BSD Now

DragonflyBSD 5.6 is out, OpenBSD Vulkan Support, bad utmp implementations in glibc and FreeBSD, OpenSSH protects itself against Side Channel attacks, ZFS vs OpenZFS, and more.

Somewhat surprisingly, OpenBSD has added the Vulkan library and ICD loader support as their newest port.

I recently released another version – 0.5.0 – of Dinit, the service manager / init system. There were a number of minor improvements, including to the build system (just running “make” or “gmake” should be enough on any of the systems which have a pre-defined configuration, no need to edit mconfig by hand), but the main features of the release were S6-compatible readiness notification, and support for updating the utmp database.

Last week, Damien Miller, a Google security researcher, and one of the popular OpenSSH and OpenBSD developers announced an update to the existing OpenSSH code that can help protect against the side-channel attacks that leak sensitive data from computer’s memory. This protection, Miller says, will protect the private keys residing in the RAM against Spectre, Meltdown, Rowhammer, and the latest RAMBleed attack.

You’ve probably heard us say a mix of “ZFS” and “OpenZFS” and an explanation is long-overdue.

OpenZFS-kmod port available, using blacklistd with NPF as fail2ban replacement, ZFS raidz expansion alpha preview 1, audio VU-meter increases CO2 footprint rant, XSAVE and compat32 kernel work for LLDB, where icons for modern X applications come from, and more.

blacklistd(8) provides an API that can be used by network daemons to communicate with a packet filter via a daemon to enforce opening and closing ports dynamically based on policy.

A couple months ago I noticed that the monitor on my workstation never power off anymore. Screensaver would go on, but DPMs (to do the poweroff) never kicked in.

Upstream describes LLDB as a next generation, high-performance debugger. It is built on top of LLVM/Clang toolchain, and features great integration with it. At the moment, it primarily supports debugging C, C++ and ObjC code, and there is interest in extending it to more languages.

If you have a traditional window manager like fvwm, one of the things it can do is iconify X windows so that they turn into icons on the root window (which would often be called the 'desktop'). Even modern desktop environments that don't iconify programs to the root window (or their desktop) may have per-program icons for running programs in their dock or taskbar. If your window manager or desktop environment can do this, you might reasonably wonder where those icons come from by default.

DragonFlyBSD's kernel optimizations pay off, differences between OpenBSD and Linux, NetBSD 2019 Google Summer of Code project list, Reducing that contention, fnaify 1.3 released, vmctl(8): CLI syntax changes, and things that Linux distributions should not do when packaging.

DragonFlyBSD lead developer Matthew Dillon has been working on a big VM rework in the name of performance and other kernel improvements recently. Here is a look at how those DragonFlyBSD 5.5-DEVELOPMENT improvements are paying off compared to DragonFlyBSD 5.4 as well as FreeBSD 12 and five Linux distribution releases. With Dillon using an AMD Ryzen Threadripper system, we used that too for this round of BSD vs. Linux performance benchmarks.

Maybe you have been reading recently about the release of OpenBSD 6.5 and wonder, "What are the differences between Linux and OpenBSD?"

We are very happy to announce The NetBSD Foundation Google Summer of Code 2019 projects:

The communiting bonding period - where students get in touch with mentors and community - started yesterday. The coding period will start from May 27 until August 19.

The opening keynote at EuroBSDCon 2016 predicted the future 10 years of BSDs. Amongst all the funny previsions, gnn@FreeBSD said that by 2026 OpenBSD will have its first implementation of SMP. Almost 3 years after this talk, that sounds like a plausible forecast... Why? Where are we? What can we do? Let's dive into the issue!

Most of OpenBSD's kernel still runs under a single lock, ze KERNEL_LOCK(). That includes most of the syscalls, most of the interrupt handlers and most of the fault handlers. Most of them, not all of them. Meaning we have collected & fixed bugs while setting up infrastructures and examples. Now this lock remains the principal responsible for the spin % you can observe in top(1) and systat(1).

In the past years, most of our efforts have been invested into the Network Stack. As I already mentioned it should be ready to be parallelized. However think we should now concentrate on removing the KERNEL_LOCK(), even if the code paths aren't performance critical.

This release finally addresses some of the problems that prevent simple running of several games.

Another blocker happens when the game expects to check the SteamAPI - either from a running Steam process, or a bundled steam_api library. OpenBSD 6.5-current now has steamworks-nosteam in ports, a stub library for Steamworks.NET that prevents games from crashing simply because an API function isn't found. The repo is here. fnaify now finds this library in /usr/local/share/steamstubs and uses it instead of the bundled (full) Steamworks.NET.dll.

The order of the arguments in the create, start, and stop commands of vmctl(8) has been changed to match a commonly expected style. Manual usage or scripting with vmctl must be adjusted to use the new syntax.

# vmctl create disk.qcow2 -s 50G

The new syntax specifies the command options before the argument:

# vmctl create -s 50G disk.qcow2

Right now I am a bit unhappy at Fedora for a specific packaging situation, so let me tell you a little story of what I, as a system administrator, would really like distributions to not do.

GPU passthrough on bhyve, confusion with used/free disk space on ZFS, OmniOS Community Edition, pfSense 2.4.4 Release p3, NetBSD 8.1 RC1, FreeNAS as your Server OS, and more.

Normally we cover news focused on KVM and sometimes Xen, but something very special has happened with their younger cousin in the BSD world, Bhyve.

All you have to do is add a line pointing the X server to the Bus ID of the passed card and the VM will boot, with acceleration and everything. He theorizes that this may not be possible on windows because of the way it looks for display devices, but it’s a solid start.

I use ZFS extensively. ZFS is my favorite file system. I write articles and give lectures about it. I work with it every day. In traditional file systems we use df(1) to determine free space on partitions. We can also use du(1) to count the size of the files in the directory. But it’s different on ZFS and this is the most confusing thing EVER. I always forget which tool reports what disk space usage! Every time somebody asks me, I need to google it. For this reason I decided to document it here - for myself - because if I can’t remember it at least I will not need to google it, as it will be on my blog, but maybe you will also benefit from this blog post if you have the same problem or you are starting your journey with ZFS.

The understanding of how ZFS is uses space and how to determine which value means what is a crucial thing. I hope thanks to this article I will finally remember it!

The OmniOS Community Edition Association is proud to announce the general availability of OmniOS - r151030.

We are pleased to announce the release of pfSense® software version 2.4.4-p3, now available for new installations and upgrades!

Due to the significant nature of the changes in 2.4.4 and later,

The NetBSD Project is pleased to announce NetBSD 8.1, the first update of the NetBSD 8 release branch. It represents a selected subset of fixes deemed important for security or stability reasons, as well as new features and enhancements.

Some highlights of the 8.1 release are:

What if you could have a server OS that had built in RAID, NAS and SAN functionality, and could manage packages, containers and VMs in a GUI? What if that server OS was also free to download and install? Wouldn’t that be kind of awesome? Wouldn’t that be FreeNAS?

FreeBSD 11.3-beta 1 is out, BSDCan 2019 recap, OpenIndiana 2019.04 is out, Overview of ZFS Pools in FreeNAS, why open source firmware is important for security, a new Opnsense release, wireguard on OpenBSD, and more.

We have released a new OpenIndiana Hipster snapshot 2019.04. The noticeable changes:

FreeNAS uses the OpenZFS (ZFS) file system, which handles both disk and volume management. ZFS offers RAID options mirror, stripe, and its own parity distribution called RAIDZ that functions like RAID5 on hardware RAID. The file system is extremely flexible and secure, with various drive combinations, checksums, snapshots, and replication all possible. For a deeper dive on ZFS technology, read the ZFS Primer section of the FreeNAS documentation.

SUGGEST LAYOUT attempts to balance usable capacity and redundancy by automatically choosing an ideal vdev layout for the number of available disks.

The goal of the root of trust should be to verify that the software installed in every component of the hardware is the software that was intended. This way you can know without a doubt and verify if hardware has been hacked. Since we have very little to no visibility into the code running in a lot of places in our hardware it is hard to do this. How do we really know that the firmware in a component is not vulnerable or that is doesn’t have any backdoors? Well we can’t. Not unless it was all open source.

This update addresses several privilege escalation issues in the access control implementation and new memory disclosure issues in Intel CPUs. We would like to thank Arnaud Cordier and Bill Marquette for the top-notch reports and coordination.

Earlier this week I imported a port for WireGuard into the OpenBSD ports tree. At the moment we have the userland daemon and the tools available. The in-kernel implementation is only available for Linux. At the time of writing there are packages available for -current.

WireGuard (cl)aims to be easier to setup and faster than OpenVPN and while I haven't been able to verify the latter, the first is certainly true...once you've figured it out. Most documentation out there is for Linux so I had to figure out the wireguardgo service and the tun parameters. But all in all, sure, it's easier. Especially the client configuration on iOS which I didn't cover here because it's essentially pkgadd libqrencode ; cat client.conf | qrencode -t ansiutf8, scan the code with the WireGuard app and you're good to go. What is particularly neat is that WireGuard on iOS supports Always-on.

Running AIX on QEMU on Linux on Windows, your NAS fleet with TrueCommand, Unleashed 1.3 is available, LLDB: CPU register inspection support extension, V7 Unix programs often not written as expected, and more.

YES it’s real!

Hundreds of thousands of FreeNAS and TrueNAS systems are deployed around the world, with many sites having dozens of systems. Managing multiple systems individually can be time-consuming. iXsystems has responded to the challenge by creating a “single pane of glass” application to simplify the scaling of data, drive management, and administration of iXsystems NAS platforms. We are proud to introduce TrueCommand.

This is the fourth release of Unleashed - an operating system fork of illumos. For more information about Unleashed itself and the download links, see our website.

Why did we decide to fork illumos? After all, there are already many illumos distributions available to choose from. We felt we could do better than any of them by taking a more aggressive stance toward compatibility and reducing cruft from code and community interactions alike.

Upstream describes LLDB as a next generation, high-performance debugger. It is built on top of LLVM/Clang toolchain, and features great integration with it. At the moment, it primarily supports debugging C, C++ and ObjC code, and there is interest in extending it to more languages.

My work continues with the two milestones from last month, plus a third that's closely related:

Yesterday I wrote that V7 ed read its terminal input in cooked mode a line at a time, which was an efficient, low-CPU design that was important on V7's small and low-power hardware. Then in comments, frankg pointed out that I was wrong about part of that, namely about how ed read its input.

Reading this section of the source code for ed taught me that it has an interesting, undocumented, and entirely characteristic little behavior. Officially, ed commands that have you enter new text have that new text terminate by a . on a line by itself:

In other words, it turns a single line with '.' into an EOF. The consequence of this is that if you type a real EOF at the start of a line, you get the same result, thus saving you one character (you use Control-D instead of '.' plus newline). This is very V7 Unix behavior, including the lack of documentation.

This is also a natural behavior in one sense. A proper program has to react to EOF here in some way, and it might as well do so by ending the input mode. It's also natural to go on to try reading from the terminal again for subsequent commands; if this was a real and persistent EOF, for example because the pty closed, you'll just get EOF again and eventually quit. V7 ed is slightly unusual here in that it deliberately converts '.' by itself to EOF, instead of signaling this in a different way, but in a way that's also the simplest approach; if you have to have some signal for each case and you're going to treat them the same, you might as well have the same signal for both cases.

Modern versions of ed appear to faithfully reimplement this convenient behavior, although they don't appear to document it. I haven't checked OpenBSD, but both FreeBSD ed and GNU ed work like this in a quick test. I haven't checked their source code to see if they implement it the same way.

36 year old UFS bug fixed, a BSD for the road, automatic upgrades with OpenBSD, DTrace ext2fs support in FreeBSD, Dedicated SSH tunnel user, upgrading VMM VMs to OpenBSD 6.5, and more.

This update eliminates a kernel stack disclosure bug in UFS/FFS directory entries that is caused by uninitialized directory entry padding written to the disk.

As regular It’s FOSS readers should know, I like diving into the world of BSDs. Recently, I came across an interesting BSD that is designed to live on a thumb drive. Let’s take a look at NomadBSD.

NomadBSD recently released version 1.2 on April 21, 2019. This means that NomadBSD is now based on FreeBSD 12.0-p3. TRIM is now enabled by default. One of the biggest changes is that the initial command-line setup was replaced with a Qt graphical interface. They also added a Qt5 tool to install NomadBSD to your hard drive. A number of fixes were included to improve graphics support. They also added support for creating 32-bit images.

I first discovered NomadBSD back in January when they released 1.2-RC1. At the time, I had been unable to install Project Trident on my laptop and was very frustrated with BSDs. I downloaded NomadBSD and tried it out. I initially ran into issues reaching the desktop, but RC2 fixed that issue. However, I was unable to get on the internet, even though I had an Ethernet cable plugged in. Luckily, I found the wifi manager in the menu and was able to connect to my wifi.

upgrade](https://www.tumfatig.net/20190426/openbsd-automatic-upgrade/)

OpenBSD 6.5 advertises for an installer improvement: rdsetroot(8) (a build-time tool) is now available for general use. Used in combination with autoinstall.8, it is now really easy to do automatic upgrades of your OpenBSD instances.

There must be a way to run onetime commands (in the manner of fw_update) to automatically run sysmerge and packages upgrades. As for now, I’d rather do it manually.

I use ssh tunneling A LOT, for everything. Yesterday, I removed the public access of my IMAP server, it’s now only available through ssh tunneling to access the daemon listening on localhost. I have plenty of daemons listening only on localhost that I can only reach through a ssh tunnel. If you don’t want to bother with ssh and redirect ports you need, you can also make a VPN (using ssh, openvpn, iked, tinc…) between your system and your server. I tend to avoid setting up VPN for the current use case as it requires more work and more maintenance than running ssh server and a ssh client.

We're running dedicated vmm(4)/vmd(8) servers to host opinionated VMs.

To get connected to the console you need to have access to the host your VM is running on. The same username and public SSH key, as provided for the VM, are used to create a local user on the host.

$ vmctl stop id [-fw]```

-c Automatically connect to the VM console.```

FreeBSD ZFS vs. ZoL performance, Dragonfly 5.4.2 has been release, containing web services with iocell, Solaris 11.4 SRU8, Problem with SSH Agent forwarding, OpenBSD 6.4 to 6.5 upgrade guide, and more.

With iX Systems having released new images of FreeBSD reworked with their ZFS On Linux code that is in development to ultimately replace their existing FreeBSD ZFS support derived from the code originally found in the Illumos source tree, here are some fresh benchmarks looking at the FreeBSD 12 performance of ZFS vs. ZoL vs. UFS and compared to Ubuntu Linux on the same system with EXT4 and ZFS.

Upgrading guide

Here's the tag commit, for what has changed from 5.4.1 to 5.4.2

After your next reboot, you can optionally update your rescue system:

As always, make sure your packages are up to date:

I'm a huge fan of the FreeBSD jails feature. It is a great system for splitting services into logical units with all the performance of the bare metal system. In fact, this very site runs in its own jail! If this is starting to sound like LXC or Docker, it might surprise you to learn that OS-level virtualization has existed for quite some time. Kudos to the Linux folks for finally getting around to it. 😛

There are plenty of options when it comes to setting up the jail system. Ezjail and Iocage seem popular, or you could do things manually. Iocage was recently rewritten in python, but was originally a set of shell scripts. That version has since been forked under the name Iocell, and I think it's pretty neat, so this tutorial will be using Iocell.

Once you have installed iocell and configured your ZFS pool, you'll need to run a few commands before creating your first jail. First, tell iocell which ZFS pool to use by issuing iocell activate $POOLNAME. Iocell will create a few datasets.

As you can imagine, your jails are contained within the /iocell/jails dataset. The /iocell/releases dataset is used for storing the next command we need to run, iocell fetch. Iocell will ask you which release you'd like to pull down. Since we're running 11.0 on the host, pick 11.0-RELEASE. Iocell will download the necessary txz files and unpack them in /iocell/releases.

Today we are releasing the SRU 8 for Oracle Solaris 11.4. It is available via 'pkg update' from the support repository or by downloading the SRU from My Oracle Support Doc ID 2433412.1.

After hacking the matrix.org website today, the attacker opened a series of GitHub issues mentioning the flaws he discovered. In one of those issues, he mentions that “complete compromise could have been avoided if developers were prohibited from using [SSH agent forwarding].”

Start by performing the pre-upgrade steps. Next, boot from the install kernel, bsd.rd: use bootable install media, or place the 6.5 version of bsd.rd in the root of your filesystem and instruct the boot loader to boot this kernel. Once this kernel is booted, choose the (U)pgrade option and follow the prompts. Apply the configuration changes and remove the old files. Finish up by upgrading the packages: pkg_add -u.

OpenBSD 6.5 has been released, mount ZFS datasets anywhere, help test upcoming NetBSD 9 branch, LibreSSL 2.9.1 is available, Bail Bond Denied Edition of FreeBSD Mastery: Jails, and one reason ed(1) was a good editor back in the days in this week’s episode.

ZFS is very flexible about mountpoints, and there are many features available to provide great flexibility.

Folks,

We have released LibreSSL 2.9.1, which will be arriving in the LibreSSL

It includes the following changes and improvements from LibreSSL 2.8.x:

The LibreSSL project continues improvement of the codebase to reflect modern,

I had a brilliant, hideous idea: to produce a charity edition of FreeBSD Mastery: Jails featuring the cover art I would use if I was imprisoned and did not have access to a real cover artist. (Never mind that I wouldn’t be permitted to release books while in jail: we creative sorts scoff at mere legal and cultural details.)

It is common to describe ed(1) as being line oriented, as opposed to screen oriented editors like vi. This is completely accurate but it is perhaps not a complete enough description for today, because ed is line oriented in a way that is now uncommon. After all, you could say that your shell is line oriented too, and very few people use shells that work and feel the same way ed does.

Introducing funlinkat(), an OpenBSD Router with AT&T U-Verse, using NetBSD on a raspberry pi, ZFS encryption is still under development, Rump kernel servers and clients tutorial, Snort on OpenBSD 6.4, and more.

One of the first syscalls which was created in Unix-like systems is unlink. In FreeBSD this syscall is number 10 (source) and in Linux, the number is dependent on the architecture but for most of them is also the tenth syscall (source). This indicated that this is one of the primary syscalls. The unlink syscall is very simple and we provide one single path to the file that we want to remove.

I upgraded to AT&T's U-verse Gigabit internet service in 2017 and it came with an Arris BGW-210 as the WiFi AP and router. The BGW-210 is not a terrible device, but I already had my own Airport Extreme APs wired throughout my house and an OpenBSD router configured with various things, so I had no use for this device. It's also a potentially-insecure device that I can't upgrade or fully disable remote control over.

Do you have an old Raspberry Pi lying around gathering dust, maybe after a recent Pi upgrade? Are you curious about BSD Unix? If you answered "yes" to both of these questions, you'll be pleased to know that the first is the solution to the second, because you can run NetBSD, as far back as the very first release, on a Raspberry Pi.

One of the big upcoming features that a bunch of people are looking forward to in ZFS is natively encrypted filesystems. This is already in the main development tree of ZFS On Linux, will likely propagate to FreeBSD (since FreeBSD ZFS will be based on ZoL), and will make it to Illumos if the Illumos people want to pull it in. People are looking forward to native encryption so much, in fact, that some of them have started using it in ZFS On Linux already, using either the development tip or one of the 0.8.0 release candidate pre-releases (ZoL is up to 0.8.0-rc3 as of now). People either doing this or planning to do this show up on the ZoL mailing list every so often.

The rump anykernel architecture allows to run highly componentized kernel code configurations in userspace processes. Coupled with the rump sysproxy facility it is possible to run loosely distributed client-server "mini-operating systems". Since there is minimum configuration and the bootstrap time is measured in milliseconds, these environments are very cheap to set up, use, and tear down on-demand.

As you may recall from previous posts, I am running an OpenBSD server on an APU2 air-cooled 3 Intel NIC box as my router/firewall for my secure home network. Given that all of my Internet traffic flows through this box, I thought it would be a cool idea to run an Intrusion Detection System (IDS) on it. Snort is the big hog of the open source world so I took a peek in the packages directory on one of the mirrors and lo and behold we have the latest & greatest version of Snort available! Thanks devs!!!