[c] segmentation fault- request help

1 Attachment(s)

Was reading c guides on dynamic memory allocation functions.. and tried to write a program to create a linked list, to back up the knowledge... and i failed miserably :p can't figure out why it keeps giving me segmentation faults >.<

i figured the scanf is the cause.. but don't understand why ;o
I'd appreciate if someone could point out whats wrong

here's a piece of the code to blame : (i attached the full source code just in case)

Code:

// new entry and head are both links to struct record type ofc new_entry=(struct record*)malloc(sizeof(struct record)); printf("Name:\n"); scanf("%s",new_entry->name); printf("Age:\n"); scanf("%i",new_entry->age); // <<-- is this correct? new_entry->next=NULL; head=new_entry;

Re: [c] segmentation fault- request help

I have no idea what causes the mentioned error, but what's your struct record like? What I mean is, do the variables, especially name, have any memory reserved?

Re: [c] segmentation fault- request help

sorry, missed out the upper part of the full source code somehow lol, here it is :

Code:

#include <stdio.h> #include <stdlib.h> struct record{ char name[15]; int age; struct record *next; };

Re: [c] segmentation fault- request help

I think you're passing the latter scanf an int instead of int*

Re: [c] segmentation fault- request help

Ya, i figured my scanf's passing argument is wrong.. but i just couldn't figure how to correct it ;/

if i tried giving it a & thingy to "age".. would give me "error: syntax error before '&' token"

Re: [c] segmentation fault- request help

Isn't it &new_entry->age

sage

http://forum.ragezone.com/f144/c-help-255086/ was locked for a reason.

...and learn moar C, -> is not the correct operator to use.

Re: sage

Quote:

Originally Posted by username1

...and learn moar C, -> is not the correct operator to use.

It's not? What is? :huh:
Now I wonder if I've forgotten the correct way only recently or if I've always done it wrong. What's the mistake with it, anyway?

Re: sage

Quote:

Originally Posted by username1

http://forum.ragezone.com/f144/c-help-255086/ was locked for a reason.

Aye, and the reason was given: incorrect topic title and he just dumped his entire program source here. Both errors are corrected now, so there's nothing wrong with this topic.

Re: [c] segmentation fault- request help

bleh, this code really IS a miserable failure..the & operator did fix that prob... i think lol.. but then it just bumped into the next one...and i am even more clueless now .. it skips the scanf for the answer line.. and makes the while loop endless.. wth >.<
to to make it short :

Code:

while(answer!='n'){ if(!head){ blah blah blah; } else{ yada yada yada; } printf("Another entry? [y/n] \n"); scanf("%c",&answer); }

once i complete the first entry..it just prints the above , skips the answer scan and proceeds asking the name/age for the next entry.. endlessly ;/

+ more segmentation faults await ahead lol..screw this forsaken code! ;p

Thanks for your help and time btw =)

Guess i'll move on .. gonna try to learn the getopt_long function ...

Re: [c] segmentation fault- request help

You say it skips scanf, but does it also skip printf? If it doesn't I'm really amazed why it behaves like that..

Re: sage

Quote:

Originally Posted by Negata

It's not? What is? :huh:
Now I wonder if I've forgotten the correct way only recently or if I've always done it wrong. What's the mistake with it, anyway?

http://www.le.ac.uk/cc/tutorials/c/ccccstrc.html#ps

...and it looks like your "blah blah blah" and "yada yada yada" are doing something to answer.

protip: A debugger is your friend.

1 Attachment(s)

Re: sage

Quote:

Originally Posted by username1

Introduction to C Programming: Computer Centre: University of Leicester

...and it looks like your "blah blah blah" and "yada yada yada" are doing something to answer.

protip: A debugger is your friend.

That tutorial is the one i've been reading =P, finished it already.
Btw, i tried to make a smaller simpler program in the same way, for testing purposes.. and it ended up the same way.. answer is skipped ;o

Here's how it ran:
Name:
Al
19
Al is 19 years old
Another entry?[y/n]
Name:
Don
17
Don is 17 years old
Another entry?[y/n]
Name:
n
15
n is 15 years old
Another entry?[y/n]
Name:
gdsgsdgds
n
gdsgsdgds is 15 years old
Another entry?[y/n]
[alex@greeny temp]$

Seems like age and answer are both written in the same block of memory... but why? ;/ (the other program ran the same way.. the loop would only end when i put 'n' during age scanning)

ps: attached the code for the smaller one.

Re: sage

Quote:

Originally Posted by username1

Introduction to C Programming: Computer Centre: University of Leicester

now if I have a struct A with a member B, an a of type A*, why is a->B not correct? That's exactly the case here and exactly what is said on the page you referred to.

I told you to use a debugger!

Quote:

Originally Posted by Negata

now if I have a struct A with a member B, an a of type A*, why is a->B not correct? That's exactly the case here and exactly what is said on the page you referred to.

I think it has something to do with scanf expecting the address of a struct member instead of its value. Anyway &(a->B) is the correct precedence, not (&a)->B.

P.S. I seem to have found your problem:

Code:

@004010EF>u,1 004010EF E9 68FFFFFF jmp 0040105C @004010EF>s @0040105C>gr MSVCRT.scanf l 3 r Name: longcat 100000 @0040109F>u,15 0040109F FF75 00 push dword ptr ss:[ebp] 004010A2 53 push ebx 004010A3 68 30104000 push 00401030 004010A8 FFD7 call edi 004010AA 53 push ebx 004010AB FF15 0C104000 call dword ptr ds:[MSVCRT.free] 004010B1 68 18104000 push 00401018 004010B6 FFD7 call edi 004010B8 8D4424 3F lea eax, dword ptr ss:[esp+3F] 004010BC 50 push eax 004010BD 68 14104000 push 00401014 004010C2 FFD6 call esi 004010C4 83C4 34 add esp, 34 004010C7 807C24 13 6E cmp byte ptr ss:[esp+13], 6E 004010CC 74 1B jz 004010E9 @0040109F>g 4010b8 longcat is 100000 years old Another entry?[y/n] @004010B8>s @004010BC>ra EAX : 0063FE3B @004010BC>d 63fe3b l 1 0063FE30 79 y @004010BC>s @004010BD>s @004010C2>p @004010C4>d 63fe3b l 1 0063FE30 0A . @004010C4>q

The carriage return from the previous input was not picked up by a previous scanf() so it becomes the input of this.

Re: [c] segmentation fault- request help

Eek.. doesn't look/sound like something i can comprehend for now, thanks though :p

Have to put C aside for a bit... got "algorithms and datastructures" exam approaching.. gonna be reading material for it