Bug in xsym.c ?

Bug in xsym.c ?

[Nick Clifton]

Hi Klee,

  I ran across a problem compiling bfd/xsym.c on a Solaris platform
  the other day.  The function bfd_sym_fetch_type_information_table_ent()
  was testing a variable called 'index' which is not defined.  It was
  not clear from the context what the test was supposed to be doing,
  but my best guess was that it was intended to check that a parameter
  called 'entry' was non-NULL.

  I created a patch like this.  Is it correct ?

Cheers
        Nick


Index: bfd/xsym.c
===================================================================
RCS file: /cvs/src/src/bfd/xsym.c,v
retrieving revision 1.3
diff -c -3 -p -w -r1.3 xsym.c
*** bfd/xsym.c	30 Nov 2002 08:39:40 -0000	1.3
--- bfd/xsym.c	5 Mar 2003 08:59:48 -0000
*************** bfd_sym_fetch_type_information_table_ent
*** 1131,1137 ****
    BFD_ASSERT (bfd_sym_valid (abfd));
    sdata = abfd->tdata.sym_data;
  
!   if (index == 0)
      return -1;
  
    if (bfd_seek (abfd, offset, SEEK_SET) < 0)
--- 1131,1137 ----
    BFD_ASSERT (bfd_sym_valid (abfd));
    sdata = abfd->tdata.sym_data;
  
!   if (entry == 0)
      return -1;
  
    if (bfd_seek (abfd, offset, SEEK_SET) < 0)

Re: Bug in xsym.c ?

[Elias Athanasopoulos]

Hi Nick, 

On Wed, Mar 05, 2003 at 09:05:25AM +0000, Nick Clifton wrote:
> Hi Klee,
> 
>   I ran across a problem compiling bfd/xsym.c on a Solaris platform
>   the other day.  The function bfd_sym_fetch_type_information_table_ent()
>   was testing a variable called 'index' which is not defined.  It was
>   not clear from the context what the test was supposed to be doing,
>   but my best guess was that it was intended to check that a parameter
>   called 'entry' was non-NULL.
> 
>   I created a patch like this.  Is it correct ?
>   
> !   if (entry == 0)
>       return -1;
>   
>     if (bfd_seek (abfd, offset, SEEK_SET) < 0)

I believe the test should be made against the 'offset' variable.

Although this is not a proof:

[anteater@neutrino bfd]$ grep bfd_sym_fetch_type_information_table_entry *.c
xsym.c:bfd_sym_fetch_type_information_table_entry (abfd, entry, offset)
xsym.c:  if (bfd_sym_fetch_type_information_table_entry (abfd, entry, tindex) < 0)
xsym.c:   if (bfd_sym_fetch_type_information_table_entry (abfd, &entry, index) < 0)

As I understand it, 'offset' should have been named 'index'.

Elias

-- 
University of Athens			I bet the human brain 
Physics Department				is a kludge --Marvin Minsky 

	

Re: Bug in xsym.c ?

[Nick Clifton]

Hi Elias,

> I believe the test should be made against the 'offset' variable.
> 
> Although this is not a proof:
> 
> [anteater@neutrino bfd]$ grep bfd_sym_fetch_type_information_table_entry *.c
> xsym.c:bfd_sym_fetch_type_information_table_entry (abfd, entry, offset)
> xsym.c:  if (bfd_sym_fetch_type_information_table_entry (abfd, entry, tindex) < 0)
> xsym.c:   if (bfd_sym_fetch_type_information_table_entry (abfd, &entry, index) < 0)
> 
> As I understand it, 'offset' should have been named 'index'.

Seems fair enough.  Klee has not responded, so I have checked in a
patch to rename 'entry' to 'offset'.

Cheers
         Nick

2003-03-12  Nick Clifton  <nickc@redhat.com>

	* xsym.c (bfd_sym_fetch_type_information_table_entry): Change
	'entry' to 'offset' in test for zero value.