When text-ops are registered via modules, we need an owner field so
they're correctly tracked. Hence, add this field to all text-ops and
correctly keep module references.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
All font definitions have been moved to font.h so remove the old
definitions and include font.h instead.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
The owner fields specifies the module that provides the code for the given
backend. The font-core is responsible of ref/unref'ing the module so its
code is always available as long as the module exists.
We cannot push this into the modules as modules _must_ never call
kmscon_module_unref(KMSCON_THIS_MODULE)! Because this might drop the last
reference and hence the function might disallocate the module. However,
this means that it cannot return because the caller is _part_ of the
module and so no longer valid.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Several parts of kmscon have huge external dependencies. However, we allow
them to be disabled during build-time. But this is not enough as it
requires distributions to choose which options to use. Therefore, we now
allow dynamicly loadable modules that can optionally be installed and
kmscon automatically picks them up.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Each font-backend now also gets a .finalize callback that is called when
the last font-user drops its font. This will allow modules to destroy
themself when the last reference is dropped.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
The new destroy callback is called whenever an entries last reference is
dropped. This will allow modules to keep reference of its code users.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
With modules we no longer need to conditionally load backends. Instead
each module will register the backend with the main application on load.
This will temporarily break pango, freetype2 and unifont backends as they
are not registered until we introduce full module support.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
We want to modularize the text-font library, so we need each backend as a
different library. This step links each of them into a different static
library so we can now link them either into the binary or into modules.
This step also makes 8x16 always compiled into the text-font backend so we
always have a working implementation even though no modules might be
loaded.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
We need the git-revision for module-version checks so provide the
infrastructure now and print it in log_init().
Note that the git-describe string is distributed with the tarballs so
end-users will not have to generate it themself. But when building from
git, the revision will be automatically updated whenever something
changes.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
"struct dirent" has always been a mess. As its size may differ between
systems, we need to dynamically allocate it. This helper does that for us.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Instead of implementing our own thread-safe backend-storage, we now use
the generic shl-registry object.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Instead of implementing a thread-safe backend-system we can now use the
new shl-registry storage module.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
The shl-register module is used to create registries of named-objects. The
main purpose for it will be for module-systems were we have different
backends implementing an interface. The registry can then be used to find
a specific module (or the default module).
No sophisticated filter/search mechanism is used as this cannot be done
efficiently in a generic way.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
The standard OpenGL surfaces are y-inverted (which we didn't implement
correctly previously when using our own gbm-surfaces). Therefore, we need
to flip the y-coordinates now as we use correct GL surfaces with the new
drm-uterm backend.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Until now we always created our buffers via gbm_bo_create() and used
surfaceless EGL contexts so we can schedule buffer-flips manually.
However, recent gbm and EGL mesa releases include gbm-surfaces that allow
us to ignore all this an let EGL manage the buffers.
This introduces some problems as we are no longer under control of the
buffers so we cannot know which buffer is active. That means, applications
that use the uterm-drm devices must redraw the screen on each frame.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
We used font-size=15 in the past but changed it long ago. However, the
documentation still used 15 so fix this now.
Reported-by: James Buren <ryu0@ymail.com>
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
We now set COLORTERM to "kmscon" so applications can identify their
controlling terminal. Note that COLORTERM is not used for terminfo/termcap
detection (unlike TERM).
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
We must pass "-p" to make "login" preserve the environment. We make sure
the environment is correctly reset before calling it so it is safe.
Furthermore, this allows us to pass additional environment variables to
the child without having /bin/login reset it.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
This property controls whether the environment should be reset before
spawning the child process. Defaults to "no" and affects whether "-p" is
passed as default argument to /bin/login.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
The colorterm property is used (if set) as COLORTERM environment variable.
Note that this variable (despite the name) is not used with
terminfo/termcap. Instead, it's solely used to identify the running
terminal.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
The wcwidth() POSIX helper returns the number of cells that a character
occupies. This is normally 1, however, if it is 0 we simply discard this
input (this should never happen as all characters that wcwidth() returns 0
for should already be handled by the VTE layer).
But if it is >1, we write the character in the first cell, set the width
correctly and mark all following characters as width=0.
Everything else is left unchanged, that means, the running application has
to be aware of wcwidth() and correctly send two backspaces to clear a
width=2 character and so on.
This behavior almost exactly matches the xterm behavior. There are some
corner cases like erasing only one part of a multi-cell character etc.
that might differ. However, these are non-standard cases that no
application should never rely on.
Anyway, if we spot any of these differences between xterm and kmscon, we
should fix kmscon to behave like xterm does.
Thanks to Shixin for looking this up in other emulators.
Reported-by: Shixin Zeng
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Multi-cell glyphs are glyphs that span across multiple horizontal cells.
The font renderers already support this. This patch fixes the console
renderers to take advantage of this.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Instead of clipping every character to a one-cell boundary, we now render
multi-cell characters correctly into multiple cells.
This does not adjust the console-renderers but only the font renderers to
provide the correct glyphs.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
wcwidth() is a POSIX function that returns the number of cells that a
wide-character occupies. The glibc function cannot be used as it depends
on the locale and we need _always_ UTF8 no matter what the locale is.
This implementation is provided by Markus Kuhn and is equivalent to
xterm's behavior.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
If we cannot write LED states to an input device, we should print proper
warnings instead of ignoring it.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
This option uses XDG_SEAT to determine the current seat. If this is NULL,
we fall back to "seat0".
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
kmscon does always pass a proper argv array so the default is never used.
However, for security reasons we shouldn't hardcode /bin/login. If some
other program uses the pty layer and forgets to pass in a suitable argv
all the time, they mind end up with opening a root shell.
Hence, use /bin/login as default.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
The wayland compositor support was some research project that turned out
to not work that well. The main problem is that the DRM planes support
cannot be easily forwarded to a child. Sub-surface support might change
that but until then we remove compositor support.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Instead of relying on --switchvt, we should always activate fake-VTs
during startup. Otherwise, we might end up with dead seats as there is no
way to wake up kmscon later.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
We tried to avoid running unused code during shutdown by moving VT
handling into a separate eloop and run it exclusively. However, with
delayed VT switching, we need to run other event-handlers to make the VT
switch work. Therefore, run all of this during shutdown and remove the
vt_eloop event-loop.
This means that we might create new seats, displays, etc. during shutdown.
This is valid and doesn't harm kmscon, but we want to avoid doing stuff
that is undone few miliseconds later. Therefore, we set a "exiting" flag
and check that before initiating most of the big handlers.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
We currently only start seats (switch to their VT) during startup but we
should also do that for hotplugged seats. Therefore, move that logic into
kmscon_seat and perform the startup after the seat is fully set up.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
We don't use the newly created objects so no need to return them.
Furthermore, we want the functions to already perform actions on the
objects so there is no guarantee the objects are still alive after
function return. Thus, simply return nothing.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Same as previous commit: Move uterm-monitor device-data handling into the
real device create/destroy functions to allow dynamic device removal.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
If we want to be able to destroy seats from anywhere, not just
uterm-monitor callbacks, we need to be able to reset the monitor-event
seat-data so we no longer access the seat memory.
Hence, move the seat-data set/reset logic into the actual seat
creation/destroy functions.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
We should handle everything --listen related in kmscon_main so pass in the
allowed-VTs mask from the caller.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
CAP_KILL will be documented in kmscon manuals, but we should actually
check for the capability during startup.
However, this requires linking with libcap or libcap-ng and we currently
want to avoid that for such a simple task.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Ignore SIGPIPE globally. We handle HUP correctly everywhere and this
signal is just annoying.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
If we autoselect a VT and the current VT is unused, use it. This is
similar to the behavior of Weston and X-Server.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
If we run in listen mode, we are supposed to take over a seat. If a seat
supports VTs (like kernel VTs or kmscon cdev fake VTs) we assume that
there is some manager for these VTs (the one who created them). Therefore,
there is no need to run kmscon in listen mode on these seats. Instead, you
should run kmscon in default mode on these seats.
We enforce this limitation because if the VT-master on those seats dies
and causes a HUP, they have no way to notify us when they startup again.
Therefore, this kind of setup is broken. Furthermore, no-one would every
want such setups.
Instead use the startup mechanism of the VT/seat-manager to start kmscon
in default mode on those seats.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
Instead of configuring everything for every kind of situation, we now
split kmscon into two modes: default-mode and listen-mode
In default-mode we run on the given seats once until we encounter a HUP or
until we are closed. It's a perfect replacement for agetty.
In listen-mode, we run only on non-VT seats and provide the full kmscon
functionality. We run as system daemon and wait for new seats and close
seats if they are destroyed. We simply provide a full system-console on
all seats.
Signed-off-by: David Herrmann <dh.herrmann@googlemail.com>
