phoronix.com
>even worse support for SMP environments
LOL
OPENBSD IS FINISHED AND BANKRUPT
Ryder Smith
Other urls found in this thread:
Grayson Hall
>Our focus is security
>Intel's SMT/Hyperthreading has been shown repeatedly to be insecure
>Rather than try and work around it, we're going to just say fuck it and disable it all
I don't see the problem.
Robert Flores
>running OpenBSD on Intel or AMD platforms
Are you fucking retarded? Linux has the best support for those platforms out of any open source operating systems. OpenBSD is only good for really old PCs or for embedded devices, and possibly servers (Linux is just as good for servers). Then NetBSD is for all the rest of the obscure shit that has no other software except the OEM loaded stuff.
Julian Smith
yes, thank you for delivering very false info.
hope you keep shilling so this secure os stays under the millenial's radar.
op is a millenial shill and everything millenials touch become literal shit.
stay away from my ware
Austin Wilson
Fuck you OP
Noah Campbell
Indeed.
OpenBSD is a meme
>Filesystem
default FS doesn't even support SSD TRIM, and OpenBSD doesn't support anything modern like ZFS or BTRFS.
In the CIA triad of Confidentiality, Integrity, and Availability, availability seems to be the one that's lacking. Who cares how hack-resistant your system is if the data you're protecting is corrupted?
That's not even getting into the volume management stuff that's missing, and the snapshots, and the everything.
"b-b-but MUH BACKUPS!!"
You do realize that if the filesystem is not secure and does not protect against bitrot and corruption, your precious backups are going to be fucked, because you'll be backing up corrupted data. Who even knows how far you'll have to roll back in order to get to a clean state?
>Security
"Only two remote holes in the default install!!!!!!!"
Yay!
I hope you realize that this literally only applies to a base system install with absolutely no packages added. In other words, not exactly representative or meaningful towards... anything really
>Sustainability
A few years ago, OpenBSD was actually in danger of shutting down because they couldn't keep the fucking lights on. How could anyone see this as a system they could rely on, when it could be in danger of ending at any time?
>Standards-compliance
"B-But OpenBSD is written in strictly standards-compliant C! Clearly that's better than muh GNU virus!"
So you're not allowed to create extensions to the standard? You should only implement the standard and nothing more? Keep in mind that this is nothing like EEE, as the GNU extensions are Free Software, with freely available source code, as opposed to proprietary shite. People should be allowed to innovate and improve things.
If you're gonna be anal about standards-compliance, then why let people make their own implementations anyway? Why not have the standards organizations make one C implementation and force everyone to use it?
Jace Edwards
>devs aren't skilled enough to work around in their outdated scheduler
LOL
Christian Nguyen
Modern OpenBSD doesn't work on a K6-2.
Justin Foster
if they disable SMT their OS will be just as bad as ReactOS
Landon Williams
SMT is a weird hack anyway.
It gives a 30% speedup only in the very best scenario, but in practice it can actually slow things down.
Carson Ortiz
the freebsd fags are seething
Jaxon Stewart
this is an openbsd thread you retard
Christopher Roberts
based
Evan Campbell
Even the linked article said if you have more than two cores, it almost always reduces performance
Ethan Baker
nah i mean after all those freebsd fiascos they're doing everything they can to discredit openbsd
Jeremiah Cooper
There it is. Always posting this one. Have a (you)
Isaiah Rogers
Guess I won't be using Atom N270s anymore
Grayson Hill
The linked article seems to confuse SMP and SMT
Carson Martin
>No ZFS
>No Jails
It's been finished and bankrupt fool.
Ethan Howard
Improve your English, ranjeet
Brody Martin
I'm going to stop using OpenBSD now.
I suggest you fellows do the same.
Gavin Collins
from what century did you come from?
Isaiah Myers
wut?
Julian Morris
freetards on full damage control
Lincoln Price
Theo protects us, don't worry.
Michael Long
no
Luis Howard
freebsd samefag retards out in force today.
>Those wishing to toggle the OpenBSD SMT support can use the new hw.smt sysctl setting on OpenBSD/AMD64 and is being extended to cover CPUs from other vendors and architectures.
>tfw 7 more side channel exploits about to come out for intel
Samuel Fisher
Linux is NEVER the answer.
Joseph Jackson
>can be disabled
wtf Theo why u do dis
Hudson Roberts
how isn't the only way to secure intel cpu to rip it out of motherboard?
Christopher Allen
OpenBSD is for safe, no speed
Jason Bailey
oh here we go. /v/ tier arguments
"muh hypterthreading"
"muh MAC"
from people who's technical thinking operates within marketing buzzword terminology.
Josiah Moore
>I'm going to stop using OpenBSD now.
ok
>I suggest you fellows do the same.
nah
Jackson Wilson
so, netbsd then?
Zachary King
Is the no SSD Trim thing true and does it really matter that much
Cameron Moore
>millennial
Why self claimed good posters are always above the age 30? Did you spend your miserable life mastering the art of posting on Jow Forums?
Asher Bennett
>tfw the one freebsd cuck who replies to this thread is offline
yea i'd say it matters. it'll shorten the life of your disk, but by how much i don't know.
Alexander Hill
tfw no nvidia support
Aaron Bailey
There's none. Intel always has been garbage, OpenBSD always pointed out that Intel is garbage and their hardware is badly designed with insane flaws.
OpenBSD is the only true defender of freedom and security in the time of information technology, no other project is even remotely that close in regards of security, stability and well design ecosystem.
On top their documentation is excellent
Jaxson Diaz
Every time I distrohop, I always end up going back to that one broken piece of trash distro which rules them all .... WINDOWS
Jason Martin
SMT (Hyper-Threading) makes things up to 25% or 30% faster *for applications written to exploit hyper-threading*.
For the typical ways in which OpenBSD is used (network server, etc.) it probably doesn't matter too much.
I suspect numerical applications would suffer, but OpenBSD is not exactly a platform for numerical work in the first place.
If you run a firewall or a web server the impact if probably negligible to null.
In fact they are pretty much saying that their scheduler is not really SMT-aware in the first place, so it probably has little effect.
Landon Hughes
NO, Intel is finished. That's another showcase, how flawed Intel CPU's are.
Luis Jones
I have this same problem desu
The only thing I want windows for is photoshop because I do not know how to pirate it on other platforms
Connor Gonzalez
>2 releases per year
Don't forget that they come with original soundtracks.
openbsd.org
Ryder Hughes
You'd have to be a real gullible sap to stick with Intel, a real fucking sucker.
Matthew Bell
>cia cat
Tyler Wilson
Landon Moore
can someone explain to me the performance benefit of multithreading?
why would x264 perform better on 1c/2t than 1c/1t
Christian Hill
If they divide the file into 2 pieces each thread can work on half. I at least think this is a simplification of how things work.
Colton Reed
why would this have better performance though if it's just two pathways to the same resources
Nathaniel Ortiz
>freeBSD
wew
>so, netbsd then?
yes, or GNU/Linux.
It's true, and pic related is why it matters.
Carter Bell
then use -current you spud
Kayden King
> Then NetBSD is for all the rest of the obscure shit that has no other software except the OEM loaded stuff.
This meme needs to end, NetBSD has fewer supported platforms than Linux, it hasn't been ported to a new ISA since 2007. Besides, 90% of obscure shit is already running linux.
Christopher Cox
Wish I could run openbsd on my laptop. My next laptop will definitely be one that runs well with openbsd
Gabriel Flores
Hyper Threading is a meme as always and was a mistake
Kevin Fisher
The idea is to have as many threads as execution units.
But if you have a single execution unit, there is no particular reason to have two threads (unless you design your program to do some computation while the other thread is blocked in a system call). It's likely to make things slower.
Generally you want 8 threads for 8 execution units. With hyper-threading, 4 cores mean you have 8 execution units.
Hyper-Threading is a way to "fake" having more cores by allowing two threads to share many systems like caches, etc. It's a cheap way to increase the number of execution units without having too many additional gates. But that's precisely the security issue; if they share caches, etc. in a way that's a bit "promiscuous", one thread can spy on another (keeping in mind that the threads might be for different users, or even different VMs).
OpenBSD is disabling Hyper-threading, not multi-threading.
Chase Lewis
/tpg/
Adrian King
Any thoughts on Qubes?
Aiden Scott
SMT is like adding a second lane to a road. Sometimes one lane is sufficient but frequently with CPUs the one lane gets backed up by a process that is taking its sweet ass time and that process will be hogging all of the resources of the core while its doing "nothing". SMT allows other tasks to pass that process and make use of the resources that aren't being utilized at that moment. If there aren't a lot of process with frequent breaks SMT can cause a backup in its own way because there's still not extra parts to handle all of the tasks simultaneously so the processor may cause one active process to wait on another active process.
Leo Watson
>It gives a 30% speedup only in the very best scenario
>Even the linked article said if you have more than two cores, it almost always reduces performance
There are few cases where SMT doesn't improve performance. 30% is like the average level of improvement in general not just in the best case. It's true that SMT doesn't always help but it helps so often that if you don't have a specific application that would benefit from disabling it you should leave it enabled. Or at least that's how it was. SMT is a very significant feature on modern chips. If Intel's implementation is really broken by design then that's going to be a huge problem, in particular with the server and enterprise segments which aren't going to appreciate the higher power usage and worse performance.
Christopher Bell
HAHAHAHAHAHAHA
Elijah Mitchell
Brayden Butler
>SMT doesn't improve performance
Hyperthreading =/= how other SMT works. You basically never get negative scaling (and the positive scaling can be much higher) on ryzen and on IBM's recent POWER chips. Intel just has a teeny tiny core that ends up with shitloads of resource contention.