4.3.20 crashes on start on Windows 10 build 9879

[Sergei]

4.3.20 crashes on start with 0xC0000005 (Access violation) on Windows 10 build 9879 x64. I could not find VBoxStartup.log ort VBox.log anywhere on my disk, but here is WinDbg diagnostics with call stack etc.

Microsoft (R) Windows Debugger Version 6.3.9600.17237 AMD64
Copyright (c) Microsoft Corporation. All rights reserved.

CommandLine: "C:\Program Files\Oracle\VirtualBox\VirtualBox.exe"
Symbol search path is: *** Invalid ***
****************************************************************************
* Symbol loading may be unreliable without a symbol search path. *
* Use .symfix to have the debugger choose a symbol path. *
* After setting your symbol path, use .reload to refresh symbol locations. *
****************************************************************************
Executable search path is:
ModLoad: 00007ff6`43820000 00007ff6`43926000 VirtualBox.exe
ModLoad: 00007fff`81540000 00007fff`816f7000 ntdll.dll
ModLoad: 00007fff`7f070000 00007fff`7f11b000 C:\WINDOWS\system32\KERNEL32.DLL
ModLoad: 00007fff`7ea70000 00007fff`7ec36000 C:\WINDOWS\system32\KERNELBASE.dll
ModLoad: 00007fff`7d0f0000 00007fff`7d16f000 C:\WINDOWS\system32\apphelp.dll
ModLoad: 00007fff`7e580000 00007fff`7e5ac000 C:\WINDOWS\SYSTEM32\bcrypt.dll
SHIMVIEW: ShimInfo(Complete)
(126c.11d8): Break instruction exception - code 80000003 (first chance)
*** ERROR: Symbol file could not be found. Defaulted to export symbols for ntdll.dll -
ntdll!LdrInitShimEngineDynamic+0x360:
00007fff`815f20c0 cc int 3
0:000> g
(126c.11d8): Access violation - code c0000005 (first chance)
First chance exceptions are reported before any exception handling.
This exception may be expected and handled.
*** ERROR: Symbol file could not be found. Defaulted to export symbols for VirtualBox.exe -
VirtualBox+0x15602:
00007ff6`43835602 8a042515000000 mov al,byte ptr [15h] ds:00000000`00000015=??
*** ERROR: Symbol file could not be found. Defaulted to export symbols for C:\WINDOWS\system32\KERNEL32.DLL -
0:000> k
Child-SP RetAddr Call Site
00000000`0017fa60 00007ff6`43827aa1 VirtualBox+0x15602
00000000`0017fb00 00007fff`7f07a6e2 VirtualBox+0x7aa1
00000000`0017fb60 00007fff`8155c9a4 KERNEL32!BaseThreadInitThunk+0x22
00000000`0017fb90 00000000`00000000 ntdll!RtlUserThreadStart+0x34

Looks like a try to read a byte from a structure via a NULL pointer. It's clear, isn't it?
A number of people on this forum want Windows 10 support. Can we expect a fixed version in near future?

[loukingjr]

Define "near future"?

A new build of Windows 10 is expected in January. Can we expect that whatever changes are coming wouldn't affect a fix in VirtualBox that comes out before then?

[mpack]

How many of those people wanting Win10 host support are willing to pay for it? ... Or are the devs supposed to get excited by the prospect of working for nothing on something that might get invalidated by the next Win10 beta?

[loukingjr]

Exactly.

[Jacob Klein]

I would be quite interested in knowing whether the incompatibility is caused by something within VirtualBox, or some bug within the Windows 10 architecture. If it's the fault of Windows 10, I could provide valuable feedback to Microsoft in order to get it fixed!

However, until an Oracle developer is willing to investigate, I guess we'll never know.
Usually, these sorts of investigations are enough to get excited about making something work.

[loukingjr]

However, until an Oracle developer is willing to investigate, I guess we'll never know.
Usually, these sorts of investigations are enough to get excited about making something work.

What makes you think they aren't? Oh, I know, because they don't keep you informed what they are working on.

[Jacob Klein]

I do not have any evidence that the Oracle or Microsoft developers have identified the problem.

However, if they did identify it, I would have thought we would either see it fixed in a VirtualBox test build (which I install and retest on every new version), or would have indicated somewhere that the incompatibility is a due to a Windows bug. Again, I do not have any further information.

I'm sorry you feel the need to get so defensive/hostile. It is the lack of communication regarding the issue, that is creating the frustration. I surely didn't mean to offend you or anyone else.

[loukingjr]

I'm sorry you feel the need to get so defensive/hostile. It is the lack of communication regarding the issue, that is creating the frustration. I surely didn't mean to offend you or anyone else.

I'm not defensive nor hostile and I'm certainly not offended. I'm also not frustrated because I'm not kept up to date. I look at all developing software as taking time to finish and never expect developers to keep me abreast of what they are working on.

[Jacob Klein]

We see it differently.
I see it as a bug/issue/problem, and am trying to solve it, but I've gone as far as I can go towards solving it.
I wish there was something more I could do, to help the other users who have the same problem, but I've provided any diagnostic data that I know of.

[loukingjr]

Maybe you don't know this but there is both a developers mailing list and an IRC developers channel you could join which might be more satisfying for you than the users forum. https://www.virtualbox.org/wiki/Community

[socratis]

Jacob, I would add that if you really want to know what the developers are working on, there are a couple more approaches that you want to look at:

1. Try the Testbuilds (https://www.virtualbox.org/wiki/Testbuilds). They might break other things, but they're supported.
2. Take a look at the Timeline (https://www.virtualbox.org/timeline). They're more than open in that regard.
3. Build VirtualBox from source (https://www.virtualbox.org/wiki/Build_instructions). Daily updated from the SVN tree, the same as above. Remember, those are the public changes, but if you look at it closely, and you decided to (let's say) build yesterday's SVN, you'd be closer to 4.4.

Just one thing; don't be an "Dad, are we there yet?" kind of a guy, because you'll accomplish to a) irritate the developers and b) delay them from actually doing the work.

[Jacob Klein]

Thanks for the replies, guys.

I use the TestBuilds (checking it daily, to see if this issue is resolved yet), and I sometimes monitor the Timeline (which is often difficult to sift through to see if this particular issue is resolved). And maybe I'll monitor the Developer mailing list a bit more. I'm not really interested in building VirtualBox myself, but I appreciate that the option exists.

If there's anything I (as a user, not as a developer), can do to further help the developers solve this or test a solution, I am of course willing.

In the meantime, I try to be as patient as I can, and I hope that you can appreciate the uneasiness it is causing the Windows 10 early adopters. So far as I know, this is the first Microsoft OS where they are actively seeking early feedback from anyone willing to install it, and this issue continues to be one of the few open issues I have, across all of my apps.

Regards,
Jacob

[loukingjr]

Jacob, what you don't seem to understand is Microsoft wants feedback on Windows 10 and any software that they include. They would have very little interest if a third party application works on Windows 10 at this time. They have their hands full I'm sure making sure the OS works correctly first. As I mentioned to you once before, they are not going to change Windows 10 to work with third party applications.

As far as VirtualBox I'm sure as has also been mentioned to you before, that Windows 10 is released before they put much effort into making VirtualBox run on it.

[Jacob Klein]

Microsoft is actively taking feedback on anything relating to Windows 10, including apps and software that no-longer function correctly, including VirtualBox. Several users have given that feedback to them already. The Microsoft Windows Insider lead is even responding to tweets about such incompatibilities, and relaying info directly to the Microsoft devs.
http://blogs.windows.com/bloggingwindow ... r-program/
https://twitter.com/GabeAul/with_replies

You and I see the situation very differently. My opinion is that this issue should be proactively fixed before the Windows 10 release, and not reactively fixed afterwards. It is still a struggle to remain patient, with no Oracle developer feedback on this issue, but I do my best to sit here and stare at the issue while the waiting game endures.

[loukingjr]

So you have said many, many times. You've posted about how you see the issue at least 5 times which to me doesn't look very patient.