10 Common Mistakes in Memory Analysis (Part 7)

Another common mistake I observe is relying on what debuggers report without double-checking. Present day debuggers, like WinDbg or GDB, are symbol-driven, they do not possess much of that semantic knowledge that a human debugger has. Also, it is better to report more than less: what is irrelevant can be skipped over by a skilled memory analyst but what looks suspicious to the problem at hand shall be double-checked to see if it is not coincidental. One example we consider here is Coincidental Symbolic Information

An application is frequently crashing. The process memory dump file shows only one thread left inside without any exception handling frames. In order to hypothesize about the probable cause the thread raw stack data is analyzed. It shows a few C++ STL calls with a custom smart pointer class and memory allocator like this:

app!std::vector<SmartPtr<ClassA>, std::allocator<SmartPtr<ClassA> > >::operator[]+

WinDbg !analyze-v command output points to this code:

app!std::bad_alloc::~bad_alloc <PERF> (app+0x0)+0
00400000 4d  dec     ebp

Raw stack data contains a few symbolic references to bad_alloc destructor too:

0012f9c0  00000100
0012f9c4  00400100 app!std::bad_alloc::~bad_alloc <PERF> (app+0x100)
0012f9c8  00000000
0012f9cc  0012f9b4
0012f9d0  00484488 app!_NULL_IMPORT_DESCRIPTOR+0x1984
0012f9d4  0012fa8c
0012f9d8  7c828290 ntdll!_except_handler3
0012f9dc  0012fa3c
0012f9e0  7c82b04a ntdll!RtlImageNtHeaderEx+0xee
0012f9e4  00482f08 app!_NULL_IMPORT_DESCRIPTOR+0x404
0012f9e8  00151ed0
0012f9ec  00484c1e app!_NULL_IMPORT_DESCRIPTOR+0x211a
0012f9f0  00000100
0012f9f4  00400100 app!std::bad_alloc::~bad_alloc <PERF> (app+0x100)

By linking all these three pieces together an engineer hypothesized that the cause of failure is memory allocation. However, careful analysis reveals all of them as a coincidental symbolic information and renders hypothesis much less plausible:

1.  The address of app!std::bad_alloc::~bad_alloc is 00400000 which coincides with the start of the main application module:

0:000> lm a 00400000
start    end        module name
00400000 004c4000   app    (no symbols)

As a consequence, its assembly language code makes no sense:

0:000> u 00400000
00400000 4d              dec     ebp
00400001 5a              pop     edx
00400002 90              nop
00400003 0003            add     byte ptr [ebx],al
00400005 0000            add     byte ptr [eax],al
00400007 000400          add     byte ptr [eax+eax],al
0040000a 0000            add     byte ptr [eax],al
0040000c ff              ???

2. All std::vector references are in fact fragments of a UNICODE string that can be dumped using du command:

0012ef14  00430056 app!std::vector<SmartPtr<ClassA>, std::allocator<SmartPtr<ClassA> > >::operator[]+0x16
0012ef18  00300038
0012ef1c  0043002e app!std::vector<SmartPtr<ClassA>, std::allocator<SmartPtr<ClassA> > >::size+0x1

0:000> du 0012ef14 l6
0012ef14  "VC80.C"

3. Raw stack data references to bad_alloc destructor are still module addresses in disguise, 00400100 or app+0×100, with nonsense assembly code:

0:000> u 00400100
00400100 50              push    eax
00400101 45              inc     ebp
00400102 0000            add     byte ptr [eax],al
00400104 4c              dec     esp
00400105 010500571aac    add     dword ptr ds:[0AC1A5700h],eax
0040010b 4a              dec     edx
0040010c 0000            add     byte ptr [eax],al
0040010e 0000            add     byte ptr [eax],al

- Dmitry Vostokov @ DumpAnalysis.org + TraceAnalysis.org -

Leave a Reply

You must be logged in to post a comment.