Re: Failed to emulate instruction sequence [ 41f646040874064c017d88eb144c89 ] at 0xbea5fcf9

From: Mario Marietto <marietto2008_at_gmail.com>
Date: Mon, 02 Dec 2024 23:33:22 UTC
--> Did you reboot or reload the module ?

I have rebooted the machine.

--> Did you use the new bhyve binary you built from the patch ?

I didn't recompile bhyve,since I don't need it.

--> Note: copying binaries around from other systems or your custom
versions will not be something I can help debug.

The system that I have on the UFS disk is not another system. It's the same
system that I have on the ZFS disk,without the last round of (latest)
packages. I'm sure that the system on the ZFS system started to present the
problem when I did :

# pkg update
# pkg upgrade

I don't have the problem on the UFS disk because I didn't do that.

I haven't done # freebsd-update on the ZFS and UFS systems for some
months,so both are stuck to 14.1-RELEASE-p5 GENERIC.



On Tue, Dec 3, 2024 at 12:09 AM Mark Peek <mp@freebsd.org> wrote:

> On Mon, Dec 2, 2024 at 2:39 PM Mario Marietto <marietto2008@gmail.com>
> wrote:
>
>> ok I've used this :
>>
>> .op_flags = VIE_OP_F_IMM8
>>
>> After the compilation is finished,I copied the compiled vmm.ko module
>> from /usr/obj/usr/src/amd64.amd64/sys/modules/vmm/vmm.ko to /boot/kernel
>> and to /boot/modules.
>>
>> Regarding the bhyve executable,I'm using two custom versions of it,called
>> bhyve-lin and bhyve-win. When I upgrade the system,they cannot change. What
>> can change is bhyve,but I don't use it. In addition,to be sure,I copied
>> bhyve-lin and bhyve-win from the system where I don't have the error. With
>> this being said,the recompilation of vmm.ko didn't solve my issue.
>>
>
> Did you reboot or reload the module? Did you use the new bhyve binary
> you built from the patch? Note: copying binaries around from other systems
> or your custom versions will not be something I can help debug. Only code
> built from a newly checked out /usr/src and applying the patch will provide
> the correct information on whether it works or not.
>
> But if it didn't solve your issue you should document how it can be
> reproduced and either 1) add it to the bug I mentioned or 2) open a new
> bug. Make sure to include information on the host system as well
> (intel/amd, etc.) and how the various guest bits (.img, .fd, etc) were
> download/created.
>
> Mark
>


-- 
Mario.