Andalu

@Dietmar Tried on a MSI Z590-A board with a Crucial P5 NVMe 500GB drive. Here what I did: - XP installed on the HD drive; - inserted the NVMe drive into the M.2 slot and installed the Microsoft NVMe driver v2.0 (NVMe drive recognized correctly); - cloned the HD drive to the NVMe drive using a bootable USB stick of Aomei Partition Assistant Pro v8.6 (prepared in Win10 with WinPE); - removed the NVMe drive from the M.2 slot and inserted into the NVMe to USB adapter; - applied the RMPrepUSB settings as per the "Tutorial.txt" and 6 files copied to the NVMe root replacing existing ones; - re-inserted the NVMe into the M.2 slot as the only drive connected to the board; - XP only boots via NTLDR.sys, since choosing to boot via the ReactOS loader I got this message: "Could not load the SYSTEM hive!". Did I make a mistake or miss something?

@Dietmar Thank you for the explanation. Tomorrow, I'll try your method.

@Dietmar Hi, could you please make a step-by-step tutorial for boards with legacy NVMe support as well?

@Dietmar A tool that performs trim on NVMe in XP would be there and it is Adata SSD tool v1.2.0 that works with the Silicon Motion driver. Although it does not accurately report smart data: the trim operation seems to be completed without problems: The question is whether how effective is the trim on the NVMe drive since I don't know a way to perform a certain verification because other tools (for example TrimCheck) also provide contradictory information.

@pappyN4 Hi I tried several times with two XP Integral Edition ISOs (with Microsoft and Silicon Motion drivers) to replicate your issue but I have encountered any problems. It worked every time I chose the NVMe disk partition on which to install XP, highlighted the "quick format" and then pressed the "ESC" key to return to the partition choice screen. In the past few days, however, while installing XP with the MS NVMe v2.0 driver once the system froze at 20% of formatting. Restarting and repeating the installation all went well. Just for fun and if you have some time, could you try the modifications I linked here? Unfortunately, the method you described in your Github guide doesn't work on my systems.

So there is no possibility to get other drivers to use in XP via emu_extender? I also have one more question To get the win8_uaspstor.sys driver to work properly with win7_storport.sys v6.1.7601.25735 I applied a modification to storpor8.sys at offsets 3BE2 and 3BF8 by changing respectively 'ScsiPortMoveMemory' to 'StorPortMoveMemory' and 'ScsiPortNotification' to 'StorPortNotification'. To avoid conflicts with drivers already in the system, I renamed stoport.sys and storpor8.sys to something else and also the hexadecimal value referring to storport.sys within storpor8.sys itself. This mod also avoids system crashes when you reboot or shutdown the system with a UAS device still connected to the USB port as already reported some time ago by @pappyN4. I have tried the UAS driver on two intel systems for two months using all my UAS devices with no "apparent" problems detected. Here is the question: what can I test to be sure that the driver thus modded can be safely applied and that it is not instead an insane modification (and wasted time)? Thanks

@George King Thanks Edit: I tried the Generic SD/MMC driver from Windows 8.0 build 8056 and it installs without issues on my old laptop (intel core2 duo): https://ibb.co/fdTYdCN

Where can I get the latest master? I'm still using the extender from February 2021 with included 80 functions and now I read that the one with 81 functions is old

The use of a USB-Ethernet converter does not allow the connection between the two PCs. It has already been discussed here and in following posts.

@Dietmar I just copied 76 files with variable size from 1.9 to 2.9GB for a total of 201GB from an external USB3 disk (connected to an 'intel' 5GB USB port) to a SATA to USB3 adapter (connected to an 'intel' 10GB USB ports) without any block. The files were collected in groups of 4 or 5 files. Each group was copied one after the other, overlapping the copy of one group while the copy of another was underway. The copy operation has never been interrupted. After about 180 GB, the copy speed was drastically reduced but at least no issue has occurred. I will do other tests to get more information. Edit1: 201GB package just copied from the sata to USB3 adapter (connected to an 'intel' 10GB USB port) to the internal 970 Evo NVMe without issues in 20 min. Edit2: 201GB package copied from the internal 970 Evo NVMe to the external NVMe to USB3.2 adapter without any block in 8 min. In both cases, files copied starting from the smaller one up to the bigger one. In my previous tests, the block occurred by copying larger files also (7/8/9/10 GB and over) without a pre-configured order. Edit3: previuos test repeated but now copying files without order and again no issue detected. This time I mounted a different NVMe disk on the USB3.2 Gen.2 adapter. Edit4: new test by copying 41 files for a total of 241GB from a NVMe to USB 3.2 Gen.2 adapter (RTL9210B chipset) to another NVMe to USB 3.2 Gen.2 adapter (RTL9210B chipset): also in this case the copy was completed without any crash in 14 min. Edit5: other test by copying at same time 41 files (241GB) from the NVMe to USB adapter (RTL9210B chipset) + 277.863 files (6.4GB) from a sata to USB adapter to another external NVMe to USB adapter (RTL9210B chipset): this time I got the insufficient resources error after the copy of the 277 files was completed while the copy of the other files stopped at 23 files and 149GB. Test repeated once again using another NVMe to USB 3.2 Gen.2 adapter (RTL9210 chipset) as destination disk and the crash occurs at same point as in the previous one, when the copy reached exactly 160.689.047.447 bytes (149GB). It's too strange, there must be something else that escapes me (clearly, the maximum number of files on the NVMe disk or in a folder has not been reached). Edit6: with a different NVMe enclosure (JMS583 chipset) as destination disk, the copy of the 41 files (241GB) stopped before at 22 files and 135GB written while the simultaneous copy of 277.863 files (6.4GB) was already completed as in the previous tests.

@Dietmar it would be interesting to understand why only once the copy of 224GB ended without problems while in the other my 5 tests the usual message "insufficient system resources exist to complete the requested service" appeared. In my tests the block occurred randomly, sometimes after a few GB while another time when there were only 20GBs upon completion of the copy of 224GB. I got the crash also with the Etron UAS driver. P.S.: tests done on Q370 and H470 systems for the intel USB3 controller with installed the ported USB3/UAS driver from Win8 by copying files from an external NVMe to USB 3.2 Gen.2 adapter (connected to a USB 10GB port) to an internal NVMe and vice versa. The copy successfully ended was on the Q370 board from the external NVMe to the internal one.

bsod 0xA5 (0x0000000D, ..., 0x4449555F, 0x0) solved on asrock H310M-HDV with acpiV5 (and no other "non-acpi" bsod detected). Thanks

@Dietmar @Damnation thanks for the info. It would have been a good alternative, much more convenient and practical thanks to its portability. With a single PCI realtek card I'm forced to disassemble and reassemble repeatedly the machines with an Intel i219 chipsets onboard.

@Dietmar Hi, do you have never used an USB to ethernet adapter for debugging? I am trying one with realtek 8151 chipset (driver 2018 installed) but even disabling the onboard LAN controller in bios and changing the IP address in the boot.ini for the TARGET system, the debugging doesn't start anyway. Can there be a solution or is it better not to waste any more time on this experiment?

@Mov AX, 0xDEAD I'm trying the compiled ACPI v4 update on some intel boards. Until now it worked on GA-H470, Asus Prime Q370 and Lenovo IH110MS. I got the BSOD 0xA5 (0x0000000D, ..., ..., 0x0) for the Asrock H310M-HDV also by installing XP from a pendrive. Here the debug log: https://anonfiles.com/p63dLbabyd/H310M_7z Edit: luckily the asrock board has a realtek 8111 chipset controller Edit2: the acpi V4 also works on Asus Prime B250 and Asrock B250M Pro4 (on the lattest the restart still doesn't work with a PS/2 keyboard connected but I noticed a little changing: the system no longer blocks now on the "windows is shutting down" message but the screen becomes black even if the system is still ON and blocked).

@Mov AX, 0xDEAD taking a look at other resources, I noticed that there is a conflict related to the PCHRESV. But I have not found any reference for the memory range involved in any other devices/resources, included the 'PCI device' indicated. Below two images: https://anonfiles.com/ZdMbV7Zexa/Resources1_png https://anonfiles.com/j3N2V6Z4x0/Resources2_png Edit: it seems a common presumed problem also for my other boards (Q370 and Q87) and probably for the others not yet checked.

@Mov AX, 0xDEAD here the shot for the resource tab: https://anonfiles.com/9aI0TaZ1x0/ResourceTab_png

@Mov AX, 0xDEAD the A5 bsod (0x03, ***, 0xC0140008, 0x494E495F) was solved with one of the first acpiv2 releases. It had worked for all my boards except the GA-H470 for which I still got the a5 bsod with code (0x11, 0x08, ***, 0x20160527), as already reported here: https://msfn.org/board/topic/183464-compiling-acpi-v20-driver-for-windows-xp-sp3-x32-only/?do=findComment&comment=1216248 Now, also this bsod is solved with the acpi.sys V4. So far, following your guide I was able to get only acpi.sys but modifying its code is still out of my reach, at least until now it is so .... never say never Thanks

@Dietmar thank you so much, It worked Also tried by installing XP-IE from scratch, no issues detected. The only yellow mark that resists every version is this: https://anonfiles.com/7ds2R2Z2x0/Yellow_mark_resource_png Not a problem, it is there since I have the GA-H470 but it never bothered me Edit: image updated

is there another acpi.sys to try? The last acpi.sys working version is from @Dietmar with DBhack. Without the hack I got A5 bsod (0x11, 0x08, ..., 0x20160527)

@Mov AX, 0xDEAD below the new log, but it's only 640.557 byte: https://anonfiles.com/Pbja4aZ3xe/H470_checked_2_TXT

@Mov AX, 0xDEAD here the checked version: https://anonfiles.com/v0yex2Zdxc/H470_checked_TXT @Damnation thanks for uploading the acpi checked version. I think the above debug could be correct.

Thanks to @Damnation @Dietmar @Mov AX, 0xDEAD (in alphabetical order ) This should be my first debugging log without (apparent) errors: https://anonfiles.com/p0YelbZ3x4/H470_TXT Since I probably made something wrong, please let me know what is missing. Edit: I didn't see any classic BSOD, the screen become black just entering "g" after the command "bu acpi!DriverEntry". The acpi.sys on the TARGET system has SHA-1: 166A17FFE3421083E81A3A7CF9E01D7CD63F52A7 Here an image from the HOST system: https://anonfiles.com/t4Q5m8Z0xf/Image_1_png

@Dietmar the WinDBg folder already contains the modded files as for the Blackwingcat msvcrt.dll which is already in the system32 folder.

@Mov AX, 0xDEAD I'll try the new advice soon, thanks.