I was troubleshooting a GPU passthrough issue and out of desperation tried resetting BIOS (clear CMOS button) on the motherboard to start fresh. This however made my SSD first not appear at all then appear but unbootable. The odd thing is if I connect it via a USB Dock (USB port) I can boot every time no issues other than slow boot speeds. Running Arch Linux OS with systemd-boot.
Not detected
Before the the reset I did sometimes have an issue with this new SSD not appearing in BIOS possibly due to bad sata connection on the SSD. This issue thus might be unrelated to resetting BIOS. The old ssd worked fine and I tried multiple ports on the motherboard with multiple sata cables but still sometimes randomly the device is not detected at boot time. I can't quite tell if it's purely random or because I adjust the connection since sometimes just resetting helps other times not.
Unbootable
Other times the device appears in the BIOS/UEFI menu but won't boot. Before clearing CMOS I would get 3 options relating to the system: 2x Linux Boot Manager and the name of SSD. Booting with SSD would not work presumably because it boot in BIOS mode (?). After clearing CMOS Linux Boot Manager options are gone. Why? I was under impression that clearing BIOS should not affect booting like this. The SATA settings are the same as before (AHCI).
What am I missing here? Seems like it's something obvious but can't turn anything up.
Try updating bios to see if that helps
That's not really an option as any newer firmware than the current cripples the cpu by disabling non-k overclocking.
It's odd how with usb I see both bios and uefi boot options but not with sata...
try a different port on the board.
its part of intels plan to force you to buy a overclocking cpu
Yeah I tried all 4 ports and even esata, not to mention that the connection hasn't changed since before clearing CMOS. Loose connection could explain no boot device but I don't see how it could explain lack of UEFI option.
I've heard about Intel's latest asshattery. I am using 3770 which is quite old now, and firmware with no overclocking has been out for a couple of years now so perhaps it's not directly related to the said asshattery.
Tried a new cable that fits more snugly which seems to have fixed the disappearing drive. There is still no UEFI boot option when connected via SATA but there is one with USB. Really a bizarre issue.
might be a corrupted uefi partition
Maybe, but then why does it boot reliably over usb? If it's corrupt it shouldn't boot regardless of connection method. I can't come up with an explanation for that.
Maybe during all of your tinkering you just borke the AHCI driver and left intact the USB driver, so the OS can now boot from USB but not from SATA. Try to troubleshoot the issue going from this hint and let me know if it was any useful, I'm courious to know the outcome of this. You can assure me 100% the after booting through USB everything works without an issue at all?
Absolutely no issues when booted, I've been booting from usb for 3 days now. I don't believe I touched anything related to drivers but that would make sense. The onlky thing is would linux drivers affect BIOS/UEFI? Shouldn't I at least see the UEFI option in bios menu?
Here are the logs from journalctl -b -p err:
Mar 31 10:36:38 workstation kernel: sd 6:0:0:0: [sda] No Caching mode page found
Mar 31 10:36:38 workstation kernel: sd 6:0:0:0: [sda] Assuming drive cache: write through
Mar 31 10:36:38 workstation kernel: hid-generic 0003:1B1C:1B13.0002: usb_submit_urb(ctrl) failed: -1
Mar 31 10:36:38 workstation kernel: hid-generic 0003:1B1C:1B13.0004: usb_submit_urb(ctrl) failed: -1
Mar 31 10:36:44 workstation NetworkManager[413]: <error> [1459417004.078144] [devices/nm-device.c:2324] nm_device_generate_connection(): (virbr0): Generated connection does not verify: bridge.forward-delay: value '1' is out of range <2-30>
Mar 31 10:37:01 workstation kernel: [drm:intel_set_pch_fifo_underrun_reporting [i915]] *ERROR* uncleared pch fifo underrun on pch transcoder B
Mar 31 10:37:01 workstation kernel: [drm:intel_pch_fifo_underrun_irq_handler [i915]] *ERROR* PCH transcoder B FIFO underrun
Mar 31 10:37:03 workstation kernel: [drm:intel_set_pch_fifo_underrun_reporting [i915]] *ERROR* uncleared pch fifo underrun on pch transcoder A
USB errors were there before with no noticeable issues.
And here's dmesg | grep ahci:
[dom@workstation ~]$ dmesg | grep ahci
[ 0.474608] ahci 0000:00:1f.2: version 3.0
[ 0.485001] ahci 0000:00:1f.2: AHCI 0001.0300 32 slots 6 ports 6 Gbps 0x3f impl SATA mode
[ 0.485006] ahci 0000:00:1f.2: flags: 64bit ncq pm led clo pio slum part ems apst
[ 0.518909] scsi host0: ahci
[ 0.519168] scsi host1: ahci
[ 0.519387] scsi host2: ahci
[ 0.519578] scsi host3: ahci
[ 0.519710] scsi host4: ahci
[ 0.519854] scsi host5: ahci
Everything looks in order, nothing is sticking out that much really (maybe only the error with the Intel HD graphic driver, but that's not relevant right now). Yeah, you should be able to see the drive in the UEFI 100% of the time if it's connected and everything works "phisically". Try to make an image of the disk, write it on a disk you know for sure it works (always detected by the motherboard) and see if it boots up. Also you may try to upgrade the SSD firmware, because it might just be an unstable one that gives you issues.
Cloned the SSD to a hard drive and exact same issue. The odd thing is that I also did a fresh install on a different disk but same system and now I am seeing the boot option for the OS that is on a disk sitting the desk. This means UEFI is storing boot loader entries on the internal memory and is not reading it from disk. Didn't know that's how it worked, seems like really stupid implementation...
Anyone have any info on this?
non. might try and disable legacy support for bios and update grub.
No option to disable legacy bios. Updating to latest bios didn't help. Not using grub so nothing to update. Will go ahead and wipe the disk and start fresh.
Never knew UEFI stores boot options on integrated memory. The real question is why it doesn't refresh those options... If someone comes along with some knowledge of what's going on I would love to know.