Jump to content

Jakob99

Member
  • Posts

    340
  • Joined

  • Last visited

  • Days Won

    1
  • Donations

    0.00 USD 
  • Country

    United States

Posts posted by Jakob99

  1. 24 minutes ago, kar1 said:

    I managed to finally create a .inf file for ACPIEx that makes it bind to a virtual ACPI device.
    I took Microsoft's loopback network adapter .inf file (msloop.inf) as example to write a new one.

    After boot the event log no longer shows that the driver couldn't load due to unavailable hardware.
    However it just silently fails to load, even if I bind it to a valid ACPI HWID (ACPI_HAL\PNP0C08).

    ACPI_HAL\PNP0C08 is the ACPI HWID that the normal 'Microsoft ACPI-Compliant System Device' is bound to.
    And acpiapic is the one that the computer itself uses, but acpiapic doesn't use need driver to work.

    I thought that adding a new legacy device with the Device Manager and giving it a new .inf that uses the same ACPI HWID could work.
    However the driver still not loading seems to hint at a bigger issue, which is that we will need to backport KMDF 1.13 to Windows 7.

    This has been done in the past with KMDF when Windows 7's KMDF 1.11 was backported to Windows XP.
    We have to essentially backport KMDF 1.13 to Windows 7, by finding out where all its files are on a Windows 8.1 OS.

    For registering the new KMDF 1.13 with Windows 7, we can  make a clean Windows 7 SP1 install, then make a snapshot of before & after the official KMDF 1.11 update is installed.
    This will reveal in which places of the registry we need to register a new unofficial version, and which kind of data we need to write.

    For the current ACPIEx driver it's uploaded here, but it still doesn't work:

    https://www.udrop.com/KqcD/ACPIEx___SpbCx-Win7-testing.zip (md5: a9367f0922e9f95a5ae1b14aca75ec58)

    ACPIEx & SpbCx-Win7-testing.zip 215.59 kB · 0 downloads

    Interesting. Idk where the KMDF files would be stored at on a Windows 8.1 and 7 installation. Maybe someone here could help out.

  2. 3 hours ago, kar1 said:

    @Jakob99 The ACPIEx.sys driver when I tried it myself didn't run at boot because it has no HWID to bind itself to.
    I don't know which HWID it's supposed to be installed for on a real Windows 8.1 system (e.g. ACPI_HAL\PNP0C08).

    Can you boot your computer into a Windows 8.1 or newer WinPE and share screenshots of the Device Manager's ACPI x64-based PC device?
    It will also help to share more screenshots about a System Device named 'System board' from the Device Manager.

    It's for the 'Parent' property and potentially some other device properties.

    But in any case, I think the only remaining part to be modified for the ACPIEx driver is the ACPIEx.inf file.
    If it gets correctly rewritten with a real Windows 8.1 HWID for ACPi, then the SpbCx driver will be able to run.

    And since Intel Serial IO drivers require the SpbCx driver as a dependency, this will make them work too.

    I got a Code 37 (Windows cannot initialize the device driver for this hardware) upon installing your Serial IO drivers. I followed the steps in the provided ReadMe, too.

    I found this Device ID in Windows 11's Device Manager (found by clicking on Serial IO A369, Details, and Parent): ACPI\PNP0A08\0, but idk if it's the correct one for ACPIex. If it's not, what would ACPIex be known as in the Device Manager?

  3. 13 hours ago, kar1 said:

    I looked it up deeper and the driver fails to load likely because of the lack of a KB3125574 Windows update:
    https://www.catalog.update.microsoft.com/Search.aspx?q=KB3125574

    This update brings a WdfLdr.sys driver that the Intel HID Event Filter relies on.
    However it should work without the KB3125574 update by taking it from a Windows 8.1 system.

    I took your Intel HID Event Filter drivers and modified them, the list is as follows:

    • Removed all invalid INF comment prefixes // and replaced them with ; instead
    • Added my compiled ntoskrn8.sys & WdfLdr.sys from Windows 8.1 to the INF files section
    • For HidEventFilter.sys & WdfLdr.sys I modified the EXE/SYS file headers to use MajorVersion 6 & MinorVersion 1
      Because it was MajorVersion 10 & MinorVersion 0 (6.1 vs. 10.0 for Win10)
    • Modified the import filename in both HidEventFilter.sys & WdfLdr.sys from ntoskrnl.exe to ntoskrn8.sys
    • Modified the __security_cookie in both SYS files & bypassed the security cookie check in driver code as well
    • Removed the Security Directory information from EXE/SYS headers for both HidEventFilter.sys & WdfLdr.sys
    • Updated all the EXE/SYS file checksums in the headers for all files
    • Created a new unsigned HidEventFilter.cat file with '6.1' OS attribute (verifies all .sys files & the .inf file)
    • Included dpinst64.exe from Microsoft (version 2009 for Windows 7 64-bit)

    Prior to installing the modified drivers, you need to remove all the previous Intel HID Event Filter drivers with DriverStore Explorer:
    - https://github.com/lostindark/DriverStoreExplorer

    This allows for removing old drivers that otherwise get reinstalled again instead of the newest modified drivers.

    The modified Intel HID Event Filter drivers (version 3):
    - https://www.udrop.com/KopH/HidEventFilter-Win7-testing-3.zip (md5: 765c950f421cd7a9b7a353e8e980487c)

    For the 0x0000007E BSOD error code, it doesn't actually mean much other than 'Windows had a trouble while running this program/driver'.
    It's a generic Windows error code rather than one that actually gets sent by the Intel HID Event Filter itself.

    Try the modified version after having removed all previous ones with DriverStore Explorer, I think it should work.

    You can also enable testsigning & easily self-sign the drivers yourself with Driver Signature Enforcement Overrider:
    https://www.softpedia.com/get/Tweak/Video-Tweak/Driver-Signature-Enforcement-Overrider.shtml

    Just sign the HidEventFilter.cat file only if needed.

     

    HidEventFilter-Win7-testing-3.zip 397.69 kB · 0 downloads

    Alright, so I tried your driver, and Intel HID Event Filter driver installs successfully with no error! Unfortunately, the touchpad still does not work as the Intel Serial IO drivers do not want to install. I tried the ones here https://www.mediafire.com/file/23np1m85w5za00a/Serial_IO_7th_Gen%2B_Win_7.zip/file but when I install them, I get a Code 31 error (Windows cannot load the driver for this hardware) or whatever that error is called. These are Skylake Serial IO drivers modded to include my 8th Gen Serial IO. Idk why this Windows 7 Skylake Serial IO driver isn't working as the Windows 8.1 Skylake drivers I modded work just fine under 8.1, so theoretically, the linked drivers should work, too! This should be the last hurdle in getting the touchpad to work under 7. I appreciate anyone's help in getting those drivers to work without producing a Code 31 error.

     

    EDIT: PCI\VEN_8086&DEV_A368&SUBSYS_08311028&REV_10 and PCI\VEN_8086&DEV_A369&SUBSYS_08311028&REV_10 (Intel Serial IO A368 and A369). Forgot to include my Device ID's for reference.

  4. On 7/19/2023 at 1:19 PM, kar1 said:

    There's a DSDT & SSDT editor (IDE) on Github for this purpose:
    https://github.com/ic005k/Xiasl (block it in the firewall to prevent its Internet check)

    Another older DSDT Editor:
    http://web.archive.org/web/20161108180259if_/http://olarila.com/dsdt/DSDTEditor_Linux_Windows.zip

    You should then be able to edit your SSDT/DSDT and load patched versions using a boot-time loader:
    https://github.com/jslegendre/ACPIPatcher

    Or you can also use Clover, OpenCore, or other loader of choice.

    Some quick tips:

    • CondRefOf function is also used to verify the return of _OSI
    • "Windows 2009" is Windows 7 & "Windows 2013" is Windows 8.1
    • Lookup "If (_OSI" or "If ((_OSI" to quickly locate the blocks that have a system check
    • WIN7 is a constant that cans also be searched
    • ACOS & ACSE variables seem to matter, they have to also be modified
    • OSYS is the variable that gets checked the most in the OS installed locks
    • Vista is named with the WLG constant
    • There's a DRS3 variable but only set for Windows 10 RS3 & later

    I made a patched version of your DSDT that makes Windows 2006 & 2009 (Vista & 7) behave as if it was Windows 8.1 (2013).

    It also returns Windows 8.1's OSYS value for Vista & 7 which will be able to unlock more ACPI objects.
    https://www.udrop.com/Kn0y/dsdt-patched.zip (md5: 4d22eb9bedc63efb618aaeef84ffb9db)

    Try using the normal version and if any problem then the 'no-optimization' one.

    I also compiled GeorgeK1ng's version of NTOSKRNL_Emu incase you need to mod the Intel HID driver.
    https://www.udrop.com/Kn0H/NTOSKRNL_Emu-GeorgeK1ng-version.zip (md5: a054a8ba9d05e587f3ac0f409d9e25e5)

    dsdt-patched.zip 423.82 kB · 1 download NTOSKRNL_Emu-GeorgeK1ng-version.zip 83.56 kB · 1 download

    Hi! I tried your modded DSDT table. The Intel HID Event Filter driver now shows up as an Unknown Device within Device Manager, however, I still get the 0x07E BSOD when attempting to install the driver. This happened with Snappy Driver Install, as well as the official Intel installer (which no longer throws "This computer does not meet the minimum requirements" when launching it). Below is what should be BSOD logs as well as the Intel driver, which is unmodded as it already supports my device and Windows 7 (as nothing has changed with the device from Skylake to 8th Gen Coffee Lake). And yes, I even tried the GeorgeKing thing, and no changes :(

    BSOD logs: https://www.mediafire.com/file/l9cpkpm0adh1k4c/73BSODEVENTFILTER.zip/file

    Intel HID Event Filter driver, unmodded: https://www.mediafire.com/file/7y8y4i36t71qtqr/HIDEventFilterDriver-1.1.1.318-Windows_RS1_Certified.zip/file

    Any help on figuring out why the driver still BSOD's even after using your modded DSDT is greatly appreciated!

  5. On 6/24/2023 at 6:41 AM, ExtremeXT said:

    Dump your ACPI tables with acpidump from iASL and send them in a zip. Changing _OSI values to allow unsupported OSes is easy enough. Then the modified table will have to be injected with something like OpenCore which can be used for Windows as well.

    Sorry for the late response. Got sidetracked by the 4th of July holiday. Anyhow, I have uploaded the ACPI table (dsdt.dsl and dsdt.aml) here: http://windowsenthusiasts1.epizy.com/Downloads/dsdt.zip

    There were dumped from the Windows 11 install on the same laptop (as opposed to the Windows 7 installation), but I see no reason these shouldn't work. If, for some reason, you need them dumped from the 7 install, please do let me know and I will do so.

    EDIT: If you get a "File not downloaded, potential security risk" message when downloading this, just ignore it and allow the download to continue. Idk if Chrome will throw this error, but Firefox can/will.

  6. I may be closer to solving this problem! It appears that the issue is this: https://winraid.level1techs.com/t/win-7-64-not-working-touchpad/94960/19?u=moline

    I can gather that the check on this laptop checks if you have Windows 8.1 or above, and if you do, the touchpad will be enabled and allow successful installation of the HID Event Filter Driver, and thus, the Serial IO drivers. If you have Windows 7, the check will fail and disable the touchpad entirely, meaning Windows 7 won't know it exists, and one of two things will happen when you try to install the HID Event Filter driver: 1. The setup.exe will tell you your computer doesn't meet the requirements (This computer does not meet the minimum requirements for installing this software), which is expected as the OS, and thus the app, do not know that the touchpad exists, or 2. you get a 7E BSOD with IntelHidEventFilter.sys identified in the stop screen when you try to install it over a HID driver from an external device, like USB Wireless Mouse.

    To fix it, the following will need to be done: https://winraid.level1techs.com/t/win-7-64-not-working-touchpad/94960/33?u=moline

    I'm still working on deciphering this. Basically, if the DSDT table is extracted, we can modify the check so that Windows 7 will pass and allow the touchpad to be seen within the OS, and allow it to function after installing the aforementioned drivers. The way to do this is commonly used for Hackintosh installs as Macintosh will fail the check no matter the version, meaning you'd have to edit the check so that Darwin (Macintosh) pass and allow the touchpad to work under it; since I'm not dealing with Hackintosh, we just have to adapt the tutorial for Windows use.

    I'll report back later.

  7. I can't believe I haven't posted this here yet, but K4sum1 managed to get AMD Radeon Graphics drivers, for use with AMD hardware such as Stoney Ridge, working under Vista, meaning we get Aero Glass now! This also opens up much newer laptops such as Acer Aspire A315-21 from around 2017-2019, meaning you don't have to go Ivy Bridge or 2015 AMD. All of the AMD Chipset drivers will work (you will need the latest one that supports Vista), but you will need to mod each INF to include your device ID, and you will also get many unsigned driver warnings, but they all work from AMD SATA to AMD USB 3.0. The AMD TPM 2.0 driver causes a BSOD (PSP 2.0 does not), so do not try to install it. The WiFi card and Bluetooth adapter will also not support Vista, so you'll either need to use USB WiFi, Realtek Ethernet, or swap out the WiFi card for a vista capable one.

    Out of curiosity, everything even works under XP except for the WiFi/Bluetooth driver as well as the graphics driver. You will, however, need a modded ACPI file. And a sidenote, I've never been able to get the VESA graphics driver to work on my Aspire A315-21, even with copying that one file manually. I do not know why it won't work despite doing it all seemingly correctly.

  8. Just now, mina7601 said:

    What you did is editing the main post and the topic title. This will not remove it. If you want your topic deleted, you have to ask a moderator.

    Ah, ok. I thought there was an option for thread creators to delete their own posts. Welp, they can go ahead and delete this since this thread is pointless now.

  9. On 3/17/2023 at 11:24 PM, Brickedandroid said:

    For now, I've tried it, but it's the newer version of Intel HD Graphics 4000 driver on my Ivy Bridge. It works partially, doesn't have any problems at all in Device Manager, but it breaks Direct3D 9 with BEX crash signature. But don't worry, Direct3D 10, 11, and OpenGL are unaffected and still worked perfectly just like normal. You only can't enjoy aero and playing videos using WMP, but you still can play games that use Direct3D 11 or higher (usually released beyond 2011).

    Ah. Does Aero rely on Direct3D9?

  10. 5 hours ago, Brickedandroid said:

    Any drivers designed for Windows 7 or higher can be backported to Vista using NTOSKRNL Emu_Extender.

    For the user-mode components, use Extended Kernel version spoofing.

    Interesting. I don't think I have quite the expertise to do this, but I know someone who may. Would this also apply to post-Ivy Bridge Intel graphics drivers such as for Haswell and later?

  11. 1 hour ago, D.Draker said:

    Technically speaking, it's an error. Windows 2000 existed only as IA-64, not x64 !

    It was a very rare OS shared only with a limited set of manufacturers,

    and only in Western Europe ! You aren't getting a copy !

    Windows Whistler build 2285 exists and it's also IA64, not x64. 

    An example is Fujitsu-Siemens, they developed an IA64 CPU earlier than Intel.

    Articles deleted or never existed in the first place. Found some hard proof .

    Scroll to the bottom and you will see - Scott Bekker is (now most likely was) editor in chief of Redmond Channel Partner magazine.,

    on July 05, 2000 he wrote "..Fujitsu-Siemens also demonstrated a 64-bit Itanium box."

    Again, Fujitsu-Siemens made an an IA64 CPU earlier than Intel, so they needed an IA64 system,

    and the system it ran was IA-64 Windows 2000 Datacenter.

    https://rcpmag.com/articles/2000/07/05/fujitsusiemens-sets-benchmark-on-datacenter-server.aspx

    And again, only for the Western European market ! 

    "Fujitsu-Siemens demonstrated its system today at Microsoft TechEd Europe in Amsterdam."

    Yeah, but regardless, it shouldn't be there as the IA-64 version was never released to the public, only to those specific testers like Fujitsu-Siemens, as you mentioned.

  12. There is nothing wrong with selecting an operating system from within Edit Profile, that works as intended, however, I think there is a mistake with it as one of the choices you can select is Windows 2000 x64 Datacenter, which, from my understanding did not exist outside of the planning stages and a few banners found within Windows "Whistler" XP (alongside the Windows 2000 Personal x64 banner). Hoping for some clarification here.

  13. 22 hours ago, SkyKuker said:

    Greetings, @Jakob99. Unfortunately, your method didn't work for me.

    Everything in the text mode setup worked perfectly until the "Setup is starting Windows 2000" portion. Right after that, the installation hangs (if not pressing F5, F6, or F7).

    If I hit F7 (as far as I've been concerned, it swaps the ACPI HAL with the standard one), the setup barely throws me a 0x7B BSOD.

    If I hit F5 (select the type of PC) and select some of them, it either hangs (ACPI types in general) or greets me with a BSOD (non-ACPI) too.

    Could you please look into this question?

    Regards, SkyKuker (Kuks).

    What's your Intel SATA driver device ID (you can find this under IDE ATA/ATAPI controllers).

  14. On 3/5/2023 at 10:02 PM, blackwingcat said:

    http://blog.livedoor.jp/blackwingcat/archives/2029423.html

    I creatd it.

    But it seems to be supported windows 2000 , so I modified inf file only.

     

    How about version 9.2 driver ? :3

    Sorry for the double post, but I just wanted to say thank you to blackwingcat for all the amazing work he has done in getting Windows 2000 to work on modern hardware such as Sandy Bridge, Ivy Bridge, and even Haswell. Thanks also goes to Realtek for keeping 2k support for their audio and Ethernet (whether they forgot about it or they really do not change much when they "update" the driver for a new Windows release) and also to ALPS Touchpad for allowing the drivers to be installed onto 2k with a simple INF mod, and also to the good folks at win-raid for supplying me with the modded BIOS that made my installation possible!

    I do have a few bug reports, however. After installing the Atheros driver, the CPU usage increases. This was not problem prior to driver installation. I think I found a temporarily fix for this, which is to disable the network adapter unless I need to browse the web with MyPal until hopefully a more viable solution can be found.

    And there is also a weird caveat with audio drivers. You MUST install any audio drivers before installing Extended Kernel/Core. This includes: Realtek audio to even USB Bluetooth (make sure to sync your Bluetooth audio device after installation so the Bluetooth audio drivers can install. I think you'll be fine if you connect to a different audio device after installing the Kernel/Core, but, this has not been tested/proven as I only have one Bluetooth Audio device that I use) and other audio devices that install drivers. If you attempt to set any of this up after setting up Extended Kernel v31 and Extended Core 16b, it will not work even though the drivers do not throw any error codes (such as code 10) in the device manager. Realtek audio stops working (the computer reverts to the no audio driver installed beep when dialog boxes pop up) as does Bluetooth audio as well as Intel Display audio (this, afaik, can only be installed after kernel/core and will always cause this error until it is fixed). Idk why this happens. I'm hoping blackwingcat can figure out a solution. In short, audio drivers such as Realtek and Bluetooth audio installed prior to Kernel/Core will work, and will continue to work even after you install the kernel/core, just don't install the Intel Display Audio until bwc finds a solution and releases a kernel update for it.

×
×
  • Create New...