Installation of 5.0.22 hosed my VMs.

[karozans]

I am running the latest version of Mac OSx. Everything was working fine with a Red Hat Linux 64 bit VM.

I upgraded virtualbox to 5.0.22. Everything worked fine.

I installed the latest version of the guest additions and it destroyed by linux VM. See attachment.

Untitled.tiff (31.28 KiB) Viewed 3469 times

I deleted my VM, then reloaded Red Hat from a clean install. Installed all the updates for Red Hat. Then installed the Guest additions CD and rebooted and it hosed my VM again.

There seems to be a bug in the latest guest additions CD.

[StreetBob]

I had the same issue with 5.0.24. I got the same screen.
Earlier today I read a post in the Linux guest section of this forum.

When it gets to the "Oh No! Something has gone wrong." screen, hit the tab key to get focus on the Log Out button, then hit enter.
When the screen goes black, hit the cmd key + F6 to get a console login.
Login then run startx.

They still need to fix the code. This is two releases with faulty code.

[frank]

In such cases we need the current guest kernel log ('dmesg') and the file /var/log/Xorg.0.log from the guest.

[bryanf]

Hello,

I've got same issue once VB Guest Additions is installed with an ISO today and VBox 5.0.24r108355. Have done several installs and now snapshots with various packages installed in various orders - problem continues.
General concept:
CentOS 7 with Gnome desktop
yum update
yum install kernel-devel
yum install gcc*

tried with updates pre VBGuestAdd.
tried VBGuestAdd without above yum installs
tried all updates then VBGuestAdd.

All fail to boot. VM will accept kill signal. However, cannot boot with error about recovery.

Screen Shot 2016-06-30 at 3.56.23 PM.png (47.05 KiB) Viewed 3383 times

Are there any logs or dmesgs stored in VBInstall.log (not that I'm aware)?
Otherwise, one is not able to get to the dmesg or logs since the VM cannot really be accessed at command line.
VM accepts a kill signal via Mac VBox GUI icon top left corner or poweroff.

Is there a method in VB to boot without X starting which may then grant terminal access?

Thank you.



FYI Update:
Devices => Insert Guest Additions Virtual CD Image ==> mounts: VBOXADDITIONS_5.0.24_108355

[bryanf]

FYI: MacBook Pro Intel with Apple Cinema display via Thunderbolt. Did not every disconnect the display but probably a good idea to have tried it.

Regular boot seems to fail. Tried running in Detachable Mode and it came up.

Part of the dmesg below.

[ 3.334962] vboxguest: module verification failed: signature and/or required key missing - tainting kernel
[ 3.348238] vgdrvHeartbeatInit: Setting up heartbeat to trigger every 2000 milliseconds
[ 3.360413] input: PC Speaker as /devices/platform/pcspkr/input/input5
[ 3.362845] input: Unspecified device as /devices/pci0000:00/0000:00:04.0/input/input6
[ 3.363210] vboxguest: misc device minor 58, IRQ 20, I/O port d040, MMIO at 00000000f0000000 (size 0x400000)
[ 3.363213] vboxguest: Successfully loaded version 5.0.24 (interface 0x00010004)
[ 3.368665] sr 1:0:0:0: Attached scsi generic sg0 type 5
[ 3.368702] sd 2:0:0:0: Attached scsi generic sg1 type 0
[ 3.412846] snd_intel8x0 0000:00:05.0: disable (unknown or VT-d) VM optimization
[ 3.432220] AES CTR mode by8 optimization enabled
[ 3.492707] XFS (sda1): Mounting V4 Filesystem
[ 3.494193] alg: No test for __gcm-aes-aesni (__driver-gcm-aes-aesni)
[ 3.503610] ppdev: user-space parallel port driver
[ 3.516917] alg: No test for crc32 (crc32-pclmul)
[ 3.520941] intel_rapl: no valid rapl domains found in package 0
[ 3.681038] Adding 3178492k swap on /dev/mapper/centos-swap. Priority:-1 extents:1 across:3178492k SSFS
[ 3.692240] snd_intel8x0 0000:00:05.0: white list rate for 1028:0177 is 48000
[ 4.460341] XFS (sda1): Ending clean mount
[ 4.460404] SELinux: initialized (dev sda1, type xfs), uses xattr
[ 6.363950] floppy0: no floppy controllers found
[ 6.363986] work still pending
[ 6.599985] type=1305 audit(1467323323.159:4): audit_pid=621 old=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:auditd_t:s0 res=1


Running in Headless Mode as well.

Screen Shot 2016-06-30 at 5.47.07 PM.png (25.68 KiB) Viewed 3378 times

[grayfox]

I have a MB Pro with the last software on the Mac and VBox side.

Today I upgraded my Fedora 24 Guest with the latest kernel (4.6.3...) and all heck broke loose.

I had to reinstall the the VBox additions *and* the latest extension pack to get any traction.

The Main Problem for me is that VB can no longer save the state of the running machine. Watching the Virtual Box
Manager while trying to save the instance shows the state changing from 'Saving' to 'Aborted' and attempting
to reboot the aborted image leads to the 'Something Went Wrong' dialog.

If I do a machine shutdown, then I can reboot the fedora instance. Not very satisfying...

Another aggravation is that the VM boots to a black screen and I have to click the mouse to see the desktop.

I might try to revert to the earlier kernel which seemed to be working fine for me...

Jerry

[grayfox]

Whoops,

I am running VBox 5.0.24 with Fedora 24 and there is definitely a problem with kernel 4.6.3-300.

I have removed the kernel and am booting kernel 4.5.7-200 and every thing seems to be working
again.

This is on an updated MacBook Pro.

Jerry

[karozans]

I had the same issue with 5.0.24. I got the same screen.
Earlier today I read a post in the Linux guest section of this forum.

When it gets to the "Oh No! Something has gone wrong." screen, hit the tab key to get focus on the Log Out button, then hit enter.
When the screen goes black, hit the cmd key + F6 to get a console login.
Login then run startx.

They still need to fix the code. This is two releases with faulty code.

Hi StreetBob. Thanks for the help on this. Following your directions I was able to get logged in fine. I notice that they released a new version of virtualbox just a few days ago, but I am still having the same problem with the new one. Hopefully they fix this soon.

[grayfox]

I found that turning off 3d acceleration cured all of my problems....

[socratis]

Moving to "Linux guests" from "Mac OS X Hosts" since the common denominator are the Linux GAs.

[karozans]

I found that turning off 3d acceleration cured all of my problems....

Hi grayfox. That worked for me as well. Seems like 3D acceleration and/or guest additions is causing this.

[wearyofallthiscrap]

FWIW, I'm running a Debian guest using kernel 4.6.x, on a Windows 10 host. Under both 5.0.22, 5.0.24, and 5.1.0, turning on 3D acceleration causes a myriad array of problems. Installing the Guest Additions immediately afterwards strangely makes no difference. That's strange to me, anyhow, but I am not an expert.

Initially, using kernel 4.6.0 under VB 5.0.[22,24], the guest would come up with a completely garbled screen (see Attachment One). Pressing Ctrl-Alt-F1 to switch to a text terminal worked, but would immediately dump an error message from vboxvideo ("bad unpin", see Attachment Two). Switching back to the X11 display would change nothing, but would trigger a new "bad unpin" error when returning to the terminal.

Upgrading the kernel and re-installing the GAs made no difference.

Installing VB 5.1.0 actually made things worse: the X11 display would be black with garbled lines, and switching to the terminal mode would properly resize the window but display nothing else. Still just black, with no response to keyboard input.

Turning off 3D acceleration in the machine settings fixed everything all at once.

Probably there's some corresponding tweaks which I should have made to the guest's X.org settings and/or kernel drivers at the same time I enabled 3D, but I have no idea what they would have been -- I've been approaching this kind of thing as "features get turned on in the machine config, and then the guest just probes stuff at startup, if it works it uses a feature, if not then not". It's probably naive of me, but often works, and it's better than editing X11 config files by hand...