Dietmar

@reboot12 Here is first try of translating the Source Code for ME from OpenBSD 79 to XP Dietmar 1. Früher poll184_14-Startpfad bleibt erhalten -> kein poll184_19-Code-10-APM-Ersatz 2. OpenBSD-POLLING-WIDX-Logik: nach jedem Poll: widx = (closed == 0 ? Wifi_MAX_RX - 1 : closed - 1) & ~7 RFH_Q0_FRBDCB_WIDX_TRG wird auch bei count == 0 geschrieben 3. OpenBSD-artiger late nic_lock: MAC_ACCESS_REQ setzen auf MAC_ACCESS_EN warten GOING_TO_SLEEP prüfen 4. OpenBSD-artiger late ME/Wake/Ownership-Touch: HAP_WAKE WAKE_ME WAKE_ME_PCIE_OWNER_EN PCI_OWN_SET OS_ALIVE INIT_DONE RFKILL_WAKE_L1A_EN L1A_NO_L0S_RX HPET 0xffff0000 Link Power Management disabled 5. 0x3b wird nicht gefälscht closed bleibt echt nur RFH/MAC/ME wird wie OpenBSD nachgetriggert

@reboot12 I found this link https://codebrowser.dev/linux/linux/drivers/net/wireless/intel/iwlwifi/mei/main.c.html What it means is very easy. ME kills the 9560 on the gemini chip platform, so for our j4105. And on the Asrock j4105-ITX I cant disable ME, the name in Bios for ME there is TXE but when I disable it, the Wlan 9560 is gone. I can see ME at work: The 2 RX packages I read above are data for arp and DHCP. After this, crazy ME brings the firmware to write 3b, so the RX ring is blocked. I also found in the original OpenBSD 79 Source Code, which I modd for POLLING for the 9560, has an own C-block, how to handle this ME. So, fun is not over but I have to build this all with the help of ChatGPT, like translation from OpenBSD C-Code to XPC-Code. Then, XP is "allowed" under ME. I also tried a dirty hack, just to jump over this 3b code. But then the 9560 driver does not install any longer Dietmar 1. iwm_nic_lock() MAC_ACCESS_REQ setzen auf MAC_ACCESS_EN warten GOING_TO_SLEEP beachten 2. iwm_apm_init() HAP_WAKE_L1A setzen INIT_DONE setzen MAC_CLOCK_READY warten DMA clock / L1_ACT_DIS setzen 3. iwm_prepare_card_hw() HW_IF_CONFIG PREPARE / NIC_READY / ME_OWN / WAKE_ME prüfen 4. iwm_nic_rx_mq_init() RFH_RXF_DMA_CFG RFH_GEN_CFG DMA_SNOOP RFH_RXF_RXQ_ACTIVE WIDX_TRG 5. iwm_power_update_device() POWER_TABLE_CMD 6. iwm_power_mac_update_mode() MAC_PM_POWER_TABLE 7. iwm_ltr_config() LTR_CONFIG

@reboot12 I need to make some more tests. Until now only 2 RX packages, because I force one ring for this. But because it is a pure POLLING driver and DHCP works already, I think already today I send to you Dietmar

Yessssssssssssaaaaaaaaaaaaaahhhh after about next 500h of work and about 1000 not working drivers, I get the driver for the Wlan 9560 on the gemini chipset with j4105 cpu to run. It is a pure polling driver, based on OpenBSD 79. The step for to bring the firmware to alive=1 I managed only with Ubuntu. Crazy, 10 different rings for TX and RX and management data. Again I was one step away from to give up Dietmar

@UsefulAGKHelper No, he disappeared suddently in November 2024, dont know why Dietmar

@reboot12 For a try it may work. The problem in this is, that the way, the Bios treated DSDT is not static. Suddently you will get Bsod. This makes me like crazy, until @Mov AX, 0xDEAD tells me, that this was the exact reason why he build his DSDT patcher at boottime Dietmar

@reboot12 For one of my AMD boards I bought a changeable case on motherboard for the Bios chip. The original chip I solder out and can put it in and out of this case as often as I want, like in former times. After this, I can modd DSDT for this board while during flashing whole bios also in minutes Dietmar

@reboot12 Yes, I did this with success with the DSDT Patcher at boot time from @Mov AX, 0xDEAD . This is a really nice tool and with it I change DSDT a lot of times in minutes. But only for XP bit 32 I think. asl does not work, because it patches only the static DSDT in registry and gives Bsod on reboot. You can ask @Damnation he succeed with grub4dos I think, but dont know if this also works with XP bit 64 Dietmar

@reboot12 This is the problem from those acpi.sys hacks. So, this way does not work until now. I remember from past, that this happens, when an acpi.sys hack also destroyed necessary answers to other devices. This takes a long run with Windbg and Ida Pro to fix this Dietmar

@reboot12

@reboot12 May be, that from other place also a jump goes direct to this bsod A5 0x02 This will with this acpi13.sys not happen any longer Dietmar https://www.upload.ee/files/19415611/acpi13.sys.html

@reboot12 Here is acpi12.sys good luck Dietmar https://www.upload.ee/files/19415565/acpi12.sys.html

@reboot12 I just notice, that there are at least 5 places for this Bsod A5 0x02, so here is acpi11.sys Dietmar https://www.upload.ee/files/19415521/acpi11.sys.html

@reboot12 Here is acpi10.sys for this Bsod, hihi Dietmar https://www.upload.ee/files/19415497/acpi10.sys.html

@reboot12 I just found, that there is a second place for this Bsod. Here is acpi9.sys good luck Dietmar https://www.upload.ee/files/19415399/acpi9.sys.html

@reboot12 nono, I just find this Bsod via acpi.pdb here is acpi8.sys Dietmar https://www.upload.ee/files/19415305/acpi8.sys.html

@reboot12 Here is a new try acpi7.sys good luck Dietmar https://www.upload.ee/files/19415187/acpi7.sys.html

@reboot12 Yes, this Bsod is a result of my acpi hacks before. I make a complete new acpi5.sys Dietmar https://www.upload.ee/files/19415026/acpi5.sys.html

@reboot12 @reboot12 File: \Windows\System32\drivers\MegaSR.sys Status: 0xC0000098 This is not an ACPI crash. It happens before acpi.sys is loaded Dietmar MegaSR.sys is a RAID/SCSI boot-start driver, and 0xC0000098 means that this boot driver is missing, corrupt, wrong architecture, or not valid for this Windows installation. If the machine does NOT really boot from a MegaSR/LSI/RAID controller, the safest fix is to disable MegaSR offline and enable the normal Windows 7 storage drivers. Boot from a Windows 7 DVD/USB or WinPE, press Shift+F10, then do: diskpart list vol exit Find the Windows partition letter. In this example I use D:. Change it if your Windows partition has another letter. reg load HKLM\OFF D:\Windows\System32\Config\SYSTEM reg query HKLM\OFF\Select Look at the value "Current". If Current = 0x1, use ControlSet001. If Current = 0x2, use ControlSet002. To be safe, patch both ControlSets: reg add HKLM\OFF\ControlSet001\Services\MegaSR /v Start /t REG_DWORD /d 4 /f reg add HKLM\OFF\ControlSet002\Services\MegaSR /v Start /t REG_DWORD /d 4 /f reg add HKLM\OFF\ControlSet001\Services\msahci /v Start /t REG_DWORD /d 0 /f reg add HKLM\OFF\ControlSet002\Services\msahci /v Start /t REG_DWORD /d 0 /f reg add HKLM\OFF\ControlSet001\Services\pciide /v Start /t REG_DWORD /d 0 /f reg add HKLM\OFF\ControlSet002\Services\pciide /v Start /t REG_DWORD /d 0 /f reg add HKLM\OFF\ControlSet001\Services\iaStorV /v Start /t REG_DWORD /d 0 /f reg add HKLM\OFF\ControlSet002\Services\iaStorV /v Start /t REG_DWORD /d 0 /f reg unload HKLM\OFF

@reboot12 MegaSR.sys is a RAID driver. I think, maybe you make a mistake and you want to load a normal AHCI driver. Or another driver special for this eMMC flash. When anything with RAID is enabled in Bios, this driver is choosen Dietmar PS: For embedded boards there is a eMMC hotfix Windows6.1-KB2732471-v2-x64.msu or you a need a modd for the Intel eMMC driver iaiosd.sys iaiosd.inf

@reboot12 Yes, this comes from hack before, now no child is found. Oh..crazy hack, child is now found always good luck Dietmar https://www.upload.ee/files/19412025/Bsod4_w7.zip.html

@reboot12 Here is next hack, good luck Dietmar https://www.upload.ee/files/19411085/Bsod_w7.zip.html

@reboot12 This Bsod come for sure from _ini in acpi.sys . Here is next try, after this I am out of ideas. The signatur is broken in any case even checksum is correct Dietmar https://www.upload.ee/files/19410857/C0000034Bsod_w7.zip.html

@reboot12 Not so easy to find the correct place with IdaPro. But I found the *.pdb file for it. The problem comes from _ini as you can see in Bsod. Now, always when _ini tells error, that something was not found with c0000034 => 00000000 Nice hack Dietmar https://www.upload.ee/files/19409225/acpi_ini_status0_only.zip.html

@reboot12 Oh..nice Here is the last update for the driver for this Dell Wyse 5070 Wlan 9560. I had crazy problems with this firmware 46, but now all is on a good way. I can see via this from scratch new build driver now my router Dietmar