Native POSIX execution (native_posix)

Overview

Using this board, a Zephyr application can be compiled together with the Zephyr kernel, creating a normal console executable that runs as a native application on the host OS, without emulation. Instead, you use native host tools for compiling, debugging, and analyzing your Zephyr application, eliminating the need for architecture-specific target hardware in the early phases of development.

This board provides a few peripherals such as an Ethernet driver and UART. See Peripherals for more information.

Host system dependencies

This port is designed to run in POSIX compatible operating systems, but it has only been tested on Linux.

Note

You must have the 32-bit C library installed in your system (in Ubuntu 16.04 install the gcc-multilib package)

Note

This port will not work in Windows Subsystem for Linux (WSL) because WSL does not support native 32-bit binaries.

Important limitations

The underlying assumptions behind this port set some limitations on what can and cannot be done. These limitations are due to the code executing natively in the host CPU without any instrumentation or means to interrupt it unless the simulated CPU is sleeping.

You can imagine the code executes in a simulated CPU which runs at an infinitely fast clock: No time passes while the CPU is running. Therefore interrupts, including timer interrupts, will not arrive while code executes, except immediately after the SW enables or unmasks them if they were pending.

This behavior is intentional, as it provides a deterministic environment to develop and debug. For more information please see the Rationale for this port and Architecture sections

Therefore these limitations apply:

  • There can not be busy wait loops in the application code that wait for something to happen without letting the CPU sleep. If busy wait loops do exist, they will behave as infinite loops and will stall the execution. For example, the following busy wait loop code, which could be interrupted on actual hardware, will stall the execution of all threads, kernel, and HW models:

    while (1){}
    

    Similarly the following code where we expect condition to be updated by an interrupt handler or another thread, will also stall the application when compiled for this port.

    volatile condition = true;
    while (condition){}
    
  • Code that depends on its own execution speed will normally not work as expected. For example, code such as shown below, will likely not work as expected:

    peripheral_x->run = true;
    
    /* Wait for a number of CPU cycles */
    for (int i = 0; i < 100; i++) NOP;
    
    /* We expect the peripheral done and ready to do something else */
    
  • This port is not meant to, and could not possibly help debug races between HW and SW, or similar timing related issues.

  • You may not use hard coded memory addresses because there is no I/O or MMU emulation.

Working around these limitations

If a busy wait loop exists, it will become evident as the application will be stalled in it. To find the loop, you can run the binary in a debugger and pause it after the execution is stuck; it will be paused in some part of that loop.

The best solution is to remove that busy wait loop, and instead use an appropriate kernel primitive to synchronize your threads. Note that busy wait loops are in general a bad coding practice as they keep the CPU executing and consuming power.

If removing the busy loop is really not an option, you may add a conditionally compiled call to k_cpu_idle() if you are waiting for an interrupt, or a call to k_busy_wait() with some small delay in microseconds. In the previous example, modifying the code as follows would work:

volatile condition = true;
while (condition) {
     #if defined(CONFIG_ARCH_POSIX)
             k_cpu_idle();
     #endif
}

How to use it

Compiling

Specify the native_posix board target to build a native POSIX application:

cd $ZEPHYR_BASE/samples/hello_world
mkdir build && cd build
cmake -GNinja -DBOARD=native_posix ..
ninja

Running

The result of the compilation is an executable (zephyr.exe) placed in the zephyr/ subdirectory of the build folder. Run the zephyr.exe executable as you would any other Linux console application.

$ zephyr/zephyr.exe
# Press Ctrl+C to exit

This executable accepts several command line options depending on the compilation configuration. You can run it with the --help command line switch to get a list of available options:

$ zephyr/zephyr.exe --help

Note that the Zephyr kernel does not actually exit once the application is finished. It simply goes into the idle loop forever. Therefore you must stop the application manually (Ctrl+C in Linux).

Application tests using the ztest framework will exit after all tests have completed.

If you want your application to gracefully finish when it reaches some point, you may add a conditionally compiled (CONFIG_ARCH_POSIX) call to posix_exit(int status) at that point.

Debugging

Since the Zephyr executable is a native application, it can be debugged and instrumented as any other native program. The program is compiled with debug information, so it can be run directly in, for example, gdb or instrumented with valgrind.

Because the execution of your Zephyr application is normally deterministic (there are no asynchronous or random components), you can execute the code multiple times and get the exact same result. Instrumenting the code does not affect its execution.

To ease debugging you may want to compile your code without optimizations (e.g., -O0) by setting CONFIG_NO_OPTIMIZATIONS.

Address Sanitizer (ASan)

You can also build Zephyr with Address Sanitizer. To do this, set CONFIG_ASAN, for example, in the application project file, or in the cmake command line invocation.

Note that you will need the ASan library installed in your system. In Debian/Ubuntu this is libasan1.

Rationale for this port

The main intents of this port are:

  • Allow functional debugging, instrumentation and analysis of the code with native tooling.
  • Allow functional regression testing, and simulations in which we have the full functionality of the code.
  • Run tests fast: several minutes of simulated time per wall time second.
  • Possibility to connect to external tools which may be able to run much faster or much slower than real time.
  • Deterministic, repeatable runs: There must not be any randomness or indeterminism (unless host peripherals are used). The result must not be affected by:
    • Debugging or instrumenting the code.
    • Pausing in a breakpoint and continuing later.
    • The host computer performance or its load.

The aim of this port is not to debug HW/SW races, missed HW programming deadlines, or issues in which an interrupt comes when it was not expected. Normally those would be debugged with a cycle accurate Instruction Set Simulator (ISS) or with a development board.

Comparison with other options

This port does not try to replace cycle accurate instruction set simulators (ISS), development boards, or QEMU, but to complement them. This port’s main aim is to meet the targets described in the previous Rationale for this port section.

Comparison of different debugging targets

Comparison of different debugging options. Note that realism has many dimensions: Having the real memory map or emulating the exact time an instruction executes is just some of it; Emulating peripherals accurately is another side.

This native port compiles your code directly to x86, with no instrumentation or monitoring code. Your code executes directly in the host CPU. That is, your code executes just as fast as it possibly can.

Simulated time is normally decoupled from real host time. The problem of how to emulate the instruction execution speed is solved by assuming that code executes in zero simulated time.

There is no I/O or MMU emulation. If you try to access memory through hardcoded addresses your binary will simply segfault. The drivers and HW models for this architecture will hide this from the application developers when it relates to those peripherals. In general this port is not meant to help developing low level drivers for target HW. But for developing application code.

Your code can be debugged, instrumented, or analyzed with all normal native development tools just like any other Linux application.

Execution is fully reproducible, you can pause it without side-effects.

How does this port compare to QEMU:

With QEMU you compile your image targeting the board which is closer to your desired board. For example an ARM based one. QEMU emulates the real memory layout of the board, loads the compiled binary and through instructions translation executes that ARM targeted binary on the host CPU. Depending on configuration, QEMU also provides models of some peripherals and, in some cases, can expose host HW as emulated target peripherals.

QEMU cannot provide any emulation of execution speed. It simply executes code as fast as it can, and lets the host CPU speed determine the emulated CPU speed. This produces highly indeterministic behavior, as the execution speed depends on the host system performance and its load.

As instructions are translated to the host architecture, and the target CPU and MMU are emulated, there is a performance penalty.

You can connect gdb to QEMU, but have few other instrumentation abilities.

Execution is not reproducible. Some bugs may be triggered only in some runs depending on the computer and its load.

How does this port compare to an ISS:

With a cycle accurate instruction set simulator you compile targeting either your real CPU/platform or a close enough relative. The memory layout is modeled and some or all peripherals too.

The simulator loads your binary, slowly interprets each instruction, and accounts for the time each instruction takes. Time is simulated and is fully decoupled from real time. Simulations are on the order of 10 to 100 times slower than real time.

Some instruction set simulators work with gdb, and may provide some extra tools for analyzing your code.

Execution is fully reproducible. You can normally pause your execution without side-effects.

Architecture

Zephyr layering in native build

Zephyr layering when built against an embedded target (left), and targeting the native_posix board (right)

This board is based on the POSIX architecture port of Zephyr. In this architecture each Zephyr thread is mapped to one POSIX pthread, but only one of these pthreads executes at a time. This architecture provides the same interface to the Kernel as other architectures and is therefore transparent for the application.

This board does not try to emulate any particular embedded CPU or SOC. The code is compiled natively for the host x86 system, as a 32-bit binary assuming pointer and integer types are 32-bits wide.

To ensure determinism when the Zephyr code is running, and to ease application debugging, the board uses a different time than real time: simulated time. This simulated time is, in principle, not linked to the host time.

The Zephyr application sees the code executing as if the CPU were running at an infinitely fast clock, and fully decoupled from the underlying host CPU speed. No simulated time passes while the application or kernel code execute.

The CPU boot is emulated by creating the Zephyr initialization thread and letting it run. This in turn may spawn more Zephyr threads. Eventually the SW will run to completion, that is, it will set the CPU back to sleep.

At this point, control is transferred back to the HW models and the simulation time can be advanced.

When the HW models raise an interrupt, the CPU wakes back up, the interrupt is handled, the SW runs until completion again, and control is transferred back to the HW models, all in zero simulated time.

If the SW unmasks a pending interrupt while running, or triggers a SW interrupt, the interrupt controller may raise the interrupt immediately depending on interrupt priorities, masking, and locking state.

About time in native_posix

Normally simulated time runs fully decoupled from the real host time and as fast as the host compute power would allow. This is desirable when running in a debugger or testing in batch, but not if interacting with external interfaces based on the real host time.

The Zephyr kernel is only aware of the simulated time as provided by the HW models. Therefore any normal Zephyr thread will also know only about simulated time.

The only link between the simulated time and the real/host time, if any, is created by the clock and timer model.

This model can be configured to slow down the execution of native_posix to real time. You can do this with the --rt and --no-rt options from the command line. The default behavior is set with CONFIG_NATIVE_POSIX_SLOWDOWN_TO_REAL_TIME. Note that all this model does is wait before raising the next system tick interrupt until the corresponding real/host time. If, for some reason, native_posix runs slower than real time, all this model can do is “catch up” as soon as possible by not delaying the following ticks. So if the host load is too high, or you are running in a debugger, you will see simulated time lagging behind the real host time. This solution ensures that normal runs are still deterministic while providing an illusion of real timeness to the observer.

When locked to real time, simulated time can also be set to run faster or slower than real time. This can be controlled with the --rt-ratio=<ratio> and -rt-drift=<drift> command line options. Note that both of these options control the same underlying mechanism, and that drift is by definition equal to ratio - 1. It is also possible to adjust this clock speed on the fly with native_rtc_adjust_clock().

In this way if, for example, --rt-ratio=2 is given, the simulated time will advance at twice the real time speed. Similarly if --rt-drift=-100e-6 is given, the simulated time will progress 100ppm slower than real time. Note that the these 2 options have no meaning when running in non real-time mode.

How simulated time and real time relate to each other

Simulated time (st) can be calculated from real time (rt) as

st = (rt - last_rt) * ratio + last_st

And vice-versa:

rt = (st - last_st) / ratio + last_rt

Where last_rt and last_st are respectively the real time and the simulated time when the last clock ratio adjustment took place.

All times are kept in microseconds.

Peripherals

The following peripherals are currently provided with this board:

Interrupt controller:
A simple yet generic interrupt controller is provided. It can nest interrupts and provides interrupt priorities. Interrupts can be individually masked or unmasked. SW interrupts are also supported.
Clock, timer and system tick model

This model provides the system tick timer. By default CONFIG_SYS_CLOCK_TICKS_PER_SEC configures it to tick every 10ms.

This peripheral driver also provides the needed functionality for this architecture-specific k_busy_wait().

Please refer to the section About time in native_posix for more information.

UART
An optional UART driver can be compiled with native_posix. For more information refer to the section UART.
Console driver:

A console driver is provided which by default is configured to:

  • Redirect any printk() write to the native host application’s stdout.
  • If the legacy shell is compiled with your application, redirect the native application’s stdin to the legacy shell for any input. For more information refer to the section Legacy shell support.
Real time clock

The real time clock model provides a model of a constantly powered clock. By default this is initialized to the host time at boot.

This RTC can also be set to start from time 0 with the --rtc-reset command line option.

It is possible to offset the RTC clock value at boot with the --rtc-offset=<offset> option, or to adjust it dynamically with the function native_rtc_offset().

After start, this RTC advances with the simulated time, and is therefore affected by the simulated time speed ratio. See About time in native_posix for more information.

The time can be queried with the functions native_rtc_gettime_us() and native_rtc_gettime(). Both accept as parameter the clock source:

  • RTC_CLOCK_BOOT: It counts the simulated time passed since boot. It is not subject to offset adjustments
  • RTC_CLOCK_REALTIME: RTC persistent time. It is affected by offset adjustments.
  • RTC_CLOCK_PSEUDOHOSTREALTIME: A version of the real host time, as if the host was also affected by the clock speed ratio and offset adjustments performed to the simulated clock and this RTC. Normally this value will be a couple of hundredths of microseconds ahead of the simulated time, depending on the host execution speed. This clock source should be used with care, as depending on the actual execution speed of native_posix and the host load, it may return a value considerably ahead of the simulated time.
Entropy device:
An entropy device based on the host random() API. This device will generate the same sequence of random numbers if initialized with the same random seed. You can change this random seed value by using the command line option: --seed=<random_seed> where the value specified is a 32-bit integer such as 97229 (decimal), 0x17BCD (hex), or 0275715 (octal).
Ethernet driver:

A simple TAP based ethernet driver is provided. The driver will create a zeth network interface to the host system. One can communicate with Zephyr via this network interface. Multiple TAP based network interfaces can be created if needed. The IP address configuration can be specified for each network interface instance. See CONFIG_ETH_NATIVE_POSIX_SETUP_SCRIPT option for more details. The Native Posix Ethernet sample app provides some use examples and more information about this driver configuration.

Note that this device can only be used with Linux hosts, and that the user needs elevated permissions.

Bluetooth controller:
It’s possible to use the host’s Bluetooth adapter as a Bluetooth controller for Zephyr. To do this the HCI device needs to be passed as a command line option to zephyr.exe. For example, to use hci0, use sudo zephyr.exe --bt-dev=hci0. Using the device requires root privileges (or the CAP_NET_ADMIN POSIX capability, to be exact) so zephyr.exe needs to be run through sudo. The chosen HCI device must be powered down and support Bluetooth Low Energy (i.e. support the Bluetooth specification version 4.0 or greater).
Display driver:

A display driver is provided that creates a window on the host machine to render display content.

This driver requires a 32-bit version of the SDL2 library on the host machine and pkg-config settings to correctly pickup the SDL2 install path and compiler flags.

On a Ubuntu 18.04 host system, for example, install the pkg-config and libsdl2-dev:i386 packages, and configure the pkg-config search path with these commands:

$ sudo apt-get install pkg-config libsdl2-dev:i386
$ export PKG_CONFIG_PATH=/usr/lib/i386-linux-gnu/pkgconfig

UART

This driver can be configured to either create and link the UART to a new pseudoterminal (i.e. /dev/pts<nbr>), or to map the UART input and output to the executable’s stdin and stdout. This is chosen by selecting either CONFIG_NATIVE_UART_0_ON_OWN_PTY or CONFIG_NATIVE_UART_0_ON_STDINOUT For interactive use with the Shell, choose the first (OWN_PTY) option. The second (STDINOUT) option can be used with the shell for automated testing, such as when piping other processes’ output to control it. This is because the shell subsystem expects access to a raw terminal, which (by default) a normal Linux terminal is not.

When CONFIG_NATIVE_UART_0_ON_OWN_PTY is chosen, the name of the newly created UART pseudo-terminal will be displayed in the console. If you want to interact with it manually, you should attach a terminal emulator to it. This can be done, for example with the command:

$ xterm -e screen /dev/<ttyn> &

where /dev/<ttyn> should be replaced with the actual TTY device.

You may also chose to automatically attach a terminal emulator to it by passing the command line option -attach_uart to the executable. The command used for attaching to the new shell can be set with the command line option -attach_uart_cmd=<"cmd">. Where the default command is given by CONFIG_NATIVE_UART_AUTOATTACH_DEFAULT_CMD. Note that the default command assumes both xterm and screen are installed in the system.

Legacy shell support

When the legacy shell subsystem is compiled with your application, the native standard input (stdin) will be redirected to the shell. You may use this shell interactively through the console, by piping another process output to it, or by feeding it a file.

When using it interactively you may want to select the option CONFIG_NATIVE_POSIX_SLOWDOWN_TO_REAL_TIME.

When feeding stdin from a pipe or file, the console driver will ensure reproducibility between runs of the process:

  • The execution of the process will be stalled while waiting for new stdin data to be ready.
  • Commands will be fed to the shell as fast as the shell can process them. To allow controlling the flow of commands to the shell, you may use the driver directive !wait <ms>.
  • When the file ends, or the pipe is closed the driver will stop attempting to read it.

Driver directives

The console driver understands a set of special commands: driver directives. These directives are captured by the console driver itself and are not forwarded to the shell. These directives are:

  • !wait <ms>: When received, the driver will pause feeding commands to the shell for <ms> milliseconds.
  • !quit: When received the driver will cause the application to gracefully exit by calling posix_exit().

Use example

For example, after you have built an application with the legacy shell, you can feed it the following set of commands through a pipe:

echo -e \
'select kernel\nuptime\n!wait 500\nuptime\n!wait 1000\nuptime\n!quit' \
| zephyr/zephyr.exe