System: Xiaomi Mi Notebook Pro 15.6 ('19 refresh w/ MX250) / TM1701
CPU: Intel i7-8550U
GPU: nVidia MX250
RAM: 16 GB DDR4 / 2400 Mhz
Storage: NVMe Samsung PM981 512 GB / FW EX7301Q / 2019.05
Motherboard model number: Unknown, but picture attached. Manufacturer signature probably beneath heat sink assembly or on the backside. I read somewhere they manufacture at Foxconn. I am reluctant to remove the heat sink because I don't have any manual and thus can not be sure to have the correct tools and replacement parts.
BIOS: XMAKB5R0P0E07 / EC: XMAKB5R0P007
Since when does the issue exist: BIOS Update from XMAKB5R0P0C07 (factory installed, but not avail from Xiaomi, a dump has been posted by a user here: https://github.com/Cr0wTom/Mi-Notebo...6/mybackup.zip) to XMAKB5R0P0E07 (official download: https://cdn.cnbj1.fds.api.mi-img.com...BIOS_P0E07.zip)
No other tampering was done. System did not fall or experience liquid damage.
Symptoms / Troubleshooting:
Troubleshooting to no effect:
Questions:
Wild assumptions:
Bonus questions:
CPU: Intel i7-8550U
GPU: nVidia MX250
RAM: 16 GB DDR4 / 2400 Mhz
Storage: NVMe Samsung PM981 512 GB / FW EX7301Q / 2019.05
Motherboard model number: Unknown, but picture attached. Manufacturer signature probably beneath heat sink assembly or on the backside. I read somewhere they manufacture at Foxconn. I am reluctant to remove the heat sink because I don't have any manual and thus can not be sure to have the correct tools and replacement parts.
BIOS: XMAKB5R0P0E07 / EC: XMAKB5R0P007
Since when does the issue exist: BIOS Update from XMAKB5R0P0C07 (factory installed, but not avail from Xiaomi, a dump has been posted by a user here: https://github.com/Cr0wTom/Mi-Notebo...6/mybackup.zip) to XMAKB5R0P0E07 (official download: https://cdn.cnbj1.fds.api.mi-img.com...BIOS_P0E07.zip)
No other tampering was done. System did not fall or experience liquid damage.
Symptoms / Troubleshooting:
- Memtest under Windows (via mdsched.exe): Fail (can fail as early as 12 minutes in), Memtest86 V9 (Free and Pro): Pass (even after 18+ hours test)
- System POSTs and boots to OS.
- Frequent (every x minutes once the system is used, if it idles system seems more stable) and wildly varying BSODs under Windows, Frequent app crashes under Linux (POP-OS 21.10, booted from USB)
- Disk read/write does not seem trustworthy. Disk image of NVMe (not mounted / read-only) with dd to USB-drive for backup purposes does not provide repeatable result, 1 to 20+ bytes changed. However, during troubleshooting I messed up my Windows partition to the point it can't boot anymore. I could test the drive in another system, and it seems physically okay (if structurally messed up, with the Windows partition inaccessible).
Troubleshooting to no effect:
- Tried to re-apply BIOS update XMAKB5R0P0E07
- CMOS Clear (https://youtu.be/i_whyUb2Pwk?t=218), there is no seperate RTC-Battery on the mainboard
- Tried to mitigate issue by allowing Windows only 8, 4, 2 GB of RAM (with msconfig), same in Linux (varios kernel boot options, including mem=8G,mem=4G,mem=2G)
- Same behavior with or without storage
- Contacted Xiaomi for support. Instructed me to only use Windows 10 Home (Chinese Edition) as OS and / or to visit mainland China for repair
Questions:
- First of all: Help?
- If there is anything else you want me to test, report, please let me know.
Wild assumptions:
- Restoring original BIOS should fix the issue. But how do I flash the BIOS dump? I guess it as easy as fptw64.exe -f mybackup.bin -bios (after unlocking). But how can I be sure the written data is correct, as read/write to any other also seems troubled?
Bonus questions:
- How do I fix the IntelME region after flash? Could the IntelME region be the reason, why the update did not work as supposed?
- Just out of curiosity: I found out XIAOMI somehow restricted TPM use to SHA1 only, even though it should support SHA256, is there any way to re-enable it? (https://wiki.archlinux.org/title/Xia...ebook_Pro_15.6) Why should they do it in the first place?