Re: [Jilc-dev] Problem

[Gopal.V]

If memory serves me right, Gaurav Vaish wrote:
>     The new jdasm is unable to deassemble a simple file - I am attaching the
> code and bytecode here.
        Actually new jdasm is appropriate here. A small typo is what is 
causing the core dump..... I'll give you three guesses ... read the 
code below and find out on you own... ;-).      
||  OLD CODE
||  -------
||          case LABEL:
||              printf("%d;code offset ",
||                      insn_offset+(Int16)(code[start+1]<<8|code[start+2]));
||  NEW CODE
||  ---------
||              case LABEL:  PRINT_OP_NAME(found);
||                       printf("\t %s;code offset ", insn_offset + \
||                                    ( ((Int16)code[start + 1]) << 0x08 | \
||                                      ((Int16)code[start + 2]) ) );

To give a hint I call it "a format string vulnerability" , LOL.
This kind of thing happens to the best of us...

        To check this kind of error, try running my original code, that
shows a GOTO instruction where the SEGFAULT occurs. So I go and check 
the GOTO's param type "LABEL". And there you are...

<stuff>
That reminds me , ever tried 

if(NULL == 0) printf("%s\n",NULL); 

That's when you start appreciating gcc .... (will this dump a core ?)
</stuff>

Gopal.V

PS: 2:40 AM .... and I get this mail saying my code doesn't work !
    swap out of deep sleep mode to deep hack mode ....... AAH !
-- 
 The difference between insanity and genius is only measured by success
 //===<=>===\\
|| GNU RULEZ ||
 \\===<=>===//