Re: [PATCH] DWARF 5 support: Handle line table and file indexes

[Simon Marchi <simon.marchi@polymtl.ca>]

On 2019-10-18 3:31 p.m., Ali Tamur wrote:
> I've done everything asked, except this bit:
>> I'd still suggest adding a gdb_assert to verify that in the case of 
> DWARF <= 4, index should be > 0.
> 
> This caused failure in multiple tests, I think the correct behaviour is to
> return nullptr when DWARF <= 4 and index = 0 (done).

Oh, right, requesting the entry 0 isn't invalid, it's just that it represents the
compilation directory, which is not explicitly represented.  We indeed return
NULL at the moment when requesting index 0 (it underflows the unsigned variable),
so you are right that we should continue doing that.

> @@ -20651,12 +20689,11 @@ dwarf_decode_line_header (sect_offset sect_off, struct dwarf2_cu *cu)
>     Returns NULL if FILE_INDEX should be ignored, i.e., it is pst->filename.  */
>  
>  static const char *
> -psymtab_include_file_name (const struct line_header *lh, int file_index,
> +psymtab_include_file_name (const struct line_header *lh, const file_entry &fe,
>  			   const struct partial_symtab *pst,
>  			   const char *comp_dir,
>  			   gdb::unique_xmalloc_ptr<char> *name_holder)

The comment above this function says:

  Return the file name of the psymtab for included file FILE_INDEX

It would need to be updated.

> @@ -24190,6 +24221,14 @@ dwarf_alloc_die (struct dwarf2_cu *cu, int num_attrs)
>  \f
>  /* Macro support.  */
>  
> +static bool
> +is_valid_file_index (int file_index, struct line_header *lh)
> +{
> +  if (lh->version >= 5)
> +    return 0 <= file_index && file_index < lh->file_names_size ();
> +  return 1 <= file_index && file_index <= lh->file_names_size ();
> +}

I think it would make sense if this was a method of line_header, since it's related to
the line_header::file_name_at method.

Simon