Re: gdb/infrun.c:2674: internal-error: insert_step_resume_breakpoint:Assertion

[Andrew Cagney <cagney at gnu dot org>]

Something here doesn't add up:

> /home/olaf/gdb-head/src/gdb/infrun.c:2674: internal-error: insert_step_resume_breakpoint: Assertion `frame_id_eq (step_frame_id, get_frame_id (step_frame))' failed.
> A problem internal to GDB has been detected,

  2674    gdb_assert (frame_id_eq (step_frame_id, get_frame_id 
(step_frame)));

> #5  0x100ea3f0 in insert_step_resume_breakpoint (step_frame=0x104556c8, ecs=0x5752)
>     at /home/olaf/gdb-head/src/gdb/infrun.c:2679

  2679    gdb_assert (step_range_start == step_range_end
  2680                || (get_frame_pc (step_frame) >= step_range_start
  2681                    && get_frame_pc (step_frame) <= step_range_end));

> #6  0x100eb930 in handle_inferior_event (ecs=0xffffd240) at /home/olaf/gdb-head/src/gdb/infrun.c:1984

  1970        if (step_range_end != 0
  1971            && stop_signal != TARGET_SIGNAL_0
  1972            && stop_pc >= step_range_start && stop_pc < 
step_range_end
  1973            && frame_id_eq (get_frame_id (get_current_frame ()), 
step_frame_id))
  1974          {
  1975            /* The inferior is about to take a signal that will 
take it
  1976               out of the single step range.  Set a breakpoint at the
  1977               current PC (which is presumably where the signal 
handler
  1978               will eventually return) and then allow the inferior to
  1979               run free.
  1980
  1981               Note that this is only needed for a signal delivered
  1982               while in the single-step range.  Nested signals 
aren't a
  1983               problem as they eventually all return.  */
  1984            insert_step_resume_breakpoint (get_current_frame (), 
ecs);
  1985          }

If we trust the internal-error message, it appears that the call could 
only have been made if the failing assertion was true :-(

Andrew