Infineon Aurix

From ErikaWiki

Revision as of 09:53, 30 May 2014 by Eguidieri (Talk | contribs)
Jump to: navigation, search

Contents

Infineon Aurix support

ERIKA Enterprise supports the Infineon AURIX family microcontrollers. The support for RT-Druid will be released soon. Currently only the TC27x AURIX family have been fully ported:

  1. Support for the HIGHTEC GCC Compiler (form both single and multicore) and TASKING compiler (only for singlecore).
  2. Support for single- and multi-stack configurations.
  3. ISR Type 1 and Type 2.
  4. ORTI support.
  5. Full Lauterbach support.
  • Supported compilers:
    • HIGHTEC GCC Compiler v4.6.2.0 (for both Single and Multicore).
    • TASKING VX-toolset for TriCore v4.0r1 (only for Singlecore).
    • WIND RIVER DIAB Compiler Rel. 5.9.2.0 (only for Singlecore).
  • Mode of operation
    • Monostack: The Monostack configuration of the ERIKA Kernel models the fact that all tasks and ISRs in the system share the same stack.
    • Multistack: Every thread can have its private stack, or it can share it with other threads.
    • Multicore: It follows the same philosophy used by ERIKA on other multicore systems and specified by Autosar: two instances of the operating systems run on the two cores, and communication between cores is performed with a few APIs and shared memory.

MCU & Board

  • The porting have been completly developed on top of:
    • TriBoard TC2x5 v2.0 equiped with a TC275TE MCU

Target Configuration and Programming

ERIKA Enterprise is configured through RT-Druid and an OIL file. Here are listed, after the information to set compiler path, the OIL fields customized for Tricore Aurix architecture.

Compiler Path

It is possible to choose the path of the compiler in three different ways:

  • PATH enviornment variable: You can put compilers bin directories in PATH and the environmet can use them from here.
  • Compiler specific environment variables:
    • HIGHTEC GCC toolchain: You can specify the compiler path for HIGHTEC GCC with the TRICORE_GCCDIR.
    • Altium TASKING toolchain: You can specify the compiler path for Altium TASKING with the TRICORE_TASKINGDIR.
    • Wind River Diab toolchain: You can specify the compiler path for Altium TASKING with the TRICORE_DIABDIR.
  • RT-Druid configuration file see RT-Druid configuration#Compiler paths:
    • preference_tricore__path_for_gnu_compiler set HIGHTEC gcc compiler path for Tricore AURIX
    • preference_tricore__path_for_tasking_compiler set Altium TASKING compiler path for Tricore AURIX
    • preference_tricore__path_for_diab_compiler set Wind River Diab compiler path for Tricore AURIX[a]

Here is an example of RT-Druid configuration file.

Compilers Notes

  • [a] RT-Druid support for Diab will be released ASAP.

CPU

CPU_DATA must be set to TRICORE. The compiler is specificed with the COMPILER_TYPE item, supported values are GNU (the default value) and TASKING.

On multicore systems, one CPU_DATA instance can exist for each core. They must have different IDs. See ERIKA multicore support for more details about multicore systems on Erika. Later in this page you will find a paragraph relative to specific multicore support developed for Tricore Aurix, according to Autosar OS 4.0 Rev 3 specifications.

Tricore AURIX support CPU clock configuration with a simple PLL driver. The target clock value in MHz can be set with the CPU_CLOCK' field. We kept the algorithm to avaluate PLL parameters simple, so it implements a best effort approach to set the right value. In any case max declared CPU clock value (i.e. 200 Mhz for TC27x family), is guaranteed to be perfectly matched.

N.B. To get the real value set you can use EE_tc27x_get_clock() API after executing StartOS, PLL configuration is done during OS start-up.

Moreover a new field to declare a custom linker script have been added to CPU_DATA block: LINKERSCRIPT = "fake.ld";

Example of a CPU_DATA section:

   CPU_DATA = TRICORE {
     CPU_CLOCK = 200.0;
     APP_SRC = "code.c";
     COMPILER_TYPE = GNU;
     MULTI_STACK = TRUE {
       IRQ_STACK = TRUE {
           SYS_SIZE = 128;
       };
     };
   };

MCU

MCU_DATA support only the TC27x value. Compiler can be specified even in MCU_DATA block and will inherited by all configured CPUs (useful in multicore project configuration).

 MCU_DATA = TRICORE {
   MODEL = TC27x;
   COMPILER_TYPE = GNU;
 };

BOARD

There is only support form TriBoard TC2x5 equiped with a TC275TE MCU. This board only provide debug interface and 8 leds, so the support is limited to the LEDs configuration/driving and, because standard ERIKA demos require it, an external button (external means that have to be soldered by the user) mounted on pin P15.8 of TC275TE (corresponding to pin 71 of PERIPHERALS (Xx02,Xx02) connector). Add the following OIL field to enable this support.

 BOARD_DATA = TRIBOARD_TC2X5;

Interrupt Handling

Due to the special implementation of the Interrupt Vector in AURIX architecture, each entry of the vector table is simply identified by it's priority value. It must be the application code that configure the service request node (SRN') with the right priority, assigning in this way the right handler.

 ISR IsrLow {
   CATEGORY = 2;
   PRIORITY = 1;
   HANDLER = "isr_low"; 
 };

The mean of the fields are:

  • CATEGORY: the type of ISR as specified by OSEK.
  • PRIORITY: The ISR priority that represent it's position inside Interrupt Vector so it has to be considered as ISR Identifier. The ENTRY field, the one usually used to declare Interrupt IDs, is still available but it's superfluous because it has to be equal to PRIORITY and, in any case, priority value take the precedence.
  • HANDLER: Declare the interrupt handler symbol. If it's not declared the the handler symbol is supposed to be equal to ISR block name (IsrLow in the example).

Trap Handling

With TriCore AURIX, for the first time, we support the options to register handler for TRAP/Exceptions. OSEK OIL do not provide any field for trap handling: it suppose that all the handling is done inside the Kernel. Usually that's what happens, but sometime can be useful have some mechanism to attach an handler for a particular class of TRAPs.

We decide to exetend ISR entry with the field TRAP that can be set to TRUE to enable TRAP handling. In this case you have to choose the TRAP class with the field LEVEL (or ENTRY). The only valid values for this configuration are:

  • TRAP_MMU (Actually useless because TC27x family don't have MMU, just a place holder)
  • TRAP_PROT (Handler for protection traps)
  • TRAP_INST (Handler for instructions traps)
  • TRAP_CONT (Handler for context traps)
  • TRAP_BUS (Handler forn bus traps)
  • TRAP_ASS (Handler for assertion traps) (please don't be silly :))
  • TRAP_SYS (Handler for system calls)
  • TRAP_NMI (Handler for NMI trap)

OIL TRAP configuration example:

 ISR trap_context {
     LEVEL = "TRAP_CONT";
     HANDLER = "EE_trap_context";   //Trap handler
     TRAP = TRUE;
 };

To define the handler in your code, you have to use the following syntax:

 TRAP(EE_CLASS_TRAPCONT, EE_trap_context) {
   while(1) {
     ; /* dummy */
   }
 }

TRAP class identifiers are the following:

  • EE_CLASS_TRAPMMU
  • EE_CLASS_TRAPPROT
  • EE_CLASS_TRAPINST
  • EE_CLASS_TRAPCONT
  • EE_CLASS_TRAPBUS
  • EE_CLASS_TRAPASS
  • EE_CLASS_TRAPSYS
  • EE_CLASS_TRAPNMI

Inside the TRAP handler TIN (Trap Identification Number) value can be accessed with EE_tc_get_TIN() function or, for HIGHTEC GNUC Compiler, with the local variable tin. For more information you can check the $(ee)/pkg/cpu/tricore/inc/ee_tc_trap.h file.

EEOPT

EEOPT is a way to specify configuration flags to the Erika build environment. EEOPTs can be specified as strings in the OS section of the OIL file. Examples:

EE_OPT = "EE_DEBUG";
EE_OPT = "__ASSERT__";

Please notice that spelling inside the OIL file includes an underscore: EE_OPT.

The only supported format for EEOPTs is a single name, which should be a valid C identifier (i.e., only Latin letters, digits, and underscore are allowed; the first character cannot be a digit). And any other format is not supported, and even if it works now, it may break in the future.

The following EEOPTs are specific of AURIX Architecture:

  • EE_DEBUG: Replace the often used DEBUG option (because it conflict with compilers DEBUG define). Enable debug compiler options, basically less optimization and debug symbols generation plus defualt TRAP handlers are implemented as busy loops instead of system reset.
  • EE_EXECUTE_FROM_RAM: When specified, a linker script is used that maps both code and data in the RAM space. Executables produced with this option can be used only together with a debugger that loads the program in memory. By default, code and constant data are mapped to Flash, data to RAM.
  • EE_SAVE_TEMP_FILES: Enable temporary files saving for the compiler, useful to debug build process and to inspect generate assembly code. It's useful only for HIGHTEC GCC compiler because for TASKING compiler is always active. For the GCC compiler it has been added the switch because the size of temporary files is huge.
  • EE_ICACHE_ENABLED: Enable the instruction cache in start-up code.
  • EE_DCACHE_ENABLED: Enable data cache in start-up code

OSEK/VDX Extensions

This Section contains information about the OSEK/VDX Extensions (or optional features) that have been implemented for the AURIX support.

Resource Managament at ISR level

This feauture is automatically enabled by RT-Druid during the configuration generation step. To specify that a Resource is used by both a Task and a ISR you need to add that resource to the corrisponding ISR object as follows:

 TASK Task1 {
   ...
   RESOURCE = "ResourceA";
 };
 
 ISR <SYMBOL> {
   PRIORITY = <PRIORITY_LEVEL>;
   CATEGORY = <ISR_TYPE>;
   RESOURCE = "ResourceA";
 };
 
 RESOURCE ResourceA { RESOURCEPROPERTY = STANDARD; };

System Timer

The OSEK/VDX standard provides support for a System Counter (a counter that is automatically linked to a hardware timer). The System Timer is used to give a coherent timing reference across the entire application.

In ERIKA Enterprise, this special counter has been named System Timer. To use it, you need to set a specific attribute in a Counter definition. Please note that only one counter for each core can be the System Timer.

A Counter which is not a System Counter must be incremented explicitly using the Autosar primitive IncrementCounter.

The following is an example OIL definition for a System Counter:

 CPU_DATA = TRICORE{
   CPU_CLOCK = 200.0;
   ...
 };
 
 COUNTER SystemTimer {
   MINCYCLE = 1;
   MAXALLOWEDVALUE = 2147483647;
   TICKSPERBASE = 1;
   TYPE = HARDWARE {
     DEVICE = "STM_SR0";
     SYSTEM_TIMER = TRUE;
     PRIORITY = 1;
   };
   SECONDSPERTICK = 0.001;
 };

The meaning of the various attributes is as follows:

  • CPU_DATA/CPU_CLOCK is used to declare the clock frequency (in MHZ).
  • COUNTER/TYPE must be set to HARDWARE, and SYSTEM_TIMER must be set to true.
  • COUNTER/TYPE/DEVICE must be a valid device that can be used for a system timer. Currently, for TRICORE only STM (System Timer Module) MCU peripheral is a valid device for system timer. Both Interrupt source of this peripheral can be set to device and allowed values are STM_SR0 and STM_SR1
  • COUNTER/TYPE/PRIORITY By default SYSTEM_TIMER is tied to smallest ISR priority (i.e. PRIORITY = 1;), but it can be overritten. Ovveride is necessay in multicore environment because priority 1 is used by the Intercore Interrupt Request.
  • COUNTER/SECONDSPERTICK is used to declare the wanted time duration of one hardware tick in seconds.

The System Timer can be attached to ALARMs as usual, as in the following example:

 ALARM AlarmExample {
   COUNTER = SystemTimer;
   ACTION  = ACTIVATETASK{
      TASK = TaskExample;
   };
 };

CPU MCU & BOARD API

In addition to AUTOSAR OS Kernel Interface, in ERIKA AURIX porting are implemented a bunch of utility functions that will be considered as part of ERIKA API. As usual they are separated in the three logical layer that compose ERIKA architecture abstraction: CPU, MCU & BOARD.

CPU API

CPU layer represent all the behaviour shared between all the families of AURIX MCUs. In this layer are declared the functions to temporary disable ENDINIT and SAFETY_ENDINIT register protection (see Infineon AURIX documentation). These functions are a little bit tricky: declaration belong to CPU layer, because AURIX architecture documentation states that every AURIX implementation has some kind of ENDINIT and SAFETY_ENDINIT protection, but delegate implementation details to each AURIX family, so functions definitions has been done inside MCU files.

  • void EE_tc_endint_disable( void ): Temporary disable ENDINT protection.
  • void EE_tc_endint_enable( void ): Re-enable ENDINT protection.
  • void EE_tc_safety_endinit_disable( void ): Temporary disable SAFETY_ENDINIT protection.
  • void EE_tc_safety_endinit_enable( void ): Re-enable SAFETY_ENDINIT protection.

MCU API

MCU layer represent the behaviour tied to an specific AURIX family of MCUs. For now only family TC27x is supported. Most part of the utilities belong to this layer:

  • void EE_tc27x_get_clock ( void ): Return CPU clock frequency in HZ.
  • void EE_tc27x_configure_clock( EE_UREG fclock ): Make the best effort to set CPU clock frequency to fclock value. It's the function used by StartOS when a CPU_CLOCK is configured in CPU_DATA.
  • void EE_tc27x_delay ( EE_UREG usec ): Implement a busy loop wait of usec micorseconds.
  • void EE_tc27x_stm_set_sr0( EE_UINT32 usec, EE_TYPEISR2PRIO intvec ): Programs STM compare register 0 to trigger an IRQ after usec microseconds. intvec is the priority tied to this source, in other words it is the Interrupt Vector Table entry that will handle STM interrupt. With intvec == 0, the correponding service request node is left unprogrammed or resetted.
  • void EE_tc27x_stm_set_sr0_next_match( EE_UINT32 usec ): Change previous programmed STM compare register 0 to trigger next IRQ after usec microseconds. To mantain fixed interrupt interval, it have to been called at the beginning of intvec handler.
  • void EE_tc27x_stm_set_sr1( EE_UINT32 usec, EE_TYPEISR2PRIO intvec ): Programs STM compare register 1 to trigger an IRQ after usec microseconds. intvec is the priority tied to this source, in other words it is the Interrupt Vector Table entry that will handle STM interrupt. With intvec == 0, the correponding service request node is left unprogrammed or resetted.
  • void EE_tc27x_stm_set_sr1_next_match( EE_UINT32 usec ): Change previous programmed STM compare register 1 to trigger next IRQ after usec microseconds. To mantain fixed interrupt interval, it have to been called at the beginning of intvec handler.

SR0 and SR1 API are both available to the user if System Timer is not configured. Otherwise only the one not used by system timer will be available.

Board API

BOARD layer represent the specific board support. There is only aminimal support for TriBoard TC2x5 v2.0 equiped with a TC275TE MCU:

  • void EE_tc2x5_leds_init( void ): Initialize the 8 boards leds.
  • void EE_tc2x5_leds_on( void ): Turns all the 8 leds.
  • void EE_tc2x5_leds_off( void ): Turns off all the 8 leds.
  • void EE_tc2x5_turn_led(enum EE_tc2x5_led_id led_id, enum EE_tc2x5_led_status onoff): Turn the status of the led led_id (led IDs are collected in an enum in the form: EE_TRIBOARD_2X5_LED_{x} with {x}=[1..8]) on (onoff == EE_TRIBOARD_2X5_LED_ON) or off (onoff == EE_TRIBOARD_2X5_LED_OFF).
  • EE_BIT EE_tc2x5_read_button( void ): read external button value
  • void EE_tc2x5_button_irq_init( EE_TYPEISR2PRIO intvec ): Configure the external button has an interrupt source and tie it to intvec priority handler.
  • void EE_tc2x5_button_irq_clear_request( void ): Clear external button interrupt request.

External button have to be connected to pin P15.8 on TC275TE corrisponding to pin 71 of PERIPHERALS (Xx02,Xx02) connector of TriBoard TC2x5.

Debugger support

Lauterbach Trace32 and ORTI support

The ERIKA build system for Infineon AURIX produces a Lauterbach Trace32 configuration file (t32.cmm) and a ORTI description file (system.orti) inside the output directory. The ORTI file is produced only if ORTI support is enabled in the OIL configuration file. To launch the Trace32 debugger for singlecore projects, please issue t32mtc from the output directory. The ERIKA build system honors the T32SYS environment variable, if set (default T32SYS ?= C:/T32)

Muticore Projects

For multicore projects, the files mentioned above are produced inside the core directories for each core, and a startup script (named tc27x_mc_start.sh under cygwin and linux or tc27x_mc_start.bat for Windows cmd) is produced in the output directory. If the build targeted AURIX Flash (the default behaviour), before executing debug you need to flash multicore images. To do that simply execute tc27x_mc_flash.sh (under Linux or Cygwin environment) or tc27x_mc_flash.bat (in Windows environment) scripts.

To run the debugger, please issue tc27x_mc_start.[sh,bat] from the output directory. The script creates up to three instances of the debugger, one for each core. Slave cores cannot be started until master core (Core 0) enable them, thing that is made early in start-up code. Nonetheless, the startup script loads all the code and the debug symbols for each core and each debugger instances. The barrier inside StartOS() comes handy to synchronize the code running on the different cores.

N.B. To let know the build system for wich Trace32 architecture (windows, windows64, pc_linux ecc.) it have to generate script for, set T32ARCH environment variable/pass it as make parameter with the right value (default T32ARCH ?= windows64).

iSYSTEM winIDEA support

ERIKA Enterprise is supported by iSYSTEM winIDEA. For more information about the ERIKA Enterprise support please chek the iSYSTEM - Erika webpage.

For applications, based on an AUTOSAR/OSEK compliant OS such as ERIKA Enterprise, the OSEK Run-Time Interface (ORTI) file is a method for describing the structure of the RTOS to the debugger. By reading in the ORTI file generated by the RTDRUID when building an ERIKA-based application, the winIDEA debugger becomes ERIKA Enterprise OS-aware.

ERIKA Enterprise OS awareness provides the following features:

- Display of OS Resources and Status

- Run-Time Analysis (Profiler Timeline) of Tasks and Interrupts (ISR Category 1 and 2)

- Analysis of CPU Utilization (Profiler Statistics) of Tasks and Interrupts (ISR Category 1 and 2)

Multicore Autosar OS Support

ERIKA support multicore environments a way before than the first Autosar Multicore OS Release. So it happened that the historical ERIKA multicore support addresses some of the same requiremets of AUTOSAR Multicore OS with a completely different policy.

In detail all the primitives calls on other core, with traditional implementation, called Remote Notifications or RN, are done asynchronously with a fire and forget policy. This approach reduce latency to minimun, but, on the other hand, don't give the opportunity to check for errors in the call site.

Autosar, that take into account a way more code consistency and reliability than efficiency, requires that all the primitives calls on others cores are synchronous giving the opportunity to caller code to correctly handle errors.

To implements Autosar OS requirements a completly new multicore dispatcher have been implemented. This have been called ERIKA Autosar RPC (Remote Procedure Call).

Both traditional and the new dispatcher are available for Infineon AURIX. You can switch between them in OIL with the REMOTENOTIFICATION oil field.

 REMOTENOTIFICATION = USE_RN;

Enable traditional Remote Notification dispatcher. It's the default value, so you don't need to write this.

 REMOTENOTIFICATION = USE_RPC;

Enable the new AUTOSAR Remote Procedure Call dispatcher.

Currently with the new RPC dispatcher you can execute on remote cores all the following primitives:

  • ActivateTask
  • ChainTask (Terminate local TASK but can schedule on remote core)
  • GetTaskState
  • SetEvent
  • GetAlarmBase
  • GetAlarm
  • SetRelAlarm
  • SetAbsAlarm
  • CancelAlarm
  • GetCounterValue
  • GetElapsedValue
  • ShutdownOS ( with new ShutdownAllCores API )

So in addition to TASK and EVENT primitives, the new muticore dispatcher support Counters and Alarms as required by Autosar specifications. Multicore awareness have been added to ShutdownOS to fulfill multicore AUTOSAR shutdown sequence.

New Multicore API

AUTOSAR expect a master/slaves multicore enviroment that naturally fit AURIX architecture. New API to start and stops slave cores specified in AUTOSAR v4.0 rev 3.0 have been added:

  • StartCore: This function starts the core specified by the parameter CoreID. The OUT parameter allows the caller to check whether the operation was successful or not. If a core is started by means of this function StartOS shall be called on the core. It is not supported to call this function after StartOS().
  • StartNonAutosarCore: The function starts the core specified by the parameter CoreID. It is allowed to call this function after StartOS(). The OUT parameter allows the caller to check whether the operation was successful or not. It is not allowed to call StartOS on cores activated by StartNonAutosarCore.
  • GetNumberOfActivatedCores: returns the number of cores activated by the StartCore function. This function might be a macro
  • GetCoreID: The function returns a unique core identifier.
  • ShutdownAllCores : After this service the OS on all AUTOSAR cores is shut down. Allowed at TASK level and ISR level and also internally by the OS. The function will never return. The function will force other cores into a shutdown.

New Spinlocks API

Support for AUTOSAR Spinlock API have been added. OIL implementation have been extended to support Spinlocks configuration.

There's two spinlocks behaviour modes: SINGLE and ORDERED. If SINGLE mode is configured a core get an error if, holding a spinlock, try to get access to another one. If ORDERED mode is configurated the core get this error only if it try to get spinlocks out of the order (see AUTOSAR documentation paragraph 7.9.29 The spinlock mechanism).

To configure spinlocks in SINGLE mode, just don't provide any order:

 SPINLOCK spinlock_1 {  };
 SPINLOCK spinlock_2 {  };
 SPINLOCK spinlock_3 {  };

To configure spinlocks in ORDERED mode provide a complete ordering:

 SPINLOCK spinlock_1 { NEXT_SPINLOCK=spinlock_2; };
 SPINLOCK spinlock_2 { NEXT_SPINLOCK=spinlock_3; };
 SPINLOCK spinlock_3 {  };


Provided AUTOSAR API to interact with Spinlocks are:

  • GetSpinlock : Tries to occupy a spin-lock variable. If the function returns, either the lock is successfully taken or an error has occurred. The spinlock mechanism is an active polling mechanism. The function does not cause a de-scheduling.
  • ReleaseSpinlock: Releases a spinlock variable that was occupied before. Before terminating a TASK all spinlock variables that have been occupied with GetSpinlock() shall be released. Before calling WaitEVENT all Spinlocks shall be released.
  • TryToGetSpinlock: Has the same functionality as GetSpinlock with the difference that if the spinlock is already occupied by a TASK on a different core the function sets the OUT parameter to TRYTOGETSPINLOCK_NOSUCCESS and returns with E_OK.

Two spinlocks implementation are provided. The trivial one that does not guarentee upper bond wait, but that can implement the try-to-get behaviour. And the queued one that prevent from starvation queuing requests, but cannot implement try-to-get behaviour (TryToGetSpinlock API is mapped into GetSpinlock).

To enable queued spinlocks just add to the OIL:

 SPINLOCKS = QUEUED;

Additional Notes

Please note that the AURIX port has not been yet completely released on this website. If you need more information about the development status of this port please contact Evidence Srl.

Personal tools