Google Answers Logo
View Question
 
Q: Port C main() function to Perl ( Answered,   2 Comments )
Question  
Subject: Port C main() function to Perl
Category: Computers > Programming
Asked by: russ_man-ga
List Price: $200.00
Posted: 14 Aug 2003 11:28 PDT
Expires: 13 Sep 2003 11:28 PDT
Question ID: 244781
Ok. Below is a C main() function that I need ported to Perl. I'm using
the Inline::C perl module so I can call the C subroutines from perl.
This shouldn't be to hard for *you* Perl Gurus :-)

typedef unsigned char u8;
typedef unsigned short u16;
typedef signed short i16;
typedef unsigned int u32;


int main(int argc, const char *argv[])
{
        u8 buf[0x20];
        int fd;
        off_t fp=0xF0000;
        const char *devmem="/dev/mem";
 
        if(sizeof(u8)!=1 || sizeof(u16)!=2 || sizeof(u32)!=4 ||
'\0'!=0)
        {
                fprintf(stderr,"%s: compiler incompatibility\n",
argv[0]);
                exit(255);
        }
 
        if(argc>=2)
                devmem=argv[1];
        if((fd=open(devmem, O_RDONLY))==-1 || lseek(fd, fp,
SEEK_SET)==-1)
        {
                perror(devmem);
                exit(1);
        }
 
        printf("# dmidecode %s\n", VERSION);
        while(fp<0xFFFFF)
        {
                if(myread(fd, buf, 0x10, devmem)==-1)
                        exit(1);
                fp+=16;

                if(memcmp(buf, "_SM_", 4)==0 && fp<0xFFFFF)
                {
                        if(myread(fd, buf+0x10, 0x10, devmem)==-1)
                                exit(1);
                        fp+=16;
 
                        if(checksum(buf, buf[0x05])
                         && memcmp(buf+0x10, "_DMI_", 5)==0
                         && checksum(buf+0x10, 0x0F))
                        {
 
                                printf("SMBIOS %u.%u present.\n",
buf[0x06], buf[0x07]);
 
                                dmi_table(fd, DWORD(buf+0x18),
WORD(buf+0x16), WORD(buf+0x1C),
                                        (buf[0x06]<<8)+buf[0x07],
argv[0], devmem);
 
                                /* dmi_table moved us far away */
                                lseek(fd, fp, SEEK_SET);
                        }
                }
                else if(memcmp(buf, "_DMI_", 5)==0
                 && checksum(buf, 0x0F))
                {
                        printf("Legacy DMI %u.%u present.\n",
                                buf[0x0E]>>4, buf[0x0E]&0x0F);
                        dmi_table(fd, DWORD(buf+0x08), WORD(buf+0x06),
WORD(buf+0x0C),
                               
((buf[0x0E]&0xF0)<<4)+(buf[0x0E]&0x0F), argv[0], devmem);
 
                        /* dmi_table moved us far away */
                        lseek(fd, fp, SEEK_SET);
                }
        }
 
        if(close(fd)==-1)
        {
                perror(devmem);
                exit(1);
        }

        return 0;
}


-Russell

Request for Question Clarification by studboy-ga on 14 Aug 2003 21:13 PDT
Hi,

Can you give me the #include at the top of this main/C file?
Also, are you running on Linux or Solaris?  I need a way to test this.

Thanks

Clarification of Question by russ_man-ga on 18 Aug 2003 11:36 PDT
Sorry for not responding right a-way I had problems with my google
account.  Anyway I'm running RedHat 9.0 and the headers are below.

#include <sys/types.h>
#include <sys/stat.h>
#include <stdio.h>
#include <unistd.h>
#include <fcntl.h>
#include <string.h>
#include <stdlib.h>
#include <errno.h>

Clarification of Question by russ_man-ga on 25 Aug 2003 07:33 PDT
Hello, 

The reason I want this ported is because it will be the first step in
creating a free CAPN module DMI::Decode. I would like to use the
module like so:

$dmi = new DMI::Decode;

$serial_number  = $dmi->serial_number();

etc…..


The problem is that I don’t know C well enough to port it to perl. I
thinking I can cut & paste the C subroutines and use Inline::C to call
them.  So their for I only need the main() function to be in perl.
Please, if you fill like this will not work or if there is a better
way I really appreciate any input.

Thanks,

-Russell

Request for Question Clarification by jeanluis-ga on 25 Aug 2003 17:10 PDT
I see... well I don't think you will be able to use the program as you
intend, without heavy modification to the C code. Here is how it
works:
It opens /dev/mem, and reads until it finds either SMBIOS, or legacy
DMI data. It then calls the dmi_table() which parses that data, and
prints it out in english as it goes.

So in order to get what you want, you will have to do one of two
things to the C code.
1) Modify it so it parses the whole SMBIOS (or DMI) and only returns
what you want (i.e. $dmi->serial_number parses the whole SMBIOS but
only returns the serial number, igoring almost everything, this is
computationally instensive), OR
2) Modify it so it parses the whole SMBIOS (or DMI) at init time and
stores all the data in a big fat easy to understand data structure,
then when you call a function like $dmi->serial_number it just goes to
your data struct and retuns the requested element.

In either case you will have to change the nature of the C program,
because the way the C code is setup it just prints out whatever it
comes accross, it is not really setup to return some value that you
want to know...

This means that I don't think that porting just the main function to
perl is going to be of much help... It may be a good starting point
for you to get your bearings in C->perl porting... So it might be of
some use in that respect... but it seems like it is only a small step
in the direction you want to go.

Of course I assume you are using the same version of dmidecode as I
am, I have version 2.2... Which version are you using?

p.s. still working on the port, didn't have much time over the
weekend... But I have some time tonight, we will see how far I get...
Talk to you later.
--jld

Request for Question Clarification by jeanluis-ga on 25 Aug 2003 17:14 PDT
Another thing I want to mention, I may change some of the interfaces
in the dmidecode, I assume this is ok. I will post whatever changes I
make.

p.s. It is my opinion that the dmidecode package was not written with
very good coding conventions... Just wanted to throw that out there...
:D
--jld

Clarification of Question by russ_man-ga on 26 Aug 2003 08:01 PDT
Hello Jeanluis,

I’m using the same version of dmidecode that you are.  This is just a
quest that I’m on to create a CPAN module that doesn’t exist yet.  And
yes porting the main() function is just a start. However by having you
help me get stated hopefully I’ll be able to port the rest of the C
subroutines to perl or use Inline::C modify the C subroutines so that
it stores the information in a structure  then I can access the
elements that way.


Thanks
Russell Pettway
Answer  
Subject: Re: Port C main() function to Perl
Answered By: jeanluis-ga on 26 Aug 2003 16:19 PDT
 
OK, Here is the function in perl, it ain't the best lookin code I have
ever seen... Tho perl is rarely good looking to me, I had to do some
things to get it to work, some of which are not great... For example
when the dmi_table function is called I didn't use the DWORD and WORD
macros that take into account endianess, you might want to write a
function that performs the  operation of those macros. Another thing
is that I couldn't seem to get perl to work with u8, u16, and u32...
So C functions that I called directly from perl I had to hack a little
to get them to work (ie I had to change u16 to int, and u8 to char
etc... it also didnt seem to like unsigned...) In other words perl
considers scalars signed, and if I tried to call a function that takes
an unsigned int as a param then perl wouldn't let me pass a scalar
into that param, I had to make it signed.... which kind of sucks,
maybe you can find a way around this problem... This means that I had
to change the interface to dmi_table, it is included below, if you
just copy and paste the code below into a perl script and try to run
it as is you will notice that it will fail. You will have to comment
out the dmi_table function that is in the dmidecode.c file to get it
to work. Anyhow below is all of the code I had to have in my perl
file... Try and get it to run, dont forget to comment out dmi_table in
the dmidecode.c file, and fixup the -I include paths on the first line
of the script.
Let me know how it goes, also sorry about #including .c files... I
hate that... but it made this little project a bit easier...Also
please excuse all lines that got split in half

#!/usr/bin/perl -I<basedir>/Inline-0.44
-I<basedir>/dl/Inline-0.44/blib/lib/

use Inline C;

dmimain();

sub dmimain
{ 
   $fp = 0xF0000;
   $devmem = "/dev/mem";

   if ($#ARGV == 0)
   {
      $devmem = $ARGV[0];
   }
   $fd = c_open($devmem);
   if ($fd == -1) 
   {
      die "Unable to open $devmem\n";
   }
   if (c_lseek($fd, $fp) == -1)
   {
      die "Unable to lseek $devmem\n";
   }
   print_version();
   while ($fp < 0xFFFFF)
   {
      $buf1 = q_myread($fd, 0x10, $devmem);
      $fp += 0x10;
      if (($buf1 =~ /^_SM_/) && ($fp < 0xFFFFF))
      {
         @spbuf = split(/ */,$buf1);
         for ($i = 0; $i < 0x10; $i++) 
         {
            vec($vecbuf, $i, 8) = ord($spbuf[$i]);
         }
         for ($i = 0; $i < 0x10; $i++)
         {
            vec($vecbuf, $i+0x10, 8) = ord(q_myread($fd, 1, $devmem));
            $buf2 .= chr(vec($vecbuf, $i+0x10,8));
         }
         $fp += 0x10;
         if ((q_checksum($vecbuf, ord ($spbuf[5])))
         && ($buf2 =~ /_DMI_/))
         {
#should really use DWORD, and WORD macros here!!!
            $base = (vec($vecbuf, 0x1b, 8) << 24) + (vec($vecbuf,
0x1a, 8) << 16) + (vec($vecbuf, 0x19, 8) << 8) + (vec($vecbuf, 0x18,
8));
            $len = (vec($vecbuf, 0x17, 8) << 8) + (vec($vecbuf, 0x16,
8));
            $num = (vec($vecbuf, 0x1d, 8) << 8) + (vec($vecbuf, 0x1c,
8));
            $ver = (vec($vecbuf, 0x06, 8) << 8) + (vec($vecbuf, 0x07,
8));
            printf("SMBIOS %u.%u present.\n", ord($spbuf[0x06]),
ord($spbuf[0x07]));
            dmi_table($fd, $base, $len, $num, $ver, "dmidecode",
$devmem);
            c_lseek($fd, $fp);
         }
         else 
         {
            if (($buf1 =~ /_DMI_/) && (q_checksum($vecbuf, 0xF))) 
            {
                             $base = (vec($vecbuf, 0x0b, 8) << 24) +
(vec($vecbuf, 0x0a, 8) << 16) + (vec($vecbuf, 0x09, 8) << 8) +
(vec($vecbuf, 0x08, 8));
               $len = (vec($vecbuf, 0x07, 8) << 8) + (vec($vecbuf,
0x06, 8));
               $num = (vec($vecbuf, 0x0d, 8) << 8) + (vec($vecbuf,
0x0c, 8));
               $ver = ((vec($vecbuf, 0x0e, 8) & 0xF0) << 4) +
(vec($vecbuf, 0x0e, 8) & 0x0F);
               printf("Legacy DMI %u.%u present.\n", ord($spbuf[0x0E])
>> 4, ord($spbuf[0x0E]) & 0xF);
               dmi_table($fd, $base, $len, $num, $ver, "dmidecode",
$devmem);
               c_lseek($fd, $fp);

            }
         }
      }
   }
   close(FD);
   return 0; 
} 

#here is the C code I had to write/modify to get this to work:

           __END__
           __C__


#include "dmidecode.c"
#include "util.c"

#define JLD_MAX_BUF_SIZE     0x40
static char buf[JLD_MAX_BUF_SIZE];

char* q_myread(int fd, size_t count, const char* prefix) 
{
	ssize_t r=1; 
	size_t r2=0;
	
	if (JLD_MAX_BUF_SIZE < count) 
	{
		fprintf(stderr, "count is too large, please increase
JLD_MAX_BUF_SIZE to %x\n", count+1);
		exit (1);
	}
   memset(buf, 0, JLD_MAX_BUF_SIZE);
	while(r2!=count && r!=0)
	{
		r=read(fd, buf+r2, count-r2);
		if(r==-1)
		{
			if(errno!=EINTR)
			{
				close(fd);
				perror(prefix);
				exit(1);
			}
		}
		else 
		{
			r2+=r;
		}
	}
	
	if(r2!=count)
	{
		close(fd);
		fprintf(stderr, "%s: Unexpected end of file\n", prefix);
		exit(1);
	}
	return buf;
}

char q_get_data(const char* data, int i)
{
   int q;
   for (q = 0; q < 0x10; q++) {
      printf("* %x %c\n", data[q]);
   }
   return data[i];
}

void print_data(const char* d)
{
   int i;
   for (i = 0; i < 0x20; i++) {
      printf("%x %x %c\n", i, (unsigned char) d[i], d[i]);
   }
}

int c_open(char* filename) 
{
   return open(filename, O_RDONLY);
}

int c_lseek(int fd, int fp)
{
   return lseek(fd, fp, SEEK_SET);
}

int c_close(int fd)
{
   return close(fd);
}

void print_version() 
{ 
   printf("# dmidecode %s -- ported by jeanluis-ga\n", VERSION); 
}

int q_checksum(const char *buf, size_t len)
{
	u8 sum=0;
	size_t a;
	for(a=0; a<len; a++)
   {
		sum+=buf[a];
   }
	return (sum==0);
}


void dmi_table(int fd, long base, int len, int num, int ver, const
char *pname, const char *devmem)
{
	u8 *buf;
	u8 *data;
	int i=0;
#ifdef USE_MMAP
	div_t mmbase;
	size_t psize;
	void *mmp;
#endif /* USE_MMAP */
	
	printf("%u structures occupying %u bytes.\n",
		num, len);
	printf("Table at 0x%08X.\n",
		base);
#ifdef USE_MMAP
	psize=getpagesize();
	mmbase=div(base, psize);
	/*
	 * We need PROT_WRITE for ASCII filtering in dmi_string. Do NOT
change MAP_PRIVATE
	 * to MAP_SHARED unless you also remove PROT_WRITE and disable ASCII
filtering, because
	 * we DO NOT WANT to write to /dev/mem. Thanks a lot to Gabriel Forte
for helping
	 * me hunting down an annoying bug related to this.
	 */
	mmp=mmap(0, mmbase.rem+len, PROT_READ|PROT_WRITE, MAP_PRIVATE, fd,
mmbase.quot*psize);
    if(mmp==MAP_FAILED)
    {
       perror("mmap");
       return;
    }
	buf=(u8 *)mmp+mmbase.rem;
#else /* USE_MMAP */
	if((buf=malloc(len))==NULL)
	{
		perror(pname);
		return;
	}
	if(lseek(fd, (off_t)base, SEEK_SET)==-1)
	{
		perror(devmem);
		return;
	}
	if(myread(fd, buf, len, devmem)==-1)
	{
		free(buf);
		printf("Table is unreachable, sorry. Try compiling dmidecode with
-DUSE_MMAP.\n"
			"This problem is known on the IBM T23, T30 and X30 laptops, the
Fujitsu-Siemens\n"
			"S-4582 laptop as well as IA-64 systems. If your system differ,
please report\n");
		exit(1);
	}
#endif /* USE_MMAP */
	
	data=buf;
	while(i<num && data+sizeof(struct dmi_header)<=buf+len)
	{
		u8 *next;
		struct dmi_header *h=(struct dmi_header *)data;
		
		printf("Handle 0x%04X\n\tDMI type %d, %d bytes.\n",
			HANDLE(h), h->type, h->length);
		
		/* look for the next handle */
		next=data+h->length;
		while(next-buf+1<len && (next[0]!=0 || next[1]!=0))
			next++;
		next+=2;
		if(next-buf<=len)
			dmi_decode(data, ver);
		else
			printf("\t<TRUNCATED>\n");
		
		data=next;
		i++;
	}
	
	if(i!=num)
		printf("Wrong DMI structures count: %d announced, only %d
decoded.\n",
			num, i);
	if(data-buf!=len)
		printf("Wrong DMI structures length: %d bytes announced, structures
occupy %d bytes.\n",
			len, (unsigned int)(data-buf));
	
#ifdef USE_MMAP
	if(munmap(mmp, mmbase.rem+len)==-1)
		perror("munmap");
#else /* USE_MMAP */
	free(buf);
#endif /* USE_MMAP */
}

Please try this out, and if you have ANY comments, questions,
concerns, hate mail :D, or whatever just let me know, I will try and
respond quickly...
Thanks,
 --jld

Request for Answer Clarification by russ_man-ga on 27 Aug 2003 08:40 PDT
Hi jeanluis,

Looks great! I cut & pasted the code into a file called dmidecode.pl
in the source directory of the dmidecode-2.2x and ran the script and
got the following error:

No source code in DATA section for Inline 'C' section.
 
 at /usr/lib/perl5/5.8.0/i386-linux-thread-multi/Socket.pm line 0
INIT failed--call queue aborted, <DATA> line 1.

And I'm not sure what the paths should be for this:
-I<basedir>/Inline -I<basedir>/dl/Inline-0.44/blib/lib/

-russell

Clarification of Answer by jeanluis-ga on 27 Aug 2003 09:32 PDT
Hard to say from here, I am at work now and I don't have access to
linux at the moment... BUT the two problems you have may be related.
The basedir should be the path to the directory where the perl inline
lib is. I just noticed that I made a small mistake in the post when I
swapped out my path with <basedir>. It should be:
-I<basedir>/Inline-0.44 -I<basedir>/Inline-0.44/blib/lib/

Hopefully that works better for you... Both <basedirs> should be the
same, and the result is that the first -I should point to the
directory with the inline.pm file. And the second -I should point to
the Inline .al files (going from memory here)...

In fact when I installed the Inline module I think I had to run a
script or a makefile to create the blib/lib directory... You may need
to do that also...
Again this is all from memory, and if I am dead wrong just let me
know... I can clear this up when I get home for sure...

Anyhow my hope is that once you get the include paths correct the
problem in Socket.pm will also be solved...
:D
thanks,
--jld

Request for Answer Clarification by russ_man-ga on 27 Aug 2003 11:26 PDT
My Inline.pm module is in the following directory:

/usr/lib/perl5/site_perl/5.8.0 

And the .al files I found in:

/usr/lib/perl5/site_perl/5.8.0/auto/Inline

I plugged this in and I still get the same error.

When you get home, if you could send me the paths that you used, that
would be great! I'm ready to close this out so you can get payed, I
just wanted to see it run first.

Thanks,
-Russell

Clarification of Answer by jeanluis-ga on 27 Aug 2003 18:23 PDT
The path I used was /home/chrisd/dl/Inline-0.44...
I installed the Inline module myself in my home directory, but based
on the paths you gave me it seems that you need to use the following
include dir:
-I/usr/lib/perl5/site_perl/5.8.0

However I think that path is prob already included... So you prob
don't need to add it... Which sucks... Can you send me your directory
listing for /usr/lib/perl5/5.8.0/i386-linux-thread-multi/Socket.pm
I want to see if we have the same revs of that lib...

[/usr/lib/perl5/5.8.0/i386-linux-thread-multi]: ls -la Socket.pm 
-r--r--r--    1 root     root         9793 Sep  6  2002 Socket.pm

Sorry it took so long to get back to you, the g/f made me go out to
dinner tonight... :D Anyhow we will get this figured out somehow...
Thanks,
--jld

Request for Answer Clarification by russ_man-ga on 28 Aug 2003 07:46 PDT
No problem, I'm married so I completely understand. I hardly ever get
to play! Anyhow I't looks like I'm running a newer version see below:

# cd /usr/lib/perl5/5.8.0/i386-linux-thread-multi
# ls -la Socket.pm
-r--r--r--    1 root     root         9793 Feb 18  2003 Socket.pm

-Russell

Clarification of Answer by jeanluis-ga on 28 Aug 2003 08:06 PDT
Just as I suspected... Again I am at work, but as soon as I get home
tonight, I will look around and see if I can figure out what has
changed in the Socket.pm module....

hang tight...
--jld

Clarification of Answer by jeanluis-ga on 28 Aug 2003 15:33 PDT
Hmmm there isn't much on this problem on the web...
The only thing I could find that is close is the following:

http://www.mail-archive.com/inline%40perl.org/msg00776.html

Which doesn't even resolve the issue... This calls for drastic
measures..

We are going to have to debug this thing... There are a number of
things we can try.... first thing I can think of is for you to try and
use my Socket.pm file, my advice is the following:
1) BACKUP YOUR Socket.pm file!!!
2) Then copy/paste the following test into a new file called Socket.pm
and put in the same location as your existing Socket.pm...
3) Then cross your fingers and run it...

If this fails (with the same error) then I would try to write a
*simple* perl script that attempts to use the socket module... If that
fails with the same error then we will have to drink a beer, and
really hunker down.

So let me know the results, if this fixes the problem then GREAT! If
it fixes the problem but a new problem arises then let me know and we
will keep on keeping on...

Everything after the line that says package Socket; is my Socket.pm
file (Including the line with 1; at the VERY end)

package Socket;

our($VERSION, @ISA, @EXPORT, @EXPORT_OK, %EXPORT_TAGS);
$VERSION = "1.75";

=head1 NAME

Socket, sockaddr_in, sockaddr_un, inet_aton, inet_ntoa - load the C
socket.h defines and structure manipulators

=head1 SYNOPSIS

    use Socket;

    $proto = getprotobyname('udp');
    socket(Socket_Handle, PF_INET, SOCK_DGRAM, $proto);
    $iaddr = gethostbyname('hishost.com');
    $port = getservbyname('time', 'udp');
    $sin = sockaddr_in($port, $iaddr);
    send(Socket_Handle, 0, 0, $sin);

    $proto = getprotobyname('tcp');
    socket(Socket_Handle, PF_INET, SOCK_STREAM, $proto);
    $port = getservbyname('smtp', 'tcp');
    $sin = sockaddr_in($port,inet_aton("127.1"));
    $sin = sockaddr_in(7,inet_aton("localhost"));
    $sin = sockaddr_in(7,INADDR_LOOPBACK);
    connect(Socket_Handle,$sin);

    ($port, $iaddr) = sockaddr_in(getpeername(Socket_Handle));
    $peer_host = gethostbyaddr($iaddr, AF_INET);
    $peer_addr = inet_ntoa($iaddr);

    $proto = getprotobyname('tcp');
    socket(Socket_Handle, PF_UNIX, SOCK_STREAM, $proto);
    unlink('/tmp/usock');
    $sun = sockaddr_un('/tmp/usock');
    connect(Socket_Handle,$sun);

=head1 DESCRIPTION

This module is just a translation of the C F<socket.h> file.
Unlike the old mechanism of requiring a translated F<socket.ph>
file, this uses the B<h2xs> program (see the Perl source distribution)
and your native C compiler.  This means that it has a 
far more likely chance of getting the numbers right.  This includes
all of the commonly used pound-defines like AF_INET, SOCK_STREAM, etc.

Also, some common socket "newline" constants are provided: the
constants C<CR>, C<LF>, and C<CRLF>, as well as C<$CR>, C<$LF>, and
C<$CRLF>, which map to C<\015>, C<\012>, and C<\015\012>.  If you do
not want to use the literal characters in your programs, then use
the constants provided here.  They are not exported by default, but
can
be imported individually, and with the C<:crlf> export tag:

    use Socket qw(:DEFAULT :crlf);

In addition, some structure manipulation functions are available:

=over 4

=item inet_aton HOSTNAME

Takes a string giving the name of a host, and translates that to an
opaque string (if programming in C, struct in_addr). Takes arguments
of both the 'rtfm.mit.edu' type and '18.181.0.24'. If the host name
cannot be resolved, returns undef.  For multi-homed hosts (hosts with
more than one address), the first address found is returned.

For portability do not assume that the result of inet_aton() is 32
bits wide, in other words, that it would contain only the IPv4 address
in network order.

=item inet_ntoa IP_ADDRESS

Takes a string (an opaque string as returned by inet_aton(),
or a v-string representing the four octets of the IPv4 address in
network order) and translates it into a string of the form 'd.d.d.d'
where the 'd's are numbers less than 256 (the normal human-readable
four dotted number notation for Internet addresses).

=item INADDR_ANY

Note: does not return a number, but a packed string.

Returns the 4-byte wildcard ip address which specifies any
of the hosts ip addresses.  (A particular machine can have
more than one ip address, each address corresponding to
a particular network interface. This wildcard address
allows you to bind to all of them simultaneously.)
Normally equivalent to inet_aton('0.0.0.0').

=item INADDR_BROADCAST

Note: does not return a number, but a packed string.

Returns the 4-byte 'this-lan' ip broadcast address.
This can be useful for some protocols to solicit information
from all servers on the same LAN cable.
Normally equivalent to inet_aton('255.255.255.255').

=item INADDR_LOOPBACK

Note - does not return a number.

Returns the 4-byte loopback address.  Normally equivalent
to inet_aton('localhost').

=item INADDR_NONE

Note - does not return a number.

Returns the 4-byte 'invalid' ip address.  Normally equivalent
to inet_aton('255.255.255.255').

=item sockaddr_family SOCKADDR

Takes a sockaddr structure (as returned by pack_sockaddr_in(),
pack_sockaddr_un() or the perl builtin functions getsockname() and
getpeername()) and returns the address family tag.  It will match the
constant AF_INET for a sockaddr_in and AF_UNIX for a sockaddr_un.  It
can be used to figure out what unpacker to use for a sockaddr of
unknown type.

=item sockaddr_in PORT, ADDRESS

=item sockaddr_in SOCKADDR_IN

In a list context, unpacks its SOCKADDR_IN argument and returns an
array
consisting of (PORT, ADDRESS).  In a scalar context, packs its (PORT,
ADDRESS) arguments as a SOCKADDR_IN and returns it.  If this is
confusing,
use pack_sockaddr_in() and unpack_sockaddr_in() explicitly.

=item pack_sockaddr_in PORT, IP_ADDRESS

Takes two arguments, a port number and an opaque string, IP_ADDRESS
(as returned by inet_aton(), or a v-string).  Returns the sockaddr_in
structure with those arguments packed in with AF_INET filled in.  For
Internet domain sockets, this structure is normally what you need for
the arguments in bind(), connect(), and send(), and is also returned
by getpeername(), getsockname() and recv().

=item unpack_sockaddr_in SOCKADDR_IN

Takes a sockaddr_in structure (as returned by pack_sockaddr_in()) and
returns an array of two elements: the port and an opaque string
representing the IP address (you can use inet_ntoa() to convert the
address to the four-dotted numeric format).  Will croak if the
structure does not have AF_INET in the right place.

=item sockaddr_un PATHNAME

=item sockaddr_un SOCKADDR_UN

In a list context, unpacks its SOCKADDR_UN argument and returns an
array
consisting of (PATHNAME).  In a scalar context, packs its PATHNAME
arguments as a SOCKADDR_UN and returns it.  If this is confusing, use
pack_sockaddr_un() and unpack_sockaddr_un() explicitly.
These are only supported if your system has E<lt>F<sys/un.h>E<gt>.

=item pack_sockaddr_un PATH

Takes one argument, a pathname. Returns the sockaddr_un structure with
that path packed in with AF_UNIX filled in. For unix domain sockets,
this
structure is normally what you need for the arguments in bind(),
connect(), and send(), and is also returned by getpeername(),
getsockname() and recv().

=item unpack_sockaddr_un SOCKADDR_UN

Takes a sockaddr_un structure (as returned by pack_sockaddr_un())
and returns the pathname.  Will croak if the structure does not
have AF_UNIX in the right place.

=back

=cut

use Carp;
use warnings::register;

require Exporter;
use XSLoader ();
@ISA = qw(Exporter);
@EXPORT = qw(
	inet_aton inet_ntoa
	sockaddr_family
	pack_sockaddr_in unpack_sockaddr_in
	pack_sockaddr_un unpack_sockaddr_un
	sockaddr_in sockaddr_un
	INADDR_ANY INADDR_BROADCAST INADDR_LOOPBACK INADDR_NONE
	AF_802
	AF_AAL
	AF_APPLETALK
	AF_CCITT
	AF_CHAOS
	AF_CTF
	AF_DATAKIT
	AF_DECnet
	AF_DLI
	AF_ECMA
	AF_GOSIP
	AF_HYLINK
	AF_IMPLINK
	AF_INET
	AF_INET6
	AF_ISO
	AF_KEY
	AF_LAST
	AF_LAT
	AF_LINK
	AF_MAX
	AF_NBS
	AF_NIT
	AF_NS
	AF_OSI
	AF_OSINET
	AF_PUP
	AF_ROUTE
	AF_SNA
	AF_UNIX
	AF_UNSPEC
	AF_USER
	AF_WAN
	AF_X25
	IOV_MAX
	MSG_BCAST
	MSG_BTAG
	MSG_CTLFLAGS
	MSG_CTLIGNORE
	MSG_CTRUNC
	MSG_DONTROUTE
	MSG_DONTWAIT
	MSG_EOF
	MSG_EOR
	MSG_ERRQUEUE
	MSG_ETAG
	MSG_FIN
	MSG_MAXIOVLEN
	MSG_MCAST
	MSG_NOSIGNAL
	MSG_OOB
	MSG_PEEK
	MSG_PROXY
	MSG_RST
	MSG_SYN
	MSG_TRUNC
	MSG_URG
	MSG_WAITALL
	MSG_WIRE
	PF_802
	PF_AAL
	PF_APPLETALK
	PF_CCITT
	PF_CHAOS
	PF_CTF
	PF_DATAKIT
	PF_DECnet
	PF_DLI
	PF_ECMA
	PF_GOSIP
	PF_HYLINK
	PF_IMPLINK
	PF_INET
	PF_INET6
	PF_ISO
	PF_KEY
	PF_LAST
	PF_LAT
	PF_LINK
	PF_MAX
	PF_NBS
	PF_NIT
	PF_NS
	PF_OSI
	PF_OSINET
	PF_PUP
	PF_ROUTE
	PF_SNA
	PF_UNIX
	PF_UNSPEC
	PF_USER
	PF_WAN
	PF_X25
	SCM_CONNECT
	SCM_CREDENTIALS
	SCM_CREDS
	SCM_RIGHTS
	SCM_TIMESTAMP
	SHUT_RD
	SHUT_RDWR
	SHUT_WR
	SOCK_DGRAM
	SOCK_RAW
	SOCK_RDM
	SOCK_SEQPACKET
	SOCK_STREAM
	SOL_SOCKET
	SOMAXCONN
	SO_ACCEPTCONN
	SO_ATTACH_FILTER
	SO_BACKLOG
	SO_BROADCAST
	SO_CHAMELEON
	SO_DEBUG
	SO_DETACH_FILTER
	SO_DGRAM_ERRIND
	SO_DONTLINGER
	SO_DONTROUTE
	SO_ERROR
	SO_FAMILY
	SO_KEEPALIVE
	SO_LINGER
	SO_OOBINLINE
	SO_PASSCRED
	SO_PASSIFNAME
	SO_PEERCRED
	SO_PROTOCOL
	SO_PROTOTYPE
	SO_RCVBUF
	SO_RCVLOWAT
	SO_RCVTIMEO
	SO_REUSEADDR
	SO_REUSEPORT
	SO_SECURITY_AUTHENTICATION
	SO_SECURITY_ENCRYPTION_NETWORK
	SO_SECURITY_ENCRYPTION_TRANSPORT
	SO_SNDBUF
	SO_SNDLOWAT
	SO_SNDTIMEO
	SO_STATE
	SO_TYPE
	SO_USELOOPBACK
	SO_XOPEN
	SO_XSE
	UIO_MAXIOV
);
@EXPORT_OK = qw(CR LF CRLF $CR $LF $CRLF

	       IPPROTO_TCP
	       TCP_KEEPALIVE
	       TCP_MAXRT
	       TCP_MAXSEG
	       TCP_NODELAY
	       TCP_STDURG);

%EXPORT_TAGS = (
    crlf    => [qw(CR LF CRLF $CR $LF $CRLF)],
    all     => [@EXPORT, @EXPORT_OK],
);

BEGIN {
    sub CR   () {"\015"}
    sub LF   () {"\012"}
    sub CRLF () {"\015\012"}
}

*CR   = \CR();
*LF   = \LF();
*CRLF = \CRLF();

sub sockaddr_in {
    if (@_ == 6 && !wantarray) { # perl5.001m compat; use this && die
	my($af, $port, @quad) = @_;
	warnings::warn "6-ARG sockaddr_in call is deprecated" 
	    if warnings::enabled();
	pack_sockaddr_in($port, inet_aton(join('.', @quad)));
    } elsif (wantarray) {
	croak "usage:   (port,iaddr) = sockaddr_in(sin_sv)" unless @_ == 1;
        unpack_sockaddr_in(@_);
    } else {
	croak "usage:   sin_sv = sockaddr_in(port,iaddr))" unless @_ == 2;
        pack_sockaddr_in(@_);
    }
}

sub sockaddr_un {
    if (wantarray) {
	croak "usage:   (filename) = sockaddr_un(sun_sv)" unless @_ == 1;
        unpack_sockaddr_un(@_);
    } else {
	croak "usage:   sun_sv = sockaddr_un(filename)" unless @_ == 1;
        pack_sockaddr_un(@_);
    }
}

sub AUTOLOAD {
    my($constname);
    ($constname = $AUTOLOAD) =~ s/.*:://;
    croak "&Socket::constant not defined" if $constname eq 'constant';
    my ($error, $val) = constant($constname);
    if ($error) {
	croak $error;
    }
    *$AUTOLOAD = sub { $val };
    goto &$AUTOLOAD;
}

XSLoader::load 'Socket', $VERSION;

1;

Request for Answer Clarification by russ_man-ga on 29 Aug 2003 08:55 PDT
Nope, I still get the same error; If you could remove the follwoing
line:

#!/usr/bin/perl -I<basedir>/Inline-0.44
-I<basedir>/dl/Inline-0.44/blib/lib/

And see if you get the same error. This will help to determin if my
paths are correct. For sum reason I think it's Inline.

thanks,
-Russell

Clarification of Answer by jeanluis-ga on 29 Aug 2003 09:25 PDT
Hmm, I tend to agree. I can try this tonight when I get home, I am not
sure if I was clear in one of my last posts. I basically said that I
don't think you need any -I paths on the #!perl line... So if you
haven't removed those two paths competely from your script then you
should do so... I don't think it will have any impact on the issue for
you however... I can try this when I get home but I don't have inline
installed on my machine, I had to intall it myself... Of course this
means that you could try and install Inline 0.44 on your machine and
try that out... You can download it here:
"cpan Inline 0.44"
http://cpan.org/authors/id/I/IN/INGY/Inline-0.44.tar.gz

Just unzip/tar it to your home dir, and then run the Makefile.pl, then
I think you need to make it... that will create the blib/lib directory
and fill it up with all the correct files... Then you will need to try
and run this guy with the infamous -I's from above...

When I get home I will look into this further... Let me know if you
make any progress...
Thanks for hangin in there,
--jld

Clarification of Answer by jeanluis-ga on 30 Aug 2003 11:03 PDT
Hello again, sorry it took so long to get back to you, my internet
gateway got nailed by that dang virus that everyone loves so much
these days, anyhow I am back online now and ready to play.... As you
requested I removed the -I's from my script and it didn't run...
because I don't have the Inline libs installed in the main perl
path... I am running Mandrake Linux which is a little different than
RedHat (Even tho it uses RH as a baseline...) Anyhow have you had any
success installing Inline 0.44 and running with that?

Keep up in the loop, I want to get this guy figured out..
Thanks,
    --jld

Request for Answer Clarification by russ_man-ga on 02 Sep 2003 07:47 PDT
Sorry, I didn’t think about till after I posted it. But your right it
should be in my @INC… Well not sure what’s going on. I’ll try some
simple scripts to confirm my configuration is working. I’ll post the
results in a little while.

Thanks 
Russell

Clarification of Answer by jeanluis-ga on 03 Sep 2003 14:48 PDT
Any luck yet?
Let me know how things are going. I just want to be sure we have this
thing figured out.
Thanks,
--jld

Request for Answer Clarification by russ_man-ga on 04 Sep 2003 15:28 PDT
Sorry, I just got back from Vacation yesterday, It was one of those
deals were I wasn’t going and at the last minute I went. So anyhow I
basically cut and pasted Inline C code from the Cook-book just to be
sure my Inline in working properly. All went well so I’m a little
confused as to what the problem really is.

-Russell

Clarification of Answer by jeanluis-ga on 04 Sep 2003 16:10 PDT
It could be the case that the code you tested with didn't use the
Socket.pm module... You might want to use that module and actually
make a call into it at the very least... If that works, but my code
still doesn't work you might want to try and install Inline 0.44 as I
suggested above.

Let me know how things go.
--jld

Request for Answer Clarification by russ_man-ga on 05 Sep 2003 07:52 PDT
Ok I tried both things. I extracted the Inline-044 to my home
directory and setup the -I includes. Second I tested the Socket.pm
module with the following code:


#tcp-client
( $them, $port ) = @ARGV;

$port = 23 unless $port;
$them = 'localhost' unless $them;
$AF_INET = 2;
$SOCK_STREAM = 1;

$SIG{'INT'} = 'dokill';
sub dokill {
    kill 9,$child if $child;
}

$sockaddr = 'S n a4 x8';

#chop($hostname = `hostname`);
($name,$aliases,$proto) = getprotobyname('tcp');
($name,$aliases,$port) = getservbyname($port,'tcp')
    unless $port =~ /^\d+$/;;
($name,$aliases,$type,$len,$thisaddr) =
	gethostbyname($hostname);
($name,$aliases,$type,$len,$thataddr) = gethostbyname($them);

$this = pack($sockaddr, $AF_INET, 0, $thisaddr);
$that = pack($sockaddr, $AF_INET, $port, $thataddr);

if (socket(S, $AF_INET, $SOCK_STREAM, $proto)) {
    print "socket ok\n";
}
else {
    die $!;
}

if (bind(S, $this)) {
    print "bind ok\n";
}
else {
    die $!;
}

if (connect(S,$that)) {
    print "connect ok\n";
}
else {
    die $!;
}

select(S); $| = 1; select(STDOUT);
$a=<S>;print "$a";
while( $b=<STDIN> ) {
print S "$b\n";
$a=<S>;print "$a";
}
exit 1 ;


According to the code above my Socket.pm module is working fine. But I
still get the same exact error. And extracting Inline and setting up
the -I includes still didn't work either.

-Russell

Clarification of Answer by jeanluis-ga on 05 Sep 2003 08:36 PDT
Ok, I have one question right now:
Are you ***100%*** sure that you actually ran with Inline 0.44 and not
the default one installed on your system.

The 1st thing I would do is verify that you were running Inline 0.44
from your home directory.

let me know when you are 100% sure of this
thanks,
  --jld

Request for Answer Clarification by russ_man-ga on 08 Sep 2003 07:16 PDT
The line I used is this. 

#!/usr/bin/perl -I /home/russell/Inline-0.44  -I /home/russell/Inline-0.44/blib/lib

This is after I ran perl Makefile.PL and make.

I think that it's using the right Inline.

Thanks,
Russell Pettway

Clarification of Answer by jeanluis-ga on 08 Sep 2003 17:51 PDT
Ok... So then have you tried it with Inline-0.44 and the Socket.pm
file posted above?

--jld

Clarification of Answer by jeanluis-ga on 08 Sep 2003 19:07 PDT
Ok, here is a new direction.... If my last suggestion doesn't work...

First off: your include path only needs the following:
-I /home/russell/Inline-0.44/blib/lib

you should remove the other include path from your #! line for this to
even have a chance at working...

Next open the file /home/russell/Inline-0.44/blib/lib/Inline.pm in
your favorite text editor.

Do a search for "Socket" in the Inline.pm file.
You should get 3 hits on the word Socket, the first one is
require Socket;

Just above the read_DATA function add these 2 lines of code:
sub CR   () {"\015"}
sub LF   () {"\012"}

Then ERASE the line
require Socket;

a few lines down, you will find the line:
	my ($CR, $LF) = (&Socket::CR, &Socket::LF);

You want to replace that line with the following line:
	my ($CR, $LF) = (CR, LF);

These few changes basically remove the Socket dependancy from
Inline...

Let me know how this works for you... I am starting to run out of
tricks so hopefully this works...

Let me know,
thanks --jld

Request for Answer Clarification by russ_man-ga on 09 Sep 2003 07:27 PDT
Ok, I put...

#!/usr/bin/perl -I /home/russell/Inline-0.44/blib/lib at the top of
the file.

I got the same error.

So I made the changes to Inline.pm like you said.

The same error.

So I did perl Makefile.PL and then make in the Inline directory.

Still the same error.

I'm going to close this out so you can get paid if you don't have
anymore suggestions.

Thanks,
Russell

Clarification of Answer by jeanluis-ga on 09 Sep 2003 08:30 PDT
Hmm... Just try one more thing to put my mind at ease... open the file
/home/russell/Inline-0.44/blib/lib/Inline.pm in a text editor and put
in some syntax errors... Like
askljdhsdlkjhsdaflkjsdfhlsdakjfh on a line of code (not in the
comments). Then run your script with only the -I
/home/russell/Inline-0.44/blib/lib include.

If your script fails because of the inserted syntax error, then ok I
don't know where to go. But if the script fails for the same old
reason then it is attempting to use the Inline module from your system
include. Which means that you need to either remove (temporarly) the
Inline.pm module in your system path. Or figure out some other way to
run the Inline.pm module in your home blib dir.

I haven't given up yet... :p
--jld

Request for Answer Clarification by russ_man-ga on 09 Sep 2003 11:15 PDT
Well I removed Inline from @INC. So it is now using
/home/russell/Inline-0.44/blib/lib

Now I get a differant error: 

# ./dmidecode.pl
No source code in DATA section for Inline 'C' section.
 
 at ./dmidecode.pl line 0
INIT failed--call queue aborted, <DATA> line 1.

Russell

Clarification of Answer by jeanluis-ga on 09 Sep 2003 15:50 PDT
OK! Finally some progress! Now in your system include path there
should also be a denter.pm file... It seems like this file may also
need to be temp removed from your system include path...

Another thing to try is the following in your dmidecode.pl script the
first few lines are as follows (with line numbers inserted):
1 #!/...perl...
2 
3 use Inline C;
4 
5 dmimain();

Try to insert the following code on line number 4 if the above does
not work:
Inline->init();

Sorry this is taking so long, it is damn hard debugging from this
vantage point... Glad you are hangin in there w/ me... Let me know how
this works out for you...

Thanks,
--jld

Request for Answer Clarification by russ_man-ga on 11 Sep 2003 11:00 PDT
I'm getting the same error:


No source code in DATA section for Inline 'C' section.
 
 at dmidecode.pl line 0
INIT failed--call queue aborted, <DATA> line 1.


thanks,
-russell

Request for Answer Clarification by russ_man-ga on 11 Sep 2003 11:07 PDT
I changed 

use Inline C; 

to 

use Inline C => 'END';

And it starts to compile but dies the error is below.

/usr/bin/perl /usr/lib/perl5/5.8.0/ExtUtils/xsubpp  -typemap
/usr/lib/perl5/5.8.0/ExtUtils/typemap   dmidecode_pl_b1a3.xs >
dmidecode_pl_b1a3.xsc && mv dmidecode_pl_b1a3.xsc dmidecode_pl_b1a3.c
gcc -c  -I/root/dmidecode-2.2 -D_REENTRANT -D_GNU_SOURCE
-DTHREADS_HAVE_PIDS -DDEBUGGING -fno-strict-aliasing
-I/usr/local/include -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64
-I/usr/include/gdbm -O2 -march=i386 -mcpu=i686 -g   -DVERSION=\"0.00\"
-DXS_VERSION=\"0.00\" -fPIC
"-I/usr/lib/perl5/5.8.0/i386-linux-thread-multi/CORE"  
dmidecode_pl_b1a3.c
dmidecode_pl_b1a3.c:19: syntax error before "void"
make: *** [dmidecode_pl_b1a3.o] Error 1

This came from the out.make file.

thanks,
-russell

Clarification of Answer by jeanluis-ga on 12 Sep 2003 18:10 PDT
OK... In order to debug this one I will need your help... 
You need to goto the dir you were building in and open the following
file in a text editor:
_Inline/build/dmidecode_pl_b1a3/dmidecode_pl_b1a3.c

Please post the first 30 lines of that file here...

Basically the problem is that there is some kind of syntax error in
the C code... and the way to figure out the problem is to look at the
C code that gcc is tring to compile.

let me know
--jld
Comments  
Subject: Re: Port C main() function to Perl
From: kveldulf-ga on 21 Aug 2003 23:01 PDT
 
A few questions:

Does this have to be a literal port? or can it just probe for the
device? The reason I ask, is that perl doesn't have anything like the
sizeof operator.

Where does the myread function come from?  I don't think is part of
the Standard lib.
Subject: Re: Port C main() function to Perl
From: jeanluis-ga on 22 Aug 2003 17:32 PDT
 
I will port this to perl in the next few days as I get time... maybe
monday, but I also run Linux, and have the dmi source code that you
got your function from. I don't know if I have the same version that
you have, but I don't think that matters, because the dmi lib compiles
and runs with the function you provided (I checked)... The only
question I have is why you would want this function in perl?
Regardless of your reasons, I will post an answer within a week...
Thanks,
   --jld

Important Disclaimer: Answers and comments provided on Google Answers are general information, and are not intended to substitute for informed professional medical, psychiatric, psychological, tax, legal, investment, accounting, or other professional advice. Google does not endorse, and expressly disclaims liability for any product, manufacturer, distributor, service or service provider mentioned or any opinion expressed in answers or comments. Please read carefully the Google Answers Terms of Service.

If you feel that you have found inappropriate content, please let us know by emailing us at answers-support@google.com with the question ID listed above. Thank you.
Search Google Answers for
Google Answers  


Google Home - Answers FAQ - Terms of Service - Privacy Policy