[Pnet-developers] [bugs #11212] Marshalling bug - code pointer in struct marshalled to unmanaged becomes invalid

[Ilyak]

This mail is an automated notification from the bugs tracker
 of the project: DotGNU Portable.NET.

/**************************************************************************/
[bugs #11212] Latest Modifications:

Changes by: 
                Ilyak <address@hidden>
'Date: 
                Sun 12/12/2004 at 09:47 (GMT)

------------------ Additional Follow-up Comments ----------------------------
(gdb) run call.exe
Starting program: /usr/local/bin/ilrun call.exe
[Thread debugging using libthread_db enabled]
[New Thread 1073879904 (LWP 1429)]
First time
Proceed to emitter
Proceed to callback
Second time
Proceed to emitter

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 1073879904 (LWP 1429)]
0x0000ffe9 in ?? ()
(gdb) bt
#0  0x0000ffe9 in ?? ()
#1  0x405fb675 in EmitCallback () from /home/ilyak/bug/libcall.so
#2  0x080d78d3 in ffi_call_SYSV () at src/x86/sysv.S:59
#3  0x080d7899 in ffi_raw_call (cif=0x0, fn=0x405fb658 <EmitCallback>,
    rvalue=0x13, fake_avalue=0xffe9) at src/x86/ffi.c:515
#4  0x08094b8e in _ILCVMInterpreter (thread=0x8157e70) at cvm_call.c:889
#5  0x0804c3d8 in _ILCallMethod (thread=0x8157e70, method=0x40296578,
    unpack=0x804c25c <_ILCallUnpackVResult>, result=0xbffff710, isCtor=0,
    _this=0x0, pack=0x804bfa4 <_ILCallPackVParams>, userData=0xbffff720)
    at call.c:706
#6  0x0804c6ad in ILExecThreadCallV (thread=0x8157e70, method=0x40296578,
    result=0xbffff710, args=0xbffff720) at call.c:928
#7  0x0804bb6c in main (argc=2, argv=0xbffff7d4) at ilrun.c:494


Not quite a NULL, while sure this is not a real pointer :)






/**************************************************************************/
[bugs #11212] Full Item Snapshot:

URL: <http://savannah.gnu.org/bugs/?func=detailitem&item_id=11212>
Project: DotGNU Portable.NET
Submitted by: Ilyak
On: Sat 12/04/2004 at 08:04

Category:  None
Severity:  5 - Average
Item Group:  None
Resolution:  None
Privacy:  Public
Assigned to:  None
Status:  Open


Summary:  Marshalling bug - code pointer in struct marshalled to unmanaged 
becomes invalid

Original Submission:  There is an example that shows problem: If i have a 
struct with Delegate in it, and i pass ref ThisStruct to unmanaged method 
twice, on second pass Delegate becomes invalid and ilrun crashes.

Mono crashes even on first pass there :))

Follow-up Comments
------------------


-------------------------------------------------------
Date: Sun 12/12/2004 at 09:47       By: Ilyak <Ilyak>
(gdb) run call.exe
Starting program: /usr/local/bin/ilrun call.exe
[Thread debugging using libthread_db enabled]
[New Thread 1073879904 (LWP 1429)]
First time
Proceed to emitter
Proceed to callback
Second time
Proceed to emitter

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 1073879904 (LWP 1429)]
0x0000ffe9 in ?? ()
(gdb) bt
#0  0x0000ffe9 in ?? ()
#1  0x405fb675 in EmitCallback () from /home/ilyak/bug/libcall.so
#2  0x080d78d3 in ffi_call_SYSV () at src/x86/sysv.S:59
#3  0x080d7899 in ffi_raw_call (cif=0x0, fn=0x405fb658 <EmitCallback>,
    rvalue=0x13, fake_avalue=0xffe9) at src/x86/ffi.c:515
#4  0x08094b8e in _ILCVMInterpreter (thread=0x8157e70) at cvm_call.c:889
#5  0x0804c3d8 in _ILCallMethod (thread=0x8157e70, method=0x40296578,
    unpack=0x804c25c <_ILCallUnpackVResult>, result=0xbffff710, isCtor=0,
    _this=0x0, pack=0x804bfa4 <_ILCallPackVParams>, userData=0xbffff720)
    at call.c:706
#6  0x0804c6ad in ILExecThreadCallV (thread=0x8157e70, method=0x40296578,
    result=0xbffff710, args=0xbffff720) at call.c:928
#7  0x0804bb6c in main (argc=2, argv=0xbffff7d4) at ilrun.c:494


Not quite a NULL, while sure this is not a real pointer :)

-------------------------------------------------------
Date: Sun 12/12/2004 at 09:16       By: Ilyak <Ilyak>
I remember that when i traced this bug with gdb, i end up rith segfault when 
trying to execute code at something like 0x00000006.

And, anyway, if i'm getting a NULL for the delegate, why don't i get 
NullPointerException? (I get hang instead)
And I beleive this behavior is invalid anyway, both calls should succeed, 
should not them?

-------------------------------------------------------
Date: Sat 12/11/2004 at 14:45       By: Gopal.V <t3rmin4t0r>
I added WriteLines for cbi.callback in the code and this is what it threw up.

First time
call+CallbackDelegate
Proceed to emitter
Proceed to callback
Second time
Uncaught exception: System.NullReferenceException: The value 'null' was found 
where an instance of an object was required

So you're getting a C# null for the delegate there.






File Attachments
-------------------

-------------------------------------------------------
Date: Sat 12/04/2004 at 08:04  Name: bug.tar.gz  Size: 684B   By: Ilyak
Tiny bug example. Run build.sh to make.
http://savannah.gnu.org/bugs/download.php?item_id=11212&amp;item_file_id=1942






For detailed info, follow this link:
<http://savannah.gnu.org/bugs/?func=detailitem&item_id=11212>

_______________________________________________
  Message sent via/by Savannah
  http://savannah.gnu.org/