Frequent and annoying Wine bug

[Robbert Xerox]

Hi , i know this is not a wine-bug channel, but this
bug is affecting quite some apps, and seems to affect
more and more ...
At least five apps from wine-bug mailinglist throw up
an exception which either reads: "Assertion Failed
!bogus context in Local_Unwind()" or "in
Exception_Handler". As a reference a filed 2 bugs, and
commented another one where you can read the
+relay,+seh logs:
 Bug 3097
 Bug 3127
 Bug 3139

Some googling around makes me believe that this is a
Borland specific error. I tried to install Demo
Borland compiler in wine to see if i could reproduce
this error, but unfortunately it fails with exactly
the same error.

all programs show the same pattern before an exception
is thrown: they call LoadstringA() ,with something
like 
 L"Failed to get data for '%s'" or L"Access
violation at address %p in module '%s'. %s of address
%p" See the bugs for more info.

I think this bug is really affecting quite some more
apps then the ones mentioned in wine-bugs so would be
great if anyone could shine a light on this, as of how
to solve the bug. Regards



		
___________________________________________________________ 
How much free photo storage do you get? Store your holiday 
snaps for FREE with Yahoo! Photos http://uk.photos.yahoo.com

[Alexandre Julliard]

Robbert Xerox <xerox_xerox2000(a)yahoo.co.uk> writes:
> I think this bug is really affecting quite some more
> apps then the ones mentioned in wine-bugs so would be
> great if anyone could shine a light on this, as of how
> to solve the bug. Regards

I tried several of the apps you mention and they all work fine here,
so I suspect some sort of glibc/kernel issue. You should provide more
details on your setup, and ask other people who see the problem to
provide details too so that we can see if there's a correlation with
kernel versions etc.

-- 
Alexandre Julliard
julliard(a)winehq.org

[Francois Gouget]

Hi,

Robbert Xerox wrote:
> Hi , i know this is not a wine-bug channel, but this
> bug is affecting quite some apps, and seems to affect
> more and more ...
> At least five apps from wine-bug mailinglist throw up
> an exception which either reads: "Assertion Failed
> !bogus context in Local_Unwind()" or "in
> Exception_Handler".

I have seen a bug pretty much like this in an application and here's 
what Mike Hearn had to say about it:
> There were actually two bugs here:
>
> * Despite what the unit tests indicate, IShellFolder::BindToObject
>  with a NULL pidl can sometimes work. I guess it depends on the exact
>  class implementation. In this case I fixed it with an application
>  specific hack, we just return the desktop folder (it should probably
>  return an addreffed This).
>
> * The app apparently relies on an old Win98 specific bug where
>  SFGAO_FILESYSTEM is set on My Computer. This bug is documented here:
>
> http://msdn.microsoft.com/msdnmag/issues/05/06/CAtWork/default.aspx
>
> at the bottom. But it's not present in Windows NT, as our unit tests
>  actually prove (we should pay more attention to 98/NT differences!!):

http://test.winehq.com/data/200506231000/98_PaulVriensW98SEFull/shell32:shlf...
> It can be fixed with the patch to wine.inf, if you then do "make
>  install" in the tools directory and re-run wineprefixcreate.

I've attached the patch so you can try it. That's not code I can comment 
on but maybe Mike will when he's back from his not-quite-vacations.

-- 
Francois Gouget
fgouget(a)codeweavers.com


Index: dlls/shell32/shlfolder.c
===================================================================
RCS file: /var/cvs/wine/dlls/shell32/shlfolder.c,v
retrieving revision 1.104
diff -u -p -r1.104 shlfolder.c
--- dlls/shell32/shlfolder.c	20 Jul 2005 10:29:05 -0000	1.104
+++ dlls/shell32/shlfolder.c	28 Jul 2005 21:33:11 -0000
@@ -272,6 +272,15 @@ HRESULT SHELL32_BindToChild (LPCITEMIDLI
     HRESULT hr;
     LPITEMIDLIST pidlChild;
 
+    TRACE("pidlRoot=%p, pathRoot=%s, pidlComplete=%p, riid=%s, ppvOut=%p\n",
+          pidlRoot, pathRoot, pidlComplete, debugstr_guid(riid), ppvOut);
+
+    if (!pidlComplete)
+    {
+        SHELL32_CoCreateInitSF(_ILCreateDesktop(), pathRoot, NULL, &CLSID_ShellDesktop, riid, ppvOut);
+        return S_OK;
+    }
+    
     if (!pidlRoot || !ppvOut || !pidlComplete || !pidlComplete->mkid.cb)
         return E_INVALIDARG;
 
Index: tools/wine.inf
===================================================================
RCS file: /var/cvs/wine/tools/wine.inf,v
retrieving revision 1.37
diff -u -p -r1.37 wine.inf
--- tools/wine.inf	27 Jul 2005 15:42:40 -0000	1.37
+++ tools/wine.inf	28 Jul 2005 21:30:35 -0000
@@ -100,6 +100,8 @@ HKCR,TypeLib\{00020430-0000-0000-C000-00
 HKCR,TypeLib\{00020430-0000-0000-C000-000000000046}\2.0,,,"OLE Automation"
 HKCR,TypeLib\{00020430-0000-0000-C000-000000000046}\2.0\0\win32,,,"stdole2.tlb"
 HKCR,TypeLib\{00020430-0000-0000-C000-000000000046}\2.0\FLAGS,,,"0"
+; FIXME: this is to emulate a Windows 98 bug which is gone in NT: very few apps need it
+HKCR,CLSID\{20d04fe0-3aea-1069-a2d8-08002b30309d}\ShellFolder,"Attributes",0x10001,f0000154
 
 [ControlClass]
 HKLM,System\CurrentControlSet\Control\Class\{4d36e978-e325-11ce-bfc1-08002be10318},,,"Ports (COM & LPT)"

[Robbert Xerox]

Hi, late response, been on holiday :). The patch from 
francios gougetdoesn't fix the error. I dug a little
deeper into this:I had a knoppix live cd lying around
and there bug is not present. A program like Neatimage
(www.neatimage.com) startsup just fine there.  I also
did some tracing to see where the difference is. I'm
quite noobish in this so correct me plz if i'm wrong:

On knoppix running neatimage fine i see this:

0009:Ret  user32.LoadStringA() retval=0000001b
ret=00527b86
0009:Call
kernel32.RaiseException(0eedfade,00000001,00000007,4067f930)
ret=004aefac
0009:Call ntdll.RtlRaiseException(4067f7d8)
ret=404a5393 fs=1007
 eax=4067f7d8 ebx=40569244 ecx=00000000 edx=00000018
esi=4067f94c edi=4067f808
 ebp=4067f834 esp=4067f7d8 ds=002b es=002b gs=0000
flags=00200216
trace:seh:EXC_RtlRaiseException code=eedfade flags=1
addr=0x404a5300
trace:seh:EXC_RtlRaiseException  info[0]=004aefac
trace:seh:EXC_RtlRaiseException  info[1]=411d3068
trace:seh:EXC_RtlRaiseException  info[2]=411d2fb0
trace:seh:EXC_RtlRaiseException  info[3]=411d3058
trace:seh:EXC_RtlRaiseException  info[4]=4067f974
trace:seh:EXC_RtlRaiseException  info[5]=4067f964
trace:seh:EXC_RtlRaiseException  info[6]=4067f94c
trace:seh:EXC_RtlRaiseException  eax=4067f7d8
ebx=40569244 ecx=00000000 edx=00000018 esi=4067f94c
edi=4067f808
trace:seh:EXC_RtlRaiseException  ebp=4067f834
esp=4067f7d8 cs=0023 ds=002b es=002b fs=1007 gs=0000
flags=00200216
trace:seh:EXC_CallHandler calling handler at 0x538243
code=eedfade flags=1
0009:Call
ntdll.RtlUnwind(4067f990,0053916b,4067f7d8,00000000)
ret=0053916b fs=1007
 eax=4067f990 ebx=4067f7d8 ecx=00549d68 edx=4067f7d8
esi=0000001c edi=4067f7d8
 ebp=4067f0a4 esp=4067f040 ds=002b es=002b gs=0000
flags=00200206
trace:seh:EXC_RtlUnwind code=eedfade flags=3
trace:seh:EXC_CallHandler calling handler at 0x538243
code=eedfade flags=3
trace:seh:EXC_CallHandler handler returned 1
trace:seh:EXC_CallHandler calling handler at
0x401ae2c0 code=eedfade flags=3
trace:seh:EXC_CallHandler handler returned 1
0009:Ret  ntdll.RtlUnwind() retval=00000000
ret=0053916b fs=1007
 eax=00000000 ebx=4067f7d8 ecx=00549d68 edx=4067f7d8
esi=0000001c edi=4067f7d8
 ebp=4067f0a4 esp=4067f040 ds=002b es=002b gs=0000
flags=00200206

0009:Call advapi32.RegSetValueExA(0000005c,411d3058
"StartCount",00000000,00000004,4067f97c,00000004)
ret=004aef22

Looks to me as if an exception is thrown, wine is
handling the exception and the program continues fine

On my fc3 computer i see this:

000b:Ret  user32.LoadStringA() retval=0000001b
ret=00527b86
000b:Call
kernel32.RaiseException(0eedfade,00000001,00000007,7fa9f93c)
ret=004aefac
000b:Call ntdll.RtlRaiseException(7fa9f824)
ret=7fcfe544 fs=003b
 eax=7fcec305 ebx=7fd5bacc ecx=00000000 edx=0eedfade
esi=7fa9f958 edi=7fa9f854
 ebp=7fa9f880 esp=7fa9f824 ds=007b es=007b gs=0033
flags=00200212
trace:seh:__regs_RtlRaiseException code=eedfade
flags=1 addr=0x7fcfe4e0
trace:seh:__regs_RtlRaiseException  info[0]=004aefac
trace:seh:__regs_RtlRaiseException  info[1]=7e4c32f4
trace:seh:__regs_RtlRaiseException  info[2]=7e4c323c
trace:seh:__regs_RtlRaiseException  info[3]=7e4c32e4
trace:seh:__regs_RtlRaiseException  info[4]=7fa9f980
trace:seh:__regs_RtlRaiseException  info[5]=7fa9f970
trace:seh:__regs_RtlRaiseException  info[6]=7fa9f958
trace:seh:__regs_RtlRaiseException  eax=7fcec305
ebx=7fd5bacc ecx=00000000 edx=0eedfade esi=7fa9f958
edi=7fa9f854
trace:seh:__regs_RtlRaiseException  ebp=7fa9f880
esp=7fa9f824 cs=0073 ds=007b es=007b fs=003b gs=0033
flags=00200212
trace:seh:EXC_CallHandler calling handler at 0x538243
code=eedfade flags=1
000b:Call
kernel32.GetModuleFileNameA(00000000,7fa9f0c8,00000080)
ret=00541424
000b:Call
ntdll.RtlAllocateHeap(7fdb0000,00000000,00000100)
ret=7fd06d22
000b:Ret  ntdll.RtlAllocateHeap() retval=7fdef278
ret=7fd06d22
000b:Call
ntdll.LdrLockLoaderLock(00000000,00000000,7fa9efdc)
ret=7fd1638f
000b:Ret  ntdll.LdrLockLoaderLock() retval=00000000
ret=7fd1638f
000b:Call
ntdll.LdrFindEntryForAddress(00400000,7fa9efd8)
ret=7fd163a1
000b:Ret  ntdll.LdrFindEntryForAddress()
retval=00000000 ret=7fd163a1
000b:Call ntdll.LdrUnlockLoaderLock(00000000,0000000b)
ret=7fd163e5
000b:Ret  ntdll.LdrUnlockLoaderLock() retval=00000000
ret=7fd163e5
000b:Call
ntdll.RtlUnicodeToMultiByteN(7fa9f0c8,00000080,7fa9efdc,7fdef278,00000052)
ret=7fcff857
000b:Ret  ntdll.RtlUnicodeToMultiByteN()
retval=00000000 ret=7fcff857
000b:Call
ntdll.RtlFreeHeap(7fdb0000,00000000,7fdef278)
ret=7fd06d4a
000b:Ret  ntdll.RtlFreeHeap() retval=00000001
ret=7fd06d4a
000b:Ret  kernel32.GetModuleFileNameA()
retval=00000029 ret=00541424
000b:Call kernel32.GetVersion() ret=005413a7
000b:Call ntdll.RtlGetVersion(7fa9eef8) ret=7fd3d280
000b:Ret  ntdll.RtlGetVersion() retval=00000000
ret=7fd3d280
000b:Ret  kernel32.GetVersion() retval=c0000a04
ret=005413a7
000b:Call kernel32.GetCurrentThreadId() ret=005413c6
000b:Ret  kernel32.GetCurrentThreadId()
retval=0000000b ret=005413c6
000b:Call
user32.EnumThreadWindows(0000000b,00541388,7fa9f0b8)
ret=005413cc
000b:Call kernel32.94(7f8909a0) ret=7f84adbc
000b:Ret  kernel32.94() retval=0000000b ret=7f84adbc
000b:Call
ntdll.RtlAllocateHeap(7fdb0000,00000000,00000080)
ret=7f84b762
000b:Ret  ntdll.RtlAllocateHeap() retval=7fdef278
ret=7f84b762
000b:Call
ntdll.RtlFreeHeap(7fdb0000,00000000,7fdef278)
ret=7f8502e6
000b:Ret  ntdll.RtlFreeHeap() retval=00000001
ret=7f8502e6
000b:Ret  user32.EnumThreadWindows() retval=00000001
ret=005413cc
000b:Call user32.MessageBoxA(00000000,0057ff94
"Assertion failed: !\"bogus context in
_ExceptionHandler()\", file xx.cpp, line
3072",7fa9f0e4 "NeatImage.exe",00012010) ret=0054146f

Looks to me as if an exception is thrown, now the
program itself tries to handle the exception, and the
program exits with a window abnormal termination.

My question really is why the exception is handled in
two different ways, what could be the cause of this?

If this helps: one guy on the buglist got this error
after upgrading from Fedora Core 1 with vanilla kernel
2.6.10 (which worked
fine) to Fedora Core 3 with kernel 2.6.10-1.766_FC3



--- Francois Gouget <fgouget(a)codeweavers.com> wrote:
> Hi,
>
> Robbert Xerox wrote:
> > Hi , i know this is not a wine-bug channel, but
> > this
> > bug is affecting quite some apps, and seems to
> > affect
> > more and more ...
> > At least five apps from wine-bug mailinglist throw
> > up
> > an exception which either reads: "Assertion Failed
> > !bogus context in Local_Unwind()" or "in
> > Exception_Handler".
>
> I have seen a bug pretty much like this in an
> application and here's 
> what Mike Hearn had to say about it:
> > There were actually two bugs here:
> >
> > * Despite what the unit tests indicate,
> > IShellFolder::BindToObject
> >  with a NULL pidl can sometimes work. I guess it
> > depends on the exact
> >  class implementation. In this case I fixed it
> > with an application
> >  specific hack, we just return the desktop folder
> > (it should probably
> >  return an addreffed This).
> >
> > * The app apparently relies on an old Win98
> > specific bug where
> >  SFGAO_FILESYSTEM is set on My Computer. This bug
> > is documented here:

http://msdn.microsoft.com/msdnmag/issues/05/06/CAtWork/default.aspx
> > at the bottom. But it's not present in Windows
>
> NT, as our unit tests
> > actually prove (we should pay more attention to
> > 98/NT differences!!):

http://test.winehq.com/data/200506231000/98_PaulVriensW98SEFull/shell32:shlf...
> > It can be fixed with the patch to wine.inf, if
>
> you then do "make
> > install" in the tools directory and re-run
> > wineprefixcreate.
>
> I've attached the patch so you can try it. That's
> not code I can comment 
> on but maybe Mike will when he's back from his
> not-quite-vacations.
>
> -- 
> Francois Gouget
> fgouget(a)codeweavers.com
> > Index: dlls/shell32/shlfolder.c

===================================================================
> RCS file: /var/cvs/wine/dlls/shell32/shlfolder.c,v
> retrieving revision 1.104
> diff -u -p -r1.104 shlfolder.c
> --- dlls/shell32/shlfolder.c	20 Jul 2005 10:29:05
> -0000	1.104
> +++ dlls/shell32/shlfolder.c	28 Jul 2005 21:33:11
> -0000
> @@ -272,6 +272,15 @@ HRESULT SHELL32_BindToChild
> (LPCITEMIDLI
>      HRESULT hr;
>      LPITEMIDLIST pidlChild;
>
> +    TRACE("pidlRoot=%p, pathRoot=%s,
> pidlComplete=%p, riid=%s, ppvOut=%p\n",
> +          pidlRoot, pathRoot, pidlComplete,
> debugstr_guid(riid), ppvOut);
> +
> +    if (!pidlComplete)
> +    {
> +        SHELL32_CoCreateInitSF(_ILCreateDesktop(),
> pathRoot, NULL, &CLSID_ShellDesktop, riid, ppvOut);
> +        return S_OK;
> +    }
> +    
>      if (!pidlRoot || !ppvOut || !pidlComplete ||
> !pidlComplete->mkid.cb)
>          return E_INVALIDARG;
>
> Index: tools/wine.inf

===================================================================
> RCS file: /var/cvs/wine/tools/wine.inf,v
> retrieving revision 1.37
> diff -u -p -r1.37 wine.inf
> --- tools/wine.inf	27 Jul 2005 15:42:40 -0000	1.37
> +++ tools/wine.inf	28 Jul 2005 21:30:35 -0000
> @@ -100,6 +100,8 @@
> HKCR,TypeLib\{00020430-0000-0000-C000-00

HKCR,TypeLib\{00020430-0000-0000-C000-000000000046}\2.0,,,"OLE
> Automation"

HKCR,TypeLib\{00020430-0000-0000-C000-000000000046}\2.0\0\win32,,,"stdole2.tlb"
>

HKCR,TypeLib\{00020430-0000-0000-C000-000000000046}\2.0\FLAGS,,,"0"
> +; FIXME: this is to emulate a Windows 98 bug which
> is gone in NT: very few apps need it

+HKCR,CLSID\{20d04fe0-3aea-1069-a2d8-08002b30309d}\ShellFolder,"Attributes",0x10001,f0000154
> [ControlClass]

HKLM,System\CurrentControlSet\Control\Class\{4d36e978-e325-11ce-bfc1-08002be10318},,,"Ports
> (COM & LPT)"

___________________________________________________________ 
To help you stay safe and secure online, we've developed the all new Yahoo! Security Centre. http://uk.security.yahoo.com