Proton gaining easy anti cheat and battle eye support

ChadD

Supreme [H]ardness
Joined
Feb 8, 2016
Messages
6,312
https://www.phoronix.com/scan.php?page=news_item&px=BattlEye-Proton-Steam-Deck

Looks like both of the big anti cheats that where messing up proton compatibility should be smooth out fairly soon.

This should fold in the few top 10/100 steam titles that where no goes.

Thought about posting this in the Steam Deck section... however as this will instantly trickle up to desktop Linux I figure its more just general (good) news.

I still find these companies solutions distasteful in their implementation... and I'm not sure what changes they will have to make to allow them to operate under Linux, I mean battle eye is pretty upfront about hooking into your windows kernel system to scan your system. Linux isn't going to allow that so I am interested to see how they square the circle.
 
This is opt in... so no idea if all these games will get fixed before the deck launch. I just saw this though in the phoronix forums and though I would repost here.
Posted by Teggs in their forum;

This could lead to all but one Red/Borked rating leaving the Top 100 on ProtonDB.

Currently, Borked ratings and their reported reasons are:

#3 - PUBG - BattlEye
#4 - Apex Legends - EAC
#8 - Destiny 2 - BattlEye
#10 - Tom Clancy's Rainbow Six Siege - BattlEye
#18 - Dead By Daylight - EAC
#30 - DayZ - BattlEye
#52 - Smite - EAC
#61 - Black Desert Online - EAC
#65 - Mount and Blade - Bannerlord - Multiplayer - BattlEye
#67 - Paladins - EAC
#72 - Fall Guys: Ultimate Knockout - EAC
#73 - Hunt: Showdown - EAC
#84 - Conqueror's Blade - other issue
 
Cute.
It's like bringing a WW1 tank into a modern battlefield.

These engines do nothing against machine learning cheats that don't even run on your OS/PC. The ones the good streamers use, you know, shroud and co with their physically impossible reaction times that nutswingers vouch for because 'he doesn't cheat'.
 
I still find these companies solutions distasteful in their implementation... and I'm not sure what changes they will have to make to allow them to operate under Linux, I mean battle eye is pretty upfront about hooking into your windows kernel system to scan your system. Linux isn't going to allow that so I am interested to see how they square the circle.
A while ago Valve has implemented a feature into the Linux kernel that allows for things like anti-cheat. I forget what it was called but I imagine that's what Valve is doing now. If developers don't like it they could just port their games to Linux?
 
  • Like
Reactions: ChadD
like this
A while ago Valve has implemented a feature into the Linux kernel that allows for things like anti-cheat. I forget what it was called but I imagine that's what Valve is doing now. If developers don't like it they could just port their games to Linux?
If the 0.1% of gamers on desktop Linux don't like it, they could just port their pc to windows?
 
If the 0.1% of gamers on desktop Linux don't like it, they could just port their pc to windows?
1% according to the latest steam survey results. That’s a whole order of magnitude higher than your guess!

I’m actually really curious to see if those numbers go up after the steam deck releases and once steamos 3 is available. Would be neat if other companies do prebuilt SFF systems using it.
 
This is great news. I don't play online anymore, but I hear that as a major reason why PC gamers haven't switched to Linux.

With Steam Deck coming soon, and improvements to Proton, Linux gaming has never been more viable.

At least for single player games, I've only had maybe 2 or 3 games not work out of the dozens I've tried over the past few years.

Sometimes there are small problems, but you just have to look on ProtonDB and usually there is some command you need or to edit a config file. For the most part, though, games just work.
 
If the 0.1% of gamers on desktop Linux don't like it, they could just port their pc to windows?
So that third party companies can just hook whatever they want into your hardware extraction layer... and scan everything you have in ram and perhaps take a poke around your storage drives.
Windows users put up with a lot of BS that's for sure.

The entire point I was making was this battle eye software tells you exactly what its going to do on your system... its hooks itself at the kernel level into your system and lists every single thing you have in ram... it even reports snips back home for processes it doesn't recognize. (Heuristic ram scanning) It also by their own admission breaks out of whatever sandboxed bit of drive its in and uses its kernel level hooks to scan drives.

They are now saying they are going to make that work under Linux. SO my take away is this... none of that level of intrusion is required for actual anti cheat to work. Which means they have been doing all that stuff under Windows for years now, and anti cheat was just an excuse. 100% no matter what bits Valve adds to the kernel battle eye is not going to be able to go roaming through RAM or drives under Linux. The Kernel team would never ever approve such a security flaw. (which would mean if Valve intended to do that it would have to be their own custom kernel build) Valve has said this will work for upstream proton meaning no custom kernel... So yes I am interested to see how they pull off Linux compatibility. IMO if they get it working under Linux without doing all the questionable things they do under Windows... windows users should actually be pissed off. Just my thoughts.
 
1% according to the latest steam survey results. That’s a whole order of magnitude higher than your guess!

I’m actually really curious to see if those numbers go up after the steam deck releases and once steamos 3 is available. Would be neat if other companies do prebuilt SFF systems using it.

I'm looking forward to seeing if anyone starts selling "steam machines" again. I mean we all know MS basically gives windows away to OEMs, and make it painful for companies not shipping windows. Guess it really depends how the deck goes. I doubt we see a ton of steam os shipping machines at the same time Decks start shipping. Who knows though if Steam sells all the decks it can make next year its possible we start seeing pre installed SteamOS as an option. I just hope Valve makes a easy to install version of SOS 3... so average users can easily give it a run. I'm not sure how hard they will try to make it easy to install dual boot type setups ect. Being arch though I have some hope... I mean they could basically just take some of the tools used by distros like Manjaro put a Valve polish on them and go pretty easy.
 
They are now saying they are going to make that work under Linux. SO my take away is this... none of that level of intrusion is required for actual anti cheat to work.
BattlEye has supported Linux (and Mac) native for years, the news is that they will allow Proton, which wasn't previously possible.

https://www.gamingonlinux.com/2021/...team-deck-will-be-opt-in-like-easy-anti-cheat

However, Easy Anti-Cheat I believe was for Windows only (or at least only free on Windows) so Epic's announcement is that they are providing a free SDK to Linux native, Proton, and also Mac.

https://www.gamingonlinux.com/2021/...-anti-cheat-for-linux-including-wine-a-proton

To address your point, kernel level anti-cheat *could* catch some stuff that a user space app could not. For example, I've heard about people using hacked mouse drivers for aim-bots, which you would need kernel level access to discover.

That said, there are hacks out there where people flash the firmware on a mouse and hack it at the hardware level which is almost undetectable (at least with software, you could use heuristics, but you could do that in user space). Not to mention AI-style cheats, which these intrusive anti-cheat software can't catch anyway.

It's kind of a losing battle, but I do understand why they do it (at least from the game developer side) because rampant cheating can totally kill a multiplayer game.
 
I'm looking forward to seeing if anyone starts selling "steam machines" again.
It's probably unnecessary. Steam OS 3 will replace Big Picture mode in Steam, meaning any Windows/Linux/Mac PC can use the same interface as the Steam Deck. And it's not too hard to configure your PC to boot up into Big Picture mode if you really want that.

After the big failure of Steam Machines, I'm not sure any companies want to take that gamble. But maybe some will make their own Steam OS 3 handheld sort of like Android. And I hope Valve releases the ISOs so people can make their own machines. That would be the best thing I think.
 
I wonder what you could do with one of those partially-disabled PS5 SOC motherboards...
You're in luck, because Tom's Hardware has a review. Unfortunately, the iGPU is disabled, and you only get a PCIe 2.0 x4 slot for your video card, and two SATA ports for storage (no m.2).
 
It's probably unnecessary. Steam OS 3 will replace Big Picture mode in Steam, meaning any Windows/Linux/Mac PC can use the same interface as the Steam Deck. And it's not too hard to configure your PC to boot up into Big Picture mode if you really want that.

After the big failure of Steam Machines, I'm not sure any companies want to take that gamble. But maybe some will make their own Steam OS 3 handheld sort of like Android. And I hope Valve releases the ISOs so people can make their own machines. That would be the best thing I think.

Steam OS does have a couple cool features beyond just a redesigned big picture though. I'm not sure if they will matter much to desktop type users. Having said that there are plenty of companies already making handhelds.
Steam OS is supposed to switch over to a KDE desktop mode when you plug in an external monitor. I know its not the main reason to pick up a handheld. Still I think a small NUC type machine (or other company handheld) running SteamOS sounds pretty cool. Imagine a NUC or non Steam Handheld where SOS remembers that when it sees Sony A9G as the display device to use big picture... and when it sees Samsung G9 it switches to a KDE desktop.
Of course you can do all that with whatever distro you want if you know what your doing... still it would be cool to see actual useful consumer devices shipping, beyond just the deck anyway.
 
I guess what I'm getting at is that if you want a Linux gaming PC, you can build that now and have a full desktop experience. Or run the new Big Picture mode (when it comes out) if you want to use in the living room or whatever. Having a dedicated Steam box just doesn't make sense.

But for a handheld or portable PCs, maybe there is a market there. For example, a company like GPD could adopt Steam OS on their new handheld, maybe save money, and have a better experience for consumers. I think that scenario is way more likely and useful for everyone involved.
 
I guess what I'm getting at is that if you want a Linux gaming PC, you can build that now and have a full desktop experience. Or run the new Big Picture mode (when it comes out) if you want to use in the living room or whatever. Having a dedicated Steam box just doesn't make sense.

But for a handheld or portable PCs, maybe there is a market there. For example, a company like GPD could adopt Steam OS on their new handheld, maybe save money, and have a better experience for consumers. I think that scenario is way more likely and useful for everyone involved.
Windows has always been a pretty shit handheld OS. As much as MS has tried to cater to touch screens and the like... ime anyway such devices have always been just a bit annoying. Valve imo has the right idea... big picture when your on the little screen. KDE mode on an external. I imagine it will be smart enough to know a TV when it sees it and stay in BP. I look forward to real world reviews and to see how much polish it has for the masses.
 
Valve imo has the right idea... big picture when your on the little screen. KDE mode on an external. I imagine it will be smart enough to know a TV when it sees it and stay in BP. I look forward to real world reviews and to see how much polish it has for the masses.
I saw a video where they had BP on a monitor, so there must be a way to switch it (maybe it asks?) not sure they could do it automatically but that would be cool.
 
So I can't seem to find that video, but I did find an email from Gaben confirming that you can tab out of BP on the handheld to get to the KDE desktop, so I imagine the same button would work when docked.

cv51vij8i7c71.png
 
  • Like
Reactions: N4CR
like this
As I understood it... once you connect a second display device it automatically makes the second display the primary desktop running KDE. SOS3 is basically Valve Arch with KDE... it just boots into big picture when on the device display. I would assume they will have a secondary display behavior option... and I can't imagine they wouldn't think to include remember device. If at launch the deck doesn't remember display settings for individual displays that would be a shame as Linux is capable of doing just that. I would be surprised if it wasn't working day one. Hopefully flawlessly... and I don't know I don't work for valve but if I did, I would try and include a data file with as many consumer device IDs as possible so the first default on a computer monitor would be desktop and big picture on TVs. They couldn't include every possible option... but they could likely cover 95% of the market with a simple device pref file.
 
As I understood it... once you connect a second display device it automatically makes the second display the primary desktop running KDE.
Yes. That was my understanding as well. What I was commenting on is whether there is an option or setting to let the external monitor/TV go to BP mode by default and not the KDE desktop.

And then the second part is whether it would remember the setting, particularly like if you have a computer monitor and also a TV, could you have a setting unique to each display? I think that is an open question.
 
Hopefully this means Proton compatibility for the aimbots is not far behind.
 
Last edited:
If the 0.1% of gamers on desktop Linux don't like it, they could just port their pc to windows?
Yes? It's the same argument against not porting console games to PC. People want to play the game but want to do it on their platform of choice. Don't port console games and you get emulators. Don't port Windows games to Linux and you get Battle Eye in Proton. With Valve behind this, you aren't getting something made with someone's spare time. Valve is taking Steam on Linux very serious, so that game developers do as well.
 
I eventually want to be able to play the majority of games in my libraries without having to use Windows. The fact that this is another step in making that a possibility is great.

Already toying around with the idea of putting a dedicated Linux drive in my system to dual boot from and see how "satisfied" I am with switching to Linux full time, even for gaming.
 
A while ago Valve has implemented a feature into the Linux kernel that allows for things like anti-cheat. I forget what it was called but I imagine that's what Valve is doing now. If developers don't like it they could just port their games to Linux?
There were some patches submitted about a year ago related to system call emulation. The short version is that Wine needed a low-overhead mechanism to trap and emulate system calls for Windows binaries that directly issued (Windows) system calls instead of going through the Windows API. As a result, a feature was added to prctl() in Linux kernel version 5.11.

The changes introduced didn't have anything to do with adding anti-cheat features, but they potentially help Wine run Windows games that have builtin cheat-detection. That's because programs with cheat-detection or DRM are more likely to perform system calls directly instead of use the Windows APIs as they're supposed to. Those that use the APIs were already handled by Wine.

Here are some links for anyone interested in the low-level bits:

https://lwn.net/Articles/826313/
https://lkml.org/lkml/2020/11/18/815
 
Just don't get the level of privacy intrusion... if someone is going to play competitively, they should have the means to invest in hardware capable of real-time asymmetrical memory encryption with regular key refreshes. We've had this technology for how long? What would prevent this from working?
 
Back
Top