On Fri, 19 Oct 2018, Larry Gadallah wrote:
> I have observed the same kind of problem on a Dell Optiplex 990. I tried 
> an upgrade to 6.4 via bsd.rd and the system boots the kernel and then 
> reboots a few seconds later. Booting the bsd.rd or bsd.mp kernels 
> provides the same result.

Larry, you're saying that both bsd.rd and bsd.mp reboot within 'a few 
second'?  What _exactly_ is the last line of output seen?

(In this sort of circumstance, video-recording the screen such as with a 
smartphone and then reviewing that frame-by-frame may be the easiest way 
to see how far it got and whether it output any sort of 'trap' message.)

Anyway, this appears to be a different problem than what Tracy is 
reporting, where bsd.rd works just fine, but bsd.mp (and bsd.sp) reboot 
early.


> I made sure that HT was turned off in the (most recent) BIOS, and I 
> paved the box and reinstalled 6.3 via CD without incident. I tried an 
> upgrade to 6.4 and got the same results.

Did you run any 6.4 snaps, or only 6.3 and/or 6.3-stable/syspatch?  There 
are no dmesg reports from anyone else with that hardware (or even just 
that CPU) after your 6.3 report, so you may be the first person to be 
running a post-6.3 kernel on that hardware...which leaves a 6 month window 
in which a change could have broken things.

Nothing sticks out in the dmesg (it doesn't have the CPU features to 
enable PCID use, for example) and the early crash rules out several 
changes.


k> The only odd thing I can think of with this system is the installation
> is on the second hard drive. The first hard drive is disabled in the 
> BIOS.

The kernel doesn't access the drive until later while probing the device 
tree, so I don't think that's the issue.


Philip Guenther

Reply via email to