Windows guest frequently aborts

[AledH]

I'm running VBox 4.3.18 on a Fedora 20 x64 host (KDE), but I'm getting frequent aborts of a Windows 7 guest. At the moment it usually happens with a few minutes of starting the VM.

Looking in the log file shows nothing much of interest at the end, as far as I can see. The log is attached.
If I run VBoxSDL from a terminal then I see this when it aborts:

Oracle VM VirtualBox SDL GUI version 4.3.18_RPMFusion
(C) 2005-2014 Oracle Corporation
All rights reserved.

[xcb] Unknown request in queue while dequeuing
[xcb] Most likely this is a multi-threaded client and XInitThreads has not been called
[xcb] Aborting, sorry about that.
VBoxSDL: xcb_io.c:179: dequeue_pending_request: Assertion `!xcb_xlib_unknown_req_in_deq' failed.
Aborted

Unfortunately, it is now getting to the point of becoming unusable.
Any suggestions?

[Perryg]

Code: Select all

VirtualBox VM 4.3.18_RPMFusion r96516 linux.amd64

00:00:00.051423 Executable: /usr/lib64/virtualbox/VBoxSDL

We do not support the RPMFusion fork of VirtualBox. They do.

Don't rely on VBoxSDL it is not supported either. See Users manual.

[AledH]

OK.
I don't use VBoxSDL normally - I only tried that to see if it would show any further diagnostic info as the logs didn't come up with anything.

[AledH]

I have now moved from the RPMFusion build to the RPM one on VirtualBox website, and the same thing happens.
In this case the VM has been running quiescently for a few hours, but when I came to use it again it crashed out.

$ VirtualBox --startvm "Windows 7"
Qt WARNING: libpng warning: iCCP: known incorrect sRGB profile [repeated several times]
VirtualBox: xcb_conn.c:186: write_vec: Assertion `!c->out.queue_len' failed.
Aborted (core dumped)

Again, log file doesn't show anything of great interest (attached).

I got a core dump from this, and the backtrace is:

#0 0x00000033e7235877 in __GI_raise (sig=sig@entry=6) at ../nptl/sysdeps/unix/sysv/linux/raise.c:56
#1 0x00000033e7236f68 in __GI_abort () at abort.c:89
#2 0x00000033e722e7d6 in __assert_fail_base (fmt=0x33e737bb48 "%s%s%s:%u: %s%sAssertion `%s' failed.\n%n",
assertion=assertion@entry=0x33ea2176eb "!c->out.queue_len", file=file@entry=0x33ea2176e0 "xcb_conn.c", line=line@entry=186,
function=function@entry=0x33ea217704 "write_vec") at assert.c:92
#3 0x00000033e722e882 in __GI___assert_fail (assertion=0x33ea2176eb "!c->out.queue_len", file=0x33ea2176e0 "xcb_conn.c", line=186,
function=0x33ea217704 "write_vec") at assert.c:101
#4 0x00000033ea20a183 in _xcb_conn_wait () from /lib64/libxcb.so.1
#5 0x00000033ea20a4c3 in _xcb_out_send () from /lib64/libxcb.so.1
#6 0x00000033ea20ac57 in _xcb_out_flush_to () from /lib64/libxcb.so.1
#7 0x00000033ea20acf8 in xcb_take_socket () from /lib64/libxcb.so.1
#8 0x00000033e9e42536 in require_socket (dpy=0x7f77587e11e0) at xcb_io.c:68
#9 0x00000033e9e42db9 in _XFlush (dpy=0x7f77587e11e0) at xcb_io.c:511
#10 0x00000033e9e45805 in _XGetRequest (dpy=dpy@entry=0x7f77587e11e0, type=type@entry=43 '+', len=len@entry=4) at XlibInt.c:1735
#11 0x00000033e9e3ea1b in XSync (dpy=dpy@entry=0x7f77587e11e0, discard=discard@entry=0) at Sync.c:43
#12 0x00000033f9848cda in dri2XcbSwapBuffers (pdraw=0x7f772c926fb0, remainder=<optimized out>, divisor=0, target_msc=0, dpy=0x7f77587e11e0)
at dri2_glx.c:817
#13 dri2SwapBuffers (pdraw=0x7f772c926fb0, target_msc=0, divisor=0, remainder=<optimized out>, flush=<optimized out>) at dri2_glx.c:855
#14 0x00007f7737ba3a72 in renderspuVBoxPresentCompositionGeneric () from /usr/lib/virtualbox/VBoxOGLrenderspu.so
#15 0x00007f7737ba6130 in renderspuWinCmdThreadProc () from /usr/lib/virtualbox/VBoxOGLrenderspu.so
#16 0x0000003cecef575c in rtThreadMain () from /usr/lib/virtualbox/VBoxRT.so
#17 0x0000003cecf6ad5c in rtThreadNativeMain(void*) () from /usr/lib/virtualbox/VBoxRT.so
#18 0x00000033e7a07ee5 in start_thread (arg=0x7f7760167700) at pthread_create.c:309
#19 0x00000033e72f4b8d in clone () at ../sysdeps/unix/sysv/linux/x86_64/clone.S:111

[Perryg]

Check the power management in the guest. Make sure everything is shut off.

[AledH]

I turned power saving options off in Win guest. Crash still occurs, and I can usually get it to happen only after minutes of the VM being started. It does seems to occur when interacting with the VM (in this case I'm just entering random data into MS Excel) but if I just power the VM up and leave it, it will carry on running for a lot longer.

[Perryg]

All I know to do at this point is to raise a ticket at bugtracker