blob: 8b7d62b867118eed6804dddedccb3b529e96b720 [file] [log] [blame]
/*
* Copyright (c) 2010-2014 Wind River Systems, Inc.
*
* SPDX-License-Identifier: Apache-2.0
*/
/**
* @file
* @brief Kernel initialization module
*
* This module contains routines that are used to initialize the kernel.
*/
#include <offsets_short.h>
#include <zephyr/kernel.h>
#include <zephyr/sys/printk.h>
#include <zephyr/debug/stack.h>
#include <zephyr/random/random.h>
#include <zephyr/linker/sections.h>
#include <zephyr/toolchain.h>
#include <zephyr/kernel_structs.h>
#include <zephyr/device.h>
#include <zephyr/init.h>
#include <zephyr/linker/linker-defs.h>
#include <ksched.h>
#include <kthread.h>
#include <string.h>
#include <zephyr/sys/dlist.h>
#include <kernel_internal.h>
#include <zephyr/drivers/entropy.h>
#include <zephyr/logging/log_ctrl.h>
#include <zephyr/tracing/tracing.h>
#include <stdbool.h>
#include <zephyr/debug/gcov.h>
#include <kswap.h>
#include <zephyr/timing/timing.h>
#include <zephyr/logging/log.h>
#include <zephyr/pm/device_runtime.h>
#include <zephyr/internal/syscall_handler.h>
LOG_MODULE_REGISTER(os, CONFIG_KERNEL_LOG_LEVEL);
BUILD_ASSERT(CONFIG_MP_NUM_CPUS == CONFIG_MP_MAX_NUM_CPUS,
"CONFIG_MP_NUM_CPUS and CONFIG_MP_MAX_NUM_CPUS need to be set the same");
/* the only struct z_kernel instance */
__pinned_bss
struct z_kernel _kernel;
#ifdef CONFIG_PM
__pinned_bss atomic_t _cpus_active;
#endif
/* init/main and idle threads */
K_THREAD_PINNED_STACK_DEFINE(z_main_stack, CONFIG_MAIN_STACK_SIZE);
struct k_thread z_main_thread;
#ifdef CONFIG_MULTITHREADING
__pinned_bss
struct k_thread z_idle_threads[CONFIG_MP_MAX_NUM_CPUS];
static K_KERNEL_PINNED_STACK_ARRAY_DEFINE(z_idle_stacks,
CONFIG_MP_MAX_NUM_CPUS,
CONFIG_IDLE_STACK_SIZE);
static void z_init_static_threads(void)
{
STRUCT_SECTION_FOREACH(_static_thread_data, thread_data) {
z_setup_new_thread(
thread_data->init_thread,
thread_data->init_stack,
thread_data->init_stack_size,
thread_data->init_entry,
thread_data->init_p1,
thread_data->init_p2,
thread_data->init_p3,
thread_data->init_prio,
thread_data->init_options,
thread_data->init_name);
thread_data->init_thread->init_data = thread_data;
}
#ifdef CONFIG_USERSPACE
STRUCT_SECTION_FOREACH(k_object_assignment, pos) {
for (int i = 0; pos->objects[i] != NULL; i++) {
k_object_access_grant(pos->objects[i],
pos->thread);
}
}
#endif /* CONFIG_USERSPACE */
/*
* Non-legacy static threads may be started immediately or
* after a previously specified delay. Even though the
* scheduler is locked, ticks can still be delivered and
* processed. Take a sched lock to prevent them from running
* until they are all started.
*
* Note that static threads defined using the legacy API have a
* delay of K_FOREVER.
*/
k_sched_lock();
STRUCT_SECTION_FOREACH(_static_thread_data, thread_data) {
k_timeout_t init_delay = Z_THREAD_INIT_DELAY(thread_data);
if (!K_TIMEOUT_EQ(init_delay, K_FOREVER)) {
thread_schedule_new(thread_data->init_thread,
init_delay);
}
}
k_sched_unlock();
}
#else
#define z_init_static_threads() do { } while (false)
#endif /* CONFIG_MULTITHREADING */
extern const struct init_entry __init_start[];
extern const struct init_entry __init_EARLY_start[];
extern const struct init_entry __init_PRE_KERNEL_1_start[];
extern const struct init_entry __init_PRE_KERNEL_2_start[];
extern const struct init_entry __init_POST_KERNEL_start[];
extern const struct init_entry __init_APPLICATION_start[];
extern const struct init_entry __init_end[];
enum init_level {
INIT_LEVEL_EARLY = 0,
INIT_LEVEL_PRE_KERNEL_1,
INIT_LEVEL_PRE_KERNEL_2,
INIT_LEVEL_POST_KERNEL,
INIT_LEVEL_APPLICATION,
#ifdef CONFIG_SMP
INIT_LEVEL_SMP,
#endif /* CONFIG_SMP */
};
#ifdef CONFIG_SMP
extern const struct init_entry __init_SMP_start[];
#endif /* CONFIG_SMP */
/*
* storage space for the interrupt stack
*
* Note: This area is used as the system stack during kernel initialization,
* since the kernel hasn't yet set up its own stack areas. The dual purposing
* of this area is safe since interrupts are disabled until the kernel context
* switches to the init thread.
*/
K_KERNEL_PINNED_STACK_ARRAY_DEFINE(z_interrupt_stacks,
CONFIG_MP_MAX_NUM_CPUS,
CONFIG_ISR_STACK_SIZE);
extern void idle(void *unused1, void *unused2, void *unused3);
#ifdef CONFIG_OBJ_CORE_SYSTEM
static struct k_obj_type obj_type_cpu;
static struct k_obj_type obj_type_kernel;
#ifdef CONFIG_OBJ_CORE_STATS_SYSTEM
static struct k_obj_core_stats_desc cpu_stats_desc = {
.raw_size = sizeof(struct k_cycle_stats),
.query_size = sizeof(struct k_thread_runtime_stats),
.raw = z_cpu_stats_raw,
.query = z_cpu_stats_query,
.reset = NULL,
.disable = NULL,
.enable = NULL,
};
static struct k_obj_core_stats_desc kernel_stats_desc = {
.raw_size = sizeof(struct k_cycle_stats) * CONFIG_MP_MAX_NUM_CPUS,
.query_size = sizeof(struct k_thread_runtime_stats),
.raw = z_kernel_stats_raw,
.query = z_kernel_stats_query,
.reset = NULL,
.disable = NULL,
.enable = NULL,
};
#endif /* CONFIG_OBJ_CORE_STATS_SYSTEM */
#endif /* CONFIG_OBJ_CORE_SYSTEM */
/* LCOV_EXCL_START
*
* This code is called so early in the boot process that code coverage
* doesn't work properly. In addition, not all arches call this code,
* some like x86 do this with optimized assembly
*/
/**
* @brief equivalent of memset() for early boot usage
*
* Architectures that can't safely use the regular (optimized) memset very
* early during boot because e.g. hardware isn't yet sufficiently initialized
* may override this with their own safe implementation.
*/
__boot_func
void __weak z_early_memset(void *dst, int c, size_t n)
{
(void) memset(dst, c, n);
}
/**
* @brief equivalent of memcpy() for early boot usage
*
* Architectures that can't safely use the regular (optimized) memcpy very
* early during boot because e.g. hardware isn't yet sufficiently initialized
* may override this with their own safe implementation.
*/
__boot_func
void __weak z_early_memcpy(void *dst, const void *src, size_t n)
{
(void) memcpy(dst, src, n);
}
/**
* @brief Clear BSS
*
* This routine clears the BSS region, so all bytes are 0.
*/
__boot_func
void z_bss_zero(void)
{
if (IS_ENABLED(CONFIG_SKIP_BSS_CLEAR)) {
return;
}
z_early_memset(__bss_start, 0, __bss_end - __bss_start);
#if DT_NODE_HAS_STATUS(DT_CHOSEN(zephyr_ccm), okay)
z_early_memset(&__ccm_bss_start, 0,
(uintptr_t) &__ccm_bss_end
- (uintptr_t) &__ccm_bss_start);
#endif
#if DT_NODE_HAS_STATUS(DT_CHOSEN(zephyr_dtcm), okay)
z_early_memset(&__dtcm_bss_start, 0,
(uintptr_t) &__dtcm_bss_end
- (uintptr_t) &__dtcm_bss_start);
#endif
#if DT_NODE_HAS_STATUS(DT_CHOSEN(zephyr_ocm), okay)
z_early_memset(&__ocm_bss_start, 0,
(uintptr_t) &__ocm_bss_end
- (uintptr_t) &__ocm_bss_start);
#endif
#ifdef CONFIG_CODE_DATA_RELOCATION
extern void bss_zeroing_relocation(void);
bss_zeroing_relocation();
#endif /* CONFIG_CODE_DATA_RELOCATION */
#ifdef CONFIG_COVERAGE_GCOV
z_early_memset(&__gcov_bss_start, 0,
((uintptr_t) &__gcov_bss_end - (uintptr_t) &__gcov_bss_start));
#endif /* CONFIG_COVERAGE_GCOV */
}
#ifdef CONFIG_LINKER_USE_BOOT_SECTION
/**
* @brief Clear BSS within the bot region
*
* This routine clears the BSS within the boot region.
* This is separate from z_bss_zero() as boot region may
* contain symbols required for the boot process before
* paging is initialized.
*/
__boot_func
void z_bss_zero_boot(void)
{
z_early_memset(&lnkr_boot_bss_start, 0,
(uintptr_t)&lnkr_boot_bss_end
- (uintptr_t)&lnkr_boot_bss_start);
}
#endif /* CONFIG_LINKER_USE_BOOT_SECTION */
#ifdef CONFIG_LINKER_USE_PINNED_SECTION
/**
* @brief Clear BSS within the pinned region
*
* This routine clears the BSS within the pinned region.
* This is separate from z_bss_zero() as pinned region may
* contain symbols required for the boot process before
* paging is initialized.
*/
#ifdef CONFIG_LINKER_USE_BOOT_SECTION
__boot_func
#else
__pinned_func
#endif /* CONFIG_LINKER_USE_BOOT_SECTION */
void z_bss_zero_pinned(void)
{
z_early_memset(&lnkr_pinned_bss_start, 0,
(uintptr_t)&lnkr_pinned_bss_end
- (uintptr_t)&lnkr_pinned_bss_start);
}
#endif /* CONFIG_LINKER_USE_PINNED_SECTION */
#ifdef CONFIG_STACK_CANARIES
#ifdef CONFIG_STACK_CANARIES_TLS
extern __thread volatile uintptr_t __stack_chk_guard;
#else
extern volatile uintptr_t __stack_chk_guard;
#endif /* CONFIG_STACK_CANARIES_TLS */
#endif /* CONFIG_STACK_CANARIES */
/* LCOV_EXCL_STOP */
__pinned_bss
bool z_sys_post_kernel;
static int do_device_init(const struct init_entry *entry)
{
const struct device *dev = entry->dev;
int rc = 0;
if (entry->init_fn.dev != NULL) {
rc = entry->init_fn.dev(dev);
/* Mark device initialized. If initialization
* failed, record the error condition.
*/
if (rc != 0) {
if (rc < 0) {
rc = -rc;
}
if (rc > UINT8_MAX) {
rc = UINT8_MAX;
}
dev->state->init_res = rc;
}
}
dev->state->initialized = true;
if (rc == 0) {
/* Run automatic device runtime enablement */
(void)pm_device_runtime_auto_enable(dev);
}
return rc;
}
/**
* @brief Execute all the init entry initialization functions at a given level
*
* @details Invokes the initialization routine for each init entry object
* created by the INIT_ENTRY_DEFINE() macro using the specified level.
* The linker script places the init entry objects in memory in the order
* they need to be invoked, with symbols indicating where one level leaves
* off and the next one begins.
*
* @param level init level to run.
*/
static void z_sys_init_run_level(enum init_level level)
{
static const struct init_entry *levels[] = {
__init_EARLY_start,
__init_PRE_KERNEL_1_start,
__init_PRE_KERNEL_2_start,
__init_POST_KERNEL_start,
__init_APPLICATION_start,
#ifdef CONFIG_SMP
__init_SMP_start,
#endif /* CONFIG_SMP */
/* End marker */
__init_end,
};
const struct init_entry *entry;
for (entry = levels[level]; entry < levels[level+1]; entry++) {
const struct device *dev = entry->dev;
int result;
sys_trace_sys_init_enter(entry, level);
if (dev != NULL) {
result = do_device_init(entry);
} else {
result = entry->init_fn.sys();
}
sys_trace_sys_init_exit(entry, level, result);
}
}
int z_impl_device_init(const struct device *dev)
{
if (dev == NULL) {
return -ENOENT;
}
STRUCT_SECTION_FOREACH_ALTERNATE(_deferred_init, init_entry, entry) {
if (entry->dev == dev) {
return do_device_init(entry);
}
}
return -ENOENT;
}
#ifdef CONFIG_USERSPACE
static inline int z_vrfy_device_init(const struct device *dev)
{
K_OOPS(K_SYSCALL_OBJ_INIT(dev, K_OBJ_ANY));
return z_impl_device_init(dev);
}
#include <zephyr/syscalls/device_init_mrsh.c>
#endif
extern void boot_banner(void);
/**
* @brief Mainline for kernel's background thread
*
* This routine completes kernel initialization by invoking the remaining
* init functions, then invokes application's main() routine.
*/
__boot_func
static void bg_thread_main(void *unused1, void *unused2, void *unused3)
{
ARG_UNUSED(unused1);
ARG_UNUSED(unused2);
ARG_UNUSED(unused3);
#ifdef CONFIG_MMU
/* Invoked here such that backing store or eviction algorithms may
* initialize kernel objects, and that all POST_KERNEL and later tasks
* may perform memory management tasks (except for
* k_mem_map_phys_bare() which is allowed at any time)
*/
z_mem_manage_init();
#endif /* CONFIG_MMU */
z_sys_post_kernel = true;
z_sys_init_run_level(INIT_LEVEL_POST_KERNEL);
#if defined(CONFIG_STACK_POINTER_RANDOM) && (CONFIG_STACK_POINTER_RANDOM != 0)
z_stack_adjust_initialized = 1;
#endif /* CONFIG_STACK_POINTER_RANDOM */
boot_banner();
void z_init_static(void);
z_init_static();
/* Final init level before app starts */
z_sys_init_run_level(INIT_LEVEL_APPLICATION);
z_init_static_threads();
#ifdef CONFIG_KERNEL_COHERENCE
__ASSERT_NO_MSG(arch_mem_coherent(&_kernel));
#endif /* CONFIG_KERNEL_COHERENCE */
#ifdef CONFIG_SMP
if (!IS_ENABLED(CONFIG_SMP_BOOT_DELAY)) {
z_smp_init();
}
z_sys_init_run_level(INIT_LEVEL_SMP);
#endif /* CONFIG_SMP */
#ifdef CONFIG_MMU
z_mem_manage_boot_finish();
#endif /* CONFIG_MMU */
extern int main(void);
(void)main();
/* Mark non-essential since main() has no more work to do */
z_thread_essential_clear(&z_main_thread);
#ifdef CONFIG_COVERAGE_DUMP
/* Dump coverage data once the main() has exited. */
gcov_coverage_dump();
#endif /* CONFIG_COVERAGE_DUMP */
} /* LCOV_EXCL_LINE ... because we just dumped final coverage data */
#if defined(CONFIG_MULTITHREADING)
__boot_func
static void init_idle_thread(int i)
{
struct k_thread *thread = &z_idle_threads[i];
k_thread_stack_t *stack = z_idle_stacks[i];
size_t stack_size = K_KERNEL_STACK_SIZEOF(z_idle_stacks[i]);
#ifdef CONFIG_THREAD_NAME
#if CONFIG_MP_MAX_NUM_CPUS > 1
char tname[8];
snprintk(tname, 8, "idle %02d", i);
#else
char *tname = "idle";
#endif /* CONFIG_MP_MAX_NUM_CPUS */
#else
char *tname = NULL;
#endif /* CONFIG_THREAD_NAME */
z_setup_new_thread(thread, stack,
stack_size, idle, &_kernel.cpus[i],
NULL, NULL, K_IDLE_PRIO, K_ESSENTIAL,
tname);
z_mark_thread_as_started(thread);
#ifdef CONFIG_SMP
thread->base.is_idle = 1U;
#endif /* CONFIG_SMP */
}
void z_init_cpu(int id)
{
init_idle_thread(id);
_kernel.cpus[id].idle_thread = &z_idle_threads[id];
_kernel.cpus[id].id = id;
_kernel.cpus[id].irq_stack =
(K_KERNEL_STACK_BUFFER(z_interrupt_stacks[id]) +
K_KERNEL_STACK_SIZEOF(z_interrupt_stacks[id]));
#ifdef CONFIG_SCHED_THREAD_USAGE_ALL
_kernel.cpus[id].usage = &_kernel.usage[id];
_kernel.cpus[id].usage->track_usage =
CONFIG_SCHED_THREAD_USAGE_AUTO_ENABLE;
#endif
#ifdef CONFIG_PM
/*
* Increment number of CPUs active. The pm subsystem
* will keep track of this from here.
*/
atomic_inc(&_cpus_active);
#endif
#ifdef CONFIG_OBJ_CORE_SYSTEM
k_obj_core_init_and_link(K_OBJ_CORE(&_kernel.cpus[id]), &obj_type_cpu);
#ifdef CONFIG_OBJ_CORE_STATS_SYSTEM
k_obj_core_stats_register(K_OBJ_CORE(&_kernel.cpus[id]),
_kernel.cpus[id].usage,
sizeof(struct k_cycle_stats));
#endif
#endif
}
/**
*
* @brief Initializes kernel data structures
*
* This routine initializes various kernel data structures, including
* the init and idle threads and any architecture-specific initialization.
*
* Note that all fields of "_kernel" are set to zero on entry, which may
* be all the initialization many of them require.
*
* @return initial stack pointer for the main thread
*/
__boot_func
static char *prepare_multithreading(void)
{
char *stack_ptr;
/* _kernel.ready_q is all zeroes */
z_sched_init();
#ifndef CONFIG_SMP
/*
* prime the cache with the main thread since:
*
* - the cache can never be NULL
* - the main thread will be the one to run first
* - no other thread is initialized yet and thus their priority fields
* contain garbage, which would prevent the cache loading algorithm
* to work as intended
*/
_kernel.ready_q.cache = &z_main_thread;
#endif /* CONFIG_SMP */
stack_ptr = z_setup_new_thread(&z_main_thread, z_main_stack,
K_THREAD_STACK_SIZEOF(z_main_stack),
bg_thread_main,
NULL, NULL, NULL,
CONFIG_MAIN_THREAD_PRIORITY,
K_ESSENTIAL, "main");
z_mark_thread_as_started(&z_main_thread);
z_ready_thread(&z_main_thread);
z_init_cpu(0);
return stack_ptr;
}
__boot_func
static FUNC_NORETURN void switch_to_main_thread(char *stack_ptr)
{
#ifdef CONFIG_ARCH_HAS_CUSTOM_SWAP_TO_MAIN
arch_switch_to_main_thread(&z_main_thread, stack_ptr, bg_thread_main);
#else
ARG_UNUSED(stack_ptr);
/*
* Context switch to main task (entry function is _main()): the
* current fake thread is not on a wait queue or ready queue, so it
* will never be rescheduled in.
*/
z_swap_unlocked();
#endif /* CONFIG_ARCH_HAS_CUSTOM_SWAP_TO_MAIN */
CODE_UNREACHABLE; /* LCOV_EXCL_LINE */
}
#endif /* CONFIG_MULTITHREADING */
__boot_func
void __weak z_early_rand_get(uint8_t *buf, size_t length)
{
static uint64_t state = (uint64_t)CONFIG_TIMER_RANDOM_INITIAL_STATE;
int rc;
#ifdef CONFIG_ENTROPY_HAS_DRIVER
const struct device *const entropy = DEVICE_DT_GET_OR_NULL(DT_CHOSEN(zephyr_entropy));
if ((entropy != NULL) && device_is_ready(entropy)) {
/* Try to see if driver provides an ISR-specific API */
rc = entropy_get_entropy_isr(entropy, buf, length, ENTROPY_BUSYWAIT);
if (rc > 0) {
length -= rc;
buf += rc;
}
}
#endif /* CONFIG_ENTROPY_HAS_DRIVER */
while (length > 0) {
uint32_t val;
state = state + k_cycle_get_32();
state = state * 2862933555777941757ULL + 3037000493ULL;
val = (uint32_t)(state >> 32);
rc = MIN(length, sizeof(val));
z_early_memcpy((void *)buf, &val, rc);
length -= rc;
buf += rc;
}
}
/**
*
* @brief Initialize kernel
*
* This routine is invoked when the system is ready to run C code. The
* processor must be running in 32-bit mode, and the BSS must have been
* cleared/zeroed.
*
* @return Does not return
*/
__boot_func
FUNC_NO_STACK_PROTECTOR
FUNC_NORETURN void z_cstart(void)
{
/* gcov hook needed to get the coverage report.*/
gcov_static_init();
/* initialize early init calls */
z_sys_init_run_level(INIT_LEVEL_EARLY);
/* perform any architecture-specific initialization */
arch_kernel_init();
LOG_CORE_INIT();
#if defined(CONFIG_MULTITHREADING)
z_dummy_thread_init(&_thread_dummy);
#endif /* CONFIG_MULTITHREADING */
/* do any necessary initialization of static devices */
z_device_state_init();
/* perform basic hardware initialization */
z_sys_init_run_level(INIT_LEVEL_PRE_KERNEL_1);
#if defined(CONFIG_SMP)
arch_smp_init();
#endif
z_sys_init_run_level(INIT_LEVEL_PRE_KERNEL_2);
#ifdef CONFIG_STACK_CANARIES
uintptr_t stack_guard;
z_early_rand_get((uint8_t *)&stack_guard, sizeof(stack_guard));
__stack_chk_guard = stack_guard;
__stack_chk_guard <<= 8;
#endif /* CONFIG_STACK_CANARIES */
#ifdef CONFIG_TIMING_FUNCTIONS_NEED_AT_BOOT
timing_init();
timing_start();
#endif /* CONFIG_TIMING_FUNCTIONS_NEED_AT_BOOT */
#ifdef CONFIG_MULTITHREADING
switch_to_main_thread(prepare_multithreading());
#else
#ifdef ARCH_SWITCH_TO_MAIN_NO_MULTITHREADING
/* Custom ARCH-specific routine to switch to main()
* in the case of no multi-threading.
*/
ARCH_SWITCH_TO_MAIN_NO_MULTITHREADING(bg_thread_main,
NULL, NULL, NULL);
#else
bg_thread_main(NULL, NULL, NULL);
/* LCOV_EXCL_START
* We've already dumped coverage data at this point.
*/
irq_lock();
while (true) {
}
/* LCOV_EXCL_STOP */
#endif /* ARCH_SWITCH_TO_MAIN_NO_MULTITHREADING */
#endif /* CONFIG_MULTITHREADING */
/*
* Compiler can't tell that the above routines won't return and issues
* a warning unless we explicitly tell it that control never gets this
* far.
*/
CODE_UNREACHABLE; /* LCOV_EXCL_LINE */
}
#ifdef CONFIG_OBJ_CORE_SYSTEM
static int init_cpu_obj_core_list(void)
{
/* Initialize CPU object type */
z_obj_type_init(&obj_type_cpu, K_OBJ_TYPE_CPU_ID,
offsetof(struct _cpu, obj_core));
#ifdef CONFIG_OBJ_CORE_STATS_SYSTEM
k_obj_type_stats_init(&obj_type_cpu, &cpu_stats_desc);
#endif /* CONFIG_OBJ_CORE_STATS_SYSTEM */
return 0;
}
static int init_kernel_obj_core_list(void)
{
/* Initialize kernel object type */
z_obj_type_init(&obj_type_kernel, K_OBJ_TYPE_KERNEL_ID,
offsetof(struct z_kernel, obj_core));
#ifdef CONFIG_OBJ_CORE_STATS_SYSTEM
k_obj_type_stats_init(&obj_type_kernel, &kernel_stats_desc);
#endif /* CONFIG_OBJ_CORE_STATS_SYSTEM */
k_obj_core_init_and_link(K_OBJ_CORE(&_kernel), &obj_type_kernel);
#ifdef CONFIG_OBJ_CORE_STATS_SYSTEM
k_obj_core_stats_register(K_OBJ_CORE(&_kernel), _kernel.usage,
sizeof(_kernel.usage));
#endif /* CONFIG_OBJ_CORE_STATS_SYSTEM */
return 0;
}
SYS_INIT(init_cpu_obj_core_list, PRE_KERNEL_1,
CONFIG_KERNEL_INIT_PRIORITY_OBJECTS);
SYS_INIT(init_kernel_obj_core_list, PRE_KERNEL_1,
CONFIG_KERNEL_INIT_PRIORITY_OBJECTS);
#endif /* CONFIG_OBJ_CORE_SYSTEM */