<div dir="ltr">Sorry about the delay, I&#39;m still unable to confirm if the patch fixed the issue. I should be able to give an affirmative yea or nay by tomorrow. I have a bunch of projects due and didn&#39;t get enough time to test it with a clean build, but I already did the build so all that is left is to test it tomorrow.<div>

<br></div><div style>Sorry again,</div><div style>Ribhi</div></div><div class="gmail_extra"><br><br><div class="gmail_quote">On Thu, Apr 18, 2013 at 10:10 AM, Michael Thayer <span dir="ltr">&lt;<a href="mailto:michael.thayer@oracle.com" target="_blank">michael.thayer@oracle.com</a>&gt;</span> wrote:<br>

<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Hello Ribhi,<br>
<br>
Any update here?  The crash that you reported after applying my patch looked like it was in a completely different area of the VirtualBox code, so I still have hope that my patch fixes the issue.  I would be particularly interested if you happened to prove me wrong on that, but no worries if it is not currently as important for you as it was.<br>


<br>
Regards,<br>
<br>
Michael<div class="im"><br>
<br>
On 16/04/13 23:06, Ribhi Kamal wrote:<br>
</div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="im">
I did a quick test against V4.2.12 and, I&#39;m sorry, it did not work. The<br>
crash seems to have moved somewhere else.<br>
<br>
Steps:<br>
1- Boot up the VM with dual monitors (physical and virtual)<br>
2- Enter fullscreen<br>
3- Exit the VM from within fullscreen view.<br>
<br>
There is a small chance that the crash happened due to my changes or<br>
changes in V4.2.12 (vs V4.2.8). I&#39;m still on campus now, but once I&#39;m<br>
home or work I&#39;ll do a debug build a test it again against an unmodified<br>
V4.2.8. For now, attached is the dump analysis.<br>
<br>
Thanks,<br>
Ribhi<br>
<br>
<br>
<br>
On Tue, Apr 16, 2013 at 2:51 PM, Michael Thayer<br></div><div class="im">
&lt;<a href="mailto:michael.thayer@oracle.com" target="_blank">michael.thayer@oracle.com</a> &lt;mailto:<a href="mailto:michael.thayer@oracle.com" target="_blank">michael.thayer@oracle.<u></u>com</a>&gt;&gt; wrote:<br>
<br>
    Hello Ribhi,<br>
<br>
    Does this patch look good to you?<br>
<br>
    Regards,<br>
<br>
    Michael<br>
<br></div>
    Index: src/VBox/Frontends/VirtualBox/<u></u>__src/runtime/UIFrameBuffer.<u></u>cpp<br>
    ==============================<u></u>__============================<u></u>==__=======<br>
    --- src/VBox/Frontends/VirtualBox/<u></u>__src/runtime/UIFrameBuffer.<u></u>cpp<br>
    (revision 85072)<br>
    +++ src/VBox/Frontends/VirtualBox/<u></u>__src/runtime/UIFrameBuffer.<u></u>cpp<div class="im"><br>
    (revision 85073)<br>
    @@ -158,24 +158,19 @@<br>
              return E_FAIL;<br>
<br>
          NOREF(uScreenId);<br>
    -    /* A resize event can happen while we are switching machine<br>
    view classes,<br>
    -     * but we synchronise afterwards so that shouldn&#39;t be a<br>
    problem. We must<br>
    -     * temporarily remove the framebuffer in Display though while<br>
    switching<br>
    -     * to respect the thread synchronisation logic (see<br>
    UIFrameBuffer.h). */<br>
    +    *pbFinished = FALSE;<br>
    +    lock();  /* See comment in setView(). */<br>
          if (m_pMachineView)<br></div>
              QApplication::postEvent(m___<u></u>pMachineView,<div class="im"><br>
                                      new UIResizeEvent(uPixelFormat, pVRAM,<br>
                                                        uBitsPerPixel,<br>
    uBytesPerLine,<br>
                                                        uWidth, uHeight));<br>
          else<br>
    -    {<br>
              /* Report to the VM thread that we finished resizing and<br>
    rely on the<br>
               * synchronisation when the new view is attached. */<br>
              *pbFinished = TRUE;<br>
    -        return S_OK;<br>
    -    }<br>
    +    unlock();<br>
<br>
    -    *pbFinished = FALSE;<br>
          return S_OK;<br>
      }<br>
<br>
    @@ -200,9 +195,11 @@<br>
              return E_POINTER;<br>
          *pbSupported = TRUE;<br>
<br>
    -    if (!m_pMachineView)<br>
    -        return S_OK;<br></div>
    -    QSize screen = m_pMachineView-&gt;maxGuestSize()<u></u>__;<div class="im"><br>
    +    lock();  /* See comment in setView(). */<br>
    +    QSize screen;<br>
    +    if (m_pMachineView)<br></div>
    +        screen = m_pMachineView-&gt;maxGuestSize()<u></u>__;<div class="im"><br>
    +    unlock();<br>
          if (   (screen.width() != 0)<br>
              &amp;&amp; (uWidth &gt; (ULONG)screen.width())<br>
              &amp;&amp; (uWidth &gt; (ULONG)width()))<br>
    @@ -249,7 +246,10 @@<br>
              reg += rect;<br>
              ++ rects;<br>
          }<br></div>
    -    QApplication::postEvent(m___<u></u>pMachineView, new<div class="im"><br>
    UISetRegionEvent(reg));<br>
    +    lock();  /* See comment in setView(). */<br>
    +    if (m_pMachineView)<br></div>
    +        QApplication::postEvent(m___<u></u>pMachineView, new<div class="im"><br>
    UISetRegionEvent(reg));<br>
    +    unlock();<br>
<br>
          return S_OK;<br>
      }<br>
    @@ -271,6 +271,12 @@<br>
<br></div>
      void UIFrameBuffer::setView(__<u></u>UIMachineView * pView)<div class="im"><br>
      {<br>
    +    /* We are not supposed to use locking for things which are done<br>
    +     * on the GUI thread.  Unfortunately I am not clever enough to<br>
    +     * understand the original author&#39;s wise synchronisation logic<br>
    +     * so I will do it anyway. */<br>
    +    lock();<br>
          m_pMachineView = pView;<br>
          m_WinId = (m_pMachineView &amp;&amp; m_pMachineView-&gt;viewport()) ?<br></div>
    (LONG64)m_pMachineView-&gt;__<u></u>viewport()-&gt;winId() : 0;<div><div class="h5"><br>
    +    unlock();<br>
<br>
      }<br>
<br>
<br>
    On 28/03/13 01:30, Ribhi Kamal wrote:<br>
<br>
        It seems that UIMachineView::maxGuestSize() continues to execute<br>
        while<br>
        some other thread/process destroys the UIMachineView object. To test<br>
        this out, I put in a hack in UIMachineView to basically SpinLock<br>
        until<br>
        any existing maxGuestSize exits. Then it sets a flag using a static<br>
        variable to prevent maxGuestSize from using any member variables<br>
        after<br>
        the view has been destroyed. When the view is recreated, the<br>
        flag is reset.<br>
<br>
        I&#39;ve attached is my hack (based on 4.2.10), it works pretty well and<br>
        seems to stop the crash. I hope this will help you put in a<br>
        better fix<br>
        in the future.<br>
<br>
        Just one question, Which process/thread executes maxGuestSize?<br>
        An EMT<br>
        thread?<br>
<br>
        Cheers!<br>
<br>
        fyi, There are cases where this hack will not work (in theory)<br>
        so please<br>
        don&#39;t use it.<br>
<br>
<br>
        On Sat, Mar 16, 2013 at 4:09 PM, Ribhi Kamal &lt;<a href="mailto:rbhkamal@gmail.com" target="_blank">rbhkamal@gmail.com</a><br>
        &lt;mailto:<a href="mailto:rbhkamal@gmail.com" target="_blank">rbhkamal@gmail.com</a>&gt;<br></div></div><div class="im">
        &lt;mailto:<a href="mailto:rbhkamal@gmail.com" target="_blank">rbhkamal@gmail.com</a> &lt;mailto:<a href="mailto:rbhkamal@gmail.com" target="_blank">rbhkamal@gmail.com</a>&gt;&gt;&gt; wrote:<br>
<br>
             Just happened while switching from full screen back to<br>
        normal view.<br>
             This is something new, only in 4.2, because the virtual<br>
        machine used<br>
             to crash only while closing it so it wasn&#39;t a big deal.<br>
        I&#39;ll open a<br>
             bug once I reproduce it with the released binaries... don&#39;t<br>
        wait.<br>
<br>
             Meanwhile, please let me know if you need any additional<br>
             information/testing.<br>
<br>
             Thanks,<br>
             Ribhi<br>
<br>
<br>
<br>
             On Fri, Mar 15, 2013 at 8:26 PM, Ribhi Kamal<br>
        &lt;<a href="mailto:rbhkamal@gmail.com" target="_blank">rbhkamal@gmail.com</a> &lt;mailto:<a href="mailto:rbhkamal@gmail.com" target="_blank">rbhkamal@gmail.com</a>&gt;<br></div><div class="im">
             &lt;mailto:<a href="mailto:rbhkamal@gmail.com" target="_blank">rbhkamal@gmail.com</a> &lt;mailto:<a href="mailto:rbhkamal@gmail.com" target="_blank">rbhkamal@gmail.com</a>&gt;&gt;&gt; wrote:<br>
<br>
                 &quot;Unfortunately, I can&#39;t find the log files&quot;<br>
<br>
                 Obviously that is not true, I uploaded the logs to my<br>
        dropbox<br>
<br>
<br>
                 On Fri, Mar 15, 2013 at 8:25 PM, Ribhi Kamal<br>
        &lt;<a href="mailto:rbhkamal@gmail.com" target="_blank">rbhkamal@gmail.com</a> &lt;mailto:<a href="mailto:rbhkamal@gmail.com" target="_blank">rbhkamal@gmail.com</a>&gt;<br></div>
                 &lt;mailto:<a href="mailto:rbhkamal@gmail.com" target="_blank">rbhkamal@gmail.com</a><div class="im"><br>
        &lt;mailto:<a href="mailto:rbhkamal@gmail.com" target="_blank">rbhkamal@gmail.com</a>&gt;&gt;&gt; wrote:<br>
<br>
                     I&#39;ve been seeing a crash when closing<br>
        VirtualBox.exe that is<br>
                     almost never reproducible. Few days ago I managed<br>
        to get a<br>
                     crash dump and ran the analysis, see below. At the<br>
        time of<br>
                     the crash, I was closing the virtual machine after<br>
        it had<br>
                     been running for ~24 hours. Unfortunately, I can&#39;t<br>
        find the<br>
                     log files<br>
<br>
                     I&#39;m using the following:<br>
                     VirtualBox 4.2.8 (Cross compiled with VS2010-SP1 on<br>
        windows<br>
                     7 64bit, Target Host = x86)<br>
                     Host Win7 32bit<br>
                     Guest Linux 2.6 32bit<br>
                     Build Type: Release<br>
<br>
                     Please let me know if you have any questions.<br>
<br>
                     Thanks,<br>
                     Ribhi<br>
<br>
                     Log files:<br>
                     Successful:<br></div>
        <a href="https://www.dropbox.com/s/__xrvcr8sud4z63ia/Success.log" target="_blank">https://www.dropbox.com/s/__<u></u>xrvcr8sud4z63ia/Success.log</a><br>
        &lt;<a href="https://www.dropbox.com/s/xrvcr8sud4z63ia/Success.log" target="_blank">https://www.dropbox.com/s/<u></u>xrvcr8sud4z63ia/Success.log</a>&gt;<br>
                     Crash:<br>
        <a href="https://www.dropbox.com/s/__p5pslbt3sl9cpeo/Crash.log" target="_blank">https://www.dropbox.com/s/__<u></u>p5pslbt3sl9cpeo/Crash.log</a><div class="im"><br>
        &lt;<a href="https://www.dropbox.com/s/p5pslbt3sl9cpeo/Crash.log" target="_blank">https://www.dropbox.com/s/<u></u>p5pslbt3sl9cpeo/Crash.log</a>&gt;<br>
<br>
<br>
                     0:000&gt; !analyze -v -f<br>
<br></div>
        ******************************<u></u>__****************************<u></u>**__*******************<div class="im"><br>
<br>
                     *<br>
                     *<br>
                     *                        Exception<br>
                     Analysis                                   *<br>
                     *<br>
                     *<br>
<br></div>
        ******************************<u></u>__****************************<u></u>**__*******************<div class="im"><br>
<br>
<br>
                     GetPageUrlData failed, server returned HTTP status 404<br>
                     URL requested:<br></div>
        <a href="http://watson.microsoft.com/__StageOne/VirtualBox_exe/4_2_8___0/51420e3b/unknown/0_0_0_0/__bbbbbbb4/80000003/00000000.__htm?Retriage=1" target="_blank">http://watson.microsoft.com/__<u></u>StageOne/VirtualBox_exe/4_2_8_<u></u>__0/51420e3b/unknown/0_0_0_0/_<u></u>_bbbbbbb4/80000003/00000000.__<u></u>htm?Retriage=1</a><br>


        &lt;<a href="http://watson.microsoft.com/StageOne/VirtualBox_exe/4_2_8_0/51420e3b/unknown/0_0_0_0/bbbbbbb4/80000003/00000000.htm?Retriage=1" target="_blank">http://watson.microsoft.com/<u></u>StageOne/VirtualBox_exe/4_2_8_<u></u>0/51420e3b/unknown/0_0_0_0/<u></u>bbbbbbb4/80000003/00000000.<u></u>htm?Retriage=1</a>&gt;<br>


<br>
<br>
                     FAULTING_IP:<br>
                     VirtualBox!UIMachineView::__<u></u>maxGuestSize+18<br>
<br>
        [c:\vboxbuild\virtualbox\4.2._<u></u>_8\src\src\vbox\frontends\__<u></u>virtualbox\src\runtime\__<u></u>uimachineview.cpp<div><div class="h5"><br>
                     @ 702]<br>
                     *0145ed68 f00fc70f        lock cmpxchg8b qword ptr<br>
        [edi] *<br>
<br>
<br>
                     EXCEPTION_RECORD:  ffffffff -- (.exr<br>
        0xffffffffffffffff)<br>
                     ExceptionAddress: 00000000<br>
                         ExceptionCode: 80000003 (Break instruction<br>
        exception)<br>
                        ExceptionFlags: 00000000<br>
                     NumberParameters: 0<br>
<br>
                     FAULTING_THREAD:  00000ef4<br>
<br>
                     DEFAULT_BUCKET_ID:  STATUS_BREAKPOINT<br>
<br>
                     PROCESS_NAME:  VirtualBox.exe<br>
<br>
                     ERROR_CODE: (NTSTATUS) 0x80000003 - {EXCEPTION}<br>
          Breakpoint<br>
                     A breakpoint has been reached.<br>
<br>
                     EXCEPTION_CODE: (HRESULT) 0x80000003 <a href="tel:%282147483651" value="+12147483651" target="_blank">(2147483651</a><br>
        &lt;tel:%282147483651&gt;<br>
                     &lt;tel:%282147483651&gt;) - One or more arguments are<br>
        invalid<br>
<br>
<br>
                     MOD_LIST: &lt;ANALYSIS/&gt;<br>
<br>
                     NTGLOBALFLAG:  0<br>
<br>
                     APPLICATION_VERIFIER_FLAGS:  0<br>
<br>
                     ADDITIONAL_DEBUG_TEXT:  Followup set based on attribute<br>
                     [Is_ChosenCrashFollowupThread] from Frame:[0] on<br>
                     thread:[PSEUDO_THREAD]<br>
<br>
                     LAST_CONTROL_TRANSFER:  from 0143bea0 to 0145ed68<br>
<br>
                     PRIMARY_PROBLEM_CLASS:  STATUS_BREAKPOINT<br>
<br>
                     BUGCHECK_STR:<br></div></div>
                     APPLICATION_FAULT_STATUS___<u></u>BREAKPOINT_NULL_POINTER_READ<br>
<br>
                     STACK_TEXT:<br>
                     03a3ca98 0145ed68<br>
        virtualbox!UIMachineView::__<u></u>maxGuestSize+0x18<br>
                     03a3cab0 0143bea0<br>
                     virtualbox!UIFrameBuffer::__<u></u>VideoModeSupported+0x30<br>
                     03a3cac8 6927c724 vboxc!__<u></u>vmmdevVideoModeSupported+0x74<br>
                     03a3caec 690b4edc vboxdd!vmmdevRequestHandler+__<u></u>0xecc<div class="im"><br>
                     03a3fb5c 72f198d1 vboxvmm!IOMIOPortWrite+0x91<br>
                     03a3fb84 72f0891f<br></div>
        vboxvmm!__<u></u>HWACCMR3RestartPendingIOInstr+<u></u>__0xcf<br>
                     03a3fba4 72ea303c<br>
        vboxvmm!__<u></u>emR3ExecuteIOInstruction+0x1c<br>
                     03a3fc78 72ea3589 vboxvmm!emR3HwaccmHandleRC+__<u></u>0x189<div class="im"><br>
                     03a3fc8c 72ea3788 vboxvmm!emR3HwAccExecute+0x168<br>
                     03a3fcb0 72ea0d84 vboxvmm!EMR3ExecuteVM+0x274<br>
                     03a3fcd8 72efb2aa<br></div>
        vboxvmm!__<u></u>vmR3EmulationThreadWithId+__<u></u>0x45a<br>
                     03a3fcf8 72efb2f4 vboxvmm!vmR3EmulationThread+__<u></u>0x14<div class="im"><br>
                     03a3fd0c 69ca1523 vboxrt!rtThreadMain+0x33<br>
                     03a3fd38 69ce539b vboxrt!rtThreadNativeMain+0x6b<br>
                     03a3fd58 6bb6c556 msvcr100!_endthreadex+0x3f<br>
                     03a3fd90 6bb6c600 msvcr100!_endthreadex+0xce<br></div>
                     03a3fd9c 76b4ed6c kernel32!BaseThreadInitThunk+_<u></u>_0xe<br>
                     03a3fda8 7722377b ntdll!__RtlUserThreadStart+__<u></u>0x70<div class="im"><br>
                     03a3fde8 7722374e ntdll!_RtlUserThreadStart+0x1b<br>
<br>
<br>
                     STACK_COMMAND:  .cxr 0000000003A3C7B4 ; kb ; dds<br>
        3a3ca98 ; kb<br>
<br>
                     FOLLOWUP_IP:<br></div>
                     VirtualBox!UIMachineView::__<u></u>maxGuestSize+0<br>
<br>
        [c:\vboxbuild\virtualbox\4.2._<u></u>_8\src\src\vbox\frontends\__<u></u>virtualbox\src\runtime\__<u></u>uimachineview.cpp<div class="im"><br>
                     @ 701]<br>
                     0145ed50 83ec0c          sub     esp,0Ch<br>
<br>
                     FAULTING_SOURCE_CODE:<br>
                         697:<br>
        RT_MAKE_U64(maxSize.height(),<br>
                     maxSize.width()));<br>
                         698: }<br>
                         699:<br>
                         700: QSize UIMachineView::maxGuestSize()<br>
                      &gt;  701: {<br>
                         702:     uint64_t u64Size =<br></div>
                     ASMAtomicReadU64(&amp;m___<u></u>u64MaxGuestSize);<div class="im"><br>
                         703:     return QSize(int(RT_HI_U32(u64Size)),<br>
                     int(RT_LO_U32(u64Size)));<br>
                         704: }<br>
                         705:<br>
                         706: QSize UIMachineView::guestSizeHint()<br>
<br>
<br>
                     SYMBOL_NAME:<br></div>
          virtualbox!UIMachineView::__<u></u>maxGuestSize+0<div class="im"><br>
<br>
                     FOLLOWUP_NAME:  MachineOwner<br>
<br>
                     MODULE_NAME: VirtualBox<br>
<br>
                     IMAGE_NAME:  VirtualBox.exe<br>
<br>
                     DEBUG_FLR_IMAGE_TIMESTAMP:  51420e3b<br>
<br>
                     FAILURE_BUCKET_ID:<br>
<br></div>
        STATUS_BREAKPOINT_80000003___<u></u>VirtualBox.exe!UIMachineView::<u></u>__maxGuestSize<br>
<br>
<br>
                     BUCKET_ID:<br>
<br>
        APPLICATION_FAULT_STATUS___<u></u>BREAKPOINT_NULL_POINTER_READ__<u></u>_virtualbox!UIMachineView::__<u></u>maxGuestSize+0<br>
<br>
<br>
                     WATSON_STAGEONE_URL:<br>
        <a href="http://watson.microsoft.com/__StageOne/VirtualBox_exe/4_2_8___0/51420e3b/unknown/0_0_0_0/__bbbbbbb4/80000003/00000000.__htm?Retriage=1" target="_blank">http://watson.microsoft.com/__<u></u>StageOne/VirtualBox_exe/4_2_8_<u></u>__0/51420e3b/unknown/0_0_0_0/_<u></u>_bbbbbbb4/80000003/00000000.__<u></u>htm?Retriage=1</a><div class="im">

<br>
        &lt;<a href="http://watson.microsoft.com/StageOne/VirtualBox_exe/4_2_8_0/51420e3b/unknown/0_0_0_0/bbbbbbb4/80000003/00000000.htm?Retriage=1" target="_blank">http://watson.microsoft.com/<u></u>StageOne/VirtualBox_exe/4_2_8_<u></u>0/51420e3b/unknown/0_0_0_0/<u></u>bbbbbbb4/80000003/00000000.<u></u>htm?Retriage=1</a>&gt;<br>


<br>
<br>
                     Followup: MachineOwner<br>
                     ---------<br>
</div></blockquote>
<br>
<br>
-- <br><div class="HOEnZb"><div class="h5">
ORACLE Deutschland B.V. &amp; Co. KG   Michael Thayer<br>
Werkstrasse 24                     VirtualBox engineering<br>
71384 Weinstadt, Germany           mailto:<a href="mailto:michael.thayer@oracle.com" target="_blank">michael.thayer@oracle.<u></u>com</a><br>
<br>
Hauptverwaltung: Riesstr. 25, D-80992 München<br>
Registergericht: Amtsgericht München, HRA 95603<br>
Geschäftsführer: Jürgen Kunz<br>
<br>
Komplementärin: ORACLE Deutschland Verwaltung B.V.<br>
Hertogswetering 163/167, 3543 AS Utrecht, Niederlande<br>
Handelsregister der Handelskammer Midden-Niederlande, Nr. 30143697<br>
Geschäftsführer: Alexander van der Ven, Astrid Kepper, Val Maher<br>
</div></div></blockquote></div><br><br clear="all"><div><br></div>-- <br>-- Ribhi
</div>