Downgrade kernel failure[SOLVED]

[apiaio]

Two days ago I upgraded gentoo-sources 6.6.8 to 6.18.3 . Because virtual machine fails to start in virtualbox I want downgrade back to 6.6.8 version.
But kernel compilation ends with error

Code: Select all

  CC      arch/x86/realmode/rm/wakemain.o
In file included from ./include/uapi/linux/posix_types.h:5,
                 from ./include/uapi/linux/types.h:14,
                 from ./include/linux/types.h:6,
                 from arch/x86/realmode/rm/wakeup.h:11,
                 from arch/x86/realmode/rm/wakemain.c:2:
./include/linux/stddef.h:11:9: error: cannot use keyword ‘false’ as enumeration constant
   11 |         false   = 0,
      |         ^~~~~
./include/linux/stddef.h:11:9: note: ‘false’ is a keyword with ‘-std=c23’ onwards
./include/linux/types.h:35:33: error: ‘bool’ cannot be defined via ‘typedef’
   35 | typedef _Bool                   bool;
      |                                 ^~~~
./include/linux/types.h:35:33: note: ‘bool’ is a keyword with ‘-std=c23’ onwards
./include/linux/types.h:35:1: error: useless type name in empty declaration [-Werror]
   35 | typedef _Bool                   bool;
      | ^~~~~~~
cc1: all warnings being treated as errors
make[5]: *** [scripts/Makefile.build:243: arch/x86/realmode/rm/wakemain.o] Error 1
make[4]: *** [arch/x86/realmode/Makefile:22: arch/x86/realmode/rm/realmode.bin] Error 2
make[3]: *** [scripts/Makefile.build:480: arch/x86/realmode] Error 2
make[2]: *** [scripts/Makefile.build:480: arch/x86] Error 2
make[1]: *** [/usr/src/linux-6.6.8-gentoo/Makefile:1913: .] Error 2
make: *** [Makefile:234: __sub-make] Error 2

I have sys-devel/gcc-15.2.1_p20251122:15::gentoo installed.

Any ideas about this?

[Hu]

This is effectively the same as sys-kernel/gentoo-sources-5.15.189 compile failure with gcc-15.2.0: error: cannot use keyword ‘false’ as enumeration constant, despite the different kernel version cited in that title. gcc-15 cannot compile old kernels as released. Build using a contemporary gcc, or build a kernel that has been patched to work with gcc-15.

Why are you trying to rebuild 6.6.8? If it worked for you before, and you still have it installed, you can boot it without building it. If you don't have it installed, is there a reason that you are not using the latest 6.6.x, which as of this writing is the long-term 6.6.119?

[NeddySeagoon]

apiaio,

What Hu said answers your question but I suspect it won't fix your problem.

I need

Code: Select all

kvm.enable_virt_at_load=0

on the kernel command line with VBox 7.2.4.
That works with 6.17.3. I've not updated to 6.18.x yet.

Whatever, going back is not a long term solution. You need a way forward.

[apiaio]

Why are you trying to rebuild 6.6.8? If it worked for you before, and you still have it installed, you can boot it without building it. If you don't have it installed, is there a reason that you are not using the latest 6.6.x, which as of this writing is the long-term 6.6.119?

After time I need make some settings in kernel. Especially this one

Code: Select all

# cat /usr/src/linux/.config|grep CONFIG_EXTRA_FIRMWARE 
CONFIG_EXTRA_FIRMWARE="rtlwifi/rtl8188eufw.bin"
CONFIG_EXTRA_FIRMWARE_DIR="/lib/firmware"

@NeddySeagoon

Code: Select all

kvm.enable_virt_at_load=0

Could you please explain it in more detail?

[NeddySeagoon]

apiaio,

VirtualBox will not start if KVM support is running already.

If you look in a VM log file I expect it contains Guru Meditation. That means that its in a state waiting for ever.
For me, it was waiting to be able to enable virtualisation, which was already started.

Code: Select all

kvm.enable_virt_at_load=0

tells the kernel not to start virtualisation when the code loads and VBox is happy.

[apiaio]

If I understand correctly I should it apply in GRUB. But I use GRUB in NixOS. I boot 5 OS at least. Does exist any configuration file in Gentoo where can I put it in?

[pietinger]

If I understand correctly I should it apply in GRUB. But I use GRUB in NixOS. I boot 5 OS at least. Does exist any configuration file in Gentoo where can I put it in?

You have always only two choices:
1. A bootmanager/-loader or UEFI itself gives some kernel command line parameters to the kernel, or
2. You compile some kernel command line parameters into the kernel (because you use a manually configured kernel) ->

Code: Select all

Processor type and features  --->
    [*] Built-in kernel command line
    (kvm.enable_virt_at_load=0)

[NeddySeagoon]

apiaio,

It goes in /etc/default/grub to have grubs automation find it

Code: Select all

...
# Boot the default entry this many seconds after the menu is displayed
#GRUB_TIMEOUT=5
#GRUB_TIMEOUT_STYLE=menu

# Append parameters to the linux kernel command line
GRUB_CMDLINE_LINUX="ro net.ifnames=0"
...

Add kvm.enable_virt_at_load=0 to GRUB_CMDLINE_LINUX=
I don't normally use grub but that's from my server.
You may need to uncomment GRUB_CMDLINE_LINUX= too.

Now remake grub.cfg.

grub has a boot time editor too.
Select the kernel you wish to beet then press 'e' for edit.
Read the on screen instructions.
Go down to the kernel line.
Add kvm.enable_virt_at_load=0 to the end.
Allow the boot to continue.

This is a one time change. Only the in memory grub.cfg is changed.

[apiaio]

I've tried every possible way.

-compilation in kernel
-NixOS grub . The configuration there is completely different from other distributions. So I used
-Ubuntu grub editing /etc/default/grub aswell
-and a boot time editor. It reports "unknown command kvm.enable_virt_at_load=0"

The issue persists - virtual machines fail to start in virtualbox. When I boot into kernel 6.6.8, everything is fine.

[NeddySeagoon]

apiaio,

We need to see a VirtualBox log file so we con stop guessing.

It may be too big for a post, so use a pastebin service.

My VirtualBox VMs are in ~/'VirtualBox VMs'
Picking ~/'VirtualBox VMs'/'7.2.0 Testing'/ I have a Logs directory that contains

Code: Select all

~ $ ls VirtualBox\ VMs/7.2.0\ Testing/Logs/
VBox.log  VBox.log.1  VBox.log.2  VBox.log.3

The newest log is VBox.log.

Code: Select all

wgetpaste -s 0x0 ./VirtualBox\ VMs/7.2.0\ Testing/Logs/VBox.log

Tells https://0x0.st/P8fK.log
Feel free to look at my log.

Can you do that for one of your virtual machines after a failed start please?
The error we need to see will be in the log.

[apiaio]

Sorry for the late reply, I’ve been away for a few days.

Because I don't remember what state I left the last installation in, I removed all old files of VBox, and compiled command line parameters into the kernel and this is the log

http://0x0.st/PPGG.log

[NeddySeagoon]

apiaio,

At the end of your log it says

Code: Select all

00:00:00.668008 Changing the VM state from 'CREATED' to 'POWERING_ON'
00:00:00.668659 AIOMgr: Endpoints without assigned bandwidth groups:
00:00:00.668664 AIOMgr:     /home/miro/.VirtualBox/Machines/Win10/Win10.vdi
00:00:00.668790 Changing the VM state from 'POWERING_ON' to 'RUNNING'
00:00:00.668795 Console: Machine state changed to 'Running'

which claims the VM is running.

The earlier error

Code: Select all

00:00:00.536039 ERROR [COM]: aRC=VBOX_E_IPRT_ERROR (0x80bb0005) aIID={6ac83d89-6ee7-4e33-8ae6-b257b2e81be8} aComponent={ConsoleWrap} aText={The VBoxGuestPropSvc service call failed with the error VERR_HGCM_SERVICE_NOT_FOUND}, preserve=false aResultDetail=-2900

is harmless.

[apiaio]

Actually, this thread wasn't started for linux version downgrading but because the virtual machines weren't starting.
As NeddySeagoon wrote

Whatever, going back is not a long term solution. You need a way forward.

I was looking for solution for a few days. I finally found it.

While in version 6.6.8, it's enough to have kvm set

Code: Select all

grep kvm -i /usr/src/linux/.config
CONFIG_KVM_GUEST=y
CONFIG_HAVE_KVM=y
# CONFIG_KVM is not set
CONFIG_PTP_1588_CLOCK_KVM=y

settings in version 6.18.3 must be

Code: Select all

CONFIG_CMDLINE="kvm.enable_virt_at_load=0"
CONFIG_KVM_COMMON=y
CONFIG_HAVE_KVM_PFNCACHE=y
CONFIG_HAVE_KVM_IRQCHIP=y
CONFIG_HAVE_KVM_IRQ_ROUTING=y
CONFIG_HAVE_KVM_DIRTY_RING=y
CONFIG_HAVE_KVM_DIRTY_RING_TSO=y
CONFIG_HAVE_KVM_DIRTY_RING_ACQ_REL=y
CONFIG_KVM_MMIO=y
CONFIG_KVM_ASYNC_PF=y
CONFIG_HAVE_KVM_MSI=y
CONFIG_HAVE_KVM_READONLY_MEM=y
CONFIG_HAVE_KVM_CPU_RELAX_INTERCEPT=y
CONFIG_KVM_VFIO=y
CONFIG_KVM_GENERIC_DIRTYLOG_READ_PROTECT=y
CONFIG_KVM_GENERIC_PRE_FAULT_MEMORY=y
CONFIG_KVM_COMPAT=y
CONFIG_HAVE_KVM_IRQ_BYPASS=y
CONFIG_HAVE_KVM_NO_POLL=y
CONFIG_HAVE_KVM_PM_NOTIFIER=y
CONFIG_KVM_GENERIC_HARDWARE_ENABLING=y
CONFIG_KVM_GENERIC_MMU_NOTIFIER=y
CONFIG_KVM_ELIDE_TLB_FLUSH_IF_YOUNG=y
CONFIG_KVM_MMU_LOCKLESS_AGING=y
CONFIG_KVM_GUEST_MEMFD=y
CONFIG_KVM_X86=y
CONFIG_KVM=y
CONFIG_KVM_INTEL=y
# CONFIG_KVM_AMD is not set
CONFIG_KVM_IOAPIC=y
CONFIG_KVM_SMM=y
CONFIG_KVM_HYPERV=y
# CONFIG_KVM_XEN is not set
CONFIG_KVM_MAX_NR_VCPUS=1024

I hope this helps someone.

[NeddySeagoon]

apiaio,

Code: Select all

CONFIG_CMDLINE="kvm.enable_virt_at_load=0" 

is the same as adding kvm.enable_virt_at_load=0 to the kernel command line.

I'm pleased that you are on the road forward again.

[apiaio]

apiaio,

Code: Select all

CONFIG_CMDLINE="kvm.enable_virt_at_load=0" 

is the same as adding kvm.enable_virt_at_load=0 to the kernel command line.

I know. As previously stated, I am booting Gentoo via another system, so it seemed more appropriate to build this command into the kernel.

I’d also like to clarify that those using an AMD processor should have the following set:

Code: Select all

#CONFIG_KVM_INTEL  is not set
CONFIG_KVM_AMD=y