Scientific Computing

Windows natively or with Windows Subsystem for Linux supports USB-serial devices. The device must be recognized in Windows Device Manager.

Plug the USB-serial adapter into the Windows PC. Look in Windows Device Manager under Ports to see the COM port number. See troubleshooting notes below if it doesn’t show there.

The device must show in Windows Device Manager → USB Devices → Ports.

The serial device baud rate must be consistent between device and PC. If the baud rate is incorrect, either no text or garbled text will be seen.

Connect to serial devices using PuTTY or use WSL.

Windows Subsystem for Linux USB-serial

Command line screen or PuTTY can be used in WSL.

Configure USB-serial adapters in WSL by adding the WSL username to “dialout” group:

adduser $(whoami) dialout

If so, try in WSL terminal, (assuming device is on COM 5 for this example):

chmod 666 /dev/ttyS5

The usbipd program can help debug and connect devices in Windows and WSL.

Troubleshooting

The COM port number can change upon plugging in the same device, especially if replugging into a different physical USB port on the PC. If plugging in a different unit of the same type of device, it may likely also get a new COM port number. Distinct devices of the same model, even if sequentially plugged into the same USB port may get different COM port numbers.

If the device doesn’t show up in Device Manager → Ports, see if it was mistakenly enabled as a Mouse or Human Interface Device. If so, unplug and replug your device.

If it still fails to show up as a Port, instead showing up as mouse or HID, try right-clicking and Disabling the device and unplug/plug it once more.

If it still fails, maybe the Windows device driver is missing. Try the device in a native Linux PC and see if the device works there.

PuTTY is a terminal emulator that also works well for serial port connections. Currently, PuTTY can’t do file transfer over serial links with XMODEM, YMODEM, or ZMODEM. To do file transfer over serial links, use minicom.

PuTTY is available on Linux like:

apt install putty

macOS PuTTY install

On macOS, PuTTY is obtained by:

brew install putty

If PuTTY doesn’t link on macOS due to “pterm” name conflict, create an alias by finding the putty binary like:

alias putty="$(brew --prefix putty)/bin/putty"

Windows PuTTY install

On Windows, PuTTY can be installed with winget:

winget install PuTTY.PuTTY

PuTTY connection

Start PuTTY in Terminal command line or from the Windows Start Menu.

putty

which opens the PuTTY GUI.

Select the serial line (say /dev/ttyUSB0 for Linux, COM5 for Windows) and baud rate (say 115200).

Determine the serial port of a USB-serial adapter like:

• Linux: ls /dev/ttyUSB*
• macOS: ls /dev/tty.usbserial*

The screen terminal multiplexer program is useful for connecting to serial port devices.

On Linux, including WSL:

apt install screen

configure USB-serial adapters in Linux or WSL by adding the username to “dialout” group:

adduser $(whoami) dialout

On Homebrew:

brew install screen

A serial port connection is established at 115200 baud like:

screen /dev/ttyUSB0 115200

List the USB-serial adapter ports:

• Linux: ls /dev/ttyUSB*
• macOS: ls /dev/tty.usbserial*

Alternative: minicom is a serial terminal program with file transfer capability.

Matlab redirects the I/O streams of binaries compiled with GFortran by setting runtime environment variables for streams “stdout”, “stderr”, and “stdin”. This can lead to unexpected behavior of programs run from Matlab with built-in functions like system() as well as external language interfaces run from Matlab scripts like Java ProcessBuilder, Python subprocess, etc.

Disable I/O stream capture before the external process is started in the Matlab script:

setenv("GFORTRAN_STDOUT_UNIT", "6");

setenv("GFORTRAN_STDERR_UNIT", "0");

setenv("GFORTRAN_STDIN_UNIT", "5");

Runtime environment variables affect GCC programs including C and C++ programs.

If given a link that is suspect, or troubleshooting behavior of a website that is having trouble doing a redirect, check the HTTP header for a redirect. Curl does this with the –head option.

curl --head example.invalid

This will return the HTTP header, which will show if there is a redirect, and where it is redirecting to.

HTTP/2 301
...
Location: https://www.example.invalid

Some web servers behave differently to a HEAD request than a GET request. To see the behavior of a GET request, use the –location option to follow redirects.

curl --location --silent --dump-header - -o /dev/null example.invalid

--silent

suppresses the progress bar

--dump-header -

dumps the header to standard out

-o /dev/null

discards the body

The RISC-V based Raspberry Pi Pico microcontroller is commonly used for analog and digital control, optionally with WiFi or Bluetooth wireless connectivity. While the full Raspberry Pi boards have a general purpose ARM CPU with enough storage and RAM capable of being used as a PC, the Pico is not commonly used for general purpose computing.

The multi-core CPU of the Pi Pico is capable of running parallel multi-threaded programs. The Pico Cortex-M or RISC-V Hazard3 CPUs lacks a memory management unit (MMU). Without an MMU, the Pico is not so effective at running a preemptive multitasking OS like a full Linux distribution. However, it is possible to run a minimal Linux kernel and user space. Provided a Pico board with enough memory, it’s possible to run a minimal Linux distribution on the Pico. The analog and digital I/O pins can be used for VGA output from the Pico.

The Jetson Nano board has gone through several generations. Older Jetson Nano boards may be stuck on an older unsupported OS. If compatible with the specific Nano version hardware, the NVIDIA Jetpack SDK may be used to install a newer OS. Select a Jetson container suitable for the desired task and hardware.

Operating systems require a certain minimum ISA microarchitecture level to run for each CPU vendor family. The levels are arbitrarily defined by the vendor as a collection of features, for example AVX2 or AVX512.

A C++-based cpuid library can detect across operating systems if the Intel CPU supports a certain ISA level.

On macOS, Terminal commands can be specified to run as a Rosetta CPU arch translated binary using the arch command. This is useful for running or testing x86_64 software on Apple Silicon Macs. See “man arch” for more details.

arch -x86_64 <command>

For example to run cmake as an x86_64 binary on an Apple Silicon Mac:

arch -x86_64 cmake -B build

The CMake variable CMAKE_SYSTEM_PROCESSOR or simply

arch -x86_64 uname -m

can be used to check the architecture of the current process.

Universal or x86_64 binaries required

When using

arch -x86_64 <command>

to run a program, the program must be a universal binary or an x86_64 binary. If not, the error message will be similar to:

arch: posix_spawnp: <command>: Bad CPU type in executable

For programs like CMake that are designed for cross-compiling, command CMake to build for x86_64 by setting the CMAKE_OSX_ARCHITECTURES variable to x86_64:

cmake -B build -DCMAKE_OSX_ARCHITECTURES=x86_64

cmake --build build

The Fortran standard does not define a specific Fortran module file format. Each compiler vendor has a unique incompatible Fortran module file format. Fortran module files are not portable between different compilers or even different versions of the same compiler.

The per-compiler examples below assume Fortran source file “example.f90”:

module dummy

real, parameter :: pi = 4*atan(1.0)

contains

real function tau()
  tau = 2*pi
end function tau

end module dummy

LLVM Flang

The LLVM Flang .mod files generated are legal Fortran syntax – they are text files. The .mod format gives the version number.

Build:

gfortran -c example.f90

This generates Fortran module file “dummy.mod” and object file “example.o”.

Get the .mod file header:

head -n1 dummy.mod

The output starts like:

!mod$ v1

GNU Fortran (GFortran)

The GFortran header version is defined in module.cc as variable “MOD_VERSION”. GNU Fortran “gfortran” .mod files are GZIP and is not documented

Build a Fortran module source file like:

gfortran -c example.f90

This generates Fortran module file “dummy.mod” and object file “example.o”.

Extract the .mod file header like:

gunzip -c dummy.mod | head -n1

The output starts like:

GFORTRAN module version ‘15’ created from …

If you get:

gzip: not in gzip format

the .mod file is probably from another compiler vendor e.g. Intel oneAPI.

Intel oneAPI (ifx)

Intel oneAPI .mod files are a proprietary binary format. It is possible to determine the version of the .mod file by using od to look at the first 2 bytes of the .mod file.

od -N4 -d dummy.mod

The first number is like “13” and is the module format version. This version may change over time as oneAPI internals change. The second number is the update version, which is fixed at “1”.

NVHPC and AOCC

NVIDIA HPC SDK (NVHPC) and AOCC compilers generate .mod files that are text files. The format for legacy Flang module files is distinct from LLVMFlang Fortran module files.

Create the .mod file like:

nvfortran -c example.f90

# or

flang -c example.f90

generates a text file, beginning with the version number.

head -n1 dummy.mod

the output is like:

V34 :0x24 dummy

Cray Fortran

By default, Cray Fortran stores uppercase DUMMY.mod filenames. This can be made lowercase with the ftn -ef flag. The Cray Fortran .mod format is proprietary, but the version number might be seen like:

ftn -c example.f90

head -n2 DUMMY.mod

Related: Fortran submodule file naming