Created attachment 205591 [details]
BTX halted screenshot
and the initiated reboot, the Systems stops booting with
„BTX halted“ and a CPU-Register dump.
In the previous screen where BTX enumerates the drives, you can see the following text:
„read 1 from 0 to 0xdd78f340, error: 0x1“
This messages comes twice.
The machine is a Dell PowerEdge R515 with a LSI-RAID Controller (mps). The system is on a Hardware-RAID1 with UFS. The machine runs FreeBSD since several years and FBSD12 is an upgraded Version of FBSD11. FBSD12 ran fine all the time since its release.
The disks seems to be ok and error free. I’ve booted a 11.2 live disk, gpart list|status|show shows no errors or unusual things on the disk layout. „gpart recover“ says everything’s ok and nothing is to be done. I also can mount the 12.0 ufs partition in the live system and access the data there. fsck is also ok and says there’s no error. Though i doubt there is a hardware error.
Booting the FreeBSD-12.0-RELEASE-amd64-disc1-ISO fails with the same error. Booting the FreeBSD-12.0-RELEASE-amd64-disc1-IMG (for USB-Stick) is booting but i have to modifiy the Boot-Settings to UEFI to make it boot (this is a known issue since the release of 12.0). I’m booting the ISO-Images with idrac virtual drives, the USB-Image is copied to an SD-Card and booted from there (but as idrac-Emulated disk).
I have to correct the
The RAID1 which contains the systems is fortunately installed with GPT-Layout.
I’ve reduced now the Swap-Partition, created a new EFI-Partition and wrote the EFI-Loader boot1.efifat into the EFI-Partition.
Then i switched the Bootmode from BIOS (Legacy) to UEFI.
This makes the Server booting again and resolves the issue.
It seems in FreeBSD 12 is booting in BIOS- (Legacy-) Mode a bit broken.