Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.os.linux.misc > #13311 > unrolled thread

pcap problem

Back to article view | Back to comp.os.linux.misc

Contents

  pcap problem "Bill Cunningham" <nospam@nspam.invalid> - 2015-01-05 21:30 -0500
    Re: pcap problem John McCue <jmccue@jmclin0.bstnma.east.verizon.net> - 2015-01-06 18:09 +0000
      Re: pcap problem "Bill Cunningham" <nospam@nspam.invalid> - 2015-01-06 14:54 -0500
        Re: pcap problem Rich <rich@example.invalid> - 2015-01-06 20:05 +0000
          Re: pcap problem Richard Kettlewell <rjk@greenend.org.uk> - 2015-01-06 20:11 +0000
            Re: pcap problem Rich <rich@example.invalid> - 2015-01-06 21:22 +0000
            Re: pcap problem "Bill Cunningham" <nospam@nspam.invalid> - 2015-01-06 18:08 -0500
              Re: pcap problem Richard Kettlewell <rjk@greenend.org.uk> - 2015-01-07 09:20 +0000
        Re: pcap problem Mark Blair <mblair@quantumdata.com> - 2015-01-06 14:14 -0600
          Re: pcap problem "Bill Cunningham" <nospam@nspam.invalid> - 2015-01-06 18:05 -0500
          Re: pcap problem "Bill Cunningham" <nospam@nspam.invalid> - 2015-01-06 18:12 -0500
            Re: pcap problem Rich <rich@example.invalid> - 2015-01-06 23:37 +0000
              Re: pcap problem "Bill Cunningham" <nospam@nspam.invalid> - 2015-01-06 18:44 -0500

#13311 — pcap problem

    This is the code I have so far and I run into a segmentation fault. The 
debugger says it has something to do with "activate".

#include <stdio.h>
#include <pcap/pcap.h>
#include <stdlib.h>

int main()
{
    pcap_t *pp;
    char *perr;
    int stat;
    int datalink;
    char errbuf[PCAP_ERRBUF_SIZE];
    pp = pcap_create("handle", errbuf);
    if (pp = NULL) {
        fputs(errbuf, stderr);
        exit(1);
    }
    stat = pcap_activate(pp);
    if (stat != 0) {
        perr = pcap_geterr(pp);
        printf("%s\n", perr);
        exit(1);
    }
    datalink = pcap_datalink(pp);
    if (datalink = PCAP_ERROR_NOT_ACTIVATED) {
        fputs("not activated error\n", stderr);
        exit(1);
    }
    puts("fine");
    return 0;
}

Bill

[toc] | [next] | [standalone]

#13319

Bill Cunningham <nospam@nspam.invalid> wrote:
>    This is the code I have so far and I run into a segmentation fault. The 
> debugger says it has something to do with "activate".

Check your if statements :)
Not sure if that is your issue but best to validate the
if statements, esp the equal ones.

<snip>

John

[toc] | [prev] | [next] | [standalone]

#13322

"John McCue" <jmccue@jmclin0.bstnma.east.verizon.net> wrote in

> Check your if statements :)
> Not sure if that is your issue but best to validate the
> if statements, esp the equal ones.

    The thing is I know nothing hardly about this API. Am I even using the 
functions as they should be used. Where pointers are asked for should I be 
passing them and not what is being pointed too? What do you mean validate 
IFs ?

Bill

[toc] | [prev] | [next] | [standalone]

#13323

Bill Cunningham <nospam@nspam.invalid> wrote:

> "John McCue" <jmccue@jmclin0.bstnma.east.verizon.net> wrote in

> > Check your if statements :)
> > Not sure if that is your issue but best to validate the
> > if statements, esp the equal ones.

> ...
> What do you mean validate IFs ?

He means that you are using "assignment" operators, not "comparison"
operators, in most of your if statements.  I.e.:

       if (pp = NULL) {

This assigns NULL to the variable pp.

You likely want:

      if (pp == NULL) {

which checks pp for equality to NULL.

Note, a way to get the compiler to complain about this style for you is
to reverse the two items in the if.  Write it this way:

  if (NULL == pp) {
or
  if (5 == variable) {

Then, if you leave off an = sign, you'll get a compile error about
invalid assignment.

[toc] | [prev] | [next] | [standalone]

#13324

Rich <rich@example.invalid> writes:
> He means that you are using "assignment" operators, not "comparison"
> operators, in most of your if statements.  I.e.:
>
>        if (pp = NULL) {
>
> This assigns NULL to the variable pp.
>
> You likely want:
>
>       if (pp == NULL) {
>
> which checks pp for equality to NULL.
>
> Note, a way to get the compiler to complain about this style for you is
> to reverse the two items in the if.  Write it this way:
>
>   if (NULL == pp) {
> or
>   if (5 == variable) {
>
> Then, if you leave off an = sign, you'll get a compile error about
> invalid assignment.

Compile with ‘gcc -Wall -Wextra -Werror’ and the compiler will spot the
error automatically, without any need to write the code in a bizarre
way.

-- 
http://www.greenend.org.uk/rjk/

[toc] | [prev] | [next] | [standalone]

#13326

Richard Kettlewell <rjk@greenend.org.uk> wrote:
> Rich <rich@example.invalid> writes:
> > He means that you are using "assignment" operators, not "comparison"
> > operators, in most of your if statements.  I.e.:
> >
> >        if (pp = NULL) {
> >
> > This assigns NULL to the variable pp.
> >
> > You likely want:
> >
> >       if (pp == NULL) {
> >
> > which checks pp for equality to NULL.
> >
> > Note, a way to get the compiler to complain about this style for you is
> > to reverse the two items in the if.  Write it this way:
> >
> >   if (NULL == pp) {
> > or
> >   if (5 == variable) {
> >
> > Then, if you leave off an = sign, you'll get a compile error about
> > invalid assignment.

> Compile with ?gcc -Wall -Wextra -Werror? and the compiler will spot the
> error automatically, without any need to write the code in a bizarre
> way.

Did not know those switches would warn about that.  Thanks.

[toc] | [prev] | [next] | [standalone]

#13328

"Richard Kettlewell" <rjk@greenend.org.uk> wrote in message 
news:wwvtx03acuj.fsf@l1AntVDjLrnP7Td3DQJ8ynzIq3lJMueXf87AxnpFoA.invalid...

> Compile with 'gcc -Wall -Wextra -Werror' and the compiler will spot the
> error automatically, without any need to write the code in a bizarre
> way.

    Well I've never heard of '-Wextra'. Evidentually I need to start 
using -Wall more. I wasn't sure if this library would be on topic over in 
clc. It must be being C is concerned.

Bill

[toc] | [prev] | [next] | [standalone]

#13332

"Bill Cunningham" <nospam@nspam.invalid> writes:
> "Richard Kettlewell" <rjk@greenend.org.uk> wrote:
>> Compile with 'gcc -Wall -Wextra -Werror' and the compiler will spot
>> the error automatically, without any need to write the code in a
>> bizarre way.
>
> Well I've never heard of '-Wextra'.

I recommend spending some time reading the compiler manual section on
warnings.  (The same goes for anyone else attempting to program in C who
doesn’t know about these warnings options.)

-- 
http://www.greenend.org.uk/rjk/

[toc] | [prev] | [next] | [standalone]

#13325

On 1/6/2015 1:54 PM, Bill Cunningham wrote:
> "John McCue" <jmccue@jmclin0.bstnma.east.verizon.net> wrote in
>
>> Check your if statements :)
>> Not sure if that is your issue but best to validate the
>> if statements, esp the equal ones.
>
>      The thing is I know nothing hardly about this API. Am I even using the
> functions as they should be used. Where pointers are asked for should I be
> passing them and not what is being pointed too? What do you mean validate
> IFs ?
>
> Bill
>
>
You've been posting to comp.lang.c for years! Wasn't John McCue's clue 
enough? You're doing an *ASSIGNMENT* ("=") when you should be doing a 
boolean equals ("==") test!

[toc] | [prev] | [next] | [standalone]

#13327

"Mark Blair" <mblair@quantumdata.com> wrote in message 
news:m8hfn8$ado$1@dont-email.me...

> You've been posting to comp.lang.c for years! Wasn't John McCue's clue 
> enough? You're doing an *ASSIGNMENT* ("=") when you should be doing a 
> boolean equals ("==") test!

    You seem to be reading what you want to see in posts. Have I not said 
"It's been a while..." Have I not said before. "I have quite many things to 
do."? What are you trying to say other than you haven't been paying 
attention for years? Finally someone answers over here. Well in clc they've 
pointed things out to me and I've changed it and proceeded a bit further in 
the code.

Bill

[toc] | [prev] | [next] | [standalone]

#13329

"Mark Blair" <mblair@quantumdata.com> wrote in message 
news:m8hfn8$ado$1@dont-email.me...

> "...Wasn't John McCue's clue enough?..."

Clue? What clue? He said, "Check your if statements :)
*Not sure if that is your issue* but best to validate the
if statements, esp the equal ones."

    That sounds to me like he's a little unsure also. I wouldn't call that 
any intentional clue.

Bill

[toc] | [prev] | [next] | [standalone]

#13330

Bill Cunningham <nospam@nspam.invalid> wrote:

> "Mark Blair" <mblair@quantumdata.com> wrote in message 
> news:m8hfn8$ado$1@dont-email.me...

> > "...Wasn't John McCue's clue enough?..."

> Clue? What clue? He said, "Check your if statements :)
> *Not sure if that is your issue* but best to validate the
> if statements, esp the equal ones."

>     That sounds to me like he's a little unsure also. I wouldn't call
> that any intentional clue.

> Bill

He was trying to give you a hint (i.e., to scrutinize your if
statements) rather than handing you the answer on a silver platter.

When I first glanced at your post, I missed the assignment vs. equals
issue.  When Jon McCue's message arrived, I looked again, and
immediately saw the issue (hadn't looked closely enough for the first
glance).

[toc] | [prev] | [next] | [standalone]

#13331

"Rich" <rich@example.invalid> wrote in message 
news:m8hrk3$r1k$2@dont-email.me...

> He was trying to give you a hint (i.e., to scrutinize your if
> statements) rather than handing you the answer on a silver platter.
>
> When I first glanced at your post, I missed the assignment vs. equals
> issue.  When Jon McCue's message arrived, I looked again, and
> immediately saw the issue (hadn't looked closely enough for the first
> glance).

    Ok I probably typed one thinking it was the other. I haven't dealt with 
C for a while. The code now gets to where there's a problem with either 
pcap_activate or the "datalink" function. I am going to give this -Wall a 
try. And maybe -Wextra that's a new one on me.

Bill

[toc] | [prev] | [standalone]

Back to top | Article view | comp.os.linux.misc

csiph-web