue.c, a (when POLL was
enabled) a similar loop was needlessly used to try to retry the
k_poll() call after a spurious failure. But k_poll() does not fail
spuriously, so the loop was removed.
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
2020-03-05 15:18:14 -08:00
|
|
|
k_timeout_t timeout)
|
2017-10-02 10:53:06 -07:00
|
|
|
{
|
2023-09-27 11:20:28 +00:00
|
|
|
K_OOPS(K_SYSCALL_OBJ(msgq, K_OBJ_MSGQ));
|
|
|
|
|
K_OOPS(K_SYSCALL_MEMORY_READ(data, msgq->msg_size));
|
2017-10-02 10:53:06 -07:00
|
|
|
|
2021-03-29 10:54:23 -04:00
|
|
|
return z_impl_k_msgq_put(msgq, data, timeout);
|
2017-10-02 10:53:06 -07:00
|
|
|
}
|
2024-01-24 17:35:04 +08:00
|
|
|
#include <zephyr/syscalls/k_msgq_put_mrsh.c>
|
2025-07-09 22:18:09 +01:00
|
|
|
|
2025-07-30 13:43:18 +08:00
|
|
|
static inline int z_vrfy_k_msgq_put_front(struct k_msgq *msgq, const void *data)
|
2025-07-09 22:18:09 +01:00
|
|
|
{
|
|
|
|
|
K_OOPS(K_SYSCALL_OBJ(msgq, K_OBJ_MSGQ));
|
|
|
|
|
K_OOPS(K_SYSCALL_MEMORY_READ(data, msgq->msg_size));
|
|
|
|
|
|
2025-07-30 13:43:18 +08:00
|
|
|
return z_impl_k_msgq_put_front(msgq, data);
|
2025-07-09 22:18:09 +01:00
|
|
|
}
|
|
|
|
|
#include <zephyr/syscalls/k_msgq_put_front_mrsh.c>
|
2024-03-08 12:00:10 +01:00
|
|
|
#endif /* CONFIG_USERSPACE */
|
2017-10-02 10:53:06 -07:00
|
|
|
|
2019-06-19 07:30:50 -04:00
|
|
|
void z_impl_k_msgq_get_attrs(struct k_msgq *msgq, struct k_msgq_attrs *attrs)
|
2018-03-19 20:02:40 +05:30
|
|
|
{
|
2019-06-19 07:30:50 -04:00
|
|
|
attrs->msg_size = msgq->msg_size;
|
|
|
|
|
attrs->max_msgs = msgq->max_msgs;
|
|
|
|
|
attrs->used_msgs = msgq->used_msgs;
|
2018-03-19 20:02:40 +05:30
|
|
|
}
|
|
|
|
|
|
|
|
|
|
#ifdef CONFIG_USERSPACE
|
2021-03-29 10:54:23 -04:00
|
|
|
static inline void z_vrfy_k_msgq_get_attrs(struct k_msgq *msgq,
|
2019-08-13 12:58:38 -07:00
|
|
|
struct k_msgq_attrs *attrs)
|
2018-03-19 20:02:40 +05:30
|
|
|
{
|
2023-09-27 11:20:28 +00:00
|
|
|
K_OOPS(K_SYSCALL_OBJ(msgq, K_OBJ_MSGQ));
|
|
|
|
|
K_OOPS(K_SYSCALL_MEMORY_WRITE(attrs, sizeof(struct k_msgq_attrs)));
|
2021-03-29 10:54:23 -04:00
|
|
|
z_impl_k_msgq_get_attrs(msgq, attrs);
|
2018-03-19 20:02:40 +05:30
|
|
|
}
|
2024-01-24 17:35:04 +08:00
|
|
|
#include <zephyr/syscalls/k_msgq_get_attrs_mrsh.c>
|
2024-03-08 12:00:10 +01:00
|
|
|
#endif /* CONFIG_USERSPACE */
|
2018-03-19 20:02:40 +05:30
|
|
|
|
kernel/timeout: Make timeout arguments an opaque type
Add a k_timeout_t type, and use it everywhere that kernel API
functions were accepting a millisecond timeout argument. Instead of
forcing milliseconds everywhere (which are often not integrally
representable as system ticks), do the conversion to ticks at the
point where the timeout is created. This avoids an extra unit
conversion in some application code, and allows us to express the
timeout in units other than milliseconds to achieve greater precision.
The existing K_MSEC() et. al. macros now return initializers for a
k_timeout_t.
The K_NO_WAIT and K_FOREVER constants have now become k_timeout_t
values, which means they cannot be operated on as integers.
Applications which have their own APIs that need to inspect these
vs. user-provided timeouts can now use a K_TIMEOUT_EQ() predicate to
test for equality.
Timer drivers, which receive an integer tick count in ther
z_clock_set_timeout() functions, now use the integer-valued
K_TICKS_FOREVER constant instead of K_FOREVER.
For the initial release, to preserve source compatibility, a
CONFIG_LEGACY_TIMEOUT_API kconfig is provided. When true, the
k_timeout_t will remain a compatible 32 bit value that will work with
any legacy Zephyr application.
Some subsystems present timeout (or timeout-like) values to their own
users as APIs that would re-use the kernel's own constants and
conventions. These will require some minor design work to adapt to
the new scheme (in most cases just using k_timeout_t directly in their
own API), and they have not been changed in this patch, instead
selecting CONFIG_LEGACY_TIMEOUT_API via kconfig. These subsystems
include: CAN Bus, the Microbit display driver, I2S, LoRa modem
drivers, the UART Async API, Video hardware drivers, the console
subsystem, and the network buffer abstraction.
k_sleep() now takes a k_timeout_t argument, with a k_msleep() variant
provided that works identically to the original API.
Most of the changes here are just type/configuration management and
documentation, but there are logic changes in mempool, where a loop
that used a timeout numerically has been reworked using a new
z_timeout_end_calc() predicate. Also in queue.c, a (when POLL was
enabled) a similar loop was needlessly used to try to retry the
k_poll() call after a spurious failure. But k_poll() does not fail
spuriously, so the loop was removed.
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
2020-03-05 15:18:14 -08:00
|
|
|
int z_impl_k_msgq_get(struct k_msgq *msgq, void *data, k_timeout_t timeout)
|
unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
{
|
kernel/timeout: Make timeout arguments an opaque type
Add a k_timeout_t type, and use it everywhere that kernel API
functions were accepting a millisecond timeout argument. Instead of
forcing milliseconds everywhere (which are often not integrally
representable as system ticks), do the conversion to ticks at the
point where the timeout is created. This avoids an extra unit
conversion in some application code, and allows us to express the
timeout in units other than milliseconds to achieve greater precision.
The existing K_MSEC() et. al. macros now return initializers for a
k_timeout_t.
The K_NO_WAIT and K_FOREVER constants have now become k_timeout_t
values, which means they cannot be operated on as integers.
Applications which have their own APIs that need to inspect these
vs. user-provided timeouts can now use a K_TIMEOUT_EQ() predicate to
test for equality.
Timer drivers, which receive an integer tick count in ther
z_clock_set_timeout() functions, now use the integer-valued
K_TICKS_FOREVER constant instead of K_FOREVER.
For the initial release, to preserve source compatibility, a
CONFIG_LEGACY_TIMEOUT_API kconfig is provided. When true, the
k_timeout_t will remain a compatible 32 bit value that will work with
any legacy Zephyr application.
Some subsystems present timeout (or timeout-like) values to their own
users as APIs that would re-use the kernel's own constants and
conventions. These will require some minor design work to adapt to
the new scheme (in most cases just using k_timeout_t directly in their
own API), and they have not been changed in this patch, instead
selecting CONFIG_LEGACY_TIMEOUT_API via kconfig. These subsystems
include: CAN Bus, the Microbit display driver, I2S, LoRa modem
drivers, the UART Async API, Video hardware drivers, the console
subsystem, and the network buffer abstraction.
k_sleep() now takes a k_timeout_t argument, with a k_msleep() variant
provided that works identically to the original API.
Most of the changes here are just type/configuration management and
documentation, but there are logic changes in mempool, where a loop
that used a timeout numerically has been reworked using a new
z_timeout_end_calc() predicate. Also in queue.c, a (when POLL was
enabled) a similar loop was needlessly used to try to retry the
k_poll() call after a spurious failure. But k_poll() does not fail
spuriously, so the loop was removed.
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
2020-03-05 15:18:14 -08:00
|
|
|
__ASSERT(!arch_is_in_isr() || K_TIMEOUT_EQ(timeout, K_NO_WAIT), "");
|
2016-11-09 19:45:19 -05:00
|
|
|
|
2019-06-19 07:30:50 -04:00
|
|
|
k_spinlock_key_t key;
|
2016-10-05 17:32:01 -04:00
|
|
|
struct k_thread *pending_thread;
|
unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
int result;
|
2024-12-16 17:12:30 -08:00
|
|
|
bool resched = false;
|
unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
|
2019-06-19 07:30:50 -04:00
|
|
|
key = k_spin_lock(&msgq->lock);
|
|
|
|
|
|
2021-03-26 12:39:53 +01:00
|
|
|
SYS_PORT_TRACING_OBJ_FUNC_ENTER(k_msgq, get, msgq, timeout);
|
|
|
|
|
|
2021-03-29 10:03:49 -04:00
|
|
|
if (msgq->used_msgs > 0U) {
|
unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
/* take first available message from queue */
|
2024-04-29 14:49:11 +02:00
|
|
|
(void)memcpy((char *)data, msgq->read_ptr, msgq->msg_size);
|
2019-06-19 07:30:50 -04:00
|
|
|
msgq->read_ptr += msgq->msg_size;
|
|
|
|
|
if (msgq->read_ptr == msgq->buffer_end) {
|
|
|
|
|
msgq->read_ptr = msgq->buffer_start;
|
unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
}
|
2019-06-19 07:30:50 -04:00
|
|
|
msgq->used_msgs--;
|
unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
|
kernel: wait_q: add z_unpend_first_thread_locked() and migrate callers
Replace z_unpend_first_thread() with z_unpend_first_thread_locked() and
migrate every caller across the kernel. The old function dropped the
scheduler spinlock before returning, exposing a race window between
its caller's "arch_thread_return_value_set + z_ready_thread" pair and
a still-in-flight timeout handler that could ready the thread first --
the woken thread might then run on another CPU and see an uninitialized
swap_retval. Pre-1b8c7a3 the dticks-cancel check made the handler bail;
here we fix it cleanly by requiring the caller to hold _sched_spinlock
across the entire wake, so the handler is blocked for the duration and
runs as a no-op afterwards.
z_unpend_first_thread_locked() requires the caller to be inside a
locked region and must be paired with z_sched_ready_locked() (and
whatever return-value setup is needed) under the same lock acquisition.
Sites migrated:
Simple "set retval [+ swap_data] and ready" callers use the existing
z_sched_wake() convenience wrapper, refactored to use the new
helper internally:
sem (give, reset), mem_slab (free), stack (push),
condvar (signal, broadcast), msgq (purge),
queue (cancel_wait, queue_insert, append_list),
futex (wake).
Sites that need additional setup on the woken thread use
LOCK_SCHED_SPINLOCK + z_unpend_first_thread_locked() + custom wake:
mutex (unlock -- needs the thread reference to track new owner),
msgq put / get (needs memcpy into the receiver's swap_data
buffer before the return value is set).
The dticks-cancel check in z_thread_timeout() is left in place; it is
no longer load-bearing once z_abort_thread_timeout() has no callers,
and is removed in the next commit.
Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
2026-05-27 18:28:32 -04:00
|
|
|
/* sanity-check write_ptr in case we hand the slot to a sender */
|
|
|
|
|
__ASSERT_NO_MSG((msgq->write_ptr >= msgq->buffer_start) &&
|
|
|
|
|
(msgq->write_ptr <= (msgq->buffer_end - 1)) &&
|
|
|
|
|
((size_t)(uintptr_t)(msgq->buffer_end - msgq->write_ptr) >=
|
|
|
|
|
msgq->msg_size));
|
|
|
|
|
|
2026-08-10 12:21:47 -07:00
|
|
|
/* handle first thread waiting to write (if any). Done
|
|
|
|
|
* atomically under the scheduler's spinlock so we read the
|
kernel: wait_q: add z_unpend_first_thread_locked() and migrate callers
Replace z_unpend_first_thread() with z_unpend_first_thread_locked() and
migrate every caller across the kernel. The old function dropped the
scheduler spinlock before returning, exposing a race window between
its caller's "arch_thread_return_value_set + z_ready_thread" pair and
a still-in-flight timeout handler that could ready the thread first --
the woken thread might then run on another CPU and see an uninitialized
swap_retval. Pre-1b8c7a3 the dticks-cancel check made the handler bail;
here we fix it cleanly by requiring the caller to hold _sched_spinlock
across the entire wake, so the handler is blocked for the duration and
runs as a no-op afterwards.
z_unpend_first_thread_locked() requires the caller to be inside a
locked region and must be paired with z_sched_ready_locked() (and
whatever return-value setup is needed) under the same lock acquisition.
Sites migrated:
Simple "set retval [+ swap_data] and ready" callers use the existing
z_sched_wake() convenience wrapper, refactored to use the new
helper internally:
sem (give, reset), mem_slab (free), stack (push),
condvar (signal, broadcast), msgq (purge),
queue (cancel_wait, queue_insert, append_list),
futex (wake).
Sites that need additional setup on the woken thread use
LOCK_SCHED_SPINLOCK + z_unpend_first_thread_locked() + custom wake:
mutex (unlock -- needs the thread reference to track new owner),
msgq put / get (needs memcpy into the receiver's swap_data
buffer before the return value is set).
The dticks-cancel check in z_thread_timeout() is left in place; it is
no longer load-bearing once z_abort_thread_timeout() has no callers,
and is removed in the next commit.
Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
2026-05-27 18:28:32 -04:00
|
|
|
* sender's swap_data and complete the wake before any
|
|
|
|
|
* racing in-flight timeout handler can wake the sender.
|
|
|
|
|
*/
|
|
|
|
|
LOCK_SCHED_SPINLOCK {
|
|
|
|
|
pending_thread = z_unpend_first_thread_locked(&msgq->wait_q);
|
|
|
|
|
if (pending_thread != NULL) {
|
|
|
|
|
SYS_PORT_TRACING_OBJ_FUNC_BLOCKING(k_msgq, get, msgq, timeout);
|
|
|
|
|
|
|
|
|
|
/* add the sender's pending message to the queue */
|
|
|
|
|
(void)memcpy(msgq->write_ptr,
|
|
|
|
|
(char *)pending_thread->base.swap_data,
|
|
|
|
|
msgq->msg_size);
|
|
|
|
|
msgq->write_ptr += msgq->msg_size;
|
|
|
|
|
if (msgq->write_ptr == msgq->buffer_end) {
|
|
|
|
|
msgq->write_ptr = msgq->buffer_start;
|
|
|
|
|
}
|
|
|
|
|
msgq->used_msgs++;
|
2021-03-26 12:39:53 +01:00
|
|
|
|
kernel: wait_q: add z_unpend_first_thread_locked() and migrate callers
Replace z_unpend_first_thread() with z_unpend_first_thread_locked() and
migrate every caller across the kernel. The old function dropped the
scheduler spinlock before returning, exposing a race window between
its caller's "arch_thread_return_value_set + z_ready_thread" pair and
a still-in-flight timeout handler that could ready the thread first --
the woken thread might then run on another CPU and see an uninitialized
swap_retval. Pre-1b8c7a3 the dticks-cancel check made the handler bail;
here we fix it cleanly by requiring the caller to hold _sched_spinlock
across the entire wake, so the handler is blocked for the duration and
runs as a no-op afterwards.
z_unpend_first_thread_locked() requires the caller to be inside a
locked region and must be paired with z_sched_ready_locked() (and
whatever return-value setup is needed) under the same lock acquisition.
Sites migrated:
Simple "set retval [+ swap_data] and ready" callers use the existing
z_sched_wake() convenience wrapper, refactored to use the new
helper internally:
sem (give, reset), mem_slab (free), stack (push),
condvar (signal, broadcast), msgq (purge),
queue (cancel_wait, queue_insert, append_list),
futex (wake).
Sites that need additional setup on the woken thread use
LOCK_SCHED_SPINLOCK + z_unpend_first_thread_locked() + custom wake:
mutex (unlock -- needs the thread reference to track new owner),
msgq put / get (needs memcpy into the receiver's swap_data
buffer before the return value is set).
The dticks-cancel check in z_thread_timeout() is left in place; it is
no longer load-bearing once z_abort_thread_timeout() has no callers,
and is removed in the next commit.
Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
2026-05-27 18:28:32 -04:00
|
|
|
arch_thread_return_value_set(pending_thread, 0);
|
|
|
|
|
z_sched_ready_locked(pending_thread);
|
|
|
|
|
resched = true;
|
unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
result = 0;
|
kernel/timeout: Make timeout arguments an opaque type
Add a k_timeout_t type, and use it everywhere that kernel API
functions were accepting a millisecond timeout argument. Instead of
forcing milliseconds everywhere (which are often not integrally
representable as system ticks), do the conversion to ticks at the
point where the timeout is created. This avoids an extra unit
conversion in some application code, and allows us to express the
timeout in units other than milliseconds to achieve greater precision.
The existing K_MSEC() et. al. macros now return initializers for a
k_timeout_t.
The K_NO_WAIT and K_FOREVER constants have now become k_timeout_t
values, which means they cannot be operated on as integers.
Applications which have their own APIs that need to inspect these
vs. user-provided timeouts can now use a K_TIMEOUT_EQ() predicate to
test for equality.
Timer drivers, which receive an integer tick count in ther
z_clock_set_timeout() functions, now use the integer-valued
K_TICKS_FOREVER constant instead of K_FOREVER.
For the initial release, to preserve source compatibility, a
CONFIG_LEGACY_TIMEOUT_API kconfig is provided. When true, the
k_timeout_t will remain a compatible 32 bit value that will work with
any legacy Zephyr application.
Some subsystems present timeout (or timeout-like) values to their own
users as APIs that would re-use the kernel's own constants and
conventions. These will require some minor design work to adapt to
the new scheme (in most cases just using k_timeout_t directly in their
own API), and they have not been changed in this patch, instead
selecting CONFIG_LEGACY_TIMEOUT_API via kconfig. These subsystems
include: CAN Bus, the Microbit display driver, I2S, LoRa modem
drivers, the UART Async API, Video hardware drivers, the console
subsystem, and the network buffer abstraction.
k_sleep() now takes a k_timeout_t argument, with a k_msleep() variant
provided that works identically to the original API.
Most of the changes here are just type/configuration management and
documentation, but there are logic changes in mempool, where a loop
that used a timeout numerically has been reworked using a new
z_timeout_end_calc() predicate. Also in queue.c, a (when POLL was
enabled) a similar loop was needlessly used to try to retry the
k_poll() call after a spurious failure. But k_poll() does not fail
spuriously, so the loop was removed.
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
2020-03-05 15:18:14 -08:00
|
|
|
} else if (K_TIMEOUT_EQ(timeout, K_NO_WAIT)) {
|
unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
/* don't wait for a message to become available */
|
|
|
|
|
result = -ENOMSG;
|
|
|
|
|
} else {
|
2021-03-26 12:39:53 +01:00
|
|
|
SYS_PORT_TRACING_OBJ_FUNC_BLOCKING(k_msgq, get, msgq, timeout);
|
|
|
|
|
|
unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
/* wait for get message success or timeout */
|
2025-01-07 12:00:43 -05:00
|
|
|
_current->base.swap_data = data;
|
2021-03-26 12:39:53 +01:00
|
|
|
|
|
|
|
|
result = z_pend_curr(&msgq->lock, key, &msgq->wait_q, timeout);
|
|
|
|
|
SYS_PORT_TRACING_OBJ_FUNC_EXIT(k_msgq, get, msgq, timeout, result);
|
|
|
|
|
return result;
|
unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
}
|
|
|
|
|
|
2021-03-26 12:39:53 +01:00
|
|
|
SYS_PORT_TRACING_OBJ_FUNC_EXIT(k_msgq, get, msgq, timeout, result);
|
|
|
|
|
|
2024-12-16 17:12:30 -08:00
|
|
|
if (resched) {
|
|
|
|
|
z_reschedule(&msgq->lock, key);
|
|
|
|
|
} else {
|
|
|
|
|
k_spin_unlock(&msgq->lock, key);
|
|
|
|
|
}
|
unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
|
|
|
|
|
return result;
|
|
|
|
|
}
|
|
|
|
|
|
2017-10-02 10:53:06 -07:00
|
|
|
#ifdef CONFIG_USERSPACE
|
2021-03-29 10:54:23 -04:00
|
|
|
static inline int z_vrfy_k_msgq_get(struct k_msgq *msgq, void *data,
|
kernel/timeout: Make timeout arguments an opaque type
Add a k_timeout_t type, and use it everywhere that kernel API
functions were accepting a millisecond timeout argument. Instead of
forcing milliseconds everywhere (which are often not integrally
representable as system ticks), do the conversion to ticks at the
point where the timeout is created. This avoids an extra unit
conversion in some application code, and allows us to express the
timeout in units other than milliseconds to achieve greater precision.
The existing K_MSEC() et. al. macros now return initializers for a
k_timeout_t.
The K_NO_WAIT and K_FOREVER constants have now become k_timeout_t
values, which means they cannot be operated on as integers.
Applications which have their own APIs that need to inspect these
vs. user-provided timeouts can now use a K_TIMEOUT_EQ() predicate to
test for equality.
Timer drivers, which receive an integer tick count in ther
z_clock_set_timeout() functions, now use the integer-valued
K_TICKS_FOREVER constant instead of K_FOREVER.
For the initial release, to preserve source compatibility, a
CONFIG_LEGACY_TIMEOUT_API kconfig is provided. When true, the
k_timeout_t will remain a compatible 32 bit value that will work with
any legacy Zephyr application.
Some subsystems present timeout (or timeout-like) values to their own
users as APIs that would re-use the kernel's own constants and
conventions. These will require some minor design work to adapt to
the new scheme (in most cases just using k_timeout_t directly in their
own API), and they have not been changed in this patch, instead
selecting CONFIG_LEGACY_TIMEOUT_API via kconfig. These subsystems
include: CAN Bus, the Microbit display driver, I2S, LoRa modem
drivers, the UART Async API, Video hardware drivers, the console
subsystem, and the network buffer abstraction.
k_sleep() now takes a k_timeout_t argument, with a k_msleep() variant
provided that works identically to the original API.
Most of the changes here are just type/configuration management and
documentation, but there are logic changes in mempool, where a loop
that used a timeout numerically has been reworked using a new
z_timeout_end_calc() predicate. Also in queue.c, a (when POLL was
enabled) a similar loop was needlessly used to try to retry the
k_poll() call after a spurious failure. But k_poll() does not fail
spuriously, so the loop was removed.
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
2020-03-05 15:18:14 -08:00
|
|
|
k_timeout_t timeout)
|
2017-10-02 10:53:06 -07:00
|
|
|
{
|
2023-09-27 11:20:28 +00:00
|
|
|
K_OOPS(K_SYSCALL_OBJ(msgq, K_OBJ_MSGQ));
|
|
|
|
|
K_OOPS(K_SYSCALL_MEMORY_WRITE(data, msgq->msg_size));
|
2017-10-02 10:53:06 -07:00
|
|
|
|
2021-03-29 10:54:23 -04:00
|
|
|
return z_impl_k_msgq_get(msgq, data, timeout);
|
2017-10-02 10:53:06 -07:00
|
|
|
}
|
2024-01-24 17:35:04 +08:00
|
|
|
#include <zephyr/syscalls/k_msgq_get_mrsh.c>
|
2024-03-08 12:00:10 +01:00
|
|
|
#endif /* CONFIG_USERSPACE */
|
2017-10-02 10:53:06 -07:00
|
|
|
|
2019-06-19 07:30:50 -04:00
|
|
|
int z_impl_k_msgq_peek(struct k_msgq *msgq, void *data)
|
2018-11-09 21:04:36 -08:00
|
|
|
{
|
2019-06-19 07:30:50 -04:00
|
|
|
k_spinlock_key_t key;
|
2018-11-09 21:04:36 -08:00
|
|
|
int result;
|
|
|
|
|
|
2019-06-19 07:30:50 -04:00
|
|
|
key = k_spin_lock(&msgq->lock);
|
|
|
|
|
|
2021-03-29 10:03:49 -04:00
|
|
|
if (msgq->used_msgs > 0U) {
|
2018-11-09 21:04:36 -08:00
|
|
|
/* take first available message from queue */
|
2024-04-29 14:49:11 +02:00
|
|
|
(void)memcpy((char *)data, msgq->read_ptr, msgq->msg_size);
|
2018-11-09 21:04:36 -08:00
|
|
|
result = 0;
|
|
|
|
|
} else {
|
|
|
|
|
/* don't wait for a message to become available */
|
|
|
|
|
result = -ENOMSG;
|
|
|
|
|
}
|
|
|
|
|
|
2021-03-26 12:39:53 +01:00
|
|
|
SYS_PORT_TRACING_OBJ_FUNC(k_msgq, peek, msgq, result);
|
|
|
|
|
|
2019-06-19 07:30:50 -04:00
|
|
|
k_spin_unlock(&msgq->lock, key);
|
2018-11-09 21:04:36 -08:00
|
|
|
|
|
|
|
|
return result;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
#ifdef CONFIG_USERSPACE
|
2021-03-29 10:54:23 -04:00
|
|
|
static inline int z_vrfy_k_msgq_peek(struct k_msgq *msgq, void *data)
|
2018-11-09 21:04:36 -08:00
|
|
|
{
|
2023-09-27 11:20:28 +00:00
|
|
|
K_OOPS(K_SYSCALL_OBJ(msgq, K_OBJ_MSGQ));
|
|
|
|
|
K_OOPS(K_SYSCALL_MEMORY_WRITE(data, msgq->msg_size));
|
2018-11-09 21:04:36 -08:00
|
|
|
|
2021-03-29 10:54:23 -04:00
|
|
|
return z_impl_k_msgq_peek(msgq, data);
|
2018-11-09 21:04:36 -08:00
|
|
|
}
|
2024-01-24 17:35:04 +08:00
|
|
|
#include <zephyr/syscalls/k_msgq_peek_mrsh.c>
|
2024-03-08 12:00:10 +01:00
|
|
|
#endif /* CONFIG_USERSPACE */
|
2018-11-09 21:04:36 -08:00
|
|
|
|
2022-12-28 13:06:16 +01:00
|
|
|
int z_impl_k_msgq_peek_at(struct k_msgq *msgq, void *data, uint32_t idx)
|
|
|
|
|
{
|
|
|
|
|
k_spinlock_key_t key;
|
|
|
|
|
int result;
|
|
|
|
|
uint32_t bytes_to_end;
|
|
|
|
|
uint32_t byte_offset;
|
|
|
|
|
char *start_addr;
|
|
|
|
|
|
|
|
|
|
key = k_spin_lock(&msgq->lock);
|
|
|
|
|
|
|
|
|
|
if (msgq->used_msgs > idx) {
|
|
|
|
|
bytes_to_end = (msgq->buffer_end - msgq->read_ptr);
|
|
|
|
|
byte_offset = idx * msgq->msg_size;
|
|
|
|
|
start_addr = msgq->read_ptr;
|
|
|
|
|
/* check item available in start/end of ring buffer */
|
|
|
|
|
if (bytes_to_end <= byte_offset) {
|
|
|
|
|
/* Tweak the values in case */
|
|
|
|
|
byte_offset -= bytes_to_end;
|
|
|
|
|
/* wrap-around is required */
|
|
|
|
|
start_addr = msgq->buffer_start;
|
|
|
|
|
}
|
|
|
|
|
(void)memcpy(data, start_addr + byte_offset, msgq->msg_size);
|
|
|
|
|
result = 0;
|
|
|
|
|
} else {
|
|
|
|
|
/* don't wait for a message to become available */
|
|
|
|
|
result = -ENOMSG;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
SYS_PORT_TRACING_OBJ_FUNC(k_msgq, peek, msgq, result);
|
|
|
|
|
|
|
|
|
|
k_spin_unlock(&msgq->lock, key);
|
|
|
|
|
|
|
|
|
|
return result;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
#ifdef CONFIG_USERSPACE
|
|
|
|
|
static inline int z_vrfy_k_msgq_peek_at(struct k_msgq *msgq, void *data, uint32_t idx)
|
|
|
|
|
{
|
2023-09-27 11:20:28 +00:00
|
|
|
K_OOPS(K_SYSCALL_OBJ(msgq, K_OBJ_MSGQ));
|
|
|
|
|
K_OOPS(K_SYSCALL_MEMORY_WRITE(data, msgq->msg_size));
|
2022-12-28 13:06:16 +01:00
|
|
|
|
|
|
|
|
return z_impl_k_msgq_peek_at(msgq, data, idx);
|
|
|
|
|
}
|
2024-01-24 17:35:04 +08:00
|
|
|
#include <zephyr/syscalls/k_msgq_peek_at_mrsh.c>
|
2024-03-08 12:00:10 +01:00
|
|
|
#endif /* CONFIG_USERSPACE */
|
2022-12-28 13:06:16 +01:00
|
|
|
|
2019-06-19 07:30:50 -04:00
|
|
|
void z_impl_k_msgq_purge(struct k_msgq *msgq)
|
unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
{
|
2019-06-19 07:30:50 -04:00
|
|
|
k_spinlock_key_t key;
|
2024-12-16 17:12:30 -08:00
|
|
|
bool resched = false;
|
unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
|
2019-06-19 07:30:50 -04:00
|
|
|
key = k_spin_lock(&msgq->lock);
|
|
|
|
|
|
2021-03-26 12:39:53 +01:00
|
|
|
SYS_PORT_TRACING_OBJ_FUNC(k_msgq, purge, msgq);
|
|
|
|
|
|
2016-09-22 13:59:00 -04:00
|
|
|
/* wake up any threads that are waiting to write */
|
kernel: wait_q: add z_unpend_first_thread_locked() and migrate callers
Replace z_unpend_first_thread() with z_unpend_first_thread_locked() and
migrate every caller across the kernel. The old function dropped the
scheduler spinlock before returning, exposing a race window between
its caller's "arch_thread_return_value_set + z_ready_thread" pair and
a still-in-flight timeout handler that could ready the thread first --
the woken thread might then run on another CPU and see an uninitialized
swap_retval. Pre-1b8c7a3 the dticks-cancel check made the handler bail;
here we fix it cleanly by requiring the caller to hold _sched_spinlock
across the entire wake, so the handler is blocked for the duration and
runs as a no-op afterwards.
z_unpend_first_thread_locked() requires the caller to be inside a
locked region and must be paired with z_sched_ready_locked() (and
whatever return-value setup is needed) under the same lock acquisition.
Sites migrated:
Simple "set retval [+ swap_data] and ready" callers use the existing
z_sched_wake() convenience wrapper, refactored to use the new
helper internally:
sem (give, reset), mem_slab (free), stack (push),
condvar (signal, broadcast), msgq (purge),
queue (cancel_wait, queue_insert, append_list),
futex (wake).
Sites that need additional setup on the woken thread use
LOCK_SCHED_SPINLOCK + z_unpend_first_thread_locked() + custom wake:
mutex (unlock -- needs the thread reference to track new owner),
msgq put / get (needs memcpy into the receiver's swap_data
buffer before the return value is set).
The dticks-cancel check in z_thread_timeout() is left in place; it is
no longer load-bearing once z_abort_thread_timeout() has no callers,
and is removed in the next commit.
Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
2026-05-27 18:28:32 -04:00
|
|
|
while (z_sched_wake(&msgq->wait_q, -ENOMSG, NULL)) {
|
2024-12-16 17:12:30 -08:00
|
|
|
resched = true;
|
unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
}
|
|
|
|
|
|
2019-06-19 07:30:50 -04:00
|
|
|
msgq->used_msgs = 0;
|
|
|
|
|
msgq->read_ptr = msgq->write_ptr;
|
2016-09-22 13:59:00 -04:00
|
|
|
|
2024-12-16 17:12:30 -08:00
|
|
|
if (resched) {
|
|
|
|
|
z_reschedule(&msgq->lock, key);
|
|
|
|
|
} else {
|
|
|
|
|
k_spin_unlock(&msgq->lock, key);
|
|
|
|
|
}
|
unified: initial unified kernel implementation
Summary of what this includes:
initialization:
Copy from nano_init.c, with the following changes:
- the main thread is the continuation of the init thread, but an idle
thread is created as well
- _main() initializes threads in groups and starts the EXE group
- the ready queues are initialized
- the main thread is marked as non-essential once the system init is
done
- a weak main() symbol is provided if the application does not provide a
main() function
scheduler:
Not an exhaustive list, but basically provide primitives for:
- adding/removing a thread to/from a wait queue
- adding/removing a thread to/from the ready queue
- marking thread as ready
- locking/unlocking the scheduler
- instead of locking interrupts
- getting/setting thread priority
- checking what state (coop/preempt) a thread is currenlty running in
- rescheduling threads
- finding what thread is the next to run
- yielding/sleeping/aborting sleep
- finding the current thread
threads:
- Add operationns on threads, such as creating and starting them.
standardized handling of kernel object return codes:
- Kernel objects now cause _Swap() to return the following values:
0 => operation successful
-EAGAIN => operation timed out
-Exxxxx => operation failed for another reason
- The thread's swap_data field can be used to return any additional
information required to complete the operation, such as the actual
result of a successful operation.
timeouts:
- same as nano timeouts, renamed to simply 'timeouts'
- the kernel is still tick-based, but objects take timeout values in
ms for forward compatibility with a tickless kernel.
semaphores:
- Port of the nanokernel semaphores, which have the same basic behaviour
as the microkernel ones. Semaphore groups are not yet implemented.
- These semaphores are enhanced in that they accept an initial count and a
count limit. This allows configuring them as binary semaphores, and also
provisioning them without having to "give" the semaphore multiple times
before using them.
mutexes:
- Straight port of the microkernel mutexes. An init function is added to
allow defining them at runtime.
pipes:
- straight port
timers:
- amalgamation of nano and micro timers, with all functionalities
intact.
events:
- re-implementation, using semaphores and workqueues.
mailboxes:
- straight port
message queues:
- straight port of microkernel FIFOs
memory maps:
- straight port
workqueues:
- Basically, have all APIs follow the k_ naming rule, and use the _timeout
subsystem from the unified kernel directory, and not the _nano_timeout
one.
stacks:
- Port of the nanokernel stacks. They can now have multiple threads
pending on them and threads can wait with a timeout.
LIFOs:
- Straight port of the nanokernel LIFOs.
FIFOs:
- Straight port of the nanokernel FIFOs.
Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com>
Peter Mitsis <peter.mitsis@windriver.com>
Allan Stephens <allan.stephens@windriver.com>
Benjamin Walsh <benjamin.walsh@windriver.com>
Change-Id: Id3cadb3694484ab2ca467889cfb029be3cd3a7d6
Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 18:55:39 -04:00
|
|
|
}
|
2017-10-02 10:53:06 -07:00
|
|
|
|
|
|
|
|
#ifdef CONFIG_USERSPACE
|
2021-03-29 10:54:23 -04:00
|
|
|
static inline void z_vrfy_k_msgq_purge(struct k_msgq *msgq)
|
userspace: Support for split 64 bit arguments
System call arguments, at the arch layer, are single words. So
passing wider values requires splitting them into two registers at
call time. This gets even more complicated for values (e.g
k_timeout_t) that may have different sizes depending on configuration.
This patch adds a feature to gen_syscalls.py to detect functions with
wide arguments and automatically generates code to split/unsplit them.
Unfortunately the current scheme of Z_SYSCALL_DECLARE_* macros won't
work with functions like this, because for N arguments (our current
maximum N is 10) there are 2^N possible configurations of argument
widths. So this generates the complete functions for each handler and
wrapper, effectively doing in python what was originally done in the
preprocessor.
Another complexity is that traditional the z_hdlr_*() function for a
system call has taken the raw list of word arguments, which does not
work when some of those arguments must be 64 bit types. So instead of
using a single Z_SYSCALL_HANDLER macro, this splits the job of
z_hdlr_*() into two steps: An automatically-generated unmarshalling
function, z_mrsh_*(), which then calls a user-supplied verification
function z_vrfy_*(). The verification function is typesafe, and is a
simple C function with exactly the same argument and return signature
as the syscall impl function. It is also not responsible for
validating the pointers to the extra parameter array or a wide return
value, that code gets automatically generated.
This commit includes new vrfy/msrh handling for all syscalls invoked
during CI runs. Future commits will port the less testable code.
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
2019-08-06 13:34:31 -07:00
|
|
|
{
|
2023-09-27 11:20:28 +00:00
|
|
|
K_OOPS(K_SYSCALL_OBJ(msgq, K_OBJ_MSGQ));
|
2021-03-29 10:54:23 -04:00
|
|
|
z_impl_k_msgq_purge(msgq);
|
userspace: Support for split 64 bit arguments
System call arguments, at the arch layer, are single words. So
passing wider values requires splitting them into two registers at
call time. This gets even more complicated for values (e.g
k_timeout_t) that may have different sizes depending on configuration.
This patch adds a feature to gen_syscalls.py to detect functions with
wide arguments and automatically generates code to split/unsplit them.
Unfortunately the current scheme of Z_SYSCALL_DECLARE_* macros won't
work with functions like this, because for N arguments (our current
maximum N is 10) there are 2^N possible configurations of argument
widths. So this generates the complete functions for each handler and
wrapper, effectively doing in python what was originally done in the
preprocessor.
Another complexity is that traditional the z_hdlr_*() function for a
system call has taken the raw list of word arguments, which does not
work when some of those arguments must be 64 bit types. So instead of
using a single Z_SYSCALL_HANDLER macro, this splits the job of
z_hdlr_*() into two steps: An automatically-generated unmarshalling
function, z_mrsh_*(), which then calls a user-supplied verification
function z_vrfy_*(). The verification function is typesafe, and is a
simple C function with exactly the same argument and return signature
as the syscall impl function. It is also not responsible for
validating the pointers to the extra parameter array or a wide return
value, that code gets automatically generated.
This commit includes new vrfy/msrh handling for all syscalls invoked
during CI runs. Future commits will port the less testable code.
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
2019-08-06 13:34:31 -07:00
|
|
|
}
|
2024-01-24 17:35:04 +08:00
|
|
|
#include <zephyr/syscalls/k_msgq_purge_mrsh.c>
|
userspace: Support for split 64 bit arguments
System call arguments, at the arch layer, are single words. So
passing wider values requires splitting them into two registers at
call time. This gets even more complicated for values (e.g
k_timeout_t) that may have different sizes depending on configuration.
This patch adds a feature to gen_syscalls.py to detect functions with
wide arguments and automatically generates code to split/unsplit them.
Unfortunately the current scheme of Z_SYSCALL_DECLARE_* macros won't
work with functions like this, because for N arguments (our current
maximum N is 10) there are 2^N possible configurations of argument
widths. So this generates the complete functions for each handler and
wrapper, effectively doing in python what was originally done in the
preprocessor.
Another complexity is that traditional the z_hdlr_*() function for a
system call has taken the raw list of word arguments, which does not
work when some of those arguments must be 64 bit types. So instead of
using a single Z_SYSCALL_HANDLER macro, this splits the job of
z_hdlr_*() into two steps: An automatically-generated unmarshalling
function, z_mrsh_*(), which then calls a user-supplied verification
function z_vrfy_*(). The verification function is typesafe, and is a
simple C function with exactly the same argument and return signature
as the syscall impl function. It is also not responsible for
validating the pointers to the extra parameter array or a wide return
value, that code gets automatically generated.
This commit includes new vrfy/msrh handling for all syscalls invoked
during CI runs. Future commits will port the less testable code.
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
2019-08-06 13:34:31 -07:00
|
|
|
|
2021-03-29 10:54:23 -04:00
|
|
|
static inline uint32_t z_vrfy_k_msgq_num_free_get(struct k_msgq *msgq)
|
userspace: Support for split 64 bit arguments
System call arguments, at the arch layer, are single words. So
passing wider values requires splitting them into two registers at
call time. This gets even more complicated for values (e.g
k_timeout_t) that may have different sizes depending on configuration.
This patch adds a feature to gen_syscalls.py to detect functions with
wide arguments and automatically generates code to split/unsplit them.
Unfortunately the current scheme of Z_SYSCALL_DECLARE_* macros won't
work with functions like this, because for N arguments (our current
maximum N is 10) there are 2^N possible configurations of argument
widths. So this generates the complete functions for each handler and
wrapper, effectively doing in python what was originally done in the
preprocessor.
Another complexity is that traditional the z_hdlr_*() function for a
system call has taken the raw list of word arguments, which does not
work when some of those arguments must be 64 bit types. So instead of
using a single Z_SYSCALL_HANDLER macro, this splits the job of
z_hdlr_*() into two steps: An automatically-generated unmarshalling
function, z_mrsh_*(), which then calls a user-supplied verification
function z_vrfy_*(). The verification function is typesafe, and is a
simple C function with exactly the same argument and return signature
as the syscall impl function. It is also not responsible for
validating the pointers to the extra parameter array or a wide return
value, that code gets automatically generated.
This commit includes new vrfy/msrh handling for all syscalls invoked
during CI runs. Future commits will port the less testable code.
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
2019-08-06 13:34:31 -07:00
|
|
|
{
|
2023-09-27 11:20:28 +00:00
|
|
|
K_OOPS(K_SYSCALL_OBJ(msgq, K_OBJ_MSGQ));
|
2021-03-29 10:54:23 -04:00
|
|
|
return z_impl_k_msgq_num_free_get(msgq);
|
userspace: Support for split 64 bit arguments
System call arguments, at the arch layer, are single words. So
passing wider values requires splitting them into two registers at
call time. This gets even more complicated for values (e.g
k_timeout_t) that may have different sizes depending on configuration.
This patch adds a feature to gen_syscalls.py to detect functions with
wide arguments and automatically generates code to split/unsplit them.
Unfortunately the current scheme of Z_SYSCALL_DECLARE_* macros won't
work with functions like this, because for N arguments (our current
maximum N is 10) there are 2^N possible configurations of argument
widths. So this generates the complete functions for each handler and
wrapper, effectively doing in python what was originally done in the
preprocessor.
Another complexity is that traditional the z_hdlr_*() function for a
system call has taken the raw list of word arguments, which does not
work when some of those arguments must be 64 bit types. So instead of
using a single Z_SYSCALL_HANDLER macro, this splits the job of
z_hdlr_*() into two steps: An automatically-generated unmarshalling
function, z_mrsh_*(), which then calls a user-supplied verification
function z_vrfy_*(). The verification function is typesafe, and is a
simple C function with exactly the same argument and return signature
as the syscall impl function. It is also not responsible for
validating the pointers to the extra parameter array or a wide return
value, that code gets automatically generated.
This commit includes new vrfy/msrh handling for all syscalls invoked
during CI runs. Future commits will port the less testable code.
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
2019-08-06 13:34:31 -07:00
|
|
|
}
|
2024-01-24 17:35:04 +08:00
|
|
|
#include <zephyr/syscalls/k_msgq_num_free_get_mrsh.c>
|
userspace: Support for split 64 bit arguments
System call arguments, at the arch layer, are single words. So
passing wider values requires splitting them into two registers at
call time. This gets even more complicated for values (e.g
k_timeout_t) that may have different sizes depending on configuration.
This patch adds a feature to gen_syscalls.py to detect functions with
wide arguments and automatically generates code to split/unsplit them.
Unfortunately the current scheme of Z_SYSCALL_DECLARE_* macros won't
work with functions like this, because for N arguments (our current
maximum N is 10) there are 2^N possible configurations of argument
widths. So this generates the complete functions for each handler and
wrapper, effectively doing in python what was originally done in the
preprocessor.
Another complexity is that traditional the z_hdlr_*() function for a
system call has taken the raw list of word arguments, which does not
work when some of those arguments must be 64 bit types. So instead of
using a single Z_SYSCALL_HANDLER macro, this splits the job of
z_hdlr_*() into two steps: An automatically-generated unmarshalling
function, z_mrsh_*(), which then calls a user-supplied verification
function z_vrfy_*(). The verification function is typesafe, and is a
simple C function with exactly the same argument and return signature
as the syscall impl function. It is also not responsible for
validating the pointers to the extra parameter array or a wide return
value, that code gets automatically generated.
This commit includes new vrfy/msrh handling for all syscalls invoked
during CI runs. Future commits will port the less testable code.
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
2019-08-06 13:34:31 -07:00
|
|
|
|
2021-03-29 10:54:23 -04:00
|
|
|
static inline uint32_t z_vrfy_k_msgq_num_used_get(struct k_msgq *msgq)
|
userspace: Support for split 64 bit arguments
System call arguments, at the arch layer, are single words. So
passing wider values requires splitting them into two registers at
call time. This gets even more complicated for values (e.g
k_timeout_t) that may have different sizes depending on configuration.
This patch adds a feature to gen_syscalls.py to detect functions with
wide arguments and automatically generates code to split/unsplit them.
Unfortunately the current scheme of Z_SYSCALL_DECLARE_* macros won't
work with functions like this, because for N arguments (our current
maximum N is 10) there are 2^N possible configurations of argument
widths. So this generates the complete functions for each handler and
wrapper, effectively doing in python what was originally done in the
preprocessor.
Another complexity is that traditional the z_hdlr_*() function for a
system call has taken the raw list of word arguments, which does not
work when some of those arguments must be 64 bit types. So instead of
using a single Z_SYSCALL_HANDLER macro, this splits the job of
z_hdlr_*() into two steps: An automatically-generated unmarshalling
function, z_mrsh_*(), which then calls a user-supplied verification
function z_vrfy_*(). The verification function is typesafe, and is a
simple C function with exactly the same argument and return signature
as the syscall impl function. It is also not responsible for
validating the pointers to the extra parameter array or a wide return
value, that code gets automatically generated.
This commit includes new vrfy/msrh handling for all syscalls invoked
during CI runs. Future commits will port the less testable code.
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
2019-08-06 13:34:31 -07:00
|
|
|
{
|
2023-09-27 11:20:28 +00:00
|
|
|
K_OOPS(K_SYSCALL_OBJ(msgq, K_OBJ_MSGQ));
|
2021-03-29 10:54:23 -04:00
|
|
|
return z_impl_k_msgq_num_used_get(msgq);
|
userspace: Support for split 64 bit arguments
System call arguments, at the arch layer, are single words. So
passing wider values requires splitting them into two registers at
call time. This gets even more complicated for values (e.g
k_timeout_t) that may have different sizes depending on configuration.
This patch adds a feature to gen_syscalls.py to detect functions with
wide arguments and automatically generates code to split/unsplit them.
Unfortunately the current scheme of Z_SYSCALL_DECLARE_* macros won't
work with functions like this, because for N arguments (our current
maximum N is 10) there are 2^N possible configurations of argument
widths. So this generates the complete functions for each handler and
wrapper, effectively doing in python what was originally done in the
preprocessor.
Another complexity is that traditional the z_hdlr_*() function for a
system call has taken the raw list of word arguments, which does not
work when some of those arguments must be 64 bit types. So instead of
using a single Z_SYSCALL_HANDLER macro, this splits the job of
z_hdlr_*() into two steps: An automatically-generated unmarshalling
function, z_mrsh_*(), which then calls a user-supplied verification
function z_vrfy_*(). The verification function is typesafe, and is a
simple C function with exactly the same argument and return signature
as the syscall impl function. It is also not responsible for
validating the pointers to the extra parameter array or a wide return
value, that code gets automatically generated.
This commit includes new vrfy/msrh handling for all syscalls invoked
during CI runs. Future commits will port the less testable code.
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
2019-08-06 13:34:31 -07:00
|
|
|
}
|
2024-01-24 17:35:04 +08:00
|
|
|
#include <zephyr/syscalls/k_msgq_num_used_get_mrsh.c>
|
userspace: Support for split 64 bit arguments
System call arguments, at the arch layer, are single words. So
passing wider values requires splitting them into two registers at
call time. This gets even more complicated for values (e.g
k_timeout_t) that may have different sizes depending on configuration.
This patch adds a feature to gen_syscalls.py to detect functions with
wide arguments and automatically generates code to split/unsplit them.
Unfortunately the current scheme of Z_SYSCALL_DECLARE_* macros won't
work with functions like this, because for N arguments (our current
maximum N is 10) there are 2^N possible configurations of argument
widths. So this generates the complete functions for each handler and
wrapper, effectively doing in python what was originally done in the
preprocessor.
Another complexity is that traditional the z_hdlr_*() function for a
system call has taken the raw list of word arguments, which does not
work when some of those arguments must be 64 bit types. So instead of
using a single Z_SYSCALL_HANDLER macro, this splits the job of
z_hdlr_*() into two steps: An automatically-generated unmarshalling
function, z_mrsh_*(), which then calls a user-supplied verification
function z_vrfy_*(). The verification function is typesafe, and is a
simple C function with exactly the same argument and return signature
as the syscall impl function. It is also not responsible for
validating the pointers to the extra parameter array or a wide return
value, that code gets automatically generated.
This commit includes new vrfy/msrh handling for all syscalls invoked
during CI runs. Future commits will port the less testable code.
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
2019-08-06 13:34:31 -07:00
|
|
|
|
2024-03-08 12:00:10 +01:00
|
|
|
#endif /* CONFIG_USERSPACE */
|
kernel: Integrate object cores into kernel
Integrates object cores into the following kernel structures
sys_mem_blocks, k_mem_slab
_cpu, z_kernel
k_thread, k_timer
k_condvar, k_event, k_mutex, k_sem
k_mbox, k_msgq, k_pipe, k_fifo, k_lifo, k_stack
Signed-off-by: Peter Mitsis <peter.mitsis@intel.com>
2023-05-11 14:06:46 -04:00
|
|
|
|
|
|
|
|
#ifdef CONFIG_OBJ_CORE_MSGQ
|
kernel: obj_core: register object types from a single init point
Each kernel object type that participates in the object core framework
previously supplied its own SYS_INIT routine to initialize its
k_obj_type and to walk its static-object linker section, linking each
object core. These routines were near-identical across 11 object types
and differed only in the type id, the object struct and the obj_core
offset.
Replace that per-type boilerplate with a declarative K_OBJ_TYPE_DEFINE()
macro that emits a const descriptor into a new iterable ROM section, and
walk those descriptors once from a single SYS_INIT in obj_core.c. The
descriptor captures the type storage, type id, obj_core offset, the
static object section bounds and the object stride, which is all the
shared loop needs to initialize and link every statically defined
object.
Converted: condvar, event, fifo, lifo, mailbox, msgq, mutex, pipe, sem,
stack and timer. The non-uniform initializers (thread, mem_slab and the
internal cpu/kernel objects) are left unchanged and will be dealt with
in followup commits.
Footprint on qemu_cortex_m3 (tests/kernel/obj_core/obj_core, all types
enabled): flash 32012 -> 30880 (-1132 B), RAM unchanged. With
CONFIG_OBJ_CORE disabled the image is byte-for-byte identical.
Assisted-by: Claude:claude-opus-4-8
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
2026-06-29 20:21:47 -04:00
|
|
|
K_OBJ_TYPE_DEFINE(obj_type_msgq, k_msgq, K_OBJ_TYPE_MSGQ_ID, NULL);
|
2024-03-08 12:00:10 +01:00
|
|
|
#endif /* CONFIG_OBJ_CORE_MSGQ */
|