714 lines
19 KiB
C
714 lines
19 KiB
C
/*
|
|
* Copyright (c) 2017 Linaro Limited.
|
|
*
|
|
* SPDX-License-Identifier: Apache-2.0
|
|
*/
|
|
|
|
#include <zephyr/device.h>
|
|
#include <zephyr/init.h>
|
|
#include <zephyr/kernel.h>
|
|
#include <soc.h>
|
|
#include "arm_core_mpu_dev.h"
|
|
#include <zephyr/sys/__assert.h>
|
|
#include <zephyr/sys/math_extras.h>
|
|
#include <zephyr/sys/barrier.h>
|
|
#include <zephyr/linker/linker-defs.h>
|
|
#include <zephyr/mem_mgmt/mem_attr.h>
|
|
#include <zephyr/dt-bindings/memory-attr/memory-attr-arm.h>
|
|
|
|
#define LOG_LEVEL CONFIG_MPU_LOG_LEVEL
|
|
#include <zephyr/logging/log.h>
|
|
LOG_MODULE_DECLARE(mpu);
|
|
|
|
#define NODE_HAS_PROP_AND_OR(node_id, prop) \
|
|
DT_NODE_HAS_PROP(node_id, prop) ||
|
|
|
|
BUILD_ASSERT((DT_FOREACH_STATUS_OKAY_NODE_VARGS(
|
|
NODE_HAS_PROP_AND_OR, zephyr_memory_region_mpu) false) == false,
|
|
"`zephyr,memory-region-mpu` was deprecated in favor of `zephyr,memory-attr`");
|
|
|
|
/*
|
|
* Global status variable holding the number of HW MPU region indices, which
|
|
* have been reserved by the MPU driver to program the static (fixed) memory
|
|
* regions.
|
|
*/
|
|
static uint8_t static_regions_num;
|
|
|
|
/* Global MPU configuration at system initialization. */
|
|
static void mpu_init(void)
|
|
{
|
|
#if defined(CONFIG_SOC_FAMILY_KINETIS)
|
|
/* Enable clock for the Memory Protection Unit (MPU). */
|
|
CLOCK_EnableClock(kCLOCK_Sysmpu0);
|
|
#endif
|
|
}
|
|
|
|
/**
|
|
* Get the number of supported MPU regions.
|
|
*/
|
|
static inline uint8_t get_num_regions(void)
|
|
{
|
|
return FSL_FEATURE_SYSMPU_DESCRIPTOR_COUNT;
|
|
}
|
|
|
|
/* @brief Partition sanity check
|
|
*
|
|
* This internal function performs run-time sanity check for
|
|
* MPU region start address and size.
|
|
*
|
|
* @param part Pointer to the data structure holding the partition
|
|
* information (must be valid).
|
|
*/
|
|
static int mpu_partition_is_valid(const struct z_arm_mpu_partition *part)
|
|
{
|
|
/* Partition size must be a multiple of the minimum MPU region
|
|
* size. Start address of the partition must align with the
|
|
* minimum MPU region size.
|
|
*/
|
|
int partition_is_valid =
|
|
(part->size != 0U)
|
|
&&
|
|
((part->size &
|
|
(~(CONFIG_ARM_MPU_REGION_MIN_ALIGN_AND_SIZE - 1)))
|
|
== part->size)
|
|
&&
|
|
((part->start &
|
|
(CONFIG_ARM_MPU_REGION_MIN_ALIGN_AND_SIZE - 1)) == 0U);
|
|
|
|
return partition_is_valid;
|
|
}
|
|
|
|
/* This internal function performs MPU region initialization.
|
|
*
|
|
* Note:
|
|
* The caller must provide a valid region index.
|
|
*/
|
|
static void region_init(const uint32_t index,
|
|
const struct nxp_mpu_region *region_conf)
|
|
{
|
|
uint32_t region_base = region_conf->base;
|
|
uint32_t region_end = region_conf->end;
|
|
uint32_t region_attr = region_conf->attr.attr;
|
|
|
|
if (index == 0U) {
|
|
/* The MPU does not allow writes from the core to affect the
|
|
* RGD0 start or end addresses nor the permissions associated
|
|
* with the debugger; it can only write the permission fields
|
|
* associated with the other masters. These protections
|
|
* guarantee that the debugger always has access to the entire
|
|
* address space.
|
|
*/
|
|
__ASSERT(region_base == SYSMPU->WORD[index][0],
|
|
"Region %d base address got 0x%08x expected 0x%08x",
|
|
index, region_base, (uint32_t)SYSMPU->WORD[index][0]);
|
|
|
|
__ASSERT(region_end == SYSMPU->WORD[index][1],
|
|
"Region %d end address got 0x%08x expected 0x%08x",
|
|
index, region_end, (uint32_t)SYSMPU->WORD[index][1]);
|
|
|
|
/* Changes to the RGD0_WORD2 alterable fields should be done
|
|
* via a write to RGDAAC0.
|
|
*/
|
|
SYSMPU->RGDAAC[index] = region_attr;
|
|
|
|
} else {
|
|
SYSMPU->WORD[index][0] = region_base;
|
|
SYSMPU->WORD[index][1] = region_end;
|
|
SYSMPU->WORD[index][2] = region_attr;
|
|
SYSMPU->WORD[index][3] = SYSMPU_WORD_VLD_MASK;
|
|
}
|
|
|
|
LOG_DBG("[%02d] 0x%08x 0x%08x 0x%08x 0x%08x", index,
|
|
(uint32_t)SYSMPU->WORD[index][0],
|
|
(uint32_t)SYSMPU->WORD[index][1],
|
|
(uint32_t)SYSMPU->WORD[index][2],
|
|
(uint32_t)SYSMPU->WORD[index][3]);
|
|
|
|
}
|
|
|
|
static int region_allocate_and_init(const uint8_t index,
|
|
const struct nxp_mpu_region *region_conf)
|
|
{
|
|
/* Attempt to allocate new region index. */
|
|
if (index > (get_num_regions() - 1)) {
|
|
|
|
/* No available MPU region index. */
|
|
LOG_ERR("Failed to allocate new MPU region %u\n", index);
|
|
return -EINVAL;
|
|
}
|
|
|
|
LOG_DBG("Program MPU region at index 0x%x", index);
|
|
|
|
/* Program region */
|
|
region_init(index, region_conf);
|
|
|
|
return index;
|
|
}
|
|
|
|
#define _BUILD_REGION_CONF(reg, _ATTR) \
|
|
(struct nxp_mpu_region) { .name = (reg).dt_name, \
|
|
.base = (reg).dt_addr, \
|
|
.end = (reg).dt_addr + (reg).dt_size, \
|
|
.attr = _ATTR, \
|
|
}
|
|
|
|
/* This internal function programs the MPU regions defined in the DT when using
|
|
* the `zephyr,memory-attr = <( DT_MEM_ARM(...) )>` property.
|
|
*/
|
|
static int mpu_configure_regions_from_dt(uint8_t *reg_index)
|
|
{
|
|
const struct mem_attr_region_t *region;
|
|
size_t num_regions;
|
|
|
|
num_regions = mem_attr_get_regions(®ion);
|
|
|
|
for (size_t idx = 0; idx < num_regions; idx++) {
|
|
struct nxp_mpu_region region_conf;
|
|
|
|
switch (DT_MEM_ARM_GET(region[idx].dt_attr)) {
|
|
case DT_MEM_ARM_MPU_RAM:
|
|
region_conf = _BUILD_REGION_CONF(region[idx], REGION_RAM_ATTR);
|
|
break;
|
|
#ifdef REGION_FLASH_ATTR
|
|
case DT_MEM_ARM_MPU_FLASH:
|
|
region_conf = _BUILD_REGION_CONF(region[idx], REGION_FLASH_ATTR);
|
|
break;
|
|
#endif
|
|
#ifdef REGION_IO_ATTR
|
|
case DT_MEM_ARM_MPU_IO:
|
|
region_conf = _BUILD_REGION_CONF(region[idx], REGION_IO_ATTR);
|
|
break;
|
|
#endif
|
|
default:
|
|
/* Either the specified `ATTR_MPU_*` attribute does not
|
|
* exists or the `REGION_*_ATTR` macro is not defined
|
|
* for that attribute.
|
|
*/
|
|
LOG_ERR("Invalid attribute for the region\n");
|
|
return -EINVAL;
|
|
}
|
|
|
|
if (region_allocate_and_init((*reg_index),
|
|
(const struct nxp_mpu_region *) ®ion_conf) < 0) {
|
|
return -EINVAL;
|
|
}
|
|
|
|
(*reg_index)++;
|
|
}
|
|
|
|
return 0;
|
|
}
|
|
|
|
/**
|
|
* This internal function is utilized by the MPU driver to combine a given
|
|
* region attribute configuration and size and fill-in a driver-specific
|
|
* structure with the correct MPU region attribute configuration.
|
|
*/
|
|
static inline void get_region_attr_from_mpu_partition_info(
|
|
nxp_mpu_region_attr_t *p_attr,
|
|
const k_mem_partition_attr_t *attr, uint32_t base, uint32_t size)
|
|
{
|
|
/* in NXP MPU the base address and size are not required
|
|
* to determine region attributes
|
|
*/
|
|
(void) base;
|
|
(void) size;
|
|
|
|
p_attr->attr = attr->ap_attr;
|
|
}
|
|
|
|
/* This internal function programs an MPU region
|
|
* of a given configuration at a given MPU index.
|
|
*/
|
|
static int mpu_configure_region(const uint8_t index,
|
|
const struct z_arm_mpu_partition *new_region)
|
|
{
|
|
struct nxp_mpu_region region_conf;
|
|
|
|
LOG_DBG("Configure MPU region at index 0x%x", index);
|
|
|
|
/* Populate internal NXP MPU region configuration structure. */
|
|
region_conf.base = new_region->start;
|
|
region_conf.end = (new_region->start + new_region->size - 1);
|
|
get_region_attr_from_mpu_partition_info(®ion_conf.attr,
|
|
&new_region->attr, new_region->start, new_region->size);
|
|
|
|
/* Allocate and program region */
|
|
return region_allocate_and_init(index,
|
|
(const struct nxp_mpu_region *)®ion_conf);
|
|
}
|
|
|
|
#if defined(CONFIG_MPU_STACK_GUARD)
|
|
/* This internal function partitions the SRAM MPU region */
|
|
static int mpu_sram_partitioning(uint8_t index,
|
|
const struct z_arm_mpu_partition *p_region)
|
|
{
|
|
/*
|
|
* The NXP MPU manages the permissions of the overlapping regions
|
|
* doing the logical OR in between them, hence they can't be used
|
|
* for stack/stack guard protection. For this reason we need to
|
|
* perform a partitioning of the SRAM area in such a way that the
|
|
* guard region does not overlap with the (background) SRAM regions
|
|
* holding the default SRAM access permission configuration.
|
|
* In other words, the SRAM is split in two different regions.
|
|
*/
|
|
|
|
/*
|
|
* SRAM partitioning needs to be performed in a strict order.
|
|
* First, we program a new MPU region with the default SRAM
|
|
* access permissions for the SRAM area _after_ the stack
|
|
* guard. Note that the permissions are stored in the global
|
|
* array:
|
|
* 'mpu_config.mpu_regions[]', on 'sram_region' index.
|
|
*/
|
|
struct nxp_mpu_region added_sram_region;
|
|
|
|
added_sram_region.base = p_region->start + p_region->size;
|
|
added_sram_region.end =
|
|
mpu_config.mpu_regions[mpu_config.sram_region].end;
|
|
added_sram_region.attr.attr =
|
|
mpu_config.mpu_regions[mpu_config.sram_region].attr.attr;
|
|
|
|
if (region_allocate_and_init(index,
|
|
(const struct nxp_mpu_region *)&added_sram_region) < 0) {
|
|
return -EINVAL;
|
|
}
|
|
|
|
/* Increment, as an additional region index has been consumed. */
|
|
index++;
|
|
|
|
/* Second, adjust the original SRAM region to end at the beginning
|
|
* of the stack guard.
|
|
*/
|
|
struct nxp_mpu_region adjusted_sram_region;
|
|
|
|
adjusted_sram_region.base =
|
|
mpu_config.mpu_regions[mpu_config.sram_region].base;
|
|
adjusted_sram_region.end = p_region->start - 1;
|
|
adjusted_sram_region.attr.attr =
|
|
mpu_config.mpu_regions[mpu_config.sram_region].attr.attr;
|
|
|
|
region_init(mpu_config.sram_region,
|
|
(const struct nxp_mpu_region *)&adjusted_sram_region);
|
|
|
|
return index;
|
|
}
|
|
#endif /* CONFIG_MPU_STACK_GUARD */
|
|
|
|
/* This internal function programs a set of given MPU regions
|
|
* over a background memory area, optionally performing a
|
|
* sanity check of the memory regions to be programmed.
|
|
*/
|
|
static int mpu_configure_regions(const struct z_arm_mpu_partition regions[],
|
|
uint8_t regions_num, uint8_t start_reg_index,
|
|
bool do_sanity_check)
|
|
{
|
|
int i;
|
|
int reg_index = start_reg_index;
|
|
|
|
for (i = 0; i < regions_num; i++) {
|
|
if (regions[i].size == 0U) {
|
|
continue;
|
|
}
|
|
/* Non-empty region. */
|
|
|
|
if (do_sanity_check &&
|
|
(!mpu_partition_is_valid(®ions[i]))) {
|
|
LOG_ERR("Partition %u: sanity check failed.", i);
|
|
return -EINVAL;
|
|
}
|
|
|
|
#if defined(CONFIG_MPU_STACK_GUARD)
|
|
if (regions[i].attr.ap_attr == MPU_REGION_SU_RX) {
|
|
unsigned int key;
|
|
|
|
/* Attempt to configure an MPU Stack Guard region; this
|
|
* will require splitting of the underlying SRAM region
|
|
* into two SRAM regions, leaving out the guard area to
|
|
* be programmed afterwards.
|
|
*/
|
|
key = irq_lock();
|
|
reg_index =
|
|
mpu_sram_partitioning(reg_index, ®ions[i]);
|
|
irq_unlock(key);
|
|
}
|
|
#endif /* CONFIG_MPU_STACK_GUARD */
|
|
|
|
if (reg_index == -EINVAL) {
|
|
return reg_index;
|
|
}
|
|
|
|
reg_index = mpu_configure_region(reg_index, ®ions[i]);
|
|
|
|
if (reg_index == -EINVAL) {
|
|
return reg_index;
|
|
}
|
|
|
|
/* Increment number of programmed MPU indices. */
|
|
reg_index++;
|
|
}
|
|
|
|
return reg_index;
|
|
}
|
|
|
|
/* This internal function programs the static MPU regions.
|
|
*
|
|
* It returns the number of MPU region indices configured.
|
|
*
|
|
* Note:
|
|
* If the static MPU regions configuration has not been successfully
|
|
* performed, the error signal is propagated to the caller of the function.
|
|
*/
|
|
static int mpu_configure_static_mpu_regions(
|
|
const struct z_arm_mpu_partition static_regions[],
|
|
const uint8_t regions_num, const uint32_t background_area_base,
|
|
const uint32_t background_area_end)
|
|
{
|
|
int mpu_reg_index = static_regions_num;
|
|
|
|
/* In NXP MPU architecture the static regions are
|
|
* programmed on top of SRAM region configuration.
|
|
*/
|
|
ARG_UNUSED(background_area_base);
|
|
ARG_UNUSED(background_area_end);
|
|
|
|
mpu_reg_index = mpu_configure_regions(static_regions,
|
|
regions_num, mpu_reg_index, true);
|
|
|
|
static_regions_num = mpu_reg_index;
|
|
|
|
return mpu_reg_index;
|
|
}
|
|
|
|
/* This internal function programs the dynamic MPU regions.
|
|
*
|
|
* It returns the number of MPU region indices configured.
|
|
*
|
|
* Note:
|
|
* If the dynamic MPU regions configuration has not been successfully
|
|
* performed, the error signal is propagated to the caller of the function.
|
|
*/
|
|
static int mpu_configure_dynamic_mpu_regions(
|
|
const struct z_arm_mpu_partition dynamic_regions[],
|
|
uint8_t regions_num)
|
|
{
|
|
unsigned int key;
|
|
|
|
/*
|
|
* Programming the NXP MPU has to be done with care to avoid race
|
|
* conditions that will cause memory faults. The NXP MPU is composed
|
|
* of a number of memory region descriptors. The number of descriptors
|
|
* varies depending on the SOC. Each descriptor has a start addr, end
|
|
* addr, attribute, and valid. When the MPU is enabled, access to
|
|
* memory space is checked for access protection errors through an
|
|
* OR operation of all of the valid MPU descriptors.
|
|
*
|
|
* Writing the start/end/attribute descriptor register will clear the
|
|
* valid bit for that descriptor. This presents a problem because if
|
|
* the current program stack is in that region or if an ISR occurs
|
|
* that switches state and uses that region a memory fault will be
|
|
* triggered. Note that local variable access can also cause stack
|
|
* accesses while programming these registers depending on the compiler
|
|
* optimization level.
|
|
*
|
|
* To avoid the race condition a temporary descriptor is set to enable
|
|
* access to all of memory before the call to mpu_configure_regions()
|
|
* to configure the dynamic memory regions. After, the temporary
|
|
* descriptor is invalidated if the mpu_configure_regions() didn't
|
|
* overwrite it.
|
|
*/
|
|
key = irq_lock();
|
|
/* Use last descriptor region as temporary descriptor */
|
|
region_init(get_num_regions()-1, (const struct nxp_mpu_region *)
|
|
&mpu_config.mpu_regions[mpu_config.sram_region]);
|
|
|
|
/* Now reset the main SRAM region */
|
|
region_init(mpu_config.sram_region, (const struct nxp_mpu_region *)
|
|
&mpu_config.mpu_regions[mpu_config.sram_region]);
|
|
irq_unlock(key);
|
|
|
|
int mpu_reg_index = static_regions_num;
|
|
|
|
/* In NXP MPU architecture the dynamic regions are
|
|
* programmed on top of existing SRAM region configuration.
|
|
*/
|
|
|
|
mpu_reg_index = mpu_configure_regions(dynamic_regions,
|
|
regions_num, mpu_reg_index, false);
|
|
|
|
if (mpu_reg_index != -EINVAL) {
|
|
|
|
/* Disable the non-programmed MPU regions. */
|
|
for (int i = mpu_reg_index; i < get_num_regions(); i++) {
|
|
|
|
LOG_DBG("disable region 0x%x", i);
|
|
/* Disable region */
|
|
SYSMPU->WORD[i][0] = 0;
|
|
SYSMPU->WORD[i][1] = 0;
|
|
SYSMPU->WORD[i][2] = 0;
|
|
SYSMPU->WORD[i][3] = 0;
|
|
}
|
|
}
|
|
|
|
return mpu_reg_index;
|
|
}
|
|
|
|
/* ARM Core MPU Driver API Implementation for NXP MPU */
|
|
|
|
/**
|
|
* @brief enable the MPU
|
|
*/
|
|
void arm_core_mpu_enable(void)
|
|
{
|
|
/* Enable MPU */
|
|
SYSMPU->CESR |= SYSMPU_CESR_VLD_MASK;
|
|
|
|
/* Make sure that all the registers are set before proceeding */
|
|
barrier_dsync_fence_full();
|
|
barrier_isync_fence_full();
|
|
}
|
|
|
|
/**
|
|
* @brief disable the MPU
|
|
*/
|
|
void arm_core_mpu_disable(void)
|
|
{
|
|
/* Force any outstanding transfers to complete before disabling MPU */
|
|
barrier_dmem_fence_full();
|
|
|
|
/* Disable MPU */
|
|
SYSMPU->CESR &= ~SYSMPU_CESR_VLD_MASK;
|
|
/* Clear MPU error status */
|
|
SYSMPU->CESR |= SYSMPU_CESR_SPERR_MASK;
|
|
}
|
|
|
|
#if defined(CONFIG_USERSPACE)
|
|
|
|
static inline uint32_t mpu_region_get_base(uint32_t r_index)
|
|
{
|
|
return SYSMPU->WORD[r_index][0];
|
|
}
|
|
|
|
static inline uint32_t mpu_region_get_size(uint32_t r_index)
|
|
{
|
|
/* <END> + 1 - <BASE> */
|
|
return (SYSMPU->WORD[r_index][1] + 1) - SYSMPU->WORD[r_index][0];
|
|
}
|
|
|
|
/**
|
|
* This internal function checks if region is enabled or not.
|
|
*
|
|
* Note:
|
|
* The caller must provide a valid region number.
|
|
*/
|
|
static inline int is_enabled_region(uint32_t r_index)
|
|
{
|
|
return SYSMPU->WORD[r_index][3] & SYSMPU_WORD_VLD_MASK;
|
|
}
|
|
|
|
/**
|
|
* This internal function checks if the given buffer is in the region.
|
|
*
|
|
* Note:
|
|
* The caller must provide a valid region number.
|
|
*/
|
|
static inline int is_in_region(uint32_t r_index, uint32_t start, uint32_t size)
|
|
{
|
|
uint32_t r_addr_start;
|
|
uint32_t r_addr_end;
|
|
uint32_t end;
|
|
|
|
r_addr_start = SYSMPU->WORD[r_index][0];
|
|
r_addr_end = SYSMPU->WORD[r_index][1];
|
|
|
|
size = size == 0U ? 0U : size - 1U;
|
|
if (u32_add_overflow(start, size, &end)) {
|
|
return 0;
|
|
}
|
|
|
|
if ((start >= r_addr_start) && (end <= r_addr_end)) {
|
|
return 1;
|
|
}
|
|
|
|
return 0;
|
|
}
|
|
|
|
/**
|
|
* @brief update configuration of an active memory partition
|
|
*/
|
|
void arm_core_mpu_mem_partition_config_update(
|
|
struct z_arm_mpu_partition *partition,
|
|
k_mem_partition_attr_t *new_attr)
|
|
{
|
|
/* Find the partition. ASSERT if not found. */
|
|
uint8_t i;
|
|
uint8_t reg_index = get_num_regions();
|
|
|
|
for (i = static_regions_num; i < get_num_regions(); i++) {
|
|
if (!is_enabled_region(i)) {
|
|
continue;
|
|
}
|
|
|
|
uint32_t base = mpu_region_get_base(i);
|
|
|
|
if (base != partition->start) {
|
|
continue;
|
|
}
|
|
|
|
uint32_t size = mpu_region_get_size(i);
|
|
|
|
if (size != partition->size) {
|
|
continue;
|
|
}
|
|
|
|
/* Region found */
|
|
reg_index = i;
|
|
break;
|
|
}
|
|
__ASSERT(reg_index != get_num_regions(),
|
|
"Memory domain partition not found\n");
|
|
|
|
/* Modify the permissions */
|
|
partition->attr = *new_attr;
|
|
mpu_configure_region(reg_index, partition);
|
|
}
|
|
|
|
/**
|
|
* @brief get the maximum number of available (free) MPU region indices
|
|
* for configuring dynamic MPU partitions
|
|
*/
|
|
int arm_core_mpu_get_max_available_dyn_regions(void)
|
|
{
|
|
return get_num_regions() - static_regions_num;
|
|
}
|
|
|
|
/**
|
|
* This internal function checks if the region is user accessible or not
|
|
*
|
|
* Note:
|
|
* The caller must provide a valid region number.
|
|
*/
|
|
static inline int is_user_accessible_region(uint32_t r_index, int write)
|
|
{
|
|
uint32_t r_ap = SYSMPU->WORD[r_index][2];
|
|
|
|
if (write != 0) {
|
|
return (r_ap & MPU_REGION_WRITE) == MPU_REGION_WRITE;
|
|
}
|
|
|
|
return (r_ap & MPU_REGION_READ) == MPU_REGION_READ;
|
|
}
|
|
|
|
/**
|
|
* @brief validate the given buffer is user accessible or not
|
|
*/
|
|
int arm_core_mpu_buffer_validate(const void *addr, size_t size, int write)
|
|
{
|
|
uint8_t r_index;
|
|
|
|
/* Iterate through all MPU regions */
|
|
for (r_index = 0U; r_index < get_num_regions(); r_index++) {
|
|
if (!is_enabled_region(r_index) ||
|
|
!is_in_region(r_index, (uint32_t)addr, size)) {
|
|
continue;
|
|
}
|
|
|
|
/* For NXP MPU, priority is given to granting permission over
|
|
* denying access for overlapping region.
|
|
* So we can stop the iteration immediately once we find the
|
|
* matched region that grants permission.
|
|
*/
|
|
if (is_user_accessible_region(r_index, write)) {
|
|
return 0;
|
|
}
|
|
}
|
|
|
|
return -EPERM;
|
|
}
|
|
|
|
#endif /* CONFIG_USERSPACE */
|
|
|
|
/**
|
|
* @brief configure fixed (static) MPU regions.
|
|
*/
|
|
void arm_core_mpu_configure_static_mpu_regions(
|
|
const struct z_arm_mpu_partition static_regions[],
|
|
const uint8_t regions_num, const uint32_t background_area_start,
|
|
const uint32_t background_area_end)
|
|
{
|
|
if (mpu_configure_static_mpu_regions(static_regions, regions_num,
|
|
background_area_start,
|
|
background_area_end) == -EINVAL) {
|
|
|
|
__ASSERT(0, "Configuring %u static MPU regions failed\n",
|
|
regions_num);
|
|
}
|
|
}
|
|
|
|
/**
|
|
* @brief configure dynamic MPU regions.
|
|
*/
|
|
void arm_core_mpu_configure_dynamic_mpu_regions(
|
|
const struct z_arm_mpu_partition dynamic_regions[], uint8_t regions_num)
|
|
{
|
|
if (mpu_configure_dynamic_mpu_regions(dynamic_regions,
|
|
regions_num) == -EINVAL) {
|
|
|
|
__ASSERT(0, "Configuring %u dynamic MPU regions failed\n",
|
|
regions_num);
|
|
}
|
|
}
|
|
|
|
/* NXP MPU Driver Initial Setup */
|
|
|
|
/*
|
|
* @brief MPU default configuration
|
|
*
|
|
* This function provides the default configuration mechanism for the Memory
|
|
* Protection Unit (MPU).
|
|
*/
|
|
int z_arm_mpu_init(void)
|
|
{
|
|
uint32_t r_index;
|
|
|
|
if (mpu_config.num_regions > get_num_regions()) {
|
|
/* Attempt to configure more MPU regions than
|
|
* what is supported by hardware. As this operation
|
|
* may be executed during system (pre-kernel) initialization,
|
|
* we want to ensure we can detect an attempt to
|
|
* perform invalid configuration.
|
|
*/
|
|
__ASSERT(0,
|
|
"Request to configure: %u regions (supported: %u)\n",
|
|
mpu_config.num_regions,
|
|
get_num_regions()
|
|
);
|
|
return -1;
|
|
}
|
|
|
|
LOG_DBG("total region count: %d", get_num_regions());
|
|
|
|
arm_core_mpu_disable();
|
|
|
|
/* Architecture-specific configuration */
|
|
mpu_init();
|
|
|
|
/* Program fixed regions configured at SOC definition. */
|
|
for (r_index = 0U; r_index < mpu_config.num_regions; r_index++) {
|
|
region_init(r_index, &mpu_config.mpu_regions[r_index]);
|
|
}
|
|
|
|
/* Update the number of programmed MPU regions. */
|
|
static_regions_num = mpu_config.num_regions;
|
|
|
|
/* DT-defined MPU regions. */
|
|
if (mpu_configure_regions_from_dt(&static_regions_num) == -EINVAL) {
|
|
__ASSERT(0, "Failed to allocate MPU regions from DT\n");
|
|
return -EINVAL;
|
|
}
|
|
|
|
arm_core_mpu_enable();
|
|
|
|
return 0;
|
|
}
|