[rtl,escalation] Fix rst_ni connection of pwrmgr's receiver

Description

The alert_handler's escalation sender rst_ni is connected to rstmgr_aon_resets.rst_lc_io_div4_n[rstmgr_pkg::Domain0Sel] in top_earlgrey, but the pwrmgr's escalation_receiver is connected to rstmgr_aon_resets.rst_lc_io_div4_n[rstmgr_pkg::DomainAonSel]. Notice they are indexed by a different domain, so they are different nets.

It so happens that the behavior of the two resets (differing in the index) is almost the same, since the reset requests from pwrmgr always drives the same value for both domains. There may be minor cycle differences, but we don't expect this to cause reset divergences that can cause incorrect behavior.

Nevertheless, topearlgrey should be fixed to connect pwrmgr's rst_esc_ni to rstmgr_aon_resets.rst_lc_io_div4_n[rstmgr_pkg::Domain0Sel]. This needs to be done via the hjson files for the top.