[Bug 267028] kernel panics when booting with both (zfs,ko or vboxnetflt,ko or acpi_wmi.ko) and amdgpu.ko

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=267028

--- Comment #296 from Mark Millard <marklmi26-fbsd@yahoo.com> ---
"dmesg from crash" starts with (part of) a boot that crashes
and then also shows a boot that works. (I have diff'd the
two parts.)

The last of the crashing part's new debug messages was:

modlist_newmodule: new node addr 0xfffff80004603180 name
amdgpu_raven_vcn_bin_fw: container addr 0xfffff800039ff480, modname addr
0xffffffff82ea6025, version 1

The 0xfffff80000000007 oddity has always showed up
for a node before amdgpu_raven_vcn_bin_fw .

But all of the prior "modlist_newmodule: new node addr . . ."
messages do not show: new node addr 0xfffff80000000007

This suggests that things have been okay at allocation time
and later activity is sometimes replacing a good value with
the value 0xfffff80000000007 .

This, in turn suggests adding another check to the
debug/diagnostic patching, checking for if the values
have changed from the original allocated ones.

-- 
You are receiving this mail because:
You are the assignee for the bug.