I have something crazy for you today: an article about my test laptop Lenovo G50-70 NVRAM having become read only, subsequent investigation and troubleshooting, operational setup implications and changes, and more. Read on. https://www.dedoimedo.com/computers/lenovo-g50-read-only-nvram.html Cheers, Mrk
Holy ***. Shouldn't nvram withstand 100.000 writes? I doubt that you managed to hit the limit by installing various OSes... even if you installed or restored images by dozens every single day. Does this mean that every time the system boots/reboots the OS or the firmware writes on the NVRAM?
I don't know the internals of how it works with this particular laptop - or how many writes are done a new entry is written. But I doubt many people run 500 installations on their laptop Mrk
It does not have to do with the number of installations. If I'm not mistaken every time the system is shutdown the volative part from the NVRAM is transfered to the EEPROM memory array. https://msdn.microsoft.com/en-us/library/ms855203.aspx http://www.arlabs.com/help.html And since most users rarely leave their systems powered on 24h/day they actually write in the EEPROM thousands of times during a three year warranty period. More likely you were unfortunate and the mobo got a bad nvram module. Panagiotis
hi @Mrkvonic @pandlouk very interesting article but about what does it mean ? everytime i change a setting in the bios , a cell does die? i have a multiboot sytem , and i use always f8 to select which os has to start everytime a cell is gone ? or this happen only if i install an operation system in uefi mode? thanks
There's a limited amount of electrons in the flash memory, so you eventually erode the conductive band, and without a chemical process, you can't really replenish the missing electrons. But that's fine, because as long as the signal to noise ratio is acceptable, you can differentiate between what you classify as 0 and 1, but if not, the drive firmware will mark the particular block of cells as dead. If there's sufficient redundancy, you will not notice this, and if not, you may end up with (most likely) read only data. I am not privy to the implementation of the particular UEFI, but there could be all sorts of oddities in how data is written and what caused the corruption. Mrk
I performed what few people do - roughly 500 years worth of testing on a single machine As to how, there could be many reasons - including bad hardware. Mrk
Greetings guys. I have managed to have the same issue with a Lenovo g50-80 , 1year old with maybe 10 or 15 distros tested on it. Yesterday after installing a new Ubuntu studio distro, my nvram went read-only. Now it keeps trying to boot only from a very specific USB port. Tried clearing cmos, did nothing. Trying efibootmgr returns interrupted system call. Does anyone has a real answer ? I mean, could I possibly need a new board?
Very likely. I'm lucky that the boot device is the hard disk - so I still have all the distros available. What you could do - burn a distro to a usb drive, boot from it - but then choose "boot from local disk" in the menu options. Mrk
I agree with Mrkvonic. If it is under warranty I would use it. It seems that Lenovo and linux give these issues... could it be a bad bios/firmware that constantly writes on the nvram? or maybe a bad configured linux distro that constantly writes in the nvram? Panagiotis
