This is the mail archive of the gdb-patches@sourceware.org mailing list for the GDB project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [PATCH 6/8] Use BLOCK_ENTRY_PC to find function entry pc in infrun.c


On 2018-06-26 02:53 AM, Kevin Buettner wrote:
> find_pc_partial_function() still returns the lowest and highest address
> of a function even when that function contains non-contiguous ranges.
> In cases where greater discrimination is required, the BLOCK parameter
> may be examined to determine the actual entry PC.
> 
> This patch uses the BLOCK return value from find_pc_partial_function()
> to obtain the entry PC. If no block is found - which can happen when
> only minimal symbols are available - then the start address provided
> by find_pc_partial_function() will still be used (as before).

This one looks good to me, given how the stop_func_start field seems to
be used.  From what I understand, it is used as the entry point of the
function, therefore BLOCK_ENTRY_PC seems appropriate.

> @@ -4297,10 +4297,20 @@ fill_in_stop_func (struct gdbarch *gdbarch,
>  {
>    if (!ecs->stop_func_filled_in)
>      {
> +      const struct block *block;
> +
>        /* Don't care about return value; stop_func_start and stop_func_name
>  	 will both be 0 if it doesn't work.  */
>        find_pc_partial_function (stop_pc, &ecs->stop_func_name,
> -				&ecs->stop_func_start, &ecs->stop_func_end);
> +				&ecs->stop_func_start, &ecs->stop_func_end,
> +				&block);
> +
> +      /* If a block is returned, prefer the block's entry point instead of
> +         the lowest address of the block - these aren't necessarily the
> +	 same.  */
> +      if (block)

block != nullptr

Simon


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]