Add handle of timeout in map_reply_ie/1

[vkatsuba]

Fixing for crash:

error: crasher: initial call: ergw_context:'-port_message/2-fun-0-'/0, pid: <0.25280.231>, registered_name: [], error: {function_clause,[{gtp_v2_c,map_reply_i
e,[timeout],[{file,"/build/apps/ergw_core/src/gtp_v2_c.erl"},{line,325}]},{gtp_v2_c,map_reply_ies,1,[{file,"/build/apps/ergw_core/src/gtp_v2_c.erl"},{line,323}]},{gtp_v2_c,build_response,1,[{file,"/build/a
pps/ergw_core/src/gtp_v2_c.erl"},{line,103}]},{gtp_context,generic_error,3,[{file,"/build/apps/ergw_core/src/gtp_context.erl"},{line,741}]},{proc_lib,init_p,3,[{file,"proc_lib.erl"},{line,211}]}]}, ancesto
rs: [<0.1522.0>,ergw_socket_sup,ergw_core_sup,<0.1270.0>], message_queue_len: 0, messages: [], links: [], dictionary: [], trap_exit: false, status: running, heap_size: 610, stack_size: 28, reductions: 222;

[RoadRunnr]

NO, timeout should never arrive as cause code in that place.

[vkatsuba]

NO, timeout should never arrive as cause code in that place.

The jobs:ask/1 can return errors {error, rejected | timeout} / obs_server_tests.erl#L219. The rejected is handled in gtp_v1_c.erl#L317 and in gtp_v2_c.erl#L337 the same as provided in current PR. As possible solution we can try handle timeout in ergw_context.erl#L102 and provide our version/name of error(eg rate_limit as the reason to handle in the gtp handlers), but in any case the places for handle it will be the same. Could you please confirm that - if this would be an acceptable approach?

[vkatsuba]

Related to my last comment, was provided new implementation. @RoadRunnr please re-review.