Forgot your password?
Close
typodupeerror
Microsoft Windows Games

Microsoft Blames Windows Gaming Issues On RGB Lighting Devices (bleepingcomputer.com) 78

Posted by EditorDavid from the game-over dept.
BleepingComputer reports: Microsoft says ongoing issues causing games to crash or fail to launch after installing the August 2026 Windows updates may be caused by peripherals with RGB lighting.

As Microsoft explained when it confirmed it's investigating on Wednesday, this known issue affects games like ARC Raiders, MARVEL Tokon: Fighting Souls, and The Finals on systems running Windows 11 24H2 and 25H2... "[Peripherals or internal device components with RGB lighting features] may install drivers or code components with file names similar to inpoutx64. In systems where these drivers are found, the issue is then triggered by launching certain games," it noted.
Thanks to Slashdot reader joshuark for sharing the article.

Microsoft Blames Windows Gaming Issues On RGB Lighting Devices More | Reply

Microsoft Blames Windows Gaming Issues On RGB Lighting Devices

Comments Filter:

  • a path-traversal attack (Score:5, Insightful)

    by usedtobestine ( 7476084 ) on Saturday August 22, 2026 @11:00PM (#66302744)

    Poorly written applications by both game developers and poorly written operating systems by Microsoft.

    • Re:a path-traversal attack (Score:5, Interesting)

      by phantomfive ( 622387 ) on Saturday August 22, 2026 @11:55PM (#66302772) Journal
      Not many good programmers these days.

      No point in it, managers don't care if you do a good job, as long as you close the ticket. For some managers, if you included bugs, that's a good thing. It's more tickets that can be opened.

      • Everything is done for theater now.

      • Re: (Score:3)

        by gweihir ( 88907 )

        Not many good programmers these days.

        Yes, unfortunately. Well a big shakeup is coming and we will get liability and likely qualification requirements. The EU CRA (and also the GDPR and the DSA) are already starting that process. I guess a lot of bad coders will lose their jobs in the next 5-10 years. The same has happened in every other engineering field when society finally admitted to itself that quality matters in important technologies.

      • For some managers, if you included bugs, that's a good thing. It's more tickets that can be opened.

        It hurts me in my soul how true this really is. Gotta keep those velocity numbers up.

        • The recent trend in software companies is to fire middle managers, and the metric used to evaluate managers for firing is the velocity. So that's the reason.

          https://www.cnbc.com/2025/12/2... [cnbc.com]

          If you like your manager enough to help them, then break up tasks into smaller tasks that take an hour or so but still have 1 point of value.

      • Re: (Score:3)

        by redback ( 15527 )

        It's more tickets that can be opened.

        If you're measured on ticket closures, you need more tickets opened !

        You get what you measure.

    • Re:a path-traversal attack (Score:5, Informative)

      by AmiMoJo ( 196126 ) on Sunday August 23, 2026 @04:43AM (#66302922) Homepage Journal

      In this case it's an open source driver that is the problem. Specifically this one: https://www.highrez.co.uk/Down... [highrez.co.uk]

      The maintainer abandoned it, but everyone kept using it because it's the only real option for apps (including open source ones like Open Hardware Monitor) to access the motherboard SMBUS which controls fans and RGB lighting, among other things.

      It's not clear exactly what the cause of the crash is, but it may be related to anti-cheat tech which basically rootkits Windows and then panics when it sees certain drivers installed.

      • Re: (Score:1)

        by gweihir ( 88907 )

        Fascinating. Is driver writing on Windows so utterly messed up that writing and maintaining what is a really simple driver on the functionality side is something nobody wants to do?

        • The problem is there are many many different pieces of hardware which would have to be supported and nobody is bothering. Most of the RGB hardware uses its own interface instead of a standard. So instead this generic driver for direct hardware access has been used. The situation on Linux is actually not very different, except on Linux there are udev rules which make it possible to assign permissions to specific devices to users or groups and have more security than you get when doing the same thing on Windo

          • Re: (Score:2)

            by gweihir ( 88907 )

            So essentially the EEs have made a mess and now expect software to fix it? Well. If this driver really allows general hardware access, it is a security problem waiting to happen and should not be used at ll.

            On a related note, I do not understand the point of colorful lights in keyboards and the like. But then, I am not mentally 5 years old.

            • I don't know that it's really "the EEs" that caused the problem. I suspect it's "the regular programmers who found themselves doing embedded programming jobs". They didn't care how it could be made to work, they just wanted it to work. Most of these haven't even been on USB until recently, they were all on I2C or SPI. However there ARE some standards now and corporations should just use them like everyone else. Since they don't, the semi-pro level stuff is superior, I'm pretty sure the RGB controller I'm us

              • Re: (Score:2)

                by gweihir ( 88907 )

                I do not understand the point of colorful lights in keyboards and the like. But then, I am not mentally 5 years old.

                Sorry your heart is dead, but a lot of us still understand fun.

                If you do not find fun as an adult, you are doing it wrong. Here is a hint: Regressing yourself to a mental child is not the way to do it.

            • I did not even realize that keyboards with individual colored lights were a thing until I read the latest release notes for Terraria.

              The wiki has a good show of how the keyboard will have a flame animation if you touch lava in the game, or eyeballs when you fight some bosses. Kinda cool actually.

              https://terraria.wiki.gg/wiki/... [terraria.wiki.gg]

              Do I want it? Not really, it is more of a gimmick like the touch bar we had on MacBook Pros for a few years.

          • The problem is there are many many different pieces of hardware which would have to be supported and nobody is bothering.

            Apple does.

            • Really? Apple produces software which supports RGB LEDs on PCs? What's it called?

              • Really? Apple produces software which supports RGB LEDs on PCs? What's it called?

                Really? You think I said that Apple supplies drivers flor every piece of hardware? Argue at least honestly.

                Apple writes drivers for equipment that is used on their computers. Is your need to disagree that strong? I know you hate me, but I look at the situation that I don't mind pissing you off, comrade.

        • Who said it's messed up? This is the Linux approach to a problem: Why reinvent the wheel when someone else has written something you can use already.

          And when I said it's the Linux approach, that very much includes orphaned libraries. Linux, Windows, Unix, Cloud, it doesn't matter. This problem is so ubiquitous there's an obligatory XKCD comic: https://xkcd.com/2347/ [xkcd.com] the only difference is, the guy in Nebraska retired and no one noticed.

        • Re: (Score:2)

          by AmiMoJo ( 196126 )

          I don't think it's particularly hard or anything. The bigger issue is getting it signed by Microsoft. Because it's a very low level driver that can interact with critical parts of the hardware, like controlling your fans and power management, the bar is naturally a bit higher than a driver for some random USB device. Therefore every time it gets updated, Microsoft has to check it and sign off on it.

          There was some company that was updating it, basically just rebuilding it to keep it compatible with the lates

        • Fascinating. Is driver writing on Windows so utterly messed up that writing and maintaining what is a really simple driver on the functionality side is something nobody wants to do?

          Yes. Been a lot of trouble with Windows drivers for years.

          I never understood why they didn't take the Apple approach and write all their own drivers.

          • Re: (Score:2)

            by AmiMoJo ( 196126 )

            The best solution would have been for hardware vendors to stop using SMBUS and switch to USB. No driver needed.

            • Re: a path-traversal attack (Score:4, Interesting)

              by Qwaniton ( 166432 ) on Sunday August 23, 2026 @05:33PM (#66303678)
              Implementing this over USB is a lot more complicated than bit-banging SMBus with MMIO. That means you have to learn how modern PC hardware works and you have to think through a protocol design, think through access control semantics, God, maybe even design an API. The people writing these drivers are the people making these devices and they are not operating systems people. These devices were developed and prototyped as bit-banged gadgets built around standard simple microcontrollers running on standard electronically-simple I/O buses. This is how stuff was done on PCs too back in the MS-DOS/Windows days. This whole segment of the hardware industry got into a snag when Windows XP came out and the home PC world started moving to NT because that meant they had to start thinking about scary things like memory protection and stop treating the NT kernel like an Arduino. So, from the very beginning, they'd find these niche hacky drivers that let them bit-bang portsâ"yes, usually they just found something some small firm wrote to solve a specific problem and they shared them with each otherâ"and they'd continue writing their control apps to bit-bang the chip directly. And because this area of computing is adjacent in many ways to the world of niche industrial computing, which is where these hacky drivers usually came from, Microsoft signed the drivers supported it just enough that the show went on. The kernel team, being the most competent engineering team in all of Microsoft, likely knew that getting these people to learn kernel programming was a pipe dream and they've done a good enough job of blast-proofing the NT kernel around these hacky drivers, no one ever really had to point out that the 800 pound elephant in the middle of the room is still writing hacky VxDs for Windows 11.

            • Re: (Score:2)

              by batkiwi ( 137781 )

              Unless there's a USB standard for RGB lights connecting it via USB doesn't magically mean "no driver needed".

              USB keyboards with RGB require a specific driver for the RGB to work. The keyboard part works without a "driver" (which really isn't true, there still is a driver).

              • Re: (Score:3)

                by AmiMoJo ( 196126 )

                There doesn't need to be a standard, they can just use USB vendor defined endpoints, or a USB HID vendor defined report. Then the standard WinUSB and HID APIs work, and no driver is required. Plug it in, Windows recognizes it and attaches the appropriate generic USB driver, and the RGB control software sends whatever data is likes.

                • Re: (Score:2)

                  by gweihir ( 88907 )

                  Indeed. Could even go via USB-serial with CH340 or PL2393 and be dead easy and cheap. But a large part of the "business" view on technology is "do not pay an engineer if we can already make money off it, regardless of how broken it is". In a larger context, that mindset brings us things like climate change.

        • Everyone has written their own, that's the problem. Razer has their own, Asus has their own, Corsair has their own, EVGA has their own... Some of them partially support some of the attempts at shared standards, MS is trying to push a standard but nothing I have supports it (guess why I picked those four vendors).

          I have one application that manages RGB for my motherboard (fans and a pointless LED strip) and an ASUS mousepad (I know it's stupid), another that manages the LEDs on the video card and keyboard

      • No, in this case it's a lack of a primary first party API in the Operating System problem. RGB is ubiquitous. Why the fuck hasn't Microsoft provided an API for it? They've shoehorned every other god damn thing in and abstracted it via DirectX, so how should anyone be surprised that developers doing dodgy stuff to sync lighting to games is causing an issue.

        • Re: (Score:2)

          by AmiMoJo ( 196126 )

          To be fair if I were Microsoft I would tell motherboard manufacturers to get a grip and make it a USB device. Hanging it off the SMBUS, where critical stuff like fan control and power management also lives, is kind of insane for controlling some lights.

          If it was USB then it could just use WinUSB or even HID. No driver needed, API is very mature.

          • No driver needed?

            You're not understanding the incentive structure that led to this state of affairs in the first place. To use USB would mean that they would have to do a ton of driver work. They would have to implement these generic control surfaces and they would have to ensure that their hardware does what the USB specs say it should. You'd end up with a USB RGB controller chip.

            If you haven't noticed, RGB controller chips are a lot more electronically simple than that. They are small-project simple I/O-

            • Re: (Score:3)

              by tlhIngan ( 30335 )

              You're not understanding the incentive structure that led to this state of affairs in the first place. To use USB would mean that they would have to do a ton of driver work. They would have to implement these generic control surfaces and they would have to ensure that their hardware does what the USB specs say it should. You'd end up with a USB RGB controller chip.

              If you haven't noticed, RGB controller chips are a lot more electronically simple than that. They are small-project simple I/O-programmed chips r

            • Re: (Score:2)

              by AmiMoJo ( 196126 )

              Lots of USB devices use RGB, such a mice and keyboards. For the RGB the early implementations were using microcontrollers, which often have USB built in.

      • Re:a path-traversal attack (Score:4, Informative)

        by thecombatwombat ( 571826 ) on Sunday August 23, 2026 @02:31PM (#66303496)

        Yeah. There's a whole rabbit hole here, this abandoned driver becoming this critical piece of Windows gamer's software is a fascinating story all on its own. The short of it is Windows didn't provide a good interface to access the SMBUS, and one guy's hack to make it work got reused again and again until it's basically infrastructure and won't go away. But because it's a low level driver it can crash everything. It wasn't just abandoned the original author was like "guys, this was a wild hack, I don't want to own this for life."

        *But* it's important to clarify, it's not a matter of just users going and installing this "open source driver."

        "Poorly written applications by both game developers and poorly written operating systems by Microsoft." is still pretty accurate really.

        Tons of vendors selling this stuff, have been bundling it into their seemingly proprietary software for years now.

      • There was something in there about a kernel change breaking the driver. Which, from the looks of the page you linked, was a good thing. An unmaintained driver written using what was already an obsolete an insecure VC version, and still depends on the 2005 redist? And it provides low-level access to everything? Yuck.

        And why the hell did any game need that?

        Oh, wait a minute though. What you linked is InpOut32/x64.DLL. They're talking about InpOutx64.SYS. I don't know if they are related.

    • Poorly written applications by both game developers and poorly written operating systems by Microsoft.

      Seems strange that Windows has trouble with files with similar names.

      I mean how does that happen in the real world. Windows only checks part of a filename, then decides "Looks good to me!"

  • Will be a sleeper build, I already have a case from 2015 at the ready. And of course, no RGB

    • No RGB is a selling point for me. My case is a big black box and the only light on it shows if it is on or off as these things should be.
      • That was my thinking, but I can set the only LEDs visible through my case to show useful information about temps and load, which is nice. They aren't going away, they're too cheap and too many people like them as decoration.
  • Were these 3rd party drivers certified as WHQL compliant? Or did the users just click install when the warning message popped up about the drivers not being certified?

  • Running Windows? (Score:3)

    by PPH ( 736903 ) on Saturday August 22, 2026 @11:32PM (#66302758)

    Stick to Edison lamps. Or carbon arc lights.

  • Many are made by Chinese companies with names like MBXRZ and take 6 weeks to ship to the US to avoid tariffs. Many of these companies do *not* put much effort into making their drivers robust and well-tested.

  • rrriiiight... (Score:3, Insightful)

    by Tom ( 822 ) on Sunday August 23, 2026 @01:21AM (#66302816) Homepage Journal

    And you clowns don't think that your OS is a fragile pile of crap if it's that easy to fuck up?

    Seriously? Do you even know what the job of an OS _is_ ?

    • Hey! How come I get modded down as flamebait but you don't?

  • OMG (Score:3, Interesting)

    by Anonymous Coward on Sunday August 23, 2026 @01:36AM (#66302828)

    OMFG I hate all the lighted crap now. Sucking up power, USB slots, and contributing nothing. I spend decent effort blacking out every little led light in my bedroom for the sake of DARKNESS. Painters tape, model paint/finger nail polish all the power strips, every electronic device has some BS light.

    • Re: (Score:2)

      by dskoll ( 99328 )

      Yes, I hate all the LEDs on devices and I too would black them out or sometimes even open the device and desolder them.

      Then I discovered a magical thing called a sleep mask, and that solved all of my nighttime light problems.

    • Did you try turning them off instead of getting glue, paint, and nail polish all over everything?

  • Led lighting sucks (Score:5, Interesting)

    by boojumbadger ( 949542 ) on Sunday August 23, 2026 @02:21AM (#66302854)

    It is completely useless. I bought a box recently it has 6 fans with lighting that I shut off and lights on the ram I couldn't see how. As for Windows profitability, they'd have a lot fewer sales if it wasn't on all the prefabs. It is cheaper and easier to buy a prefab with windows than to buy the parts and put them together without. My box cam with win11 and I booted it exactly once by mistake before I wiped it an put Linux on it. Can I get the MS tax back? not on your life. Oh and the NVIDIA card has an led branding, I guess its an Aero that I can't turn off either.

    Oh well, I guess there is a cost for expediency.
    What the use is a light on my mouse that gets covered by my hand? At least I finally twigged on how to turn it off permanently using CKB-Next, except it comes back on with the box is off. How stupid is that?

    • I got a board with RGB so I decided I'd try it out. It's USB-connected so I figured it would be decent. Nope it's total shit with no direct mode, good work ASRock. But by that time I had some LED peripherals so I bought a Nollie 8ch controller which was very cheap and... I like RGB. It's convenient to see how loaded my CPU and GPU are at a glance. Internal USB hubs are practically free now so connecting them is a breeze. The kids are all right.

    • The ones in my box do nothing but support gay rights (metaphorically). Such pretty rainbows. Anyway, I guess they are harmless enough except if I am trying to sleep and they keep screaming at me.

  • InpOutx64 is a security nightmare (Score:5, Informative)

    by djgl ( 6202552 ) on Sunday August 23, 2026 @03:07AM (#66302868)

    The purpose of that driver is to bypass Windows' memory protection to access arbitrary I/O ports from user space and to map physical memory regions into the address space of a process. It is a cheap workaround for people who don't want to go through the certification process required by Microsoft for drivers nowadays.

    • Re:InpOutx64 is a security nightmare (Score:5, Insightful)

      by WarlockD ( 623872 ) on Sunday August 23, 2026 @03:47AM (#66302878)

      This is an old problem too, The old driver could, in theory, let you bit bang any ports on the system. PCI ports, PCIE ports, and the LPC. Its also an opensource singed driver and developers tend to never update whatever version of the base driver so it never keeps up with the security updates.

      I want to blame allot of people here but in the end its Microsoft's unwillingness to make some simple built in port api. Make it signed with some VID/PID. Have the vender tell the kernel what port it what's and have the kernel accept or deny it right there.

      I just don't see why Microsoft killed off the direct parallel interface. The only thing I can figure is they were to lazy to abstract the interface from their printer spooler nightmare. You don't have any of these problems with the serial.

      • Re: (Score:2)

        by djgl ( 6202552 )

        Microsoft didn't kill off the direct parallel interface. Windows 11 has the exact same parallel port support as Windows 2000. It has a kernel-level API that can be used to query the features and I/O addresses of parallel ports. But the only driver provided by Microsoft that makes use of this API is the ParallelVdm driver, which gives applications only access to the basic parallel port functionality. No bidirectional modes, no EPP, no ECP.

    • Better question. This driver is used for a specific purpose for specific hardware features that is largely ubiquitous. Why hasn't Microsoft provided a first party driver to do it and a secure API for vendors to use.

  • A job well done by Microsoft (Score:4, Funny)

    by D,Petkow ( 793457 ) on Sunday August 23, 2026 @06:38AM (#66302994)
    Ah yes, the classic Windows gaming experience: shelling out for high-end hardware, keeping your system fully updated, and then watching your favorite games immediately self-destruct because your glowing mousepad dared to talk to the kernel.

    Who could have predicted that a piece of flashing rainbow plastic would be the ultimate DRM to stop The Finals from launching? Truly, brilliant engineering.
    When your operating system is so fragile that a driver named after inpoutx64 gets stage fright from RGB LEDs, you really start to wonder if the entire Windows kernel is just held together by duct tape, corporate prayers, and ancient legacy code from 1998.
    Forget cyber threats - the real apex predator of modern PC gaming isn't ransomware; it's Microsoft's monthly patch cycle turning your rig into an expensive RGB nightlight that can't actually run video games.
    Let's face it Ransomware - would not exist, if Windows didn't exist.

    • Re: (Score:2, Flamebait)

      by quall ( 1441799 )

      I'd argue that the Linux gaming experience is much worse. Apple, let's not even get into.

      BTW it's an unmaintained and unsecure open source driver that's crashing, which was also abandoned and isn't maintained. So the problem is that all of these companies are using the same freebee driver that literally hacks into the system in order to operate, and now it no longer works.

      • Much worse on Linux? Have you acrually tried? Everything works, much better than on windows. The anticheat games which don't work, are not worth playing. I pity anyone still on windows.

      • BTW it's an unmaintained and unsecure open source driver that's crashing, which was also abandoned and isn't maintained

        I'm confused - is the driver maintained?

        • No. https://www.highrez.co.uk/Down... [highrez.co.uk] This is the driver. Note that the original authors provided sample code in VB6. All of this vendor-proprietary motherboard whizbang shit no matter the OEM was always built on hacky bit-banging code usually running off of some Ring 0 God-driver hack like InpOut32 that let userspace programs bit-bang the hardware like an MS-DOS TSR. It has always been this way, from the janky drivers for the janky hardware monitoring facilities on my SOYO motherboard from 2003 to my ASU

  • Yes. You're lighting it wrong. All your fault.

Slashdot Top Deals

In Nature there are neither rewards nor punishments, there are consequences. -- R.G. Ingersoll

Close