K4sum1

I thought that was only for .sys drivers, at least until you got to Windows 8 where it got fussy about system dll files afaik. but also idk about this case specifically. So I did that, but I get a code 28 when trying to install the new driver, it fails to install. It just prompts me to install it again right after. However if I tell it to automatically install, it actually installs, but gives me a code 10 and doesn't start. I went back to the old driver to try it on just the RTL8125, and it actually installed, and recognized the device, however it never connected. However, it also made the RTL8126 with the new driver go from code 10, to supposedly working. So I plugged the Ethernet cable into the RTL8126, and a few seconds later, I got this BSOD. So the driver still needs some work, at least to work on my machine. I know newer isn't always better here, and I'd probably trust this collection more than newer drivers as I've had them be unreliable. However I also need to get around to testing this collection so idk. For me, I would use the CFF Explorer Rebuilder to update the checksum, but this looks like it would work and be less clicks. It would be cool to have a reg file to add it to the right click or shift + right click menu for increased efficiency.

Did you happen to change anything else in your provided download? I was curious and compared it to stock newdev.dll + your change. Same file version and all. There are a multitude of differences that someone of my limited skills can't make sense of.

As cool as WiFi card is, I would like to be able to get a working driver to test and finally be able to board swap the RTL8125 PC.

I did some tests, ffprobe appears to be required for some sort of postprocessing pass done when making a mkv. I get an error about it using this command. I like to grab 720p quality with highest quality audio (opus) with metadata for archival. (I would like to improve it so it asks for width <=1280 and height <=1280 so all videos are in 720p and shorts aren't grabbed in 360p, but it works well enough.) youtube-dlp.exe -t sleep --js-runtimes quickjs -f "(bestvideo[height<=720]+bestaudio/best[height<=720])[vcodec!*=av01]" -o "%%(channel)s [%%(channel_id)s]/%%(upload_date)s - %%(title)s - (%%(duration)ss) [%%(id)s].%%(ext)s" --merge-output mkv --write-info-json --write-description --write-thumbnail --add-metadata --write-sub --all-subs --embed-subs --mtime --retries infinite "https://www.youtube.com/watch?v=7VN6OyGcaL4" Fair, but have you never wanted to drag a video onto ffplay.exe and watch it play the video without having anything installed? I think it could be useful. I'm not saying it's required to be built, but also maybe keep a full release build if you do build it.

tbf ffmpeg is an all in one library, some programs may only use ffplay. So it might be a good idea to have a normal build for those applications.

I'd like to build an application with XP compatibility, but it bundles ffmpeg. I originally tried downgrading the ffmpeg source, but I hit a bit of a brick wall when compiling and I couldn't figure out how to fix. I think I might have found a working w32pthreads.h patch though. For yt-dlp, ffprobe.exe is also required I think. I have both ffmpeg.exe and ffprobe.exe in my yt-dlp directory.

I'm curious how you worked around w32pthreads.h? It calls for multiple Vista+ functions like SRW locks and such, but I find no mention of the file in any patch.

Yeah, the topic name was fine, this should have been split to a new topic.

I would like it for my own use, in case I need it. I made this fix for it based off of the changes comparing your 1.62.1 and 1.45 patches, and it compiled, but I haven't tested it yet. https://github.com/Eclipse-Community/eRust/commit/619ed0e48e3e0a6602727dd58918a7783c2d422c I haven't tried building Mypal68 in a few days, mostly just been managing and juggling commits and building rust 20 times. I don't know what to use or where to get it, so I've just been waiting to see if you could share your tools.

I found you released the proper 1.45 patch under an issue, so you linked to the 1.35 patch until you replaced it. Also the patch linked for 1.52 only applies cleanly to 1.56. Were you considering 1.56 at some point? I ported the 1.56 patch to apply to 1.52, but I do wish I could get a clean 1.52 patch in case something's missing or I did anything wrong.

@feodor2Can you 7z up and send me your C:\DEV folder similar to how I've done this with VS2019? https://archive.org/details/funny-windows-development-kit I can't use my FWDK itself with Mypal68 as it complains about some error that I've forgotten. So I went to use EWDK, but you seem to have modified EWDK as using the ISO you linked and unpacking it into the folder results in it complaining about the Win 10 SDK being too old. I've used MSVC14 folder from EWDK and Windows Kits folder from FWDK to compile and that's worked fine from Firefox 68+, however using it to build latest Mypal68 gives me these errors. Not sure tbh, Mypal68 uses the LLVM 9 that you can download and that could be used for reference. For 68, on top of that you need a llvm-config.exe, which you can get from compiling Rust. For 78 and later, llvm-dlltool.exe is also required to be added. Also the listed minimum Clang up to 140 at least is still 9, so maybe I will just continue on and fix what comes. 102 was actually kinda painless, and 115 hasn't been too bad so far. 91 made me think it would be horrible because I had to bisect a bunch. No? I've just been working on it over the months and years. I started at 122, and basically have had a release for every major version. idk if it's possible, I think it will be a lot and very hard going by Mypal68 changes on top of extrapolating to like 140 or so if I were to try it. A lot of the reason I wanted to try this was to have a build system to build older releases that can no longer be bootstrapped. With this, I was able to build 91, and disable WebRender and have Cairo on a (relatively) recent version. https://github.com/Eclipse-Community/r3dfox/issues/32#issuecomment-4262918711

Cool, but this is missing a lot of tools like lld-link.exe making it unusable for r3dfox. If I was to aim for XP, LLVM v9 has worked up to v91, but I get a ton of warns and had to do a few code changes to make it compile. I'd need to upgrade the compiler to go much further than that.

Thank It applied cleanly, and I'm building with mostly the same config I used for modern versions, so I'm more hopeful that it compiles correctly and works than 1.52. Yes, but that section was unedited. Unless the 1.35 patch also applies to 1.45, which I guess could be the possible but also idk. fair I meant our different implementations. You just completely got rid of it while I made it well a suggestion, sending a warning(s) (it sends it three times lol) and nothing else. idk my method seems just funny taking what would be an unnecessary error into a suggestion and naming it like that. I doubt upstream Rust would want to take it, and I'm alright with my fork's method. They specifically made it fail in the first place.

Actually according to Wikipedia it's the same as 1Gbit for the same type of Cat 5e cable. For 10G, it's half the distance for Cat 6, but the same distance as 1/2.5/5 for Cat 6e. I've run 10G over Cat 5e for shorter ~25ft runs and have had no issue before. I guess it could be, however if you want high performance plus working ASPM and stuff, I don't know of any other option. Of sorts, something that I can locally host on my own network since my connection to the world is only a gigabit. (It feels weird to call that just only when so many people have a lot less, however idk how else to word it) I ran that inside a Linux VM on my PC and pointed the browser to it and ran it. iperf3 would be better, but I wasn't able to find a release that ran on (stock) XP, and I didn't feel like recompiling it for XP just to run the test. Although I might do that soon just to see the max theoretical. I think I have a X540-T2 somewhere. Will have to dig for it, but I could do a test and compare them. Good opportunity to bother with compiling iperf3 too.

I mean Firefox 149 is the latest version, and modern. r3dfox is on version 149 and soon 150. So I went to try that, and I noticed there's no link to the patch. I then checked the history and found you uploaded it, but messed up the markdown so it didn't appear as a link and I guess is invisible. Except when I copy and paste the link, GitHub just says not found. Can you reupload the patch pls. Also if you could, it would be nice to have the 1.45 patch (you only ever released 1.35 and 1.52), and a fixed 1.52 patch in case I want to build and test an older release. Also I find it funny you went and completely commented out the Cargo checksum system, while I just turned it into a warning and giving it the name, t h e s u g g e s t i o n https://github.com/Eclipse-Community/eRust/commit/4977b2926b2c8234fdfd619089b0a22466506e68 I was about to say I wasn't sure and that I never tried it, since I saw it was false when I was building. But I checked the config.toml for the newer Rust compiles I've made, and they all have been compiled with ninja. I don't have any scientific data, but I guess it works for me. I'll try building 1.62 with my config and see how it goes.

I wonder if the server got compromised and the htaccess or some other file defining access to the website was made to redirect users of certain browsers to the spam website? I think that can filter based upon user agent.

Thank you. I gave this a try. First though I noticed you included a COMPARE_8125x64.txt and here's my replies to it. Current build I decided to only do for 8125 since it (supposedly) works on it, and I figured I would need to basically remake the driver for 8126 to work with it, and make another for 8127. I didn't think it would be so relatively few code changes to combine them all. I'm not even sure what the purpose of the cat file is since there is no cat file that exists here. Also I figured I'd just ship a different inf for x86 and x64 so I didn't try to combine them. I built it with the Win 7 DDK, but I built it with a Win 7 machine. Could that be why those versions are 6 and 1? Should I try building it on XP? So I have the same issue with this driver, and this is what I found in setupapi.log for it. [2026/04/12 16:28:27 260.255] #-198 Command line processed: "C:\WINDOWS\system32\mmc.exe" "C:\WINDOWS\system32\devmgmt.msc" #I060 Set selected driver. #-019 Searching for hardware ID(s): pci\ven_10ec&dev_8125&subsys_7c911462&rev_04,pci\ven_10ec&dev_8125&subsys_7c911462,pci\ven_10ec&dev_8125&cc_020000,pci\ven_10ec&dev_8125&cc_0200 #-018 Searching for compatible ID(s): pci\ven_10ec&dev_8125&rev_04,pci\ven_10ec&dev_8125,pci\ven_10ec&cc_020000,pci\ven_10ec&cc_0200,pci\ven_10ec,pci\cc_020000,pci\cc_0200 #-019 Searching for hardware ID(s): pci\ven_10ec&dev_8125&subsys_7c911462&rev_04,pci\ven_10ec&dev_8125&subsys_7c911462,pci\ven_10ec&dev_8125&cc_020000,pci\ven_10ec&dev_8125&cc_0200 #-018 Searching for compatible ID(s): pci\ven_10ec&dev_8125&rev_04,pci\ven_10ec&dev_8125,pci\ven_10ec&cc_020000,pci\ven_10ec&cc_0200,pci\ven_10ec,pci\cc_020000,pci\cc_0200 #-019 Searching for hardware ID(s): pci\ven_10ec&dev_8125&subsys_7c911462&rev_04,pci\ven_10ec&dev_8125&subsys_7c911462,pci\ven_10ec&dev_8125&cc_020000,pci\ven_10ec&dev_8125&cc_0200 #-018 Searching for compatible ID(s): pci\ven_10ec&dev_8125&rev_04,pci\ven_10ec&dev_8125,pci\ven_10ec&cc_020000,pci\ven_10ec&cc_0200,pci\ven_10ec,pci\cc_020000,pci\cc_0200 #-019 Searching for hardware ID(s): pci\ven_10ec&dev_8125&subsys_7c911462&rev_04,pci\ven_10ec&dev_8125&subsys_7c911462,pci\ven_10ec&dev_8125&cc_020000,pci\ven_10ec&dev_8125&cc_0200 #-018 Searching for compatible ID(s): pci\ven_10ec&dev_8125&rev_04,pci\ven_10ec&dev_8125,pci\ven_10ec&cc_020000,pci\ven_10ec&cc_0200,pci\ven_10ec,pci\cc_020000,pci\cc_0200 #I022 Found "PCI\VEN_10EC&DEV_8125" in c:\new\RTL8125.inf; Device: "Realtek RTL8125 2.5GbE"; Driver: "Realtek RTL8125 2.5GbE"; Provider: "Dietmar"; Mfg: "Dietmar"; Section name: "RTL812x.ndi.NT.5.2". #I087 Driver node not trusted, rank changed from 0x00002001 to 0x0000a001. #I023 Actual install section: [RTL812x.ndi.NT.5.2]. Rank: 0x0000a001. Driver date: 04/12/2026. Version: 0.0.0.16. #-124 Doing copy-only install of "PCI\VEN_10EC&DEV_8125&SUBSYS_7C911462&REV_04\6&52F7D12&0&0048020A". #W360 An unsigned, incorrectly signed, or Authenticode(tm) signed file "c:\new\rtl8125.inf" for driver "Realtek RTL8125 2.5GbE" will be installed (Policy=Ignore). Error 0xe000022f: The third-party INF does not contain digital signature information. #-024 Copying file "c:\new\amd64\rtl8125.sys" to "C:\WINDOWS\system32\DRIVERS\rtl8125.sys". #W360 An unsigned, incorrectly signed, or Authenticode(tm) signed file "c:\new\rtl8125.inf" for driver "Realtek RTL8125 2.5GbE" will be installed (Policy=Ignore). Error 0xe000022f: The third-party INF does not contain digital signature information. #-166 Device install function: DIF_REGISTER_COINSTALLERS. #I056 Coinstallers registered. #-166 Device install function: DIF_INSTALLINTERFACES. #-011 Installing section [RTL812x.ndi.NT.5.2.Interfaces] from "c:\new\rtl8125.inf". #I054 Interfaces installed. #-166 Device install function: DIF_INSTALLDEVICE. #E161 Processing of call to class installer failed. Error 87: The parameter is incorrect. I tried testing the x86 driver on XP x86 since it didn't work for me there with the old driver, but I encountered a code 10 on both the 8125 and 8126 I'm testing with. I was also curious about more modern Windows versions, so I tested 8.1, and get a code 31 from both. Would be nice to have something working for the 8126/8127 since it has no native 7/8 driver. Also I think the compiled driver is supposed to automatically copy to the amd64 and x86 folders, but that didn't happen on my end. I had to manually copy them there.

@feodor2 I've gotten around to try building Rust 1.52 with your patch, and the patches don't apply cleanly to library/std/src/sys/windows/c.rs, fs.rs, and rwlock.rs. Looking at rwlock.rs, I see an entry for a StaticRWLock, which is something that only came later in 1.54. Can you send any other patches you may have applied to your Rust source code pls? Also can you send me your config.toml pls

I made a GitHub page for it since it helps me keep track of and organize changes better. I tried compiling as a x64 driver, and it compiles. I tried modifying the inf to install on x64, but I get the same "The parameter is incorrect error" that I got on Server 2003 and I have no idea how to fix please help. https://github.com/the-r3dacted/RTL812 X-WinXP/issues/1 8125x64.7z

I wanted to compile the driver and I decided your one might be better to do this with. However I noticed your link doesn't exist anymore and I thought I had downloaded it, but I can't find it. Can you reupload pls.

This didn't sound correct to me, so I did some testing with a 10G card that I know works with XP. Using locally hosted OpenSpeedTest to test local network speed. That's certainly more than 1.4Gbit. Also I'm not sure if it's the browser or what but these results are pretty low for being contained in my 10G network. Downloading files from my NAS with Filezilla I saw up to 600MiB/s which is the limit of the SATA SSD I have in the machine.

That's so cool, but I tried using IDA Pro before and I wasn't able to figure it out. Do you think you could mod it for me pls. I also wish they would release that WiFi driver on the forum, I know someone that could benefit from it, and I have an AX201 sitting around with no use myself.

The point isn't to connect to the outside world at high speeds, it's to communicate inside my network at those high speeds. Across computers on the same local network. Also outside of that, I mean modern boards are using this chip now. If I'm to upgrade to AM5, my top two boards use the RTL8126. So I'll want this to work. I think since Dietmar got RTL8125 working, I might be able to get RTL8126 working, but haven't tried looking at the driver. I'd need to figure out why x86 driver crashes my Ryzen system with a RTL8125 , and/or compile x64 driver to test with XP x64 to develop with. How is that kinda bad?

Tried it. and got a code 37.