[PATCH 09/13] [gdb/symtab] Keep track of processed DIEs in shard

Tom de Vries tdevries@suse.de
Fri Dec 8 12:09:05 GMT 2023


On 10/2/23 14:50, Tom de Vries via Gdb-patches wrote:
>     while (info_ptr < end_ptr)
>       {
>         sect_offset this_die = (sect_offset) (info_ptr - reader->buffer);
> +
> +      {
> +	CORE_ADDR end_prev_die
> +	  = form_addr (this_die - 1, reader->cu->per_cu->is_dwz,
> +		       reader->cu->per_cu->is_debug_types);
> +	set_parent_valid (start_cu, end_prev_die);
> +      }
> +

This turned out to cause a performance degradation of 30%, and to be 
unnecessary, so I dropped this bit.

Thanks,
- Tom


More information about the Gdb-patches mailing list