Software Diagnostics Library » Crash Dump Patterns

Archive for the ‘Crash Dump Patterns’ Category

Crash Dump Analysis Patterns (Part 77, Mac OS X)

Friday, July 20th, 2012

This is a Mac OS X / GDB counterpart to C++ Exception pattern previously described for Windows platforms:

(gdb) bt #0 0x00007fff88bd582a in __kill () #1 0x00007fff8c184a9c in abort () #2 0x00007fff852f57bc in abort_message () #3 0x00007fff852f2fcf in default_terminate () #4 0x00007fff852f3001 in safe_handler_caller () #5 0x00007fff852f305c in std::terminate () #6 0×00007fff852f4152 in __cxa_throw () #7 0×000000010e402be8 in bar () #8 0×000000010e402c99 in foo () #9 0×000000010e402cbb in main (argc=1, argv=0×7fff6e001b18)

The modeling application source code:

class Exception

{

int code;

std::string description;

public:

Exception(int _code, std::string _desc) : code(_code), description(_desc) {}

};

void bar()

{

throw new Exception(5, “Access Denied”);

}

void foo()

{

bar();

}

int main(int argc, const char * argv[])

{

foo();

return 0;

}

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

Forthcoming Training: Accelerated Mac OS X Core Dump Analysis

Posted in Core Dump Analysis, Crash Dump Patterns, GDB for WinDbg Users, Mac Crash Corner, Mac OS X, x64 Mac OS X | No Comments »

Crash Dump Analysis Patterns (Part 36, Mac OS X)

Thursday, July 19th, 2012

This is a Mac OS X / GDB counterpart to Local Buffer Overflow pattern previously described for Windows platforms. Most of the time simple mistakes in using memory and string manipulation functions are easily detected by runtime:

(gdb) bt #0 0x00007fff885e982a in __kill () #1 0x00007fff83288b6c in __abort () #2 0×00007fff8325a89f in __chk_fail () #3 0×00007fff8325a83e in __memcpy_chk () #4 0×000000010914edf3 in bar () #5 0×000000010914ee5e in foo () #6 0×000000010914ee9b in main (argc=1, argv=0×7fff68d4daf0)

This detection happens in a default optimized release version as well:

(gdb) bt #0 0x00007fff885e982a in __kill () #1 0x00007fff83288b6c in __abort () #2 0×00007fff8325a89f in __chk_fail () #3 0×00007fff8325a83e in __memcpy_chk () #4 0×000000010f59cea8 in bar [inlined] () #5 0×000000010f59cea8 in foo [inlined] () #6 0×000000010f59cea8 in main (argc=, argv=)

The more sophisticated example which overwrites stack trace without being detected involves overwriting indirectly via a pointer to a local buffer passed to the called function. In such cases we might see incorrect and truncated stack traces:

(gdb) bt #0 0x00007fff885e982a in __kill () #1 0x00007fff83288b6c in __abort () #2 0×00007fff83285070 in __stack_chk_fail () #3 0×000000010524de77 in foo () #4 0xca4000007fff64e5 in ?? ()

(gdb) bt #0 0x00007fff885e982a in __kill () #1 0x00007fff83288b6c in __abort () #2 0×00007fff83285070 in __stack_chk_fail () #3 0×0000000105ad8df7 in foo ()

Inspection of the raw stack shows ASCII-like memory values around foo symbolic reference instead of expected main and start functions:

(gdb) info r rsp rsp 0x7fff656d79d8 0x7fff656d79d8

(gdb) x/100a 0x7fff656d79d8 0x7fff656d79d8: 0x7fff83288b6c <__abort+193> 0x0 0x7fff656d79e8: 0x0 0xffffffdf 0x7fff656d79f8: 0x7fff656d7a40 0x7fff656d7a80 0x7fff656d7a08: 0x7fff83285070 <__guard_setup> 0x6675426c61636f4c 0x7fff656d7a18: 0x7265764f726566 0x0 0x7fff656d7a28: 0x0 0x0 0x7fff656d7a38: 0x0 0x73205d343336325b 0x7fff656d7a48: 0x65766f206b636174 0x776f6c6672 0x7fff656d7a58: 0x0 0x0 0x7fff656d7a68: 0x0 0x343336326d7ab0 0x7fff656d7a78: 0x0 0x7fff656d7ab0 0x7fff656d7a88: 0x105ad8df7 0xb1887b8452358ac4 0×7fff656d7a98: 0×794d000000000000 0×6769422077654e20 0×7fff656d7aa8: 0×6666754220726567 0×7265 0×7fff656d7ab8: 0×0 0×0 0×7fff656d7ac8: 0×0 0×0 0×7fff656d7ad8: 0×0 0×0 0×7fff656d7ae8: 0×0 0×0 […]

The modeling application source code:

void bar(char *buffer)

{

char data[100] = “My New Bigger Buffer”;

memcpy (buffer, data, sizeof(data));

}

void foo()

{

char data[10] = “My Buffer”;

bar(data);

}

int main(int argc, const char * argv[])

{

foo();

return 0;

}

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

Forthcoming Training: Accelerated Mac OS X Core Dump Analysis

Posted in Core Dump Analysis, Crash Dump Patterns, GDB for WinDbg Users, Mac Crash Corner, Mac OS X, x64 Mac OS X | No Comments »

Crash Dump Analysis Patterns (Part 78a, Mac OS X)

Wednesday, July 18th, 2012

This is a Mac OS X / GDB counterpart to Divide by Zero (user mode) pattern previously described for Windows platforms:

(gdb) bt #0 0×000000010d3ebe9e in bar (a=1, b=0) #1 0×000000010d3ebec3 in foo () #2 0×000000010d3ebeeb in main (argc=1, argv=0×7fff6cfeab18)

(gdb) x/i 0×000000010d3ebe9e 0×10d3ebe9e : idiv %esi

(gdb) info r rsi rsi 0×0 0

The modeling application source code:

int bar(int a, int b)

{

return a/b;

}

int foo()

{

return bar(1,0);

}

int main(int argc, const char * argv[])

{

return foo();

}

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

Forthcoming Training: Accelerated Mac OS X Core Dump Analysis

Posted in Core Dump Analysis, Crash Dump Patterns, GDB for WinDbg Users, Mac Crash Corner, Mac OS X, x64 Mac OS X | No Comments »

Crash Dump Analysis Patterns (Part 16b, Mac OS X)

Tuesday, July 17th, 2012

This is a Mac OS X / GDB counterpart to Stack Overflow (user mode) pattern previously described for Windows platforms:

(gdb) bt 10 #0 0x0000000105dafea8 in bar (i=0) #1 0x0000000105dafeb9 in bar (i=262102) #2 0x0000000105dafeb9 in bar (i=262101) #3 0x0000000105dafeb9 in bar (i=262100) #4 0x0000000105dafeb9 in bar (i=262099) #5 0x0000000105dafeb9 in bar (i=262098) #6 0x0000000105dafeb9 in bar (i=262097) #7 0x0000000105dafeb9 in bar (i=262096) #8 0x0000000105dafeb9 in bar (i=262095) #9 0x0000000105dafeb9 in bar (i=262094) (More stack frames follow...)

There are at least 262,102 frames so we don’t attempt to list them all. What we’d like to do is to get stack trace boundaries from the list of sections based on the current stack pointer address and dump the upper part of it (the stack grows from higher addresses to the lower ones) to get bottom initial stack traces:

(gdb) x $rsp 0×7fff651aeff0: 0×00000000

Because this is a stack overflow we expect RSP went out of page bounds so we expect the lowest address being 0×7fff651af000.

(gdb) maint info sections [...] Core file: `/cores/core.2763', file type mach-o-le. [...] 0x0000000105e00000->0x0000000105f00000 at 0x00035000: LC_SEGMENT. ALLOC LOAD CODE HAS_CONTENTS 0x00007fff619af000->0x00007fff651af000 at 0x00135000: LC_SEGMENT. ALLOC LOAD CODE HAS_CONTENTS 0×00007fff651af000->0×00007fff659af000 at 0×03935000: LC_SEGMENT. ALLOC LOAD CODE HAS_CONTENTS 0×00007fff659af000->0×00007fff659e4000 at 0×04135000: LC_SEGMENT. ALLOC LOAD CODE HAS_CONTENTS 0×00007fff659e4000->0×00007fff659e6000 at 0×0416a000: LC_SEGMENT. ALLOC LOAD CODE HAS_CONTENTS […]

(gdb) x/250a 0×00007fff659af000-2000 0×7fff659ae830: 0×0 0×1500000000 0×7fff659ae840: 0×7fff659ae860 0×105dafeb9 <bar+25> 0×7fff659ae850: 0×0 0×1400000000 0×7fff659ae860: 0×7fff659ae880 0×105dafeb9 <bar+25> 0×7fff659ae870: 0×0 0×1300000000 0×7fff659ae880: 0×7fff659ae8a0 0×105dafeb9 <bar+25> 0×7fff659ae890: 0×0 0×1200000000 0×7fff659ae8a0: 0×7fff659ae8c0 0×105dafeb9 <bar+25> 0×7fff659ae8b0: 0×0 0×1100000000 0×7fff659ae8c0: 0×7fff659ae8e0 0×105dafeb9 <bar+25> 0×7fff659ae8d0: 0×0 0×1000000000 0×7fff659ae8e0: 0×7fff659ae900 0×105dafeb9 <bar+25> 0×7fff659ae8f0: 0×0 0xf00000000 0×7fff659ae900: 0×7fff659ae920 0×105dafeb9 <bar+25> 0×7fff659ae910: 0×0 0xe00000000 0×7fff659ae920: 0×7fff659ae940 0×105dafeb9 <bar+25> 0×7fff659ae930: 0×0 0xd00000000 0×7fff659ae940: 0×7fff659ae960 0×105dafeb9 <bar+25> 0×7fff659ae950: 0×0 0xc00000000 0×7fff659ae960: 0×7fff659ae980 0×105dafeb9 <bar+25> 0×7fff659ae970: 0×0 0xb00000000 0×7fff659ae980: 0×7fff659ae9a0 0×105dafeb9 <bar+25> 0×7fff659ae990: 0×0 0xa00000000 0×7fff659ae9a0: 0×7fff659ae9c0 0×105dafeb9 <bar+25> 0×7fff659ae9b0: 0×0 0×900000000 0×7fff659ae9c0: 0×7fff659ae9e0 0×105dafeb9 <bar+25> 0×7fff659ae9d0: 0×0 0×800000000 0×7fff659ae9e0: 0×7fff659aea00 0×105dafeb9 <bar+25> 0×7fff659ae9f0: 0×0 0×700000000 0×7fff659aea00: 0×7fff659aea20 0×105dafeb9 <bar+25> 0×7fff659aea10: 0×0 0×600000000 0×7fff659aea20: 0×7fff659aea40 0×105dafeb9 <bar+25> 0×7fff659aea30: 0×0 0×5659b9fe0 0×7fff659aea40: 0×7fff659aea60 0×105dafeb9 <bar+25 0×7fff659aea50: 0×7fff659aea70 0×4659bd31f 0×7fff659aea60: 0×7fff659aea80 0×105dafeb9 <bar+25> 0×7fff659aea70: 0×7fff659aeaf0 0×3659b031a 0×7fff659aea80: 0×7fff659aeaa0 0×105dafeb9 <bar+25> 0×7fff659aea90: 0×7fff659af5c0 0×200000000 0×7fff659aeaa0: 0×7fff659aeac0 0×105dafeb9 <bar+25> 0×7fff659aeab0: 0×100000000 0×1659aeb18 0×7fff659aeac0: 0×7fff659aead0 0×105dafece <foo+14> 0×7fff659aead0: 0×7fff659aeaf0 0×105dafeeb <main+27> 0×7fff659aeae0: 0×7fff659aeb18 0×1 —Type to continue, or q to quit— 0×7fff659aeaf0: 0×7fff659aeb08 0×105dafe94 <start+52> 0×7fff659aeb00: 0×0 0×0 […] 0×7fff659aeff0: 0×3139336561303363 0×316235

Interesting if we set the lowest frame down and try to get register info GDB core dumps:

(gdb) frame 262102 #262102 0x0000000105dafeb9 in bar (i=1) 13 bar(i+1); (gdb) info r Segmentation fault: 11 (core dumped)

Looking its core dump show that it also experienced stack overflow:

(gdb) bt #0 0x00007fff8c1bacf0 in __sfvwrite () #1 0x00007fff8c189947 in __vfprintf () #2 0x00007fff8c184edb in vsnprintf_l () #3 0x00007fff8c1566be in __sprintf_chk () #4 0x000000010bd14d15 in print_displacement () #5 0x000000010bd10ddf in OP_E () #6 0x000000010bd13f9b in print_insn () #7 0x000000010bc164ce in length_of_this_instruction () #8 0x000000010bc9e296 in x86_analyze_prologue () #9 0x000000010bc9f1f3 in x86_frame_prev_register () #10 0x000000010bc91d70 in frame_register_unwind () #11 0x000000010bc92015 in frame_unwind_register () #12 0x000000010bc91d70 in frame_register_unwind () #13 0x000000010bc92015 in frame_unwind_register () #14 0x000000010bc91d70 in frame_register_unwind () #15 0x000000010bc92015 in frame_unwind_register () #16 0x000000010bc91d70 in frame_register_unwind () #17 0x000000010bc92015 in frame_unwind_register () #18 0x000000010bc91d70 in frame_register_unwind () #19 0x000000010bc92015 in frame_unwind_register () #20 0x000000010bc91d70 in frame_register_unwind () #21 0x000000010bc92015 in frame_unwind_register () #22 0x000000010bc91d70 in frame_register_unwind () #23 0x000000010bc92015 in frame_unwind_register () #24 0x000000010bc91d70 in frame_register_unwind () #25 0x000000010bc92015 in frame_unwind_register () #26 0x000000010bc91d70 in frame_register_unwind () #27 0x000000010bc92015 in frame_unwind_register () #28 0x000000010bc91d70 in frame_register_unwind () #29 0x000000010bc92015 in frame_unwind_register () #30 0x000000010bc91d70 in frame_register_unwind () #31 0x000000010bc92015 in frame_unwind_register () #32 0x000000010bc91d70 in frame_register_unwind () #33 0x000000010bc92015 in frame_unwind_register () #34 0x000000010bc91d70 in frame_register_unwind () #35 0x000000010bc92015 in frame_unwind_register () #36 0x000000010bc91d70 in frame_register_unwind () #37 0x000000010bc92015 in frame_unwind_register () #38 0x000000010bc91d70 in frame_register_unwind () #39 0x000000010bc92015 in frame_unwind_register () #40 0x000000010bc91d70 in frame_register_unwind () #41 0x000000010bc92015 in frame_unwind_register () #42 0x000000010bc91d70 in frame_register_unwind () #43 0x000000010bc92015 in frame_unwind_register ()

The source code of our modeling application:

void bar(int i)

{

bar(i+1);

}

void foo()

{

bar(1);

}

int main(int argc, const char * argv[])

{

foo();

return 0;

}

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

Forthcoming Training: Accelerated Mac OS X Core Dump Analysis

Posted in Core Dump Analysis, Crash Dump Patterns, GDB for WinDbg Users, Mac Crash Corner, Mac OS X, x64 Mac OS X | No Comments »

Module Patterns

Sunday, July 15th, 2012

A page to reference all different kinds of module and component related patterns is necessary, so I created this post:

I’ll update it as soon as I add more similar patterns.

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

Posted in Core Dump Analysis, Crash Dump Analysis, Crash Dump Patterns, Debugging, Malware Analysis | 2 Comments »

Crash Dump Analysis Patterns (Part 179)

Sunday, July 15th, 2012

When looking at the module list (lmv), searching for modules (.imgscan) or examining the particular module (!address, !dh) we may notice one of them as Deviant Module. The deviation may be in (but not limited to as anything is possible):

- suspicious module name

- suspicious protection

- suspicious module load address

0:005> .imgscan MZ at 00040000, prot 00000040, type 00020000 - size 1d000 MZ at 00340000, prot 00000002, type 01000000 - size 9c000 Name: iexplore.exe MZ at 02250000, prot 00000002, type 00040000 - size 2000 MZ at 023b0000, prot 00000002, type 01000000 - size b000 Name: msimtf.dll MZ at 03f80000, prot 00000002, type 00040000 - size 2000 MZ at 10000000, prot 00000004, type 00020000 - size 5000 Name: screens_dll.dll MZ at 16080000, prot 00000002, type 01000000 - size 25000 Name: mdnsNSP.dll MZ at 6ab50000, prot 00000002, type 01000000 - size 26000 Name: DSSENH.dll MZ at 6b030000, prot 00000002, type 01000000 - size 5b0000 Name: MSHTML.dll MZ at 6ba10000, prot 00000002, type 01000000 - size b4000 Name: JSCRIPT.dll MZ at 6cec0000, prot 00000002, type 01000000 - size 1b000 Name: CRYPTNET.dll MZ at 6d260000, prot 00000002, type 01000000 - size e000 Name: PNGFILTER.DLL MZ at 6d2f0000, prot 00000002, type 01000000 - size 29000 Name: msls31.dll MZ at 6d700000, prot 00000002, type 01000000 - size 30000 Name: MLANG.dll MZ at 6d740000, prot 00000002, type 01000000 - size 4d000 Name: SSV.DLL MZ at 6d7b0000, prot 00000002, type 01000000 - size c000 Name: ImgUtil.dll MZ at 6ddb0000, prot 00000002, type 01000000 - size 2f000 Name: iepeers.DLL MZ at 6df20000, prot 00000002, type 01000000 - size 33000 Name: IEShims.dll MZ at 6eb80000, prot 00000002, type 01000000 - size a94000 Name: IEFRAME.dll MZ at 703b0000, prot 00000002, type 01000000 - size 53000 Name: SWEEPRX.dll MZ at 70740000, prot 00000002, type 01000000 - size 40000 Name: SWEEPRX.dll MZ at 725a0000, prot 00000002, type 01000000 - size 12000 Name: PNRPNSP.dll MZ at 725d0000, prot 00000002, type 01000000 - size 8000 Name: WINRNR.dll MZ at 725e0000, prot 00000002, type 01000000 - size 136000 Name: MSXML3.dll MZ at 72720000, prot 00000002, type 01000000 - size c000 Name: wshbth.dll MZ at 72730000, prot 00000002, type 01000000 - size f000 Name: NAPINSP.dll MZ at 72890000, prot 00000002, type 01000000 - size 6000 Name: SensApi.dll MZ at 72ec0000, prot 00000002, type 01000000 - size 42000 Name: WINSPOOL.DRV MZ at 734b0000, prot 00000002, type 01000000 - size 6000 Name: rasadhlp.dll MZ at 736b0000, prot 00000002, type 01000000 - size 85000 Name: COMCTL32.dll MZ at 73ac0000, prot 00000002, type 01000000 - size 7000 Name: MIDIMAP.dll MZ at 73ae0000, prot 00000002, type 01000000 - size 14000 Name: MSACM32.dll MZ at 73b00000, prot 00000002, type 01000000 - size 66000 Name: audioeng.dll MZ at 73c30000, prot 00000002, type 01000000 - size 9000 Name: MSACM32.DRV MZ at 73c60000, prot 00000002, type 01000000 - size 21000 Name: AudioSes.DLL MZ at 73c90000, prot 00000002, type 01000000 - size 2f000 Name: WINMMDRV.dll MZ at 74290000, prot 00000002, type 01000000 - size bb000 Name: PROPSYS.dll MZ at 74390000, prot 00000002, type 01000000 - size f000 Name: nlaapi.dll MZ at 743a0000, prot 00000002, type 01000000 - size 4000 Name: ksuser.dll MZ at 74430000, prot 00000002, type 01000000 - size 15000 Name: Cabinet.dll MZ at 74450000, prot 00000002, type 01000000 - size 3d000 Name: OLEACC.dll MZ at 74490000, prot 00000002, type 01000000 - size 1ab000 Name: gdiplus.dll MZ at 74640000, prot 00000002, type 01000000 - size 28000 Name: MMDevAPI.DLL MZ at 74670000, prot 00000002, type 01000000 - size 32000 Name: WINMM.dll MZ at 746b0000, prot 00000002, type 01000000 - size 31000 Name: TAPI32.dll MZ at 749e0000, prot 00000002, type 01000000 - size 19e000 Name: COMCTL32.dll MZ at 74b80000, prot 00000002, type 01000000 - size 7000 Name: AVRT.dll MZ at 74ba0000, prot 00000002, type 01000000 - size 4a000 Name: RASAPI32.dll MZ at 74ce0000, prot 00000002, type 01000000 - size 3f000 Name: UxTheme.dll MZ at 74de0000, prot 00000002, type 01000000 - size 2d000 Name: WINTRUST.dll MZ at 74ea0000, prot 00000002, type 01000000 - size 14000 Name: rasman.dll MZ at 74f70000, prot 00000002, type 01000000 - size c000 Name: rtutils.dll MZ at 74f80000, prot 00000002, type 01000000 - size 5000 Name: WSHTCPIP.dll MZ at 74fb0000, prot 00000002, type 01000000 - size 21000 Name: NTMARTA.dll MZ at 75010000, prot 00000002, type 01000000 - size 3b000 Name: RSAENH.dll MZ at 75050000, prot 00000002, type 01000000 - size 5000 Name: MSIMG32.dll MZ at 75060000, prot 00000002, type 01000000 - size 15000 Name: GPAPI.dll MZ at 750a0000, prot 00000002, type 01000000 - size 46000 Name: SCHANNEL.dll MZ at 752b0000, prot 00000002, type 01000000 - size 3b000 Name: MSWSOCK.dll MZ at 75370000, prot 00000002, type 01000000 - size 45000 Name: bcrypt.dll MZ at 753f0000, prot 00000002, type 01000000 - size 5000 Name: WSHIP6.dll MZ at 75400000, prot 00000002, type 01000000 - size 8000 Name: VERSION.dll MZ at 75420000, prot 00000002, type 01000000 - size 7000 Name: CREDSSP.dll MZ at 75430000, prot 00000002, type 01000000 - size 35000 Name: ncrypt.dll MZ at 75480000, prot 00000002, type 01000000 - size 22000 Name: dhcpcsvc6.DLL MZ at 754b0000, prot 00000002, type 01000000 - size 7000 Name: WINNSI.DLL MZ at 754c0000, prot 00000002, type 01000000 - size 35000 Name: dhcpcsvc.DLL MZ at 75500000, prot 00000002, type 01000000 - size 19000 Name: IPHLPAPI.DLL MZ at 75590000, prot 00000002, type 01000000 - size 3a000 Name: slc.dll MZ at 755d0000, prot 00000002, type 01000000 - size f2000 Name: CRYPT32.dll MZ at 75740000, prot 00000002, type 01000000 - size 12000 Name: MSASN1.dll MZ at 75760000, prot 00000002, type 01000000 - size 11000 Name: SAMLIB.dll MZ at 75780000, prot 00000002, type 01000000 - size 76000 Name: NETAPI32.dll MZ at 75800000, prot 00000002, type 01000000 - size 2c000 Name: DNSAPI.dll MZ at 75a70000, prot 00000002, type 01000000 - size 5f000 Name: sxs.dll MZ at 75ad0000, prot 00000002, type 01000000 - size 2c000 Name: apphelp.dll MZ at 75b30000, prot 00000002, type 01000000 - size 14000 Name: Secur32.dll MZ at 75b50000, prot 00000002, type 01000000 - size 1e000 Name: USERENV.dll MZ at 75c90000, prot 00000002, type 01000000 - size 7000 Name: PSAPI.DLL MZ at 75ca0000, prot 00000002, type 01000000 - size c3000 Name: RPCRT4.dll MZ at 75d70000, prot 00000002, type 01000000 - size 73000 Name: COMDLG32.dll MZ at 75df0000, prot 00000002, type 01000000 - size 9000 Name: LPK.dll MZ at 75e00000, prot 00000002, type 01000000 - size dc000 Name: KERNEL32.dll MZ at 75ee0000, prot 00000002, type 01000000 - size aa000 Name: msvcrt.dll MZ at 75f90000, prot 00000002, type 01000000 - size 1e8000 Name: iertutil.dll MZ at 76180000, prot 00000002, type 01000000 - size 29000 Name: imagehlp.dll MZ at 761b0000, prot 00000002, type 01000000 - size 6000 Name: NSI.dll MZ at 761c0000, prot 00000002, type 01000000 - size 84000 Name: CLBCatQ.DLL MZ at 76250000, prot 00000002, type 01000000 - size 49000 Name: WLDAP32.dll MZ at 762a0000, prot 00000002, type 01000000 - size c6000 Name: ADVAPI32.dll MZ at 76370000, prot 00000002, type 01000000 - size 4b000 Name: GDI32.dll MZ at 763c0000, prot 00000002, type 01000000 - size 59000 Name: SHLWAPI.dll MZ at 76420000, prot 00000002, type 01000000 - size e6000 Name: WININET.dll MZ at 76510000, prot 00000002, type 01000000 - size b10000 Name: SHELL32.dll MZ at 77020000, prot 00000002, type 01000000 - size 145000 Name: ole32.dll MZ at 77170000, prot 00000002, type 01000000 - size 7d000 Name: USP10.dll MZ at 771f0000, prot 00000002, type 01000000 - size 8d000 Name: OLEAUT32.dll MZ at 77280000, prot 00000002, type 01000000 - size 18a000 Name: SETUPAPI.dll MZ at 77410000, prot 00000002, type 01000000 - size 9d000 Name: USER32.dll MZ at 774b0000, prot 00000002, type 01000000 - size 133000 Name: urlmon.dll MZ at 775f0000, prot 00000002, type 01000000 - size 127000 Name: ntdll.dll MZ at 77720000, prot 00000002, type 01000000 - size 3000 Name: Normaliz.dll MZ at 77730000, prot 00000002, type 01000000 - size 2d000 Name: WS2_32.dll MZ at 77760000, prot 00000002, type 01000000 - size 1e000 Name: IMM32.dll MZ at 77780000, prot 00000002, type 01000000 - size c8000 Name: MSCTF.dll MZ at 7c340000, prot 00000002, type 01000000 - size 56000 Name: MSVCR71.dll

0:005> !address 00040000 Usage: <unclassified> Allocation Base: 00040000 Base Address: 00040000 End Address: 0005d000 Region Size: 0001d000 Type: 00020000 MEM_PRIVATE State: 00001000 MEM_COMMIT Protect: 00000040 PAGE_EXECUTE_READWRITE

0:005> !address 10000000 Usage: <unclassified> Allocation Base: 10000000 Base Address: 10000000 End Address: 10001000 Region Size: 00001000 Type: 00020000 MEM_PRIVATE State: 00001000 MEM_COMMIT Protect: 00000004 PAGE_READWRITE

- suspicious text inside

See this case study for an example.

- suspicious import table (screen grabbing) or its absence (dynamic imports)

0:005> !dh 10000000 [...] 2330 [ 50] address [size] of Export Directory 20E0 [ 78] address [size] of Import Directory 0 [ 0] address [size] of Resource Directory 0 [ 0] address [size] of Exception Directory 0 [ 0] address [size] of Security Directory 4000 [ 34] address [size] of Base Relocation Directory 2060 [ 1C] address [size] of Debug Directory 0 [ 0] address [size] of Description Directory 0 [ 0] address [size] of Special Directory 0 [ 0] address [size] of Thread Storage Directory 0 [ 0] address [size] of Load Configuration Directory 0 [ 0] address [size] of Bound Import Directory 2000 [ 58] address [size] of Import Address Table Directory 0 [ 0] address [size] of Delay Import Directory 0 [ 0] address [size] of COR20 Header Directory 0 [ 0] address [size] of Reserved Directory […]

0:005> dps 10000000+2000 10000000+2000+58 10002000 76376101 gdi32!CreateCompatibleDC 10002004 763793d6 gdi32!StretchBlt 10002008 76377461 gdi32!CreateDIBSection 1000200c 763762a0 gdi32!SelectObject 10002010 00000000 10002014 75e4a411 kernel32!lstrcmpW 10002018 75e440aa kernel32!VirtualFree 1000201c 75e4ad55 kernel32!VirtualAlloc 10002020 00000000 10002024 77429ced user32!ReleaseDC 10002028 77423ba7 user32!NtUserGetWindowDC 1000202c 77430e21 user32!GetWindowRect 10002030 00000000 10002034 744a75e9 GdiPlus!GdiplusStartup 10002038 744976dd GdiPlus!GdipSaveImageToStream 1000203c 744cdd38 GdiPlus!GdipGetImageEncodersSize 10002040 744971cf GdiPlus!GdipDisposeImage 10002044 744a8591 GdiPlus!GdipCreateBitmapFromHBITMAP 10002048 744cdbae GdiPlus!GdipGetImageEncoders 1000204c 00000000 10002050 7707d51b ole32!CreateStreamOnHGlobal 10002054 00000000 10002058 00000000

0:000> !dh 012a0000 [...] 0 [ 0] address [size] of Export Directory 0 [ 0] address [size] of Import Directory 0 [ 0] address [size] of Resource Directory 0 [ 0] address [size] of Exception Directory 0 [ 0] address [size] of Security Directory 8000 [ FC] address [size] of Base Relocation Directory 4000 [ 1C] address [size] of Debug Directory 0 [ 0] address [size] of Description Directory 0 [ 0] address [size] of Special Directory 0 [ 0] address [size] of Thread Storage Directory 0 [ 0] address [size] of Load Configuration Directory 0 [ 0] address [size] of Bound Import Directory 0 [ 0] address [size] of Import Address Table Directory 0 [ 0] address [size] of Delay Import Directory 0 [ 0] address [size] of COR20 Header Directory 0 [ 0] address [size] of Reserved Directory […]

- suspicious path names

Age: 7, Pdb: d:\work\BekConnekt\Client_src_code_New\Release\Blackjoe_new.pdb

Debug Directories(1) Type Size Address Pointer cv 46 2094 894 Format: RSDS, guid, 1, C:\MyWork\screens_dll\Release\screens_dll.pdb

- suspicious image path (although could be just dynamic code generation for .NET assemblies)

- uninitialized image resources

0:002> lmv m C6DC start end module name 012a0000 012a9000 C6DC C (no symbols) Loaded symbol image file: C6DC.tmp Image path: C:\Users\User\AppData\Local\Temp\C6DC.tmp Image name: C6DC.tmp Timestamp: Sun May 30 20:18:32 2010 (4C02BA08) CheckSum: 00000000 ImageSize: 00009000 File version: 0.0.0.0 Product version: 0.0.0.0 File flags: 0 (Mask 0) File OS: 0 Unknown Base File type: 0.0 Unknown File date: 00000000.00000000 Translations: 0000.04b0 0000.04e4 0409.04b0 0409.04e4

- suspicious (small) image size

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

Posted in Core Dump Analysis, Crash Dump Analysis, Crash Dump Patterns, Debugging, Malware Analysis, Victimware, Victimware Analysis | 2 Comments »

Crash Dump Analysis Patterns (Part 178)

Wednesday, June 27th, 2012

One of the frequent problems is an access violation at an address that belongs to Unloaded Module. Here’s an example that recently happened on our machine during an auto-update of the popular software package so we immediately attached a debugger after seeing WER dialog box:

0:000> ~*k

. 0 Id: bc8.bcc Suspend: 1 Teb: 7efdd000 Unfrozen ChildEBP RetAddr 0035f1c4 771a0bdd ntdll!ZwWaitForMultipleObjects+0x15 0035f260 75771a2c KERNELBASE!WaitForMultipleObjectsEx+0x100 0035f2a8 75774208 kernel32!WaitForMultipleObjectsExImplementation+0xe0 0035f2c4 757980a4 kernel32!WaitForMultipleObjects+0x18 0035f330 75797f63 kernel32!WerpReportFaultInternal+0x186 0035f344 75797858 kernel32!WerpReportFault+0x70 0035f354 757977d7 kernel32!BasepReportFault+0x20 0035f3e0 77ec74df kernel32!UnhandledExceptionFilter+0x1af 0035f3e8 77ec73bc ntdll!__RtlUserThreadStart+0x62 0035f3fc 77ec7261 ntdll!_EH4_CallFilterFunc+0x12 0035f424 77eab459 ntdll!_except_handler4+0x8e 0035f448 77eab42b ntdll!ExecuteHandler2+0x26 0035f46c 77eab3ce ntdll!ExecuteHandler+0x24 0035f4f8 77e60133 ntdll!RtlDispatchException+0x127 0035f4f8 73eb2200 ntdll!KiUserExceptionDispatcher+0xf WARNING: Frame IP not in any known module. Following frames may be wrong. 0035f844 76e462fa <Unloaded_fpb.tmp>+0×12200 0035f870 76e46d3a USER32!InternalCallWinProc+0×23 0035f8e8 76e4965e USER32!UserCallWinProcCheckWow+0×109 0035f92c 76e496c5 USER32!SendMessageWorker+0×581 0035f950 7269c05c USER32!SendMessageW+0×7f 0035f9ec 7270be62 comctl32!CCSendNotify+0xc19 0035fa28 75f6f52a comctl32!SendNotify+0×36 0035fa4c 75f61d66 SHELL32!SetAppStartingCursor+0×6d 0035fa64 75f61ee2 SHELL32!CShellExecute::ExecuteNormal+0×16 0035fa78 75f61e70 SHELL32!ShellExecuteNormal+0×33 0035fa90 75f53cd0 SHELL32!ShellExecuteExW+0×62 0035fae4 003e2211 SHELL32!ShellExecuteW+0×77 0035fbc4 77e838be InstallFlashPlayer+0×2211 0035fcb4 77e83492 ntdll!RtlpFreeHeap+0xbb1 0035fcd4 757714dd ntdll!RtlFreeHeap+0×142 0035fce8 003f0324 kernel32!HeapFree+0×14 0035fd80 003f0241 InstallFlashPlayer+0×10324 0035fe10 7577339a InstallFlashPlayer+0×10241 0035fe1c 77e89ef2 kernel32!BaseThreadInitThunk+0xe 0035fe5c 77e89ec5 ntdll!__RtlUserThreadStart+0×70 0035fe74 00000000 ntdll!_RtlUserThreadStart+0×1b

1 Id: bc8.6b0 Suspend: 2 Teb: 7efda000 Unfrozen ChildEBP RetAddr 03e1f9e0 77ea2f51 ntdll!ZwWaitForMultipleObjects+0x15 03e1fb74 7577339a ntdll!TppWaiterpThread+0x33d 03e1fb80 77e89ef2 kernel32!BaseThreadInitThunk+0xe 03e1fbc0 77e89ec5 ntdll!__RtlUserThreadStart+0x70 03e1fbd8 00000000 ntdll!_RtlUserThreadStart+0x1b

2 Id: bc8.8dc Suspend: 2 Teb: 7efd7000 Unfrozen ChildEBP RetAddr 03f5fd50 77ea3352 ntdll!NtWaitForWorkViaWorkerFactory+0x12 03f5feb0 7577339a ntdll!TppWorkerThread+0x216 03f5febc 77e89ef2 kernel32!BaseThreadInitThunk+0xe 03f5fefc 77e89ec5 ntdll!__RtlUserThreadStart+0x70 03f5ff14 00000000 ntdll!_RtlUserThreadStart+0x1b

3 Id: bc8.944 Suspend: 2 Teb: 7efaf000 Unfrozen ChildEBP RetAddr 0416f8b4 77ea3352 ntdll!NtWaitForWorkViaWorkerFactory+0x12 0416fa14 7577339a ntdll!TppWorkerThread+0x216 0416fa20 77e89ef2 kernel32!BaseThreadInitThunk+0xe 0416fa60 77e89ec5 ntdll!__RtlUserThreadStart+0x70 0416fa78 00000000 ntdll!_RtlUserThreadStart+0x1b

Exception thread shows fpb.tmp module as unloaded:

0:000> lmv m fpb.tmp start end module name

Unloaded modules: 00cb0000 00d5a000 fpb.tmp Timestamp: Fri Jun 01 02:56:00 2012 (4FC82130) Checksum: 000B0CD5 ImageSize: 000AA000 73ea0000 73f15000 fpb.tmp Timestamp: Fri Jun 01 02:49:25 2012 (4FC81FA5) Checksum: 0007A7CE ImageSize: 00075000

We change the exception thread context to get registers at the time of the exception:

0:000> kv ChildEBP RetAddr Args to Child 0035f1c4 771a0bdd 00000002 0035f214 00000001 ntdll!ZwWaitForMultipleObjects+0x15 0035f260 75771a2c 0035f214 0035f288 00000000 KERNELBASE!WaitForMultipleObjectsEx+0x100 0035f2a8 75774208 00000002 7efde000 00000000 kernel32!WaitForMultipleObjectsExImplementation+0xe0 0035f2c4 757980a4 00000002 0035f2f8 00000000 kernel32!WaitForMultipleObjects+0x18 0035f330 75797f63 0035f410 00000001 00000001 kernel32!WerpReportFaultInternal+0x186 0035f344 75797858 0035f410 00000001 0035f3e0 kernel32!WerpReportFault+0x70 0035f354 757977d7 0035f410 00000001 658587c7 kernel32!BasepReportFault+0x20 0035f3e0 77ec74df 00000000 77ec73bc 00000000 kernel32!UnhandledExceptionFilter+0x1af 0035f3e8 77ec73bc 00000000 0035fe5c 77e7c530 ntdll!__RtlUserThreadStart+0x62 0035f3fc 77ec7261 00000000 00000000 00000000 ntdll!_EH4_CallFilterFunc+0x12 0035f424 77eab459 fffffffe 0035fe4c 0035f560 ntdll!_except_handler4+0x8e 0035f448 77eab42b 0035f510 0035fe4c 0035f560 ntdll!ExecuteHandler2+0x26 0035f46c 77eab3ce 0035f510 0035fe4c 0035f560 ntdll!ExecuteHandler+0x24 0035f4f8 77e60133 0135f510 0035f560 0035f510 ntdll!RtlDispatchException+0x127 0035f4f8 73eb2200 0135f510 0035f560 0035f510 ntdll!KiUserExceptionDispatcher+0xf (CONTEXT @ 0035f560) WARNING: Frame IP not in any known module. Following frames may be wrong. 0035f844 76e462fa 000201ce 0000004e 00000000 <Unloaded_fpb.tmp>+0×12200 0035f870 76e46d3a 73eb2200 000201ce 0000004e USER32!InternalCallWinProc+0×23 0035f8e8 76e4965e 00000000 73eb2200 000201ce USER32!UserCallWinProcCheckWow+0×109 0035f92c 76e496c5 013907f0 00000000 73eb2200 USER32!SendMessageWorker+0×581 0035f950 7269c05c 000201ce 0000004e 00000000 USER32!SendMessageW+0×7f 0035f9ec 7270be62 0035fa00 fffffff7 00000000 comctl32!CCSendNotify+0xc19 0035fa28 75f6f52a 000201ce 00000000 fffffff7 comctl32!SendNotify+0×36 0035fa4c 75f61d66 000201ce 00000001 00001500 SHELL32!SetAppStartingCursor+0×6d 0035fa64 75f61ee2 0035faa4 00001500 0035faa4 SHELL32!CShellExecute::ExecuteNormal+0×16 0035fa78 75f61e70 0035faa4 00001500 00000200 SHELL32!ShellExecuteNormal+0×33 0035fa90 75f53cd0 0035faa4 003fb654 003fa554 SHELL32!ShellExecuteExW+0×62 0035fae4 003e2211 000201ce 003fa554 0035fb14 SHELL32!ShellExecuteW+0×77 0035fbc4 77e838be 00da0138 77e8389a 77c467ad InstallFlashPlayer+0×2211 0035fcb4 77e83492 00000000 00da2320 00da2320 ntdll!RtlpFreeHeap+0xbb1 0035fcd4 757714dd 00da0000 00000000 00da2320 ntdll!RtlFreeHeap+0×142 0035fce8 003f0324 00da0000 00000000 003f0343 kernel32!HeapFree+0×14 0035fd80 003f0241 003e0000 00000000 010d3135 InstallFlashPlayer+0×10324 0035fe10 7577339a 7efde000 0035fe5c 77e89ef2 InstallFlashPlayer+0×10241 0035fe1c 77e89ef2 7efde000 77c46545 00000000 kernel32!BaseThreadInitThunk+0xe 0035fe5c 77e89ec5 003f02ac 7efde000 ffffffff ntdll!__RtlUserThreadStart+0×70 0035fe74 00000000 003f02ac 7efde000 00000000 ntdll!_RtlUserThreadStart+0×1b

0:000> .cxr 0035f560 eax=73eb2200 ebx=00000000 ecx=01080d68 edx=00000000 esi=73eb2200 edi=00000000 eip=73eb2200 esp=0035f848 ebp=0035f870 iopl=0 nv up ei pl nz na pe nc cs=0023 ss=002b ds=002b es=002b fs=0053 gs=002b efl=00210206 <Unloaded_fpb.tmp>+0×12200: 73eb2200 ?? ???

Then we double check that a window procedure was indeed called from that module range:

0:000> kv *** Stack trace for last set context - .thread/.cxr resets it ChildEBP RetAddr Args to Child WARNING: Frame IP not in any known module. Following frames may be wrong. 0035f844 76e462fa 000201ce 0000004e 00000000 <Unloaded_fpb.tmp>+0×12200 0035f870 76e46d3a 73eb2200 000201ce 0000004e USER32!InternalCallWinProc+0×23 0035f8e8 76e4965e 00000000 73eb2200 000201ce USER32!UserCallWinProcCheckWow+0×109 0035f92c 76e496c5 013907f0 00000000 73eb2200 USER32!SendMessageWorker+0×581 0035f950 7269c05c 000201ce 0000004e 00000000 USER32!SendMessageW+0×7f 0035f9ec 7270be62 0035fa00 fffffff7 00000000 comctl32!CCSendNotify+0xc19 0035fa28 75f6f52a 000201ce 00000000 fffffff7 comctl32!SendNotify+0×36 0035fa4c 75f61d66 000201ce 00000001 00001500 SHELL32!SetAppStartingCursor+0×6d 0035fa64 75f61ee2 0035faa4 00001500 0035faa4 SHELL32!CShellExecute::ExecuteNormal+0×16 0035fa78 75f61e70 0035faa4 00001500 00000200 SHELL32!ShellExecuteNormal+0×33 0035fa90 75f53cd0 0035faa4 003fb654 003fa554 SHELL32!ShellExecuteExW+0×62 0035fae4 003e2211 000201ce 003fa554 0035fb14 SHELL32!ShellExecuteW+0×77 0035fbc4 77e838be 00da0138 77e8389a 77c467ad InstallFlashPlayer+0×2211 0035fcb4 77e83492 00000000 00da2320 00da2320 ntdll!RtlpFreeHeap+0xbb1 00da15a0 00000000 00da1780 02971450 003e1000 ntdll!RtlFreeHeap+0×142

0:000> ub 76e462fa USER32!InternalCallWinProc+0×6: 76e462dd 68cdabbadc push 0DCBAABCDh 76e462e2 56 push esi 76e462e3 ff7518 push dword ptr [ebp+18h] 76e462e6 ff7514 push dword ptr [ebp+14h] 76e462e9 ff7510 push dword ptr [ebp+10h] 76e462ec ff750c push dword ptr [ebp+0Ch] 76e462ef 64800dca0f000001 or byte ptr fs:[0FCAh],1 76e462f7 ff5508 call dword ptr [ebp+8]

We now get a memory value pointed to by EBP+8 address:

0:000> r Last set context: eax=73eb2200 ebx=00000000 ecx=01080d68 edx=00000000 esi=73eb2200 edi=00000000 eip=73eb2200 esp=0035f848 ebp=0035f870 iopl=0 nv up ei pl nz na pe nc cs=0023 ss=002b ds=002b es=002b fs=0053 gs=002b efl=00210206 <Unloaded_fpb.tmp>+0×12200: 73eb2200 ?? ???

0:000> dp 0035f870+8 l1 0035f878 73eb2200

0:000> dd 73eb2200 73eb2200 ???????? ???????? ???????? ???????? 73eb2210 ???????? ???????? ???????? ???????? 73eb2220 ???????? ???????? ???????? ???????? 73eb2230 ???????? ???????? ???????? ???????? 73eb2240 ???????? ???????? ???????? ???????? 73eb2250 ???????? ???????? ???????? ???????? 73eb2260 ???????? ???????? ???????? ???????? 73eb2270 ???????? ???????? ???????? ????????

The value is indeed belongs to unloaded fpb.tmp module address range:

0:000> ln 73eb2200 (73eb2200) <Unloaded_fpb.tmp>+0×12200

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

Posted in Crash Dump Analysis, Crash Dump Patterns, Debugging | 2 Comments »

Individual and Enterprise Software Diagnostics Certifications

Monday, June 18th, 2012

Memory Dump Analysis Services will be administering certifications developed by Software Diagnostics Institute for memory dump and software trace analysis:

Software Diagnostics Maturity Enterprise Certification
Memory Dump Analysis Certification is available this September

Debugging TV Frames episode 0×10 contains some background information.

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

Posted in Announcements, Certification, Core Dump Analysis, Crash Dump Analysis, Crash Dump Patterns, Debugging, Software Diagnostics, Software Diagnostics Institute, Software Diagnostics Patterns, Software Engineering, Software Technical Support, Software Trace Analysis, Trace Analysis Patterns | No Comments »

Software Diagnostics Institute

Tuesday, June 12th, 2012

DumpAnalysis.org portal has been reorganized to Software Diagnostics Institute to reflect the nature of its research activities. More updates later on.

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

Posted in Announcements, Core Dump Analysis, Crash Dump Analysis, Crash Dump Patterns, Debugging, Software Behavior Patterns, Software Diagnostics, Software Diagnostics Institute, Software Diagnostics Patterns, Software Engineering, Software Trace Analysis, Trace Analysis Patterns | No Comments »

Crash Dump Analysis Patterns (Part 177)

Saturday, June 9th, 2012

Stack Trace Change is an important pattern for differential memory dump analysis, for example, when memory dumps were generated before and after a problem such as a CPU spike or hang. In the example below we have a normal expected thread stack trace from a memory dump saved before an application was reported unresponsive and another different thread stack trace after:

3 Id: 24b8.24e4 Suspend: 0 Teb: 7efa1000 Unfrozen ChildEBP RetAddr 037dfadc 75210bdd ntdll!ZwWaitForMultipleObjects+0x15 037dfb78 75791a2c KERNELBASE!WaitForMultipleObjectsEx+0x100 037dfbc0 7511086a kernel32!WaitForMultipleObjectsExImplementation+0xe0 037dfc14 00d17c1d user32!RealMsgWaitForMultipleObjectsEx+0x14d 037dfc3c 00ce161d ApplicationA!MsgWaitForMultipleObjects+0x2d 037dfc60 00cdc757 ApplicationA!WaitForSignal+0x1d 037dfc80 00cdaaf6 ApplicationA!WorkLoop+0x57 037dfca4 7579339a ApplicationA!ThreadStart+0x26 037dfcb0 77699ef2 kernel32!BaseThreadInitThunk+0xe 037dfcf0 77699ec5 ntdll!__RtlUserThreadStart+0x70 037dfd08 00000000 ntdll!_RtlUserThreadStart+0x1b

3 Id: 24b8.24e4 Suspend: 0 Teb: 7efa1000 Unfrozen ChildEBP RetAddr 037df38c 752131bb ntdll!ZwDelayExecution+0x15 037df3f4 75213a8b KERNELBASE!SleepEx+0x65 037df404 00d1670b KERNELBASE!Sleep+0xf 037df40c 00d350ef ApplicationA!Sleep+0xb 037df430 6a868aab ApplicationA!PutData+0xbf 037df444 6a8662ec ModuleA!OutputData+0x1b 037df464 00d351de ModuleA!ProcessData+0x16c 037df4a4 00ca8cb4 ApplicationA!SendData+0xbe [...]

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

Posted in Crash Dump Analysis, Crash Dump Patterns, Debugging | No Comments »

Crash Dump Analysis Patterns (Part 24, Mac OS X)

Saturday, June 9th, 2012

This is a Mac OS X / GDB counterpart to Coincidental Symbolic Information pattern previously described for Windows platforms. The idea is the same: to disassemble the address to see if the preceding instruction is a call. If it is indeed then most likely the symbolic address is a return address from past Execution Residue:

(gdb) x $rsp 0x7fff6a162a38: 0x8fab9a9c

(gdb) x/1000a 0x7fff6a162000 [...] 0x7fff6a162960: 0x7fff6a162980 0x7fff6a167922 0x7fff6a162970: 0x0 0x0 0x7fff6a162980: 0x7fff6a162a50 0×7fff8a31e716 <dyld_stub_binder_+13> 0×7fff6a162990: 0×1 0×7fff6a162b00 0×7fff6a1629a0: 0×7fff6a162b10 0×7fff6a162bc0 0×7fff6a1629b0: 0×8 0×0 […] 0×7fff6a162a00: 0×0 0×0 0×7fff6a162a10: 0×0 0×0 0×7fff6a162a20: 0×0 0×0 0×7fff6a162a30: 0×7fff6a162a60 0×7fff8fab9a9c <abort+177> 0×7fff6a162a40: 0×0 0×0 0×7fff6a162a50: 0×7fffffffffdf 0×0 […] 0×7fff6a163040: 0×35000 0×0 0×7fff6a163050: 0×35000 0×500000007 0×7fff6a163060: 0×7 0×747865745f5f 0×7fff6a163070: 0×0 0×545845545f5f 0×7fff6a163080: 0×0 0×7fff5fc01000 <__dyld_stub_binding_helper> 0×7fff6a163090: 0×22c9d 0xc00001000 0×7fff6a1630a0: 0×0 0×80000400 […]

(gdb) disass 0×7fff8a31e716 Dump of assembler code for function dyld_stub_binder_: 0×00007fff8a31e709 <dyld_stub_binder_+0>: mov 0×8(%rbp),%rdi 0×00007fff8a31e70d <dyld_stub_binder_+4>: mov 0×10(%rbp),%rsi 0×00007fff8a31e711 <dyld_stub_binder_+8>: callq 0×7fff8a31e86d <_Z21_dyld_fast_stub_entryPvl> 0×00007fff8a31e716 <dyld_stub_binder_+13>: mov %rax,%r11 0×00007fff8a31e719 <dyld_stub_binder_+16>: movdqa 0×40(%rsp),%xmm0 0×00007fff8a31e71f <dyld_stub_binder_+22>: movdqa 0×50(%rsp),%xmm1 0×00007fff8a31e725 <dyld_stub_binder_+28>: movdqa 0×60(%rsp),%xmm2 0×00007fff8a31e72b <dyld_stub_binder_+34>: movdqa 0×70(%rsp),%xmm3 0×00007fff8a31e731 <dyld_stub_binder_+40>: movdqa 0×80(%rsp),%xmm4 0×00007fff8a31e73a <dyld_stub_binder_+49>: movdqa 0×90(%rsp),%xmm5 0×00007fff8a31e743 <dyld_stub_binder_+58>: movdqa 0xa0(%rsp),%xmm6 0×00007fff8a31e74c <dyld_stub_binder_+67>: movdqa 0xb0(%rsp),%xmm7 0×00007fff8a31e755 <dyld_stub_binder_+76>: mov (%rsp),%rdi 0×00007fff8a31e759 <dyld_stub_binder_+80>: mov 0×8(%rsp),%rsi 0×00007fff8a31e75e <dyld_stub_binder_+85>: mov 0×10(%rsp),%rdx 0×00007fff8a31e763 <dyld_stub_binder_+90>: mov 0×18(%rsp),%rcx 0×00007fff8a31e768 <dyld_stub_binder_+95>: mov 0×20(%rsp),%r8 0×00007fff8a31e76d <dyld_stub_binder_+100>: mov 0×28(%rsp),%r9 0×00007fff8a31e772 <dyld_stub_binder_+105>: mov 0×30(%rsp),%rax 0×00007fff8a31e777 <dyld_stub_binder_+110>: add $0xc0,%rsp 0×00007fff8a31e77e <dyld_stub_binder_+117>: pop %rbp 0×00007fff8a31e77f <dyld_stub_binder_+118>: add $0×10,%rsp 0×00007fff8a31e783 <dyld_stub_binder_+122>: jmpq *%r11

(gdb) x/2i 0×7fff8fab9a9c 0×7fff8fab9a9c <abort+177>: mov $0×2710,%edi 0×7fff8fab9aa1 <abort+182>: callq 0×7fff8fab9c43 <usleep$nocancel>

(gdb) disass 0×7fff8fab9a9c-5 0×7fff8fab9a9c Dump of assembler code from 0×7fff8fab9a97 to 0×7fff8fab9a9c: 0×00007fff8fab9a97 <abort+172>: callq 0×7fff8fb1f54a <dyld_stub_kill> End of assembler dump.

(gdb) disass 0×7fff5fc01000 Dump of assembler code for function __dyld_stub_binding_helper: 0×00007fff5fc01000 <__dyld_stub_binding_helper+0>: add %al,(%rax) 0×00007fff5fc01002 <__dyld_stub_binding_helper+2>: add %al,(%rax) 0×00007fff5fc01004 <__dyld_stub_binding_helper+4>: add %al,(%rax) 0×00007fff5fc01006 <__dyld_stub_binding_helper+6>: add %al,(%rax) End of assembler dump.

(gdb) x/10 0×7fff5fc01000-0×10 0×7fff5fc00ff0: 0×00000000 0×00000000 0×00000000 0×00000000 0×7fff5fc01000 <__dyld_stub_binding_helper>: 0×00000000 0×00000000 0×00000000 0×00000000 0×7fff5fc01010 <__dyld_offset_to_dyld_all_image_infos>: 0×00000000 0×00000000

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

Forthcoming Training: Accelerated Mac OS X Core Dump Analysis

Posted in Core Dump Analysis, Crash Dump Patterns, GDB for WinDbg Users, Mac Crash Corner, Mac OS X, x64 Mac OS X | No Comments »

Crash Dump Analysis Patterns (Part 176)

Wednesday, June 6th, 2012

Sometimes, when an application is sluggish, periodically consumes CPU, it is possible to create a set of consecutive memory dumps of the same process to see the temporal development of any thread CPU consumption and figure out potential Spike Interval(s). For example, the following diagram was plotted from !runaway WinDbg command output for thread #1:

The 3rd and the 5th user process memory dumps in addition to increased CPU consumption also have corresponding non-waiting stack trace frames caught while executing some CPU instructions in ModuleA (not preempted with saved context). The first memory dump (yellow bar) with 437 ms user time spent out of 629 ms elapsed time also has a non-waiting stack trace but we consider it a normal application startup CPU consumption spike.

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

Posted in Crash Dump Analysis, Crash Dump Patterns | No Comments »

Crash Dump Analysis Patterns (Part 23a, Mac OS X)

Tuesday, May 29th, 2012

This is a Mac OS X / GDB counterpart to Double Free (process heap) pattern previously described for Windows platforms:

(gdb) bt #0 0x00007fff8479582a in __kill () #1 0x00007fff8e0e0a9c in abort () #2 0x00007fff8e13f84c in free () #3 0x00000001035a8ef4 in main (argc=1, argv=0x7fff631a7b20)

(gdb) x/2i 0x00000001035a8ef4-8 0x1035a8eec : mov -0×20(%rbp),%edi 0×1035a8eef : callq 0×1035a8f06

(gdb) frame 3 #3 0x00000001035a8ef4 in main (argc=1, argv=0x7fff631a7b20) at .../DoubleFree/main.c:23 23 free(p2); Current language: auto; currently minimal

(gdb) x/g $rbp-0x20 0x7fff631a7ae0: 0x00007fe6a8801400

(gdb) x/2w 0x00007fe6a8801400 0x7fe6a8801400: 0x00000000 0xb0000000

Here’s the source code of the modeling application:

int main(int argc, const char * argv[])

{

char *p1 = (char *) malloc (1024);

printf(“p1 = %p\n”, p1);

char *p2 = (char *) malloc (1024);

printf(“p2 = %p\n”, p2);

free(p2);

free(p1);

free(p2);

return 0;

}

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

Forthcoming Training: Accelerated Mac OS X Core Dump Analysis

Posted in Assembly Language, Core Dump Analysis, Crash Dump Analysis, Crash Dump Patterns, Debugging, GDB for WinDbg Users, Mac Crash Corner, Mac OS X, Software Defect Construction, x64 Mac OS X | No Comments »

Crash Dump Analysis Patterns (Part 2, Mac OS X)

Sunday, May 27th, 2012

This is a Mac OS X / GDB counterpart to Dynamic Memory Corruption (process heap) pattern previously described for Windows platforms:

(gdb) bt #0 0x00007fff8479582a in __kill () #1 0x00007fff8e0e0a9c in abort () #2 0x00007fff8e1024ac in szone_error () #3 0x00007fff8e1024e8 in free_list_checksum_botch () #4 0x00007fff8e102a7b in small_free_list_remove_ptr () #5 0x00007fff8e106bf7 in szone_free_definite_size () #6 0x00007fff8e13f789 in free () #7 0x000000010afafe23 in main (argc=1, argv=0x7fff6abaeb08)

Here’s the source code of the modeling application:

int main(int argc, const char * argv[])

{

char *p1 = (char *) malloc (1024);

printf(“p1 = %p\n”, p1);

char *p2 = (char *) malloc (1024);

printf(“p2 = %p\n”, p2);

char *p3 = (char *) malloc (1024);

printf(“p3 = %p\n”, p3);

char *p4 = (char *) malloc (1024);

printf(“p4 = %p\n”, p4);

char *p5 = (char *) malloc (1024);

printf(“p5 = %p\n”, p5);

char *p6 = (char *) malloc (1024);

printf(“p6 = %p\n”, p6);

char *p7 = (char *) malloc (1024);

printf(“p7 = %p\n”, p7);

free(p6);

free(p4);

free(p2);

printf(“Hello Crash!\n”);

strcpy(p2, “Hello Crash!”);

strcpy(p4, “Hello Crash!”);

strcpy(p6, “Hello Crash!”);

p2 = (char *) malloc (512);

printf(“p2 = %p\n”, p2);

p4 = (char *) malloc (1024);

printf(“p4 = %p\n”, p4);

p6 = (char *) malloc (512);

printf(“p6 = %p\n”, p6);

free (p7);

free (p6);

free (p5);

free (p4);

free (p3);

free (p2);

free (p1);

return 0;

}

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

Forthcoming Training: Accelerated Mac OS X Core Dump Analysis

Posted in Core Dump Analysis, Crash Dump Analysis, Crash Dump Patterns, Debugging, GDB for WinDbg Users, Mac Crash Corner, Mac OS X, Software Defect Construction, x64 Mac OS X | No Comments »

Crash Dump Analysis Patterns (Part 175)

Wednesday, May 23rd, 2012

Stored Exception pattern is mostly useful when an exception thread is not present like in this rare example:

ERROR: Unable to find system thread 9B7E ERROR: The thread being debugged has either exited or cannot be accessed ERROR: Many commands will not work properly This dump file has an exception of interest stored in it. The stored exception information can be accessed via .ecxr. ERROR: Exception C0000005 occurred on unknown thread 9B7E (95f4.9b7e): Access violation - code c0000005 (first/second chance not available)

.ecxr will not work here but the exception record is available via .exr command:

0:???> .exr -1 ExceptionAddress: 08a9ae18 (DllB.dll+0x001cae18) ExceptionCode: c0000005 (Access violation) ExceptionFlags: 00000001 NumberParameters: 1 Parameter[0]: 00000008

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

Posted in Crash Dump Analysis, Crash Dump Patterns, WinDbg Tips and Tricks | No Comments »

Crash Dump Analysis Patterns (Part 174)

Sunday, May 20th, 2012

Activity Resonance pattern is observed when two products from different vendors compete in some functional domain such malware detection. In the example below ApplicationA and AVDriverA modules belong to Vendor A and AV-B module belongs to Vendor B. Both threads are spiking threads blocking all other activity in the system:

0: kd> !running

System Processors: (0000000000000003) Idle Processors: (0000000000000000) (0000000000000000) (0000000000000000) (0000000000000000)

Prcbs Current Next 0 fffff80001845e80 fffffa8004350060 ................ 1 fffff880009c4180 fffffa80028e7060 ................

0: kd> !thread fffffa8004350060 ff THREAD fffffa8004350060 Cid 14424.14b34 Teb: 000000007efdb000 Win32Thread: fffff900c1d32c30 RUNNING on processor 0 Not impersonating DeviceMap fffff8a00148fe80 Owning Process fffffa8003d6cb30 Image: ApplicationA.exe Attached Process N/A Image: N/A Wait Start TickCount 10568630 Ticks: 0 Context Switch Count 345 LargeStack UserTime 00:02:21.360 KernelTime 01:09:32.130 Win32 Start Address ApplicationA!mainCRTStartup (0×0000000000404c1b) Stack Init fffff88006c71db0 Current fffff88006c71670 Base fffff88006c72000 Limit fffff88006c6a000 Call 0 Priority 9 BasePriority 8 UnusualBoost 0 ForegroundBoost 0 IoPriority 2 PagePriority 5 Child-SP RetAddr Call Site fffff880`06c70ec0 fffff880`0197d53c AVDriverA+0×15d69 fffff880`06c70f10 fffff880`01988556 AVDriverA+0×1453c fffff880`06c70fd0 fffff880`019886a8 AVDriverA+0×1f556 fffff880`06c71000 fffff800`0198ebfd AVDriverA+0×1f6a8 fffff880`06c71060 fffff800`019bf4f2 nt! ?? ::NNGAKEGL::`string’+0×2a6fd fffff880`06c711e0 fffff800`019c3385 nt!PspCreateThread+0×246 fffff880`06c71460 fffff800`016d28d3 nt!NtCreateThreadEx+0×25d fffff880`06c71bb0 00000000`76e61d9a nt!KiSystemServiceCopyEnd+0×13 (TrapFrame @ fffff880`06c71c20) 00000000`0008e178 00000000`74990411 ntdll!ZwCreateThreadEx+0xa 00000000`0008e180 00000000`7497cf87 wow64!whNtCreateThreadEx+0×815 00000000`0008e350 00000000`748c2776 wow64!Wow64SystemServiceEx+0xd7 00000000`0008ec10 00000000`7497d07e wow64cpu!TurboDispatchJumpAddressEnd+0×2d 00000000`0008ecd0 00000000`7497c549 wow64!RunCpuSimulation+0xa 00000000`0008ed20 00000000`76e54956 wow64!Wow64LdrpInitialize+0×429 00000000`0008f270 00000000`76e51a17 ntdll!LdrpInitializeProcess+0×17e4 00000000`0008f760 00000000`76e3c32e ntdll! ?? ::FNODOBFM::`string’+0×29220 00000000`0008f7d0 00000000`00000000 ntdll!LdrInitializeThunk+0xe

0: kd> !thread fffffa80028e7060 ff THREAD fffffa80028e7060 Cid 0dc4.0e5c Teb: 000000007efa4000 Win32Thread: 0000000000000000 RUNNING on processor 1 Not impersonating DeviceMap fffff8a000008b30 Owning Process fffffa8002817060 Image: AV-B.exe Attached Process N/A Image: N/A Wait Start TickCount 10568617 Ticks: 13 (0:00:00:00.203) Context Switch Count 1763138 UserTime 00:04:26.765 KernelTime 03:09:31.140 Win32 Start Address AV-B (0×00000000004289f2) Stack Init fffff88003b88db0 Current fffff88003b88900 Base fffff88003b89000 Limit fffff88003b83000 Call 0 Priority 15 BasePriority 15 UnusualBoost 0 ForegroundBoost 0 IoPriority 2 PagePriority 5 Child-SP RetAddr Call Site fffff880`03b88660 fffff800`019919a9 nt!ObReferenceObjectSafe+0xf fffff880`03b88690 fffff800`01991201 nt!PsGetNextProcess+0×81 fffff880`03b886e0 fffff800`019dcef6 nt!ExpGetProcessInformation+0×774 fffff880`03b88830 fffff800`019dd949 nt!ExpQuerySystemInformation+0xfb4 fffff880`03b88be0 fffff800`016d28d3 nt!NtQuerySystemInformation+0×4d fffff880`03b88c20 00000000`76e6167a nt!KiSystemServiceCopyEnd+0×13 (TrapFrame @ fffff880`03b88c20) 00000000`0118e708 00000000`74987da7 ntdll!NtQuerySystemInformation+0xa 00000000`0118e710 00000000`74988636 wow64!whNT32QuerySystemProcessInformationEx+0×93 00000000`0118e760 00000000`7498a0e9 wow64!whNtQuerySystemInformation_SpecialQueryCase+0×466 00000000`0118e800 00000000`7497cf87 wow64!whNtQuerySystemInformation+0xf1 00000000`0118e840 00000000`748c2776 wow64!Wow64SystemServiceEx+0xd7 00000000`0118f100 00000000`7497d07e wow64cpu!TurboDispatchJumpAddressEnd+0×2d 00000000`0118f1c0 00000000`7497c549 wow64!RunCpuSimulation+0xa 00000000`0118f210 00000000`76e8e707 wow64!Wow64LdrpInitialize+0×429 00000000`0118f760 00000000`76e3c32e ntdll! ?? ::FNODOBFM::`string’+0×29364 00000000`0118f7d0 00000000`00000000 ntdll!LdrInitializeThunk+0xe

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

Posted in Crash Dump Analysis, Crash Dump Patterns, Victimware, x64 Windows | No Comments »

Crash Dump Analysis Patterns (Part 173)

Sunday, May 20th, 2012

Value Adding Process is a frequently observed pattern in terminal services environments when you see one or several process names listed in each session but not necessarily required. They are usually running to provide some user experience enhancements. In such cases if observed functional problems correspond to the purpose of running additional processes we might want to eliminate them for testing and troubleshooting purposes.

0: kd> !sprocess 12 Dumping Session 12

_MM_SESSION_SPACE fffff8800e5d5000 _MMSESSION fffff8800e5d5b40 PROCESS fffffa8008d50b30 SessionId: 12 Cid: 0b04 Peb: 7fffffdc000 ParentCid: 1478 DirBase: 6bb77000 ObjectTable: fffff8a003f280b0 HandleCount: 158. Image: csrss.exe

PROCESS fffffa80030c7060 SessionId: 12 Cid: 1a48 Peb: 7fffffd8000 ParentCid: 1478 DirBase: 0a33c000 ObjectTable: fffff8a003c46c00 HandleCount: 179. Image: winlogon.exe

PROCESS fffffa8008250b30 SessionId: 12 Cid: 18c8 Peb: 7fffffdf000 ParentCid: 1a48 DirBase: 0350d000 ObjectTable: fffff8a0025b6840 HandleCount: 226. Image: LogonUI.exe

PROCESS fffffa8008b00530 SessionId: 12 Cid: 1508 Peb: 7fffffdf000 ParentCid: 02f0 DirBase: 02f65000 ObjectTable: fffff8a003b7e530 HandleCount: 197. Image: ExcitingFeatureX.exe

[...]

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

Posted in Crash Dump Analysis, Crash Dump Patterns | No Comments »

Crash Dump Analysis Patterns (Part 20d)

Saturday, May 19th, 2012

This is a specialization of Insufficient Memory (kernel pool) pattern called Memory Leak (I/O completion packets). The currently unique diagnostics this pattern provides in comparison with other kernel pool tags is that the pool allocation entries show the leaking process:

0: kd> !poolused 3 Sorting by NonPaged Pool Consumed

Pool Used: NonPaged Paged Tag Allocs Frees Diff Used Allocs Frees Diff Used Icp 1294074 42875 1251199 96642976 0 0 0 0 I/O completion packets queue on a completion ports […]

0: kd> !poolfind Icp

Scanning large pool allocation table for Tag: Icp (fffffa8013e00000 : fffffa8014100000)

*fffffa800e188260 size: 50 previous size: 40 (Allocated) Icp Process: fffffa800899dc40 *fffffa800e1882e0 size: 50 previous size: 30 (Allocated) Icp Process: fffffa800899dc40 *fffffa800e188330 size: 50 previous size: 50 (Allocated) Icp Process: fffffa800899dc40 *fffffa800e188380 size: 50 previous size: 50 (Allocated) Icp Process: fffffa800899dc40 *fffffa800e1883d0 size: 50 previous size: 50 (Allocated) Icp Process: fffffa800899dc40 *fffffa800e188420 size: 50 previous size: 50 (Allocated) Icp Process: fffffa800899dc40 *fffffa800e188470 size: 50 previous size: 50 (Allocated) Icp Process: fffffa800899dc40 *fffffa800e1884c0 size: 50 previous size: 50 (Allocated) Icp Process: fffffa800899dc40

0: kd> !process fffffa800899dc40 1 PROCESS fffffa800899dc40 SessionId: 0 Cid: 43a4 Peb: 7efdf000 ParentCid: 0412 DirBase: 09d6b000 ObjectTable: fffff8a0046c8c10 HandleCount: 1068. Image: ServiceA.exe […]

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

Posted in Crash Dump Analysis, Crash Dump Patterns, Debugging | No Comments »

Crash Dump Analysis Patterns (Part 172)

Saturday, May 19th, 2012

Recently I observed a few occurrences of a rare No Current Thread pattern in a large set of process memory dumps:

0:???> k WARNING: The debugger does not have a current process or thread WARNING: Many commands will not work ^ Illegal thread error in ‘k’

0:???> ~ WARNING: The debugger does not have a current process or thread WARNING: Many commands will not work 0 Id: 95f4.6780 Suspend: 1 Teb: 7efdd000 Unfrozen

Setting a current thread helps:

0:???> ~0s WARNING: The debugger does not have a current process or thread WARNING: Many commands will not work eax=037d0010 ebx=0002bda0 ecx=03b1a010 edx=00000007 esi=037d0010 edi=03b069fc eip=0397939f esp=0018fd98 ebp=0018fdd8 iopl=0 nv up ei pl nz na po nc cs=0023 ss=002b ds=002b es=002b fs=0053 gs=002b efl=00200202 DllA+0×939f: 0397939f 8b10 mov edx,dword ptr [eax] ds:002b:037d0010=03b1a010

0:000> k ChildEBP RetAddr WARNING: Stack unwind information not available. Following frames may be wrong. 0018fdd8 03975257 DllA+0x939f 0018fdf8 03975577 DllA+0x5257 0018fe58 772bb9a0 DllA+0x5577 0018fe78 772d9b96 ntdll!LdrpCallInitRoutine+0x14 0018ff1c 772d9a38 ntdll!LdrShutdownProcess+0x1aa 0018ff30 752279f4 ntdll!RtlExitUserProcess+0x74 0018ff44 0040625d kernel32!ExitProcessStub+0x12 0018ff5c 012528e5 Application+0x625d 0018ff88 7522339a Application!foo+0xdc88f1 0018ff94 772bbf42 kernel32!BaseThreadInitThunk+0xe 0018ffd4 772bbf15 ntdll!__RtlUserThreadStart+0x70 0018ffec 00000000 ntdll!_RtlUserThreadStart+0x1b

However, EIP of the new current thread doesn’t point to any access violation and the dereferenced address is valid:

0:000> !address 037d0010 Usage: <unclassified> Allocation Base: 037d0000 Base Address: 037d0000 End Address: 038dd000 Region Size: 0010d000 Type: 00020000 MEM_PRIVATE State: 00001000 MEM_COMMIT Protect: 00000004 PAGE_READWRITE

Also, if we inspect the raw stack data we won’t find any hidden exceptions there. So we conclude that the missing thread was exceptional. Indeed, there is a saved exception context in the process memory dump:

0:000> .exr -1 ExceptionAddress: 08a9ae18 (<Unloaded_DllB.dll>+0x001cae18) ExceptionCode: c0000005 (Access violation) ExceptionFlags: 00000001 NumberParameters: 1 Parameter[0]: 00000008

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

Posted in Crash Dump Analysis, Crash Dump Patterns, Debugging | No Comments »

Crash Dump Analysis Patterns (Part 14, Mac OS X)

Wednesday, May 9th, 2012

This is a Mac OS X / GDB counterpart to Spiking Thread pattern previously described for Windows platforms:

(gdb) info threads 4 0×00007fff85b542df in sqrt$fenv_access_off () 3 0×00007fff8616ee42 in __semwait_signal () 2 0×00007fff8616ee42 in __semwait_signal () * 1 0×00007fff8616ee42 in __semwait_signal ()

We notice a non-waiting thread and switch to it:

(gdb) thread 4 [Switching to thread 4 (core thread 3)] 0x00007fff85b542df in sqrt$fenv_access_off ()

(gdb) bt #0 0x00007fff85b542df in sqrt$fenv_access_off () #1 0×000000010cc85dc9 in thread_three (arg=0×7fff6c884ac0) #2 0×00007fff8fac68bf in _pthread_start () #3 0×00007fff8fac9b75 in thread_start ()

If we disassemble the return address for thread_three function to come back from sqrt call we see an infinite loop:

(gdb) disass 0x000000010cc85dc9 Dump of assembler code for function thread_three: 0x000000010cc85db0 <thread_three+0>: push %rbp 0×000000010cc85db1 <thread_three+1>: mov %rsp,%rbp 0×000000010cc85db4 <thread_three+4>: sub $0×10,%rsp 0×000000010cc85db8 <thread_three+8>: mov %rdi,-0×10(%rbp) 0×000000010cc85dbc <thread_three+12>: mov -0×10(%rbp),%ax 0×000000010cc85dc0 <thread_three+16>: movsd (%rax),%xmm0 0×000000010cc85dc4 <thread_three+20>: callq 0×10cc85eac <dyld_stub_sqrt> 0×000000010cc85dc9 <thread_three+25>: mov -0×10(%rbp),%rax 0×000000010cc85dcd <thread_three+29>: movsd %xmm0,(%rax) 0×000000010cc85dd1 <thread_three+33>: jmpq 0×10cc85dbc <thread_three+12> End of assembler dump.

Here’s the source code of the modeling application:

void * thread_one (void *arg)

{

while (1)

{

sleep (1);

}

return 0;

}

void * thread_two (void *arg)

{

while (1)

{

sleep (2);

}

return 0;

}

void * thread_three (void *arg)

{

while (1)

{

*(double*)arg=sqrt(*(double *)arg);

}

return 0;

}

int main(int argc, const char * argv[])

{

pthread_t threadID_one, threadID_two, threadID_three;

double result = 0xffffffff;

pthread_create (&threadID_one, NULL, thread_one, NULL);

pthread_create (&threadID_two, NULL, thread_two, NULL);

pthread_create (&threadID_three, NULL, thread_three,

&result);

pthread_join(threadID_three, NULL);

return 0;

}

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

Forthcoming Training: Accelerated Mac OS X Core Dump Analysis

May 2026
M	T	W	T	F	S	S
« Mar
				1	2	3
4	5	6	7	8	9	10
11	12	13	14	15	16	17
18	19	20	21	22	23	24
25	26	27	28	29	30	31

Archive for the ‘Crash Dump Patterns’ Category

Pages

Recent Comments

Categories

Archives

ARM64

Automated Analysis

Blogroll

Debugging Channels

Forensics

Hardware

Linux

Mac OS X

Magazines and Newspapers

Malware Analysis

Medical Diagnostics

Narratology

Related Links

Reversing

Scripting Languages

Source Code

Tracing Tools

Meta