Question

I’ve been debugging a segfault in some numerical code (C++, no dependencies on any external libraries). I’ve got it reproducing the segfault when compiled at optimization level -O0 (using gcc) and run within gdb; the backtrace I get is:

(gdb) backtrace 2
#0  0x000000000040171c in QExpFloorLog_Q_R (q=Cannot access memory at address 0x7fffff7fefec) at GlobalFunctions.hpp:11
#1  0x0000000000401b50 in MinExponentFromTheoremTwo (q=2, k_q=20 '\024', r=1122053) at GlobalFunctions.hpp:90

The signature of MinExponentFromTheoremTwo is:

unsigned char MinExponentFromTheoremTwo(const unsigned int q, const unsigned char k_q, const unsigned int r);

So, my question is: what am I to make of k_q=20 '\024'? I understand that k_q is an unsigned char, and therefore that gdb is giving me both a numerical value and an ascii representation; I have other places with unsigned char variables listed as var1=0 '\000', var2=1 '\001', etc. But k_q seems mismatched, with a value of 20 and a code of \024. Am I misunderstanding how unsigned char are used, or should this be helping me find the bug in my code?

Incidentally, I have

$ uname -a
Linux clayton-AOD255E 2.6.38-13-generic #53-Ubuntu SMP Mon Nov 28 19:33:45 UTC 2011 x86_64 x86_64 x86_64 GNU/Linux
$ g++ --version
g++ (Ubuntu/Linaro 4.5.2-8ubuntu4) 4.5.2
$ gdb --version
GNU gdb (Ubuntu/Linaro 7.2-1ubuntu11) 7.2

Thanks for any and all help!

Answer 1

The number called twenty is represented in base 10 as ’20’. In base 8, it is ’24’. \024 is the octal representation of that value.

All of the following are the same value:

20     // decimal, base-10
024    // octal, base-8
'\024' // octal, base-8
0x14   // hex, base-16