[sf-lug] K & R errata

[jim]

on my 32-bit system, ints are 4 bytes (per sizeof) 

compiled file with main with empty statement is 7102 bytes 
main() { ; } 

compiled file that declares 100 ints is 9099 bytes 
int i1, i2...i100; main() { ; }  // 7102+1997 

compiled file that defines 100 ints is 9103 bytes 
int i1=0, i2=0...i100=0; main() { ; }  // 7102+2001 

compiled file that declares an array of 100 ints is 7124 bytes 
int is[100]; main { ; }  // 7102+22 

compiled file that defines an array of 100 ints is 7576 bytes 
int is{100] = { 0,0,...0 }; main() { ; }  // 7102+474 (452 bigger) 


    next step is to analyze the structure. later, time 
for lunch. 





On Tue, 2011-07-05 at 12:44 -0700, Alison Chaiken wrote:
> jim writes:
> > "They should both get initialized to zero, which means that they
> should
> > go into the .bss section of the ELF file...."
> > JS: This is Linux-specific behavior? I.e. not necessarily true
> > for non-Linux OSes?
> 
> AFAIK all major Unices have migrated from older a.out to ELF:
> 
> http://www.freebsd.org/doc/handbook/binary-formats.html
> 
> By the way, binutils has lots of other useful stuff besides readelf,
> for example nm and objdump.     You have hours of happy man-page
> reading ahead of you!
> 
> If I didn't have a deadline, I'd make a .c file with 
> 
> int a1;
> int a2;
> . . . 
> int a100;
> 
> 
> and compare it's size to 
> 
> 
> int a[100];
> 
> 
> but I have a deadline right now.    I don't know the answer to the
> earlier question about the function declarations and memory and would
> like to explore that too, but maybe I should try to get my
> presentation finished before my business trip instead  .  .  .
> 
> -- 
> Alison Chaiken
> (650) 279-5600  (cell)
> http://www.exerciseforthereader.org/
> Spend much time at the cutting edge and you're liable to get cut.
> 
>