Bryan Steele

Dave Voutila (dv@) continues his work on multi-process device emulation for #OpenBSD vmd(8), this time moving virtio scsi (CDROM) device emulation to a separate process, further reducing the amount of code running in-process for VMs.

dv@ modified src/usr.sbin/vmd/*: vmd(8): convert virtio scsi device to a subprocess.

Break the virtio scsi device (used as a cd-rom drive) into a subprocess like the virtio block and network devices. This leaves only the entropy device (viornd) and vmmci device running in-process with the vcpus.

ok mlarkin@

This follows up on work presented by Dave at EuroBSDCon 2024.

https://www.youtube.com/watch?v=qKrxVzS96BM

https://www.openbsd.org/papers/eurobsdcon2024-vmd_multiprocessing.pdf

#mutualaidrequest #mutualaid

I don't suppose that I have any #bsd #unix #foss #infosec friends out there willing to signal boost, by chance? ​ ​

https://bsd.network/@brynet/114458997143046937

Theo de Raadt has introduced a new hw.blockcpu sysctl to #OpenBSD -current to offer more control over which CPU core types (Performance, Efficiency, and SMT) are available to schedule processes on. Modern Intel (and ARM) CPUs additionally have slower LP-E (low-power) cores that severely hinder system performance.

deraadt@ modified src/sys/*: Some new intel machines have a new 3rd tier of cpus called LP-E which are E-core (Atom) without L3 cache. These cpus are Lethargic, and it sucks when processes migrate to them.
This introduces sysctl hw.blockcpu= which takes a sequence of 4 letters.
S (for SMT), P (regular performance cpu), E (efficient cpu) generally 80% to 50% as fast), and L (lethargic cpu) which are even slower.
By setting this, you can select cpus to kick out of the scheduler. The default is SL.
The hw.smt sysctl remains for now but we will eventually delete it.
hw.smt changes and follows hw.blockcpu=S.
ok kettenis mlarkin

Heads up for #OpenBSD 7.8 users, backports for chromium, iridium, ungoogled-chromium ports have been committed, updated -stable packages should be showing up in the next few days!

chromium has historically only received updates in -current, but it appears sthen@ has managed to backport them!

https://marc.info/?l=openbsd-ports-cvs&m=177195097807729&w=2

Stefan Sperling (stsp@) has added initial support for newer "BZ" models of the Intel AX211 Wi-Fi 6e (not 7) devices found on recent laptops to the iwx(4) driver in #OpenBSD -current. 🥳

Due to an unresolved bug in Tx aggr, currently only 11a/b/g is stable, which will hopefully be resolved in-tree.

EDIT: The Tx issue for BZ devices has been fixed by stsp@, 11n/11ac should now work! ​ https://bsd.network/@stsp/116222365173857875

"Thanks to all involved for the support with review and testing of the diffs which led up to this, to Johannes Berg for help with solving various fatal firmware errors, to denis@'s company for making it possible for me to work on this project during work hours for a while, and to the OpenBSD Foundation for providing hardware."

Firmware for BZ devices is available in fw_update(8).

https://marc.info/?l=openbsd-cvs&m=177322669708761&w=2

A lot of hubbub over a remote crash, a DoS, not a privesc or RCE.

#OpenBSD never claimed to have zero remote crashes. The goal has always been to minimize attack surface, and mitigate the impact of attacks.

I believe it's worth pointing out the errata was out for 7.7/7.8 almost ~3 weeks ago, and users had a patch ~90 minutes after it was committed. ​

#mutualaidrequest #mutualaid

I don't suppose that I have any #bsd #unix #foss #infosec friends out there willing to signal boost, by chance? ​ ​

https://bsd.network/@brynet/114458997143046937

I think it's time for some #OpenBSD #79HYPE ​

After ~21 years, gcc 3.x has left the building, with the last remaining platform (OpenBSD/luna88k) ported to gcc4.

https://bsd.network/@brynet/115425313813361816

Jonathan Gray (jsg@) has updated the drm graphics drivers (inteldrm/radeondrm/amdgpu) in #OpenBSD 7.9 to Linux 6.18.y/6.18.22 from the 6.12.y longterm support version.

https://bsd.network/@brynet/116201960048161449

https://freshbsd.org/openbsd/src?q=drm&committer[]=jsg

OpenBSD now supports "Delayed hibernation" on amd64: After waiting a number of seconds (up to 24 hrs) the machine will wake from S0ix/S3 idle sleep/suspend and hibernate to disk.

https://bsd.network/@brynet/116217813921273057

The OpenBSD kernel gains a new "parking mutex".. inspired by WebKit.

https://bsd.network/@brynet/115503876824188865

A long standing ACPI issue (boot delay) that has plagued several Intel Mac models has been fixed by jcs@

https://bsd.network/@brynet/115602160298028722

OpenBSD's EFI bootloader now supports loading files from the ESP, making it easier to e.g: copy & bootstrap a ramdisk kernel.

https://bsd.network/@brynet/115630978565153559

Improved support for running OpenBSD as a guest VM on Apple Silicon machines under macOS.

https://bsd.network/@brynet/115899206016337373

OpenBSD/amd64 now supports SMP on up to 255 CPUs, such as on AMD Threadripper/EYPC.

https://bsd.network/@brynet/115899248487624689

OpenBSD iwx(4) now supports additional Intel AX211 WiFi 6/6E models, as well as 160MHz channel support! ​

https://bsd.network/@brynet/116210371257002339

https://bsd.network/@brynet/116319563256899912

Important security refinements to both pledge(2) and unveil(2), fixing several early design issues.

https://bsd.network/@brynet/116136000669207850

https://bsd.network/@brynet/116197240853794609

https://bsd.network/@brynet/116217472157803716

Plus lots more to see in 7.9! Stay tuned! ​​