my $c_file = "NO MATCH";

while (<$hfile>)
{  
    my $line = $_;
    if ($line =~ /[DWEM]\s+.*?\s+(?:c|cm|M)\:($input_pid)\s+([^\.]+\.c
++)/s)
    {
        $c_file = $2;
        my $pid = $1;

    }

        if($input_string eq $c_file)
        {
           #check with input string
          print $ofile "$line\n";  #write the output in a output file
        }

}
[download]

"Not tested" is, perhaps, a very kind way of putting it. You may have intended to match both the filename and the PID, but you failed to do the latter anywhere in the code. In addition, you're comparing '$c_file' against a non-existent variable called '$input_string'. Last of all, even if you had checked both the filename and the PID against reasonable values, you'd continue to print lines until you had another successful match - which may not be the right thing to do.

Update Modified phrasing.


--
"Language shapes the way we think, and determines what we can think about."
-- B. L. Whorf

You're reading your input file one line at a time, and only printing a line when your regex matches - while what you say you want to print is a paragraph ("\n\n"-delimited areas of your input.) There are ways to make the above work - i.e., you could set a "keep printing" flag as soon as your regex matches and reset it when you hit whatever your end condition is - but since your input is already split into paragraphs, it's much easier to use Perl's EOL variable, $/, to tell your script to read it that way.

#!/usr/bin/perl
use strict;
use warnings;

my $pid = "10570";
my $str = 'cm_child\.c';

{
    local $/ = "\n\n";
    while (<DATA>){
        print if /^[DWEM]\s+.*?\s+(?:cm?|M):$pid\s+$str/sm;
    }
}

__END__
D Fri Nov 21 14:09:41 2008  TME_BILLING_COE  cm:10570  cm_child.c(107)
+:4385 1:blrdxp-santbs:CustomerCenter
        XXX CMAP: op_custom() past op_decode, opcode: PCM_OP_CUST
 D Fri Nov 21 14:09:41 2008  TME_BILLING_COE  cm:10570 cm_child.c :92 
+1:blrdxp-santbs:CustomerCenter:0:AWT-EventQueue-0:83:1227256900:0
        op_cust_validate_customer input flist
# number of field entries allocated 20, used 3
0 PIN_FLD_POID           POID [0] 0.0.0.1 /plan -1 0
0 PIN_FLD_ACCOUNT_OBJ    POID [0] 0.0.0.1 /account -1 0
0 PIN_FLD_PAYINFO       ARRAY [1] allocated 20, used 5
1     PIN_FLD_NAME            STR [0] "Invoice1"

D Fri Nov 21 14:09:41 2008  TME_BILLING_COE  cm:10570  fm_cust_pol_pre
+p_payinfo.c:113 1:blrdxp-santbs:CustomerCenter:0:AWT-EventQueue-0:83:
+1227256900:0
        op_cust_pol_prep_payinfo input flist
# number of field entries allocated 20, used 5
0 PIN_FLD_POID           POID [0] 0.0.0.1 /payinfo/invoice -1 0
0 PIN_FLD_ACCOUNT_OBJ    POID [0] 0.0.0.1 /account -1 0
0 PIN_FLD_FLAGS           INT [0] 1
[download]

You also need to consider how you're going to deal with the metacharacters (such as '.') in the filename that you're using, since "cm_child.c" will happily match "cm_childXc" or "cm_child-c" when used as a part of the regex - but that's a different problem (one that can't be addressed since we don't know what the range of possible filenames is in your case.)