Have a fresh Centos 7 + MATE install on a 1TB SSD. Everything WAS working fine. Not sure if this is the cause but was tweaking the BIOS(Asus X99A2) attempting a firmware update and the next time I tried booting into Linux my screen freezes with the 3 following lines
EDAC sbridge: ECC is disabled. Aborting
EDAC sbridge: Couldn't find mci handler
kvm: disabled by bios
[sdb] No Caching mode page found
[sdb] Assuming drive cache: write throughI've googled quite a bit and found similar errors(this following link is for an Ubuntu install)
but no definitive answers!
Pretty sure the first line regarding ECC is bogus since my MOBO doesn't support ECC RAM anyways. I'm confused as to whether this is a BIOS setting error or a Linux install error. I went ahead and made sure my BIOS was properly updated but still get the same error. I'd like to avoid having to reformat my SSD and reinstall everything from scratch!
I'm pretty well stuck and am hoping someone has a solution. Any feedback welcome. Thanks!
22 Answers
It's a warning message from Linux kernel. If your board doesn't support ECC RAM you can disable the EDAC module creating the file /etc/modprobe.d/edac-blacklist.conf with this content:
blacklist i82975x_edac
blacklist amd64_edac_mod
blacklist skx_edac
blacklist e752x_edac
blacklist i5400_edac
blacklist i3200_edac
blacklist i5000_edac
blacklist ie31200_edac
blacklist edac_mce_amd
blacklist i3000_edac
blacklist sb_edac
blacklist pnd2_edac
blacklist i7core_edac
blacklist i7300_edac
blacklist x38_edac
blacklist i5100_edacAfter rebooting, the message shouldn't appear again...
2I got the same error running ASUS RAMPAGE V EXTREME (X99), Ubuntu Bionic, Nvidia GTX980, and after upgrading to latest BIOS 4104 (2019), but here are things to try out:
Does Num Lock work and change the LED on the keyboard? Does ctrl-alt-F2 change to a new console? This tells you that Linux/kernel is still responsive and running, and has not crashed.
It seems that the error has nothing to do with ECC, as the motherboard doesn't support ECC and the memory chips are non-ECC.
On Ubuntu, grub works ok, and I was able to drop to console by using Fail Safe mode (give it a try on a USB Stick etc).
I found out that my software RAID was still recovering, but it doesn't affect boot.
With failsafe -mode I investigated /var/log/kern.log and there was error about "NVRM: API Mismatch" suggesting the problem is related to Nvidia proprietary drivers, kernel update and DKMS.
