virtualbox.org

Discuss the 4.3.0 release.

The binaries can be downloaded here.

(The site admins usually create this topic, but somehow that was neglected this time).

Hey !

I'm working since more than 1 year with my Win7 64 bits guest (arch linux 64 bits host). I originally activated it with the activation code provided for my Laptop, I've never had trouble with that.

And now, after VirtualBox 4.3 update, my Win7 guest is asking me to activate it again. What's happening ?

I've updated virtualbox + extension pack + guest additions on all my machines.

I installed v4.3.0 last week and experienced a strange display resolution issue that only affected my Windows 7 and Windows 8.1 Preview VMs. With GAs installed (also v4.3) the guest display maxed out at something around to 1366x768 in full screen. My one XP VM and multiple Ubuntu VMs were unaffected and could snap to full screen on a 1920x1080 monitor without any issue. I have since reverted to 4.2.8 but will reinstall 4.3.0 sometime soon to see if I can duplicate the behavior. Unfortunately I idiotically neglected to save the logs at the time.

Just some feedback on the new release:

4.2.18 worked great, only after installing 4.3.0 my 64bit linux guests no longer work (they won't boot, I can't install a new 64bit machine without it telling me it can't install on a i686 system (!?!?!)).

I'm running my virtual machines on an Intel i7 920 running at 4GHz, aircooled at 60 degrees celcius with 12GB RAM on a 256GB SSD running Windows 7 Ultimate x64. I've installed my X58 chipset drivers, made sure my Windows 7 was 64bit and I even did the cold boot and bios VT CPU flag enable-and-disable 'fix' as suggested here on the forum for some crappy mainboards (set it back to on, which was the value on which everything worked just peachy before). I have just reset everything back to how it was before I installed 4.3.0.

After replaceing 4.3.0 with 4.2.18, everything works fine again... my 64bit virtual machines allocate 6GB's of RAM and they work just fine.

I don't want to sound ungrateful, but I'm glad we didn't upgrade our development machines to run Vbox 4.3.0 at the office yet. It's clearly not entirely ready for production development environments yet, but keep up the good work.. Thanks for the nice product you've made so far. I can't stand VMware or that crap Microsoft has (but please fix this nonsense with the 4.3.0 update )

ORACLE VIRTUALBOX, YOU'RE THE BEST (FOR ME)! (or 'were the best' to be absolutely correct, but I'm sure that will be fixed in the next update)

Some issues and workarounds that have arisen in other threads (before I created this one).

VERR_SSM_FIELD_NOT_CONSECUTIVE issue.
Windows 8.1 Critical Structure Corruption crash.
64bit host capability not being recognized.

j.ledun wrote:Hey !

I'm working since more than 1 year with my Win7 64 bits guest (arch linux 64 bits host). I originally activated it with the activation code provided for my Laptop, I've never had trouble with that.

And now, after VirtualBox 4.3 update, my Win7 guest is asking me to activate it again. What's happening ?

I've updated virtualbox + extension pack + guest additions on all my machines.

I've possibly found a fix/workaround for this one. It seems that VBox4.3 doesn't use the hardwareuuid from the Machine tag in the vbox file anymore.

I was able to reuse MS Office in a XP VM by adding the uuid to the Hardware tag. (It had detected hardware changes after the 4.3 upgrade).

Maybe it would also affect the hardwarechecks from Win7.

Workaround:
Before After

Could you possibly reproduce that posting in the "What changes trigger reactivation" thread in "Windows Guests"?

bluec0re wrote:I've possibly found a fix/workaround for this one. It seems that VBox4.3 doesn't use the hardwareuuid from the Machine tag in the vbox file anymore.

Thank you! This is indeed a bug, the UUID is now getting lost. Or rather the VM's UUID is no longer getting propagated to the SMBIOS machine UUID.

The HW UUID hacking (avoiding Windows re-activation) works around a regression which crept into the 4.3 development 10 months ago. A very evil silent regression obviously, as to my knowledge no one noticed this in our testing, and no beta test bug reports either.

Thanks bluec0re, your suggestion worked fine for me.

I was one of several who posted in the 4.2.18 discussion thread about Win 8 host system crashes whenever a USB device was attached to a guest VM. 4.3.0 isn't working for me either, but the symptoms are different. I'm interested in whether other people with similar host systems are experiencing this.

My host is an HP p7-1539 with AMD A10-5800k CPU, originally running Win 8 but since upgraded to Win 8.1. When I attempt to attach a USB device to a guest -- I've tried Win XP, Win 7, and Linux Mint guests -- the attach request hangs. The device status (readable hovering the cursor over the GUI device listing) changes from Busy to Captured, but the device never actually attaches (no check mark in the GUI, device never appears in the guest). Retrying the attach results in an error, e.g.:

Failed to attach the USB device SMI Corporation USB DISK [1100] to the virtual machine Minty.
USB device 'SMI Corporation USB DISK' with UUID {c90f7767-fe5e-4616-b410-6e6486b33f6d} is busy with a previous request. Please try again later.
Result Code: E_INVALIDARG (0x80070057)
Component: HostUSBDevice
Interface: IHostUSBDevice {173b4b44-d268-4334-a00d-b6521c9a740a}
Callee: IConsole {8ab7c520-2442-4b66-8d74-4ff1e195d2b6}

Neither the guest's log file nor the VboxSVC log in .VirtualBox appears to have any mention of the event, except for the error quoted above. If I unplug and re-insert the USB device on the host system, the status in VBox remains Captured. It doesn't seem to matter whether or not I upgrade Guest Additions to 4.3.

Here's the odd part, though: As a test I installed vanilla Win 8.1 (the recent RTM version available on the net, no added drivers) on both my HP machine and on a Lenovo system with an Intel CPU, and attaching USB devices with VBox 4.3 worked fine. I can't run this version of Win 8.1 on more than a trial basis, so I restored the HP system to its Win 8 factory state using recovery media -- and the problem returned! It would *appear* that the problem stems from some interaction with an HP-provided driver or utility, but I don't know what. I would expect that any HP Win 8 system with this or a similar factory software build would suffer this problem. And I am reasonably sure that 4.2.16 would still work fine on such a system (ETA: It does). Any troubleshooting tips?

Update 11/1/13: It turns out I can validate the Win 8.1 RTM with the OEM license that came with the PC, so it looks like my solution will be to reinstall Windows from that, and ditch the HP bloatware. Still don't know what caused the problem, and I still believe any user with a similar HP build will also be affected.

I can use v4.3.0 only on one of my two HP laptops!
Details are here: Re: E_NOINTERFACE (0x80004002)

> Activation ... XP

I just tried the same in a Windows7Pro guest -- followed the example given, looks like the same change made, but still get the "9 days" warning to activate Windows. More help welcome, if anyone has gotten that to work with Win7.

Because if I ever meet that guy whose voice I've become so incredibly familiar with doing the phone activation thing over and over whenever I change a VM, I may just ...
ahem. Speak harshly to him. Yeah, that's it.

I think XP activation works differently from Win7. Not just in the calculation of a hardware signature, but what it does when it detects a change. I think in XP's case it sets a flag which will cause it to keep asking for activation, even if you correct the hardware change. The flag is probably set in the registry somewhere: worth researching and reporting back. Win7 doesn't seem to work that way.

Full Screen Not Working after upgrade to 4.3

Host: OSX Mavericks on 27" 2013 iMac
VM guests: Ubuntu Studio, MS Window 8
Extensions and Guest Additions installed successfully

I upgraded from 4.2.18 to 4.3, and on the MS Windows 8 guest only, the full screen mode consumes the full display monitor (as it should), but displays the guest window as reduced size, centered on black background.

I reported this same problem when I upgraded from 4.2.16 to 4.2.18, and hoped it would be fixed - but not yet, I guess.

Does anyone else have this problem or know about it? Is it something I'm doing wrong?

Thanks