(faq.programming.win32-api): -mwindows does include -lcomdlg32.
(faq.programming.win32-no-cygwin): Update for removal of -mno-cygwin.
* overview2.sgml (ov-ex-win): Ditto.
on-disk device and we were stating it.
* path.cc (path_conv::check): Set PATH_KEPT_HANDLE.
* path.h (path_types): Add PATH_KEPT_HANDLE.
(path_conv::kept_handle): Implement.
* dtable.cc (dtable::fixup_after_exec): Specifically reject releasing
nohandle() type fhandlers rather than relying on contents of io_handle.
(dtable::fixup_after_fork): Ditto.
* fhandler_termios.cc: Add includes necessary for definitions of have_execed
and have_execed_cygwin.
(fhandler_termios::sigflush): Don't flush console input when we've started a
non-Cygwin process.
* devices.in (exists_console): Use fhandler_console::exists () rather than raw
test.
* devices.cc: Regenerate.
* fhandler.h (fhandler_console::exists): Define new function.
* fhandler_console.cc (fhandler_console::need_invisible): Use
fhandler_console::exists () rather than raw test.
* spawn.cc: Rename assitant to assistant throughout.
(child_info_spawn::worker): Simplify test for when to start a non-Cygwin
process in its own process group. Just do it whenever we start a non-Cygwin
process.
The common sim code has slightly unfinished support for these already,
but even arch ports are unable to handle these if the common header does
not define them. This is because the generated callback header includes
simple common gdb/sim headers only which causes it to skip the new ARGV
syscalls. Plus, it isn't like providing these in the common header will
break any sim targets which don't want them.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
* elf.c (_bfd_elf_map_sections_to_segments): Set INFO->user_phdrs.
* elf-nacl.c (nacl_modify_segment_map): Do nothing if INFO->user_phdrs.
(nacl_modify_program_headers): Likewise.
include/
* bfdlink.h (struct bfd_link_info): Add new member user_phdrs.
ENOENT if non-existent file got opened for reading only. Explain why.
* path.cc (path_conv::check): Stick to ENOENT if file has been opened
for informational purposes only. Add to comment.
(getpass): Rework to use /dev/tty in the first place. Install
_getpass_close_fd as pthread cleanup handler. Flush prompt explicitely.
Lock input and switch off signal input handling when reading password.
(dev_on_fs): Set bit size.
(is_dev_resident): Define new method.
* devices.in: Use different method for labelling requested name for dev
storage: prefix it with a ='. Use an actual ':' to denote an internal device.
(exists_console): Avoid repeated test.
(exists_ntdev_silent): Rename from the less-descriptive exists_ntdevs.
(dev_dev_storage): Delete unneeded define.
* gendevices: Add a flag to each line denoting whether the device lives in /dev
space.
(generate): Handle special ":" case here for internal devices. Parse
=something_dev as requiring a storage location. Tack the rest of the line back
onto the reformulated "$rest".
* devices.cc: Regenerate.
* fhandler_dev.cc (fhandler_dev::readdir): Set st_ino to device number.
* syscalls.cc (fhandler_base::stat_fixup): Ditto. Fix link count for
CD-ROM devices. Fix typo in comment.
fhandler_base::set_ino_and_dev.
* syscalls.cc (fhandler_base::stat_fixup): Ditto. Accommodate name
change throughout. Fixup link count of console devices.
(DEV_CYGDRIVE_MAJOR): Remove.
(iscygdrive_dev): Define.
* dtable.cc (fh_alloc): Accommodate above change.
* path.cc (path_conv::check): Use iscygdrive_dev to check for cygdrive.
* syscalls.cc (fhandler_base::set_ino_and_dev): Drop check for
DEV_CYGDRIVE_MAJOR from conditional.
* syscalls.cc (fhandler_base::set_ino_and_dev): New method to set
st_ino, st_dev, and st_rdev for devices in a single spot.
(fstat64): Call fhandler_base::set_ino_and_dev instead of setting
st_ino, st_dev, and st_rdev here.
(stat_worker): Ditto.