[Marek Gibney] poses an interesting puzzle. What does the following bash command line print?
( echo red ; echo green 1>&2 ) | echo blue
You’d like to think it prints three lines: red, green, and blue. But would you be surprised to find out that it can sometimes output “blue green” and sometimes just output blue. The first surprise is that it isn’t deterministic. But the second thing that is surprising is the sometimes the entire left-hand part of the line doesn’t do anything. [Chris Siebenmann] did the analysis and explains what’s going on in a recent blog post.
Before you click the link or read further, you might want to see if you can deduce what’s going on. Give up? Here’s a hint: Part of the solution hinges on the fact that echo is built into the shell.
There are a lot of ARM microcontrollers out there, and the parts from ST are featured prominently is the high-power builds we’re seeing. The STM32F4 and ~F7 are powerhouses with great support, and the STM32F0 and the other younger children of the family make for very good, low-power microcontrollers. Now, the STM32 family is getting a big brother. It runs Linux. It’s two ARM Cortex-A7 cores and one M4 core on the same chip. The STM32MP1 is the chip you want if you still can’t figure out how to waste computing cycles by blinking LEDs.
Block diagram of the STM32MP157 Image: ST
First, that Linux support. The STM32MP157C was mainlined into Linux last summer, and there is support for Android. So yes, this chip can run Linux. There is an optional 3D GPU in this family, a MIPI-DSI controller, support for HDMI-CEC, USB 2.0, and 10/100M or Gigabit Ethernet. This brings us the inevitable question of whether you can build a Raspberry Pi clone with these parts. Maybe, champ, but if you’re asking that question it’s probably not you that’s going to build one. It looks as if this chip is designed for phones, set-top boxes, and smart TVs. That doesn’t preclude a single board computer, but the biggest problem there is maintaining software support anyway.
The chip family in question all come with dual ARM Cortex-A7 processors running at a nominal 650MHz. There’s also a Cortex-M4 running at 209MHz, and the ST literature suggests that engineers are already running Linux on the A7 and an RTOS on the M4. This chip will need external memory, but DDR3 / DDR3L / LPDDR2 / LPDDR3 are supported.
This chip is only announced right now, you can’t get it on Mouser or Digikey yet, and there’s no information on pricing. However, there are two development boards available, the Evaluation board, which features 1 GB of DDR3L, 128 MB of Flash, and an 8 GB eMMC. There’s a 5.5″ display, and enough connectors to make your heart flutter. The Discovery board is a bit more cut down, and comes with a 4″ 480×800 LCD, WiFi, Bluetooth LE, and of course it comes with GPIO expansion connectors for an Arduino and Raspberry Pi. The Discovery Board is not available at this time, but it will sell for $99 USD.
Although many people think of Linux-based operating systems as graphical, really that GUI is just another application running over the bare operating system. Power users, remote administrators, and people running underpowered computers like a Raspberry Pi have a tendency to do more with command line tools. [Igor] did a FOSDEM19 presentation you can see below about how he’s providing web-like services to the command line using web servers and curl as a client.
This is subtly different from just accessing an ordinary web server via curl. The output is meant for display in the terminal. Of course, you could also hit them with a web browser, if you wanted — at least, for some of them. [Igor’s] tools include a weather reporter, a QR code encoder, information and graphs for currency and cybercurrency rates, and an online help system for programmers.
It seems like only yesterday that the Linksys WRT54G and the various open source firmware replacements for it were the pinnacle of home router hacking. But like everything else, routers have gotten smaller and faster over the last few years. The software we run on them has also gotten more advanced, and at this point we’ve got routers that you could use as a light duty Linux desktop in a pinch.
But even with no shortage of pocket-sized Linux devices in our lives, the GL-USB150 “Microrouter” that [Mason Taylor] recently brought to our attention is hard to ignore. Inside this USB flash drive sized router is a 400 MHz Qualcomm QCA9331 SoC, 64 MB of RAM, and a healthy 16 MB of storage; all for around $20 USD. Oh, and did we mention it comes with OpenWRT pre-installed? Just plug it in, and you’ve got a tiny WiFi enabled Linux computer ready to do your bidding.
On his blog [Mason] gives a quick rundown on how to get started with the GL-USB150, and details some of the experiments he’s been doing with it as part of his security research, such as using the device as a remote source for Wireshark running on his desktop. He explains that the diminutive router works just fine when plugged into a USB battery bank, offering a very discreet way to deploy a small Linux box wherever you may need it. But when plugged into a computer, things get really interesting.
If you plug the GL-USB150 into a computer, it shows up to the operating system as a USB Ethernet adapter and can be used as the primary Internet connection. All of the traffic from the computer will then be routed through the device to whatever link to the Internet its been configured to use. Depending on how you look at it, this could be extremely useful or extremely dangerous.
For one, it means that something that looks all the world like a normal USB flash drive could be covertly plugged into a computer and become a “wiretap” through which all of the network traffic is routed. That’s the bad news. On the flip side, it also means you could configure the GL-USB150 as a secure endpoint that lets you quickly and easily funnel all the computer’s traffic through a VPN or Tor without any additional setup.
In an earlier installment of Linux Fu, I mentioned how you can use inotifywait to efficiently watch for file system changes. The comments had a lot of alternative ways to do the same job, which is great. But there was one very easy-to-use tool that didn’t show up, so I wanted to talk about it. That tool is entr. It isn’t as versatile, but it is easy to use and covers a lot of common use cases where you want some action to occur when a file changes.
PostmarketOS began work on a real Linux distribution for Android phones just over 600 days ago. They recently blogged about the state of the project and ensured us that the project is definitely not dead.
PostmarketOS’ overarching goal remains a 10 year life-cycle for smartphones. We previously covered the project on Hackaday to give an introduction. Today, we’ll concern ourselves with the progress the PostmarketOS team has made.
The team admits that they’re stuck in the proof-of-concept phase, and need to break out of it. This has required foundational changes to the operating system to enable development across a wide variety of devices and processor architectures. There’s now a binary package repository powered by builds.sr.ht which will allow users to install packages for their specific device.
Other updates include fixing support for the Nexus 5 and Raspberry Pi Zero, creating support for open source hardware devices including the Pine A64-LTS and Purism Librem 5. PostmarketOS now boots on a total of 112 different devices.
We’re excited to see the PostmarketOS project making progress. With the widespread move to mobile devices, users lose control over their computing devices. PostmarketOS gives us the ability to run code that we can read and modify on these devices. It’s no small feat though. Supporting the wide variety of custom hardware in mobile devices requires a lot of effort.
While it may be a while before PostmarketOS is your daily driver, the project is well suited to building task-specific devices that require connectivity, a touch screen, and a battery. We bet a lot of Hackaday readers have a junk drawer phone that could become a project with the help of PostmarketOS.
One of the great things about Linux and similar operating systems is they are configurable. If you don’t like something, there’s a great chance you can change it easily with a few entries in a file somewhere. For example, take bash — a very popular shell by any measure. If you want a different style of command line editing, there’s an option. You want the tab key to match files regardless of case? Another option. Usually, these are set in one of your so-called profile files like .bashrc in your home directory.
As long as you are sitting in front of your single computer working, this is great. You customize your .bashrc and other files to your heart’s content and then you work in an environment that acts the way you want it to. The problem is when you have a lot of computers. Maybe you have a web server, a desktop, a firewall machine, and a few dozen Raspberry Pi computers. How do you keep all the configurations the same? Then once they are the same, how do you keep them up to date?
