[Patch, fortran] PR38324 - Wrong lbound given to allocatable components

[Tobias Burnus]

Paul Richard Thomas wrote:
> As remarked in a TODO, I think that gfc_conv_expr_descriptor should be
> doing this.  However, at this stage in the proceedings, such a fix
> would be far too invasive.
>   

Should there be a tracking bug for this TODO?

Thanks for fixing the bug!

Tobias