This PR adds Access Control to kernel objects on a per task basis to MPU
ports. The following needs to be defined in the `FreeRTOSConfig.h` to
enable this feature:
```c
```
This PR adds the following new APIs:
```c
void vGrantAccessToTask( TaskHandle_t xTask,
TaskHandle_t xTaskToGrantAccess );
void vRevokeAccessToTask( TaskHandle_t xTask,
TaskHandle_t xTaskToRevokeAccess );
void vGrantAccessToSemaphore( TaskHandle_t xTask,
SemaphoreHandle_t xSemaphoreToGrantAccess );
void vRevokeAccessToSemaphore( TaskHandle_t xTask,
SemaphoreHandle_t xSemaphoreToRevokeAccess );
void vGrantAccessToQueue( TaskHandle_t xTask,
QueueHandle_t xQueueToGrantAccess );
void vRevokeAccessToQueue( TaskHandle_t xTask,
QueueHandle_t xQueueToRevokeAccess );
void vGrantAccessToQueueSet( TaskHandle_t xTask,
QueueSetHandle_t xQueueSetToGrantAccess );
void vRevokeAccessToQueueSet( TaskHandle_t xTask,
QueueSetHandle_t xQueueSetToRevokeAccess );
void vGrantAccessToEventGroup( TaskHandle_t xTask,
EventGroupHandle_t xEventGroupToGrantAccess );
void vRevokeAccessToEventGroup( TaskHandle_t xTask,
EventGroupHandle_t xEventGroupToRevokeAccess );
void vGrantAccessToStreamBuffer( TaskHandle_t xTask,
StreamBufferHandle_t xStreamBufferToGrantAccess );
void vRevokeAccessToStreamBuffer( TaskHandle_t xTask,
StreamBufferHandle_t xStreamBufferToRevokeAccess );
void vGrantAccessToMessageBuffer( TaskHandle_t xTask,
MessageBufferHandle_t xMessageBufferToGrantAccess );
void vRevokeAccessToMessageBuffer( TaskHandle_t xTask,
MessageBufferHandle_t xMessageBufferToRevokeAccess );
void vGrantAccessToTimer( TaskHandle_t xTask,
TimerHandle_t xTimerToGrantAccess );
void vRevokeAccessToTimer( TaskHandle_t xTask,
TimerHandle_t xTimerToRevokeAccess );
```
An unprivileged task by default has access to itself only and no other
kernel object. The application writer needs to explicitly grant an
unprivileged task access to all the kernel objects it needs. The best
place to do that is before starting the scheduler when all the kernel
objects are created.
For example, let's say an unprivileged tasks needs access to a queue and
an event group, the application writer needs to do the following:
```c
vGrantAccessToQueue( xUnprivilegedTaskHandle, xQueue );
vGrantAccessToEventGroup( xUnprivilegedTaskHandle, xEventGroup );
```
The application writer MUST revoke all the accesses before deleting a
task. Failing to do so will result in undefined behavior. In the above
example, the application writer needs to make the following 2 calls
before deleting the task:
```c
vRevokeAccessToQueue( xUnprivilegedTaskHandle, xQueue );
vRevokeAccessToEventGroup( xUnprivilegedTaskHandle, xEventGroup );
```
Earlier the System Call entry from an unprivileged task looked like:
1. SVC for entering system call.
2. System call implementation.
3. SVC for exiting system call.
Now, the system call entry needs to make only one system call
and everything else is handled internally.
This PR also makes the following small changes:
1. Add one struct param for system calls with 5 parameters. This
removes the need for special handling for system calls with 5
parameters.
2. Remove raise privilege SVC when MPU wrapper v2 is used.
3. Add additional run time parameter checks to MPU wrappers
for xTaskGenericNotify and xQueueTakeMutexRecursive APIs.
These changes are tested on the following platforms:
1. STM32H743ZI (Cortex-M7)
2. STM32L152RE (Cortex-M3)
3. Nuvoton M2351 (Cortex-M23)
4. NXP LPC55S69 (Cortex-M33)
* Add runtime parameter checks
This commit adds runtime checks for function parameters
to mpu_wrappers_v2 file. The same checks are performed
in the API implementation using asserts.
Signed-off-by: kar-rahul-aws <karahulx@amazon.com>
Memory Protection Unit (MPU) Enhancements
This commit introduces a new MPU wrapper that places additional
restrictions on unprivileged tasks. The following is the list of changes
introduced with the new MPU wrapper:
1. Opaque and indirectly verifiable integers for kernel object handles:
All the kernel object handles (for example, queue handles) are now
opaque integers. Previously object handles were raw pointers.
2. Saving the task context in Task Control Block (TCB): When a task is
swapped out by the scheduler, the task's context is now saved in its
TCB. Previously the task's context was saved on its stack.
3. Execute system calls on a separate privileged only stack: FreeRTOS
system calls, which execute with elevated privilege, now use a
separate privileged only stack. Previously system calls used the
calling task's stack. The application writer can control the size of
the system call stack using new configSYSTEM_CALL_STACK_SIZE config
macro.
4. Memory bounds checks: FreeRTOS system calls which accept a pointer
and de-reference it, now verify that the calling task has required
permissions to access the memory location referenced by the pointer.
5. System call restrictions: The following system calls are no longer
available to unprivileged tasks:
- vQueueDelete
- xQueueCreateMutex
- xQueueCreateMutexStatic
- xQueueCreateCountingSemaphore
- xQueueCreateCountingSemaphoreStatic
- xQueueGenericCreate
- xQueueGenericCreateStatic
- xQueueCreateSet
- xQueueRemoveFromSet
- xQueueGenericReset
- xTaskCreate
- xTaskCreateStatic
- vTaskDelete
- vTaskPrioritySet
- vTaskSuspendAll
- xTaskResumeAll
- xTaskGetHandle
- xTaskCallApplicationTaskHook
- vTaskList
- vTaskGetRunTimeStats
- xTaskCatchUpTicks
- xEventGroupCreate
- xEventGroupCreateStatic
- vEventGroupDelete
- xStreamBufferGenericCreate
- xStreamBufferGenericCreateStatic
- vStreamBufferDelete
- xStreamBufferReset
Also, an unprivileged task can no longer use vTaskSuspend to suspend
any task other than itself.
We thank the following people for their inputs in these enhancements:
- David Reiss of Meta Platforms, Inc.
- Lan Luo, Xinhui Shao, Yumeng Wei, Zixia Liu, Huaiyu Yan and Zhen Ling
of School of Computer Science and Engineering, Southeast University,
China.
- Xinwen Fu of Department of Computer Science, University of
Massachusetts Lowell, USA.
- Yuequi Chen, Zicheng Wang, Minghao Lin of University of Colorado
Boulder, USA.
* Fix circular dependency in cmake project
Fix for https://github.com/FreeRTOS/FreeRTOS-Kernel/issues/687
In order for custom ports to also break the cycle, they must link
against freertos_kernel_include instead of freertos_kernel.
* Simplify include path
* Add croutines to the code base
* Add croutine changes to cmake, lexicon and readme
* Add croutine file to portable cmake file
* Add back more references from PR 591
Added various ...GetStaticBuffer() functions to get the buffers of statically
created objects.
---------
Co-authored-by: Paul Bartell <pbartell@amazon.com>
Co-authored-by: Nikhil Kamath <110539926+amazonKamath@users.noreply.github.com>
Co-authored-by: Gaurav Aggarwal <aggarg@amazon.com>
Both the TLS block and stack must be correctly aligned when using
picolibc. The architecture stack alignment is represented by the
portBYTE_ALIGNMENT_MASK and the TLS block alignment is provided by the
Picolibc _tls_align() inline function for Picolibc version 1.8 and
above. For older versions of Picolibc, we'll assume that the TLS block
requires the same alignment as the stack.
For downward growing stacks, this requires aligning the start of the
TLS block to the maximum of the stack alignment and the TLS
alignment. With this, both the TLS block and stack will now be
correctly aligned.
For upward growing stacks, the two areas must be aligned
independently; the TLS block is aligned from the start of the stack,
then the tls space is allocated, and then the stack is aligned above
that.
It's probably useful to know here that the linker ensures that
variables within the TLS block are assigned offsets that match their
alignment requirements. If the TLS block itself is correctly aligned,
then everything within will also be.
I have only tested the downward growing stack branch of this patch.
Signed-off-by: Keith Packard <keithpac@amazon.com>
Co-authored-by: Keith Packard <keithpac@amazon.com>
Co-authored-by: Gaurav-Aggarwal-AWS <33462878+aggarg@users.noreply.github.com>
* Pass top of stack to configINIT_TLS_BLOCK
Picolibc wants to allocate the per-task TLS block within the stack
segment, so it will need to modify the top of stack value. Pass the
pxTopOfStack variable to make this explicit.
Signed-off-by: Keith Packard <keithpac@amazon.com>
* Move newlib-specific definitions to separate file
This reduces the clutter in FreeRTOS.h caused by having newlib-specific
macros present there.
Signed-off-by: Keith Packard <keithpac@amazon.com>
* Make TLS code depend only on configUSE_C_RUNTIME_TLS_SUPPORT
Remove reference to configUSE_NEWLIB_REENTRANT as that only works
when using newlib. configUSE_C_RUNTIME_TLS_SUPPORT is always
set when configUSE_NEWLIB_REENTRANT is set, so using both was
redundant in that case.
Signed-off-by: Keith Packard <keithpac@amazon.com>
* portable-ARC: Adapt ARC support to use generalized TLS support
With generalized thread local storage (TLS) support present in the
core, the two ARC ports need to have the changes to the TCB mirrored
to them.
Signed-off-by: Keith Packard <keithpac@amazon.com>
* Add Thread Local Storage (TLS) support using Picolibc functions
This patch provides definitions of the general TLS support macros in
terms of the Picolibc TLS support functions.
Picolibc is normally configured to use TLS internally for all
variables that are intended to be task-local, so these changes are
necessary for picolibc to work correctly with FreeRTOS.
The picolibc helper functions rely on elements within the linker
script to arrange the TLS data in memory and define some symbols.
Applications wanting to use this mechanism will need changes in their
linker script when migrating to picolibc.
Signed-off-by: Keith Packard <keithpac@amazon.com>
---------
Signed-off-by: Keith Packard <keithpac@amazon.com>
Co-authored-by: Keith Packard <keithpac@amazon.com>
Co-authored-by: Gaurav-Aggarwal-AWS <33462878+aggarg@users.noreply.github.com>
* Adding in ability to support a library for freertos_config and a custom freertos_kernel_port (#558)
* Using single name definition for libraries everywhere. (#558)
* Supporting backwards compatibility with FREERTOS_CONFIG_FILE_DIRECTORY (#571)
* Removing compiler warnings for GNU and Clang. (#571)
* Added in documentation on how to consume from a main project. Added default PORT selection for native POSIX and MINGW platforms.
* Only adding freertos_config if it exists. Removing auto generation of it from a FREERTOS_CONFIG_FILE_DIRECTORY.
* Fixing clang and gnu compiler warnings.
* Adding in project information and how to compile for GNU/clang
* Fixing compiler issue with unused variable - no need to declare variable.
* Adding in compile warnings for linux builds that kernel is okay with using.
* Fixing more extra-semi-stmt clang warnings.
* Moving definition of hooks into header files if features are enabled.
* Fixing formatting with uncrustify.
* Fixing merge conflicts with main merge.
* Fixing compiler errors due to merge issues and formatting.
* Fixing Line feeds.
* Adding 'portNORETURN' into portmacros.h. Other Updates based on PR request
* Further clean-up of clang and clang-tidy issues.
* Removing compiler specific pragmas from common c files.
* Fixing missing lexicon entry and uncrustify formatting changes.
* Resolving merge issue multiple defnitions of proto for prvIdleTask
* Fixing formatting issues that are not covered by uncrustify. Use clang-tidy instead if you want this level of control.
* More uncrustify formatting issues.
* Fixing extra bracket in #if statement.
---------
Co-authored-by: Gaurav-Aggarwal-AWS <33462878+aggarg@users.noreply.github.com>
The PR #597 introduced a new config option configTICK_TYPE_WIDTH_IN_BITS
which can be defined to one of the following:
* TICK_TYPE_WIDTH_16_BITS - Tick type is 16 bit wide.
* TICK_TYPE_WIDTH_32_BITS - Tick type is 32 bit wide.
* TICK_TYPE_WIDTH_64_BITS - Tick type is 64 bit wide.
Earlier we supported 16 and 32 bit width for tick type which was
controlled using the config option configUSE_16_BIT_TICKS. The PR
tried to maintain backward compatibility by honoring
configUSE_16_BIT_TICKS. The backward compatibility did not work as
expected though, as the macro configTICK_TYPE_WIDTH_IN_BITS was used
before it was defined. This PR addresses it by ensuring that the macro
configTICK_TYPE_WIDTH_IN_BITS is defined before it is used.
Testing
1. configUSE_16_BIT_TICKS is defined to 0.
Source (function xTaskIncrementTick in tasks.c):
```
const TickType_t xConstTickCount = xTickCount + ( TickType_t ) 1;
```
Assembly:
```
109e: 4b50 ldr r3, [pc, #320] ; (11e0 <xTaskIncrementTick+0x150>)
10a0: f8d3 4134 ldr.w r4, [r3, #308] ; 0x134
10a4: 3401 adds r4, #1
10a6: f8c3 4134 str.w r4, [r3, #308] ; 0x134
```
It is clear from assembly that the tick type is 32 bit.
2. configUSE_16_BIT_TICKS is defined to 1.
Source (function xTaskIncrementTick in tasks.c):
```
const TickType_t xConstTickCount = xTickCount + ( TickType_t ) 1;
```
Assembly:
```
10e2: 4b53 ldr r3, [pc, #332] ; (1230 <xTaskIncrementTick+0x15c>)
10e4: f8b3 4134 ldrh.w r4, [r3, #308] ; 0x134
10e8: b2a4 uxth r4, r4
10ea: 3401 adds r4, #1
10ec: b2a4 uxth r4, r4
10ee: f8a3 4134 strh.w r4, [r3, #308] ; 0x134
```
It is clear from assembly that the tick type is 16 bit.
3. configTICK_TYPE_WIDTH_IN_BITS is defined to TICK_TYPE_WIDTH_16_BITS.
Source (function xTaskIncrementTick in tasks.c):
```
const TickType_t xConstTickCount = xTickCount + ( TickType_t ) 1;
```
Assembly:
```
10e2: 4b53 ldr r3, [pc, #332] ; (1230 <xTaskIncrementTick+0x15c>)
10e4: f8b3 4134 ldrh.w r4, [r3, #308] ; 0x134
10e8: b2a4 uxth r4, r4
10ea: 3401 adds r4, #1
10ec: b2a4 uxth r4, r4
10ee: f8a3 4134 strh.w r4, [r3, #308] ; 0x134
```
It is clear from assembly that the tick type is 16 bit.
4. configTICK_TYPE_WIDTH_IN_BITS is defined to TICK_TYPE_WIDTH_32_BITS.
Source (function xTaskIncrementTick in tasks.c):
```
const TickType_t xConstTickCount = xTickCount + ( TickType_t ) 1;
```
Assembly:
```
109e: 4b50 ldr r3, [pc, #320] ; (11e0 <xTaskIncrementTick+0x150>)
10a0: f8d3 4134 ldr.w r4, [r3, #308] ; 0x134
10a4: 3401 adds r4, #1
10a6: f8c3 4134 str.w r4, [r3, #308] ; 0x134
```
It is clear from assembly that the tick type is 32 bit.
5. configTICK_TYPE_WIDTH_IN_BITS is defined to TICK_TYPE_WIDTH_64_BITS.
```
#error configTICK_TYPE_WIDTH_IN_BITS set to unsupported tick type width.
```
The testing was done for GCC/ARM_CM3 port which does not support 64 bit
tick type.
6. Neither configUSE_16_BIT_TICKS nor configTICK_TYPE_WIDTH_IN_BITS
defined.
```
#error Missing definition: One of configUSE_16_BIT_TICKS and
configTICK_TYPE_WIDTH_IN_BITS must be defined in FreeRTOSConfig.h.
See the Configuration section of the FreeRTOS API documentation for
details.
```
7. Both configUSE_16_BIT_TICKS and configTICK_TYPE_WIDTH_IN_BITS defined.
```
#error Only one of configUSE_16_BIT_TICKS and
configTICK_TYPE_WIDTH_IN_BITS must be defined in FreeRTOSConfig.h.
See the Configuration section of the FreeRTOS API documentation for
details.
```
Related issue - https://github.com/FreeRTOS/FreeRTOS-Kernel/issues/628
Signed-off-by: Gaurav Aggarwal <aggarg@amazon.com>
Allow ulTaskGetIdleRunTimeCounter and ulTaskGetIdleRunTimePercent to be
used whenever configGENERATE_RUN_TIME_STATS is enabled, as this is the
only requirement for these functions to work.
listGET_OWNER_OF_NEXT_ENTRY computes `( pxConstList )->pxIndex->pxNext` after
verifying that `( pxConstList )->pxIndex` points to `xListEnd`, which due to
being a MiniListItem_t, can be shorter than a ListItem_t. Thus,
`( pxConstList )->pxIndex` is a `ListItem_t *` that extends past the end of the
`List_t` whose `xListEnd` it points to. This is fixed by accessing `pxNext`
through a `MiniListItem_t` instead.
It was possible for an unprivileged task to invoke any function with
privilege by passing it as a parameter to MPU_xTaskCreate,
MPU_xTaskCreateStatic, MPU_xTimerCreate, MPU_xTimerCreateStatic, or
MPU_xTimerPendFunctionCall.
This commit ensures that MPU_xTaskCreate and MPU_xTaskCreateStatic can
only create unprivileged tasks. It also removes the following APIs:
1. MPU_xTimerCreate
2. MPU_xTimerCreateStatic
3. MPU_xTimerPendFunctionCall
We thank Huazhong University of Science and Technology for reporting
this issue.
Signed-off-by: Gaurav Aggarwal <aggarg@amazon.com>
It was possible for a third party that had already independently gained
the ability to execute injected code to achieve further privilege
escalation by branching directly inside a FreeRTOS MPU API wrapper
function with a manually crafted stack frame. This commit removes the
local stack variable `xRunningPrivileged` so that a manually crafted
stack frame cannot be used for privilege escalation by branching
directly inside a FreeRTOS MPU API wrapper.
We thank Certibit Consulting, LLC, Huazhong University of Science and
Technology and the SecLab team at Northeastern University for reporting
this issue.
Signed-off-by: Gaurav Aggarwal <aggarg@amazon.com>
* Update of three badly terminated macro definitions
- vTaskDelayUntil() to conform to usual pattern do { ... } while(0)
- vTaskNotifyGiveFromISR() and
- vTaskGenericNotifyGiveFromISR() to remove extra terminating semicolons
- This PR addresses issues #553 and #554
* Adjust formatting of task.h
Co-authored-by: Paul Bartell <pbartell@amazon.com>
* Generalize Thread Local Storage (TLS) support
FreeRTOS's Thread Local Storage (TLS) support used variables and
functions from newlib, thereby making the TLS support specific to
newlib. This commit generalizes the TLS support so that it can be used
with other c-runtime libraries also. The default behavior for newlib
support is still kept same for backward compatibility.
The application writer would need to set configUSE_C_RUNTIME_TLS_SUPPORT
to 1 in their FreeRTOSConfig.h and define the following macros to
support TLS for a c-runtime library:
1. configTLS_BLOCK_TYPE - Type used to define the TLS block in TCB.
2. configINIT_TLS_BLOCK( xTLSBlock ) - Allocate and initialize memory
block for the task's TLS Block.
3. configSET_TLS_BLOCK( xTLSBlock ) - Switch C-Runtime's TLS Block to
point to xTLSBlock.
4. configDEINIT_TLS_BLOCK( xTLSBlock ) - Free up the memory allocated
for the task's TLS Block.
The following is an example to support TLS for picolibc:
#define configUSE_C_RUNTIME_TLS_SUPPORT 1
#define configTLS_BLOCK_TYPE void*
#define configINIT_TLS_BLOCK( xTLSBlock ) _init_tls( xTLSBlock )
#define configSET_TLS_BLOCK( xTLSBlock ) _set_tls( xTLSBlock )
#define configDEINIT_TLS_BLOCK( xTLSBlock )
Signed-off-by: Gaurav Aggarwal <aggarg@amazon.com>
* Let each stream/message can use its own sbSEND_COMPLETED
In FreeRTOS.h, set the default value of configUSE_SB_COMPLETED_CALLBACK
to zero, and add additional space for the function pointer when
the buffer created statically.
In stream_buffer.c, modify the macro of sbSEND_COMPLETED which let
the stream buffer to use its own implementation, and then add an
pointer to the stream buffer's structure, and modify the
implementation of the buffer creating and initializing
Co-authored-by: eddie9712 <qw1562435@gmail.com>
* Add supposrt for ARM CM55
* Fix file header
* Remove duplicate code
* Refactor portmacro.h
1. portmacro.h is re-factored into 2 parts - portmacrocommon.h which is
common to all ARMv8-M ports and portmacro.h which is different for
different compiler and architecture. This enables us to provide
Cortex-M55 ports without code duplication.
2. Update copy_files.py so that it copies Cortex-M55 ports correctly -
all files except portmacro.h are used from Cortex-M33 ports.
Signed-off-by: Gaurav Aggarwal <aggarg@amazon.com>
Co-authored-by: Gaurav Aggarwal <aggarg@amazon.com>
vApplicationMallocFailedHook was declared in each Heap file. which forces users to declare it and can cause problems if the prototype of the function changes.
Co-authored-by: Pierre-Noel Bouteville <pnb990@gmail.com>
* Heap improvements
This commit makes the following improvements:
1. Add a check to heap_2 to track if a memory block is allocated to the
application or not. The MSB of the size field is used for this
purpose. The same check already exists in heap_4 and heap_5. This
check prevents against double free.
2. Add a new flag configHEAP_CLEAR_MEMORY_ON_FREE to heap_2, heap_4 and
heap_5. The application writer can set it to 1 in their
FreeRTOSConfig.h to ensure that a block of memory allocated using
pvPortMalloc is cleared (i.e. set to zero) when it is freed using
vPortFree. If left undefined, configHEAP_CLEAR_MEMORY_ON_FREE
defaults to 0 for backward compatibility. We recommend setting
configHEAP_CLEAR_MEMORY_ON_FREE to 1 for better security.
3. Add a new API pvPortCalloc to heap_2, heap_4 and heap_5. This API
has the following signature:
void * pvPortCalloc( size_t xNum, size_t xSize );
It allocates memory for an array of xNum objects each of which is of
xSize and initializes all bytes in the allocated storage to zero. If
allocation succeeds, it returns a pointer to the lowest byte in the
allocated memory block. On failure, it returns a null pointer.
Signed-off-by: Gaurav Aggarwal <aggarg@amazon.com>
