zephyr/arch/x86/core/thread.c

305 lines
9.7 KiB
C
Raw Normal View History

/*
* Copyright (c) 2010-2015 Wind River Systems, Inc.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
/**
* @file
* @brief Nanokernel thread support primitives
*
* This module provides core nanokernel fiber related primitives for the IA-32
* processor architecture.
*/
#ifdef CONFIG_INIT_STACKS
#include <string.h>
#endif /* CONFIG_INIT_STACKS */
#include <toolchain.h>
#include <sections.h>
kernel/arch: consolidate tTCS and TNANO definitions There was a lot of duplication between architectures for the definition of threads and the "nanokernel" guts. These have been consolidated. Now, a common file kernel/unified/include/kernel_structs.h holds the common definitions. Architectures provide two files to complement it: kernel_arch_data.h and kernel_arch_func.h. The first one contains at least the struct _thread_arch and struct _kernel_arch data structures, as well as the struct _callee_saved and struct _caller_saved register layouts. The second file contains anything that needs what is provided by the common stuff in kernel_structs.h. Those two files are only meant to be included in kernel_structs.h in very specific locations. The thread data structure has been separated into three major parts: common struct _thread_base and struct k_thread, and arch-specific struct _thread_arch. The first and third ones are included in the second. The struct s_NANO data structure has been split into two: common struct _kernel and arch-specific struct _kernel_arch. The latter is included in the former. Offsets files have also changed: nano_offsets.h has been renamed kernel_offsets.h and is still included by the arch-specific offsets.c. Also, since the thread and kernel data structures are now made of sub-structures, offsets have to be added to make up the full offset. Some of these additions have been consolidated in shorter symbols, available from kernel/unified/include/offsets_short.h, which includes an arch-specific offsets_arch_short.h. Most of the code include offsets_short.h now instead of offsets.h. Change-Id: I084645cb7e6db8db69aeaaf162963fe157045d5a Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-11-08 10:36:50 -05:00
#include <kernel_structs.h>
#include <wait_q.h>
/* forward declaration */
#if defined(CONFIG_GDB_INFO) || defined(CONFIG_DEBUG_INFO) \
|| defined(CONFIG_X86_IAMCU)
void _thread_entry_wrapper(_thread_entry_t, void *,
void *, void *);
#endif
#if defined(CONFIG_THREAD_MONITOR)
/*
* Add a thread to the kernel's list of active threads.
*/
kernel/arch: consolidate tTCS and TNANO definitions There was a lot of duplication between architectures for the definition of threads and the "nanokernel" guts. These have been consolidated. Now, a common file kernel/unified/include/kernel_structs.h holds the common definitions. Architectures provide two files to complement it: kernel_arch_data.h and kernel_arch_func.h. The first one contains at least the struct _thread_arch and struct _kernel_arch data structures, as well as the struct _callee_saved and struct _caller_saved register layouts. The second file contains anything that needs what is provided by the common stuff in kernel_structs.h. Those two files are only meant to be included in kernel_structs.h in very specific locations. The thread data structure has been separated into three major parts: common struct _thread_base and struct k_thread, and arch-specific struct _thread_arch. The first and third ones are included in the second. The struct s_NANO data structure has been split into two: common struct _kernel and arch-specific struct _kernel_arch. The latter is included in the former. Offsets files have also changed: nano_offsets.h has been renamed kernel_offsets.h and is still included by the arch-specific offsets.c. Also, since the thread and kernel data structures are now made of sub-structures, offsets have to be added to make up the full offset. Some of these additions have been consolidated in shorter symbols, available from kernel/unified/include/offsets_short.h, which includes an arch-specific offsets_arch_short.h. Most of the code include offsets_short.h now instead of offsets.h. Change-Id: I084645cb7e6db8db69aeaaf162963fe157045d5a Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-11-08 10:36:50 -05:00
static ALWAYS_INLINE void thread_monitor_init(struct k_thread *thread)
{
unsigned int key;
key = irq_lock();
kernel/arch: consolidate tTCS and TNANO definitions There was a lot of duplication between architectures for the definition of threads and the "nanokernel" guts. These have been consolidated. Now, a common file kernel/unified/include/kernel_structs.h holds the common definitions. Architectures provide two files to complement it: kernel_arch_data.h and kernel_arch_func.h. The first one contains at least the struct _thread_arch and struct _kernel_arch data structures, as well as the struct _callee_saved and struct _caller_saved register layouts. The second file contains anything that needs what is provided by the common stuff in kernel_structs.h. Those two files are only meant to be included in kernel_structs.h in very specific locations. The thread data structure has been separated into three major parts: common struct _thread_base and struct k_thread, and arch-specific struct _thread_arch. The first and third ones are included in the second. The struct s_NANO data structure has been split into two: common struct _kernel and arch-specific struct _kernel_arch. The latter is included in the former. Offsets files have also changed: nano_offsets.h has been renamed kernel_offsets.h and is still included by the arch-specific offsets.c. Also, since the thread and kernel data structures are now made of sub-structures, offsets have to be added to make up the full offset. Some of these additions have been consolidated in shorter symbols, available from kernel/unified/include/offsets_short.h, which includes an arch-specific offsets_arch_short.h. Most of the code include offsets_short.h now instead of offsets.h. Change-Id: I084645cb7e6db8db69aeaaf162963fe157045d5a Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-11-08 10:36:50 -05:00
thread->next_thread = _kernel.threads;
_kernel.threads = thread;
irq_unlock(key);
}
#else
kernel/arch: consolidate tTCS and TNANO definitions There was a lot of duplication between architectures for the definition of threads and the "nanokernel" guts. These have been consolidated. Now, a common file kernel/unified/include/kernel_structs.h holds the common definitions. Architectures provide two files to complement it: kernel_arch_data.h and kernel_arch_func.h. The first one contains at least the struct _thread_arch and struct _kernel_arch data structures, as well as the struct _callee_saved and struct _caller_saved register layouts. The second file contains anything that needs what is provided by the common stuff in kernel_structs.h. Those two files are only meant to be included in kernel_structs.h in very specific locations. The thread data structure has been separated into three major parts: common struct _thread_base and struct k_thread, and arch-specific struct _thread_arch. The first and third ones are included in the second. The struct s_NANO data structure has been split into two: common struct _kernel and arch-specific struct _kernel_arch. The latter is included in the former. Offsets files have also changed: nano_offsets.h has been renamed kernel_offsets.h and is still included by the arch-specific offsets.c. Also, since the thread and kernel data structures are now made of sub-structures, offsets have to be added to make up the full offset. Some of these additions have been consolidated in shorter symbols, available from kernel/unified/include/offsets_short.h, which includes an arch-specific offsets_arch_short.h. Most of the code include offsets_short.h now instead of offsets.h. Change-Id: I084645cb7e6db8db69aeaaf162963fe157045d5a Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-11-08 10:36:50 -05:00
#define thread_monitor_init(thread) \
do {/* do nothing */ \
} while ((0))
#endif /* CONFIG_THREAD_MONITOR */
/**
*
* @brief Initialize a new execution thread
*
* This function is utilized to initialize all execution threads (both fiber
* and task). The 'priority' parameter will be set to -1 for the creation of
* task.
*
* This function is called by _new_thread() to initialize tasks.
*
* @param pStackMem pointer to thread stack memory
* @param stackSize size of a stack in bytes
unified/x86: add unified kernel support for x86 arch The x86 architecture port is fitted with support for the unified kernel, namely: - the interrupt exit code now calls _Swap() if the current thread is not a coop thread and if the scheduler is not locked - there is no 'task' fields in the _nanokernel anymore: _Swap() now calls _get_next_ready_thread instead - the _nanokernel.fiber field is replaced by a more sophisticated ready_q, based on the microkernel's priority-bitmap-based one - nano_private includes nano_internal.h from the unified directory - the FIBER, TASK and PREEMPTIBLE flags do not exist anymore: the thread priority drives the behaviour - the tcs uses a dlist for queuing in both ready and wait queues instead of a custom singly-linked list - other new fields in the tcs include a schedule-lock count, a back-pointer to init data (when the task is static) and a pointer to swap data, needed when a thread pending on _Swap() must be passed more then just one value (e.g. k_stack_pop() needs an error code and data) - fiberRtnValueSet() is aliased to _set_thread_return_value since it also operates on preempt threads now - _set_thread_return_value_with_data() sets the swap_data field in addition to a return value from _Swap() - convenience aliases are created for shorter names: - _current is defined as _nanokernel.current - _ready_q is defined as _nanokernel.ready_q - _Swap() sets the threads's return code to -EAGAIN before swapping out to prevent timeouts to have to set it (solves hard issues in some kernel objects). - Floating point support. Note that, in _Swap(), the register holding the thread to be swapped in has been changed from %ecx to %eax in both the legacy kernel and the unified kernel to take advantage of the fact that the return value of _get_next_ready_thread() is stored in %eax, and this avoids moving it to %ecx. Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com> Allan Stephens <allan.stephens@windriver.com> Benjamin Walsh <benjamin.walsh@windriver.com> Change-Id: I4ce2bd47bcdc62034c669b5e889fc0f29480c43b Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 16:34:35 -04:00
* @param priority thread priority
* @param options thread options: K_ESSENTIAL, K_FP_REGS, K_SSE_REGS
*
* @return N/A
*/
static void _new_thread_internal(char *pStackMem, unsigned stackSize,
int priority,
unified/x86: add unified kernel support for x86 arch The x86 architecture port is fitted with support for the unified kernel, namely: - the interrupt exit code now calls _Swap() if the current thread is not a coop thread and if the scheduler is not locked - there is no 'task' fields in the _nanokernel anymore: _Swap() now calls _get_next_ready_thread instead - the _nanokernel.fiber field is replaced by a more sophisticated ready_q, based on the microkernel's priority-bitmap-based one - nano_private includes nano_internal.h from the unified directory - the FIBER, TASK and PREEMPTIBLE flags do not exist anymore: the thread priority drives the behaviour - the tcs uses a dlist for queuing in both ready and wait queues instead of a custom singly-linked list - other new fields in the tcs include a schedule-lock count, a back-pointer to init data (when the task is static) and a pointer to swap data, needed when a thread pending on _Swap() must be passed more then just one value (e.g. k_stack_pop() needs an error code and data) - fiberRtnValueSet() is aliased to _set_thread_return_value since it also operates on preempt threads now - _set_thread_return_value_with_data() sets the swap_data field in addition to a return value from _Swap() - convenience aliases are created for shorter names: - _current is defined as _nanokernel.current - _ready_q is defined as _nanokernel.ready_q - _Swap() sets the threads's return code to -EAGAIN before swapping out to prevent timeouts to have to set it (solves hard issues in some kernel objects). - Floating point support. Note that, in _Swap(), the register holding the thread to be swapped in has been changed from %ecx to %eax in both the legacy kernel and the unified kernel to take advantage of the fact that the return value of _get_next_ready_thread() is stored in %eax, and this avoids moving it to %ecx. Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com> Allan Stephens <allan.stephens@windriver.com> Benjamin Walsh <benjamin.walsh@windriver.com> Change-Id: I4ce2bd47bcdc62034c669b5e889fc0f29480c43b Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 16:34:35 -04:00
unsigned options)
{
unsigned long *pInitialCtx;
kernel/arch: consolidate tTCS and TNANO definitions There was a lot of duplication between architectures for the definition of threads and the "nanokernel" guts. These have been consolidated. Now, a common file kernel/unified/include/kernel_structs.h holds the common definitions. Architectures provide two files to complement it: kernel_arch_data.h and kernel_arch_func.h. The first one contains at least the struct _thread_arch and struct _kernel_arch data structures, as well as the struct _callee_saved and struct _caller_saved register layouts. The second file contains anything that needs what is provided by the common stuff in kernel_structs.h. Those two files are only meant to be included in kernel_structs.h in very specific locations. The thread data structure has been separated into three major parts: common struct _thread_base and struct k_thread, and arch-specific struct _thread_arch. The first and third ones are included in the second. The struct s_NANO data structure has been split into two: common struct _kernel and arch-specific struct _kernel_arch. The latter is included in the former. Offsets files have also changed: nano_offsets.h has been renamed kernel_offsets.h and is still included by the arch-specific offsets.c. Also, since the thread and kernel data structures are now made of sub-structures, offsets have to be added to make up the full offset. Some of these additions have been consolidated in shorter symbols, available from kernel/unified/include/offsets_short.h, which includes an arch-specific offsets_arch_short.h. Most of the code include offsets_short.h now instead of offsets.h. Change-Id: I084645cb7e6db8db69aeaaf162963fe157045d5a Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-11-08 10:36:50 -05:00
/* ptr to the new task's k_thread */
struct k_thread *thread = (struct k_thread *)pStackMem;
#if (defined(CONFIG_FP_SHARING) || defined(CONFIG_GDB_INFO))
kernel/arch: consolidate tTCS and TNANO definitions There was a lot of duplication between architectures for the definition of threads and the "nanokernel" guts. These have been consolidated. Now, a common file kernel/unified/include/kernel_structs.h holds the common definitions. Architectures provide two files to complement it: kernel_arch_data.h and kernel_arch_func.h. The first one contains at least the struct _thread_arch and struct _kernel_arch data structures, as well as the struct _callee_saved and struct _caller_saved register layouts. The second file contains anything that needs what is provided by the common stuff in kernel_structs.h. Those two files are only meant to be included in kernel_structs.h in very specific locations. The thread data structure has been separated into three major parts: common struct _thread_base and struct k_thread, and arch-specific struct _thread_arch. The first and third ones are included in the second. The struct s_NANO data structure has been split into two: common struct _kernel and arch-specific struct _kernel_arch. The latter is included in the former. Offsets files have also changed: nano_offsets.h has been renamed kernel_offsets.h and is still included by the arch-specific offsets.c. Also, since the thread and kernel data structures are now made of sub-structures, offsets have to be added to make up the full offset. Some of these additions have been consolidated in shorter symbols, available from kernel/unified/include/offsets_short.h, which includes an arch-specific offsets_arch_short.h. Most of the code include offsets_short.h now instead of offsets.h. Change-Id: I084645cb7e6db8db69aeaaf162963fe157045d5a Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-11-08 10:36:50 -05:00
thread->arch.excNestCount = 0;
#endif /* CONFIG_FP_SHARING || CONFIG_GDB_INFO */
_init_thread_base(&thread->base, priority, K_PRESTART, options);
unified/x86: add unified kernel support for x86 arch The x86 architecture port is fitted with support for the unified kernel, namely: - the interrupt exit code now calls _Swap() if the current thread is not a coop thread and if the scheduler is not locked - there is no 'task' fields in the _nanokernel anymore: _Swap() now calls _get_next_ready_thread instead - the _nanokernel.fiber field is replaced by a more sophisticated ready_q, based on the microkernel's priority-bitmap-based one - nano_private includes nano_internal.h from the unified directory - the FIBER, TASK and PREEMPTIBLE flags do not exist anymore: the thread priority drives the behaviour - the tcs uses a dlist for queuing in both ready and wait queues instead of a custom singly-linked list - other new fields in the tcs include a schedule-lock count, a back-pointer to init data (when the task is static) and a pointer to swap data, needed when a thread pending on _Swap() must be passed more then just one value (e.g. k_stack_pop() needs an error code and data) - fiberRtnValueSet() is aliased to _set_thread_return_value since it also operates on preempt threads now - _set_thread_return_value_with_data() sets the swap_data field in addition to a return value from _Swap() - convenience aliases are created for shorter names: - _current is defined as _nanokernel.current - _ready_q is defined as _nanokernel.ready_q - _Swap() sets the threads's return code to -EAGAIN before swapping out to prevent timeouts to have to set it (solves hard issues in some kernel objects). - Floating point support. Note that, in _Swap(), the register holding the thread to be swapped in has been changed from %ecx to %eax in both the legacy kernel and the unified kernel to take advantage of the fact that the return value of _get_next_ready_thread() is stored in %eax, and this avoids moving it to %ecx. Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com> Allan Stephens <allan.stephens@windriver.com> Benjamin Walsh <benjamin.walsh@windriver.com> Change-Id: I4ce2bd47bcdc62034c669b5e889fc0f29480c43b Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 16:34:35 -04:00
/* static threads overwrite it afterwards with real value */
kernel/arch: consolidate tTCS and TNANO definitions There was a lot of duplication between architectures for the definition of threads and the "nanokernel" guts. These have been consolidated. Now, a common file kernel/unified/include/kernel_structs.h holds the common definitions. Architectures provide two files to complement it: kernel_arch_data.h and kernel_arch_func.h. The first one contains at least the struct _thread_arch and struct _kernel_arch data structures, as well as the struct _callee_saved and struct _caller_saved register layouts. The second file contains anything that needs what is provided by the common stuff in kernel_structs.h. Those two files are only meant to be included in kernel_structs.h in very specific locations. The thread data structure has been separated into three major parts: common struct _thread_base and struct k_thread, and arch-specific struct _thread_arch. The first and third ones are included in the second. The struct s_NANO data structure has been split into two: common struct _kernel and arch-specific struct _kernel_arch. The latter is included in the former. Offsets files have also changed: nano_offsets.h has been renamed kernel_offsets.h and is still included by the arch-specific offsets.c. Also, since the thread and kernel data structures are now made of sub-structures, offsets have to be added to make up the full offset. Some of these additions have been consolidated in shorter symbols, available from kernel/unified/include/offsets_short.h, which includes an arch-specific offsets_arch_short.h. Most of the code include offsets_short.h now instead of offsets.h. Change-Id: I084645cb7e6db8db69aeaaf162963fe157045d5a Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-11-08 10:36:50 -05:00
thread->init_data = NULL;
thread->fn_abort = NULL;
#ifdef CONFIG_THREAD_CUSTOM_DATA
/* Initialize custom data field (value is opaque to kernel) */
kernel/arch: consolidate tTCS and TNANO definitions There was a lot of duplication between architectures for the definition of threads and the "nanokernel" guts. These have been consolidated. Now, a common file kernel/unified/include/kernel_structs.h holds the common definitions. Architectures provide two files to complement it: kernel_arch_data.h and kernel_arch_func.h. The first one contains at least the struct _thread_arch and struct _kernel_arch data structures, as well as the struct _callee_saved and struct _caller_saved register layouts. The second file contains anything that needs what is provided by the common stuff in kernel_structs.h. Those two files are only meant to be included in kernel_structs.h in very specific locations. The thread data structure has been separated into three major parts: common struct _thread_base and struct k_thread, and arch-specific struct _thread_arch. The first and third ones are included in the second. The struct s_NANO data structure has been split into two: common struct _kernel and arch-specific struct _kernel_arch. The latter is included in the former. Offsets files have also changed: nano_offsets.h has been renamed kernel_offsets.h and is still included by the arch-specific offsets.c. Also, since the thread and kernel data structures are now made of sub-structures, offsets have to be added to make up the full offset. Some of these additions have been consolidated in shorter symbols, available from kernel/unified/include/offsets_short.h, which includes an arch-specific offsets_arch_short.h. Most of the code include offsets_short.h now instead of offsets.h. Change-Id: I084645cb7e6db8db69aeaaf162963fe157045d5a Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-11-08 10:36:50 -05:00
thread->custom_data = NULL;
#endif
/*
* The creation of the initial stack for the task has already been done.
* Now all that is needed is to set the ESP. However, we have been passed
* the base address of the stack which is past the initial stack frame.
* Therefore some of the calculations done in the other routines that
* initialize the stack frame need to be repeated.
*/
pInitialCtx = (unsigned long *)STACK_ROUND_DOWN(pStackMem + stackSize);
#ifdef CONFIG_THREAD_MONITOR
/*
kernel/arch: consolidate tTCS and TNANO definitions There was a lot of duplication between architectures for the definition of threads and the "nanokernel" guts. These have been consolidated. Now, a common file kernel/unified/include/kernel_structs.h holds the common definitions. Architectures provide two files to complement it: kernel_arch_data.h and kernel_arch_func.h. The first one contains at least the struct _thread_arch and struct _kernel_arch data structures, as well as the struct _callee_saved and struct _caller_saved register layouts. The second file contains anything that needs what is provided by the common stuff in kernel_structs.h. Those two files are only meant to be included in kernel_structs.h in very specific locations. The thread data structure has been separated into three major parts: common struct _thread_base and struct k_thread, and arch-specific struct _thread_arch. The first and third ones are included in the second. The struct s_NANO data structure has been split into two: common struct _kernel and arch-specific struct _kernel_arch. The latter is included in the former. Offsets files have also changed: nano_offsets.h has been renamed kernel_offsets.h and is still included by the arch-specific offsets.c. Also, since the thread and kernel data structures are now made of sub-structures, offsets have to be added to make up the full offset. Some of these additions have been consolidated in shorter symbols, available from kernel/unified/include/offsets_short.h, which includes an arch-specific offsets_arch_short.h. Most of the code include offsets_short.h now instead of offsets.h. Change-Id: I084645cb7e6db8db69aeaaf162963fe157045d5a Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-11-08 10:36:50 -05:00
* In debug mode thread->entry give direct access to the thread entry
* and the corresponding parameters.
*/
kernel/arch: consolidate tTCS and TNANO definitions There was a lot of duplication between architectures for the definition of threads and the "nanokernel" guts. These have been consolidated. Now, a common file kernel/unified/include/kernel_structs.h holds the common definitions. Architectures provide two files to complement it: kernel_arch_data.h and kernel_arch_func.h. The first one contains at least the struct _thread_arch and struct _kernel_arch data structures, as well as the struct _callee_saved and struct _caller_saved register layouts. The second file contains anything that needs what is provided by the common stuff in kernel_structs.h. Those two files are only meant to be included in kernel_structs.h in very specific locations. The thread data structure has been separated into three major parts: common struct _thread_base and struct k_thread, and arch-specific struct _thread_arch. The first and third ones are included in the second. The struct s_NANO data structure has been split into two: common struct _kernel and arch-specific struct _kernel_arch. The latter is included in the former. Offsets files have also changed: nano_offsets.h has been renamed kernel_offsets.h and is still included by the arch-specific offsets.c. Also, since the thread and kernel data structures are now made of sub-structures, offsets have to be added to make up the full offset. Some of these additions have been consolidated in shorter symbols, available from kernel/unified/include/offsets_short.h, which includes an arch-specific offsets_arch_short.h. Most of the code include offsets_short.h now instead of offsets.h. Change-Id: I084645cb7e6db8db69aeaaf162963fe157045d5a Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-11-08 10:36:50 -05:00
thread->entry = (struct __thread_entry *)(pInitialCtx -
sizeof(struct __thread_entry));
#endif
/* The stack needs to be set up so that when we do an initial switch
* to it in the middle of _Swap(), it needs to be set up as follows:
* - 4 thread entry routine parameters
* - eflags
* - eip (so that _Swap() "returns" to the entry point)
* - edi, esi, ebx, ebp, eax
*/
pInitialCtx -= 11;
kernel/arch: consolidate tTCS and TNANO definitions There was a lot of duplication between architectures for the definition of threads and the "nanokernel" guts. These have been consolidated. Now, a common file kernel/unified/include/kernel_structs.h holds the common definitions. Architectures provide two files to complement it: kernel_arch_data.h and kernel_arch_func.h. The first one contains at least the struct _thread_arch and struct _kernel_arch data structures, as well as the struct _callee_saved and struct _caller_saved register layouts. The second file contains anything that needs what is provided by the common stuff in kernel_structs.h. Those two files are only meant to be included in kernel_structs.h in very specific locations. The thread data structure has been separated into three major parts: common struct _thread_base and struct k_thread, and arch-specific struct _thread_arch. The first and third ones are included in the second. The struct s_NANO data structure has been split into two: common struct _kernel and arch-specific struct _kernel_arch. The latter is included in the former. Offsets files have also changed: nano_offsets.h has been renamed kernel_offsets.h and is still included by the arch-specific offsets.c. Also, since the thread and kernel data structures are now made of sub-structures, offsets have to be added to make up the full offset. Some of these additions have been consolidated in shorter symbols, available from kernel/unified/include/offsets_short.h, which includes an arch-specific offsets_arch_short.h. Most of the code include offsets_short.h now instead of offsets.h. Change-Id: I084645cb7e6db8db69aeaaf162963fe157045d5a Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-11-08 10:36:50 -05:00
thread->callee_saved.esp = (unsigned long)pInitialCtx;
PRINTK("\nInitial context ESP = 0x%x\n", thread->coopReg.esp);
kernel/arch: consolidate tTCS and TNANO definitions There was a lot of duplication between architectures for the definition of threads and the "nanokernel" guts. These have been consolidated. Now, a common file kernel/unified/include/kernel_structs.h holds the common definitions. Architectures provide two files to complement it: kernel_arch_data.h and kernel_arch_func.h. The first one contains at least the struct _thread_arch and struct _kernel_arch data structures, as well as the struct _callee_saved and struct _caller_saved register layouts. The second file contains anything that needs what is provided by the common stuff in kernel_structs.h. Those two files are only meant to be included in kernel_structs.h in very specific locations. The thread data structure has been separated into three major parts: common struct _thread_base and struct k_thread, and arch-specific struct _thread_arch. The first and third ones are included in the second. The struct s_NANO data structure has been split into two: common struct _kernel and arch-specific struct _kernel_arch. The latter is included in the former. Offsets files have also changed: nano_offsets.h has been renamed kernel_offsets.h and is still included by the arch-specific offsets.c. Also, since the thread and kernel data structures are now made of sub-structures, offsets have to be added to make up the full offset. Some of these additions have been consolidated in shorter symbols, available from kernel/unified/include/offsets_short.h, which includes an arch-specific offsets_arch_short.h. Most of the code include offsets_short.h now instead of offsets.h. Change-Id: I084645cb7e6db8db69aeaaf162963fe157045d5a Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-11-08 10:36:50 -05:00
PRINTK("\nstruct thread * = 0x%x", thread);
kernel/arch: consolidate tTCS and TNANO definitions There was a lot of duplication between architectures for the definition of threads and the "nanokernel" guts. These have been consolidated. Now, a common file kernel/unified/include/kernel_structs.h holds the common definitions. Architectures provide two files to complement it: kernel_arch_data.h and kernel_arch_func.h. The first one contains at least the struct _thread_arch and struct _kernel_arch data structures, as well as the struct _callee_saved and struct _caller_saved register layouts. The second file contains anything that needs what is provided by the common stuff in kernel_structs.h. Those two files are only meant to be included in kernel_structs.h in very specific locations. The thread data structure has been separated into three major parts: common struct _thread_base and struct k_thread, and arch-specific struct _thread_arch. The first and third ones are included in the second. The struct s_NANO data structure has been split into two: common struct _kernel and arch-specific struct _kernel_arch. The latter is included in the former. Offsets files have also changed: nano_offsets.h has been renamed kernel_offsets.h and is still included by the arch-specific offsets.c. Also, since the thread and kernel data structures are now made of sub-structures, offsets have to be added to make up the full offset. Some of these additions have been consolidated in shorter symbols, available from kernel/unified/include/offsets_short.h, which includes an arch-specific offsets_arch_short.h. Most of the code include offsets_short.h now instead of offsets.h. Change-Id: I084645cb7e6db8db69aeaaf162963fe157045d5a Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-11-08 10:36:50 -05:00
thread_monitor_init(thread);
}
#if defined(CONFIG_GDB_INFO) || defined(CONFIG_DEBUG_INFO) \
|| defined(CONFIG_X86_IAMCU)
/**
*
* @brief Adjust stack/parameters before invoking _thread_entry
*
* This function adjusts the initial stack frame created by _new_thread() such
* that the GDB stack frame unwinders recognize it as the outermost frame in
* the thread's stack. For targets that use the IAMCU calling convention, the
* first three arguments are popped into eax, edx, and ecx. The function then
* jumps to _thread_entry().
*
* GDB normally stops unwinding a stack when it detects that it has
* reached a function called main(). Kernel tasks, however, do not have
* a main() function, and there does not appear to be a simple way of stopping
* the unwinding of the stack.
*
* SYS V Systems:
*
* Given the initial thread created by _new_thread(), GDB expects to find a
* return address on the stack immediately above the thread entry routine
* _thread_entry, in the location occupied by the initial EFLAGS.
* GDB attempts to examine the memory at this return address, which typically
* results in an invalid access to page 0 of memory.
*
* This function overwrites the initial EFLAGS with zero. When GDB subsequently
* attempts to examine memory at address zero, the PeekPoke driver detects
* an invalid access to address zero and returns an error, which causes the
* GDB stack unwinder to stop somewhat gracefully.
*
* The initial EFLAGS cannot be overwritten until after _Swap() has swapped in
* the new thread for the first time. This routine is called by _Swap() the
* first time that the new thread is swapped in, and it jumps to
* _thread_entry after it has done its work.
*
* IAMCU Systems:
*
* There is no EFLAGS on the stack when we get here. _thread_entry() takes
* four arguments, and we need to pop off the first three into the
* appropriate registers. Instead of using the 'call' instruction, we push
* a NULL return address onto the stack and jump into _thread_entry,
* ensuring the stack won't be unwound further. Placing some kind of return
* address on the stack is mandatory so this isn't conditionally compiled.
*
* __________________
* | param3 | <------ Top of the stack
* |__________________|
* | param2 | Stack Grows Down
* |__________________| |
* | param1 | V
* |__________________|
* | pEntry | <---- ESP when invoked by _Swap() on IAMCU
* |__________________|
* | initial EFLAGS | <---- ESP when invoked by _Swap() on Sys V
* |__________________| (Zeroed by this routine on Sys V)
*
*
*
* @return this routine does NOT return.
*/
__asm__("\t.globl _thread_entry\n"
"\t.section .text\n"
"_thread_entry_wrapper:\n" /* should place this func .S file and use
* SECTION_FUNC
*/
#ifdef CONFIG_X86_IAMCU
/* IAMCU calling convention has first 3 arguments supplied in
* registers not the stack
*/
"\tpopl %eax\n"
"\tpopl %edx\n"
"\tpopl %ecx\n"
"\tpushl $0\n" /* Null return address */
#elif defined(CONFIG_GDB_INFO) || defined(CONFIG_DEBUG_INFO)
"\tmovl $0, (%esp)\n" /* zero initialEFLAGS location */
#endif
"\tjmp _thread_entry\n");
#endif /* CONFIG_GDB_INFO || CONFIG_DEBUG_INFO) || CONFIG_X86_IAMCU */
/**
*
* @brief Create a new kernel execution thread
*
* This function is utilized to create execution threads for both fiber
* threads and kernel tasks.
*
kernel/arch: consolidate tTCS and TNANO definitions There was a lot of duplication between architectures for the definition of threads and the "nanokernel" guts. These have been consolidated. Now, a common file kernel/unified/include/kernel_structs.h holds the common definitions. Architectures provide two files to complement it: kernel_arch_data.h and kernel_arch_func.h. The first one contains at least the struct _thread_arch and struct _kernel_arch data structures, as well as the struct _callee_saved and struct _caller_saved register layouts. The second file contains anything that needs what is provided by the common stuff in kernel_structs.h. Those two files are only meant to be included in kernel_structs.h in very specific locations. The thread data structure has been separated into three major parts: common struct _thread_base and struct k_thread, and arch-specific struct _thread_arch. The first and third ones are included in the second. The struct s_NANO data structure has been split into two: common struct _kernel and arch-specific struct _kernel_arch. The latter is included in the former. Offsets files have also changed: nano_offsets.h has been renamed kernel_offsets.h and is still included by the arch-specific offsets.c. Also, since the thread and kernel data structures are now made of sub-structures, offsets have to be added to make up the full offset. Some of these additions have been consolidated in shorter symbols, available from kernel/unified/include/offsets_short.h, which includes an arch-specific offsets_arch_short.h. Most of the code include offsets_short.h now instead of offsets.h. Change-Id: I084645cb7e6db8db69aeaaf162963fe157045d5a Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-11-08 10:36:50 -05:00
* The k_thread structure is carved from the "end" of the specified
* thread stack memory.
*
* @param pStackmem the pointer to aligned stack memory
* @param stackSize the stack size in bytes
* @param pEntry thread entry point routine
* @param parameter1 first param to entry point
* @param parameter2 second param to entry point
* @param parameter3 third param to entry point
unified/x86: add unified kernel support for x86 arch The x86 architecture port is fitted with support for the unified kernel, namely: - the interrupt exit code now calls _Swap() if the current thread is not a coop thread and if the scheduler is not locked - there is no 'task' fields in the _nanokernel anymore: _Swap() now calls _get_next_ready_thread instead - the _nanokernel.fiber field is replaced by a more sophisticated ready_q, based on the microkernel's priority-bitmap-based one - nano_private includes nano_internal.h from the unified directory - the FIBER, TASK and PREEMPTIBLE flags do not exist anymore: the thread priority drives the behaviour - the tcs uses a dlist for queuing in both ready and wait queues instead of a custom singly-linked list - other new fields in the tcs include a schedule-lock count, a back-pointer to init data (when the task is static) and a pointer to swap data, needed when a thread pending on _Swap() must be passed more then just one value (e.g. k_stack_pop() needs an error code and data) - fiberRtnValueSet() is aliased to _set_thread_return_value since it also operates on preempt threads now - _set_thread_return_value_with_data() sets the swap_data field in addition to a return value from _Swap() - convenience aliases are created for shorter names: - _current is defined as _nanokernel.current - _ready_q is defined as _nanokernel.ready_q - _Swap() sets the threads's return code to -EAGAIN before swapping out to prevent timeouts to have to set it (solves hard issues in some kernel objects). - Floating point support. Note that, in _Swap(), the register holding the thread to be swapped in has been changed from %ecx to %eax in both the legacy kernel and the unified kernel to take advantage of the fact that the return value of _get_next_ready_thread() is stored in %eax, and this avoids moving it to %ecx. Work by: Dmitriy Korovkin <dmitriy.korovkin@windriver.com> Allan Stephens <allan.stephens@windriver.com> Benjamin Walsh <benjamin.walsh@windriver.com> Change-Id: I4ce2bd47bcdc62034c669b5e889fc0f29480c43b Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-09-02 16:34:35 -04:00
* @param priority thread priority
* @param options thread options: K_ESSENTIAL, K_FP_REGS, K_SSE_REGS
*
*
kernel/arch: consolidate tTCS and TNANO definitions There was a lot of duplication between architectures for the definition of threads and the "nanokernel" guts. These have been consolidated. Now, a common file kernel/unified/include/kernel_structs.h holds the common definitions. Architectures provide two files to complement it: kernel_arch_data.h and kernel_arch_func.h. The first one contains at least the struct _thread_arch and struct _kernel_arch data structures, as well as the struct _callee_saved and struct _caller_saved register layouts. The second file contains anything that needs what is provided by the common stuff in kernel_structs.h. Those two files are only meant to be included in kernel_structs.h in very specific locations. The thread data structure has been separated into three major parts: common struct _thread_base and struct k_thread, and arch-specific struct _thread_arch. The first and third ones are included in the second. The struct s_NANO data structure has been split into two: common struct _kernel and arch-specific struct _kernel_arch. The latter is included in the former. Offsets files have also changed: nano_offsets.h has been renamed kernel_offsets.h and is still included by the arch-specific offsets.c. Also, since the thread and kernel data structures are now made of sub-structures, offsets have to be added to make up the full offset. Some of these additions have been consolidated in shorter symbols, available from kernel/unified/include/offsets_short.h, which includes an arch-specific offsets_arch_short.h. Most of the code include offsets_short.h now instead of offsets.h. Change-Id: I084645cb7e6db8db69aeaaf162963fe157045d5a Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-11-08 10:36:50 -05:00
* @return opaque pointer to initialized k_thread structure
*/
void _new_thread(char *pStackMem, size_t stackSize,
_thread_entry_t pEntry,
void *parameter1, void *parameter2, void *parameter3,
int priority, unsigned options)
{
_ASSERT_VALID_PRIO(priority, pEntry);
unsigned long *pInitialThread;
#ifdef CONFIG_INIT_STACKS
memset(pStackMem, 0xaa, stackSize);
#endif
/* carve the thread entry struct from the "base" of the stack */
pInitialThread =
(unsigned long *)STACK_ROUND_DOWN(pStackMem + stackSize);
/*
* Create an initial context on the stack expected by the _Swap()
* primitive.
* Given that both task and fibers execute at privilege 0, the
* setup for both threads are equivalent.
*/
/* push arguments required by _thread_entry() */
*--pInitialThread = (unsigned long)parameter3;
*--pInitialThread = (unsigned long)parameter2;
*--pInitialThread = (unsigned long)parameter1;
*--pInitialThread = (unsigned long)pEntry;
/* push initial EFLAGS; only modify IF and IOPL bits */
*--pInitialThread = (EflagsGet() & ~EFLAGS_MASK) | EFLAGS_INITIAL;
#if defined(CONFIG_GDB_INFO) || defined(CONFIG_DEBUG_INFO) \
|| defined(CONFIG_X86_IAMCU)
/*
* Arrange for the _thread_entry_wrapper() function to be called
* to adjust the stack before _thread_entry() is invoked.
*/
*--pInitialThread = (unsigned long)_thread_entry_wrapper;
#else /* defined(CONFIG_GDB_INFO) || defined(CONFIG_DEBUG_INFO) */
*--pInitialThread = (unsigned long)_thread_entry;
#endif /* defined(CONFIG_GDB_INFO) || defined(CONFIG_DEBUG_INFO) */
/*
* note: stack area for edi, esi, ebx, ebp, and eax registers can be
* left
* uninitialized, since _thread_entry() doesn't care about the values
* of these registers when it begins execution
*/
/*
kernel/arch: consolidate tTCS and TNANO definitions There was a lot of duplication between architectures for the definition of threads and the "nanokernel" guts. These have been consolidated. Now, a common file kernel/unified/include/kernel_structs.h holds the common definitions. Architectures provide two files to complement it: kernel_arch_data.h and kernel_arch_func.h. The first one contains at least the struct _thread_arch and struct _kernel_arch data structures, as well as the struct _callee_saved and struct _caller_saved register layouts. The second file contains anything that needs what is provided by the common stuff in kernel_structs.h. Those two files are only meant to be included in kernel_structs.h in very specific locations. The thread data structure has been separated into three major parts: common struct _thread_base and struct k_thread, and arch-specific struct _thread_arch. The first and third ones are included in the second. The struct s_NANO data structure has been split into two: common struct _kernel and arch-specific struct _kernel_arch. The latter is included in the former. Offsets files have also changed: nano_offsets.h has been renamed kernel_offsets.h and is still included by the arch-specific offsets.c. Also, since the thread and kernel data structures are now made of sub-structures, offsets have to be added to make up the full offset. Some of these additions have been consolidated in shorter symbols, available from kernel/unified/include/offsets_short.h, which includes an arch-specific offsets_arch_short.h. Most of the code include offsets_short.h now instead of offsets.h. Change-Id: I084645cb7e6db8db69aeaaf162963fe157045d5a Signed-off-by: Benjamin Walsh <benjamin.walsh@windriver.com>
2016-11-08 10:36:50 -05:00
* The k_thread structure is located at the "low end" of memory set
* aside for the thread's stack.
*/
_new_thread_internal(pStackMem, stackSize, priority, options);
}