[binutils-gdb] gdb: Fix remote-sim/MinGW/Darwin builds

[sergiodj+buildbot at sergiodj dot net]

*** TEST RESULTS FOR COMMIT f6cfb42730ed37bfb32cb27ef627df930f437f08 ***

Author: Pedro Alves <palves@redhat.com>
Branch: master
Commit: f6cfb42730ed37bfb32cb27ef627df930f437f08

gdb: Fix remote-sim/MinGW/Darwin builds

The recent commit e671cd59 ("Per-inferior target_terminal state, fix
PR gdb/13211, more") missed adjusting a few targets to the new
target_ops->to_interrupt interface, breaking the build for those
targets.  This fixes it.

Note: remote-sim doesn't really support async execution, so I don't
think gdbsim_interrupt is ever reached via target_interrupt.  (It is
reached via gdbsim_cntrl_c though).

The inflow.c changes are a bit ugly, but they're just doing what other
parts of the file already do to handle the same missing functions.
Targets that don't have 'kill', like mingw have their own
target_ops->to_interrupt implementation, so it's fine to make
child_interrupt be a nop.

gdb/ChangeLog:
2018-01-31  Pedro Alves  <palves@redhat.com>

	* darwin-nat.c (darwin_interrupt): Remove ptid_t parameter.
	* inflow.c (child_terminal_save_inferior): Wrap reference to
	tcgetpgrp in HAVE_TERMIOS_H.
	(child_interrupt, child_pass_ctrlc): Wrap references to signal in
	_WIN32.
	* remote-sim.c (gdbsim_interrupt): Remove ptid_t parameter and
	always iterate over all inferiors.
	(gdbsim_cntrl_c): Adjust.
	* windows-nat.c (windows_interrupt): Remove 'ptid_t' parameter.