| Age | Commit message (Collapse) | Author |
|---|
|
Tidy up vwm exit handling and stop leaking an X Region on every clean vwm_composite_paint_all().
If you noticed very long-running vwm instances hogging memory like a web browser, this is the culprit, vwm doesn't typically need much memory.
|
|
When vwm_composite_paint_all() short-circuited, occluded wasn't destroyed.
Defer the occluded region create to its time of need, which is after the
short-circuit, and followed immediately by its destruction.
|
|
This moves the console teardown back to vwm.c, trivial cleanup.
|
|
it's not being used for anything as of now.
|
|
|
|
|
|
readdir_r() has been deprecated in glibc
|
|
|
|
Sometimes `mplayer -fs` would result in an unmanaged window. It seems to be
due to an unexpected ordering of events on the window:
create notify 0x1000001
creating 0x1000001
map request 0x1000001
managing 0x1000001
configure request 0x1000001
unmap notify 0x1000001 from configure 0
unmanaging 0x1000001
configure notify 0x1000001
map notify 0x1000001 <----- this happens after the window has been unmanaged!
configure notify 0x1000001
configure request 0x1000001
configure request 0x1000001
configure request 0x1000001
configure request 0x1000001
configure notify 0x1000001
So in handling MapNotify, if the window is !managed && !override_redirect,
manage it.
This is confirmed to fix the occasionally unmanaged `mplayer -fs` window.
|
|
This was a known bug, there's a TODO sitting right there noting it.
The items array was sized very large so it never triggered and was
forgotten about.
Running `make tags` in the linux kernel source steps on it though,
because it constructs a massive argv.
This just adds a bounds check so no crash occurs in argv2xtext().
I don't see the point of allocating memory for this as the TODO's
suggested, since any such argv is unlikely to fit in the overlay
anyways.
Also shrunk the max from 1024 to 512, which is still quite large.
|
|
key: decrement key_is_grabbed release of multi-Alt
|
|
Upon releasing all keys concluding a grab the counter gets reset so this isn't
generally observed as being a problem.
Sometimes the second Alt is released by itself, restoring the origin, and the
original grabbing single Alt persists for subsequent window management
operations. It's in this situation when the bug manifests. If the final Alt
release occurred with a focused window/desktop differing from the origin, on
the final Alt release an unexpected restore to the origin occurred.
Usually this goes unnoticed, because typically the lone Alt release occurs
immediately following the other one, and the second restore to origin happens
to be idempotent.
|
|
House cleaning
|
|
If vmon hasn't seen any heirarchical changes, and we aren't forcing a redraw,
and the process information being shown isn't changed, don't bother
re-rendering the same thing the overlay already contains.
This can be further improved, like if only wchan changed and nothing else, only
redraw the wchan column for the relevant row. It gets tricky quickly though,
because a new wchan could be wider than the column currently permits for
example, then we'd need to go render all the rows to the new wchan column
width...
It's simpler to just redraw it all if anything has changed, and this stuff
doesn't generally change that frequently in practice so it's pretty effective
as-is.
|
|
The samplers may set these, but we need to clear them on every vmon_sample().
|
|
|
|
Set on window resize, clear on draw_overlay() return.
Used in combination with sample_interval check to gate HZ redraw.
Will be used to gate redraw of process monitors heirarchy as well,
in a subsequent commit.
|
|
Also moved vertical graph bars drawing to helper function
|
|
We currently only use the width, and XTextExtents does substantially more crap
per character in Xorg's xlib. XTextWidth is not just a wrapper around it, it's
a specialized subset implementation.
|
|
Trying this out now that there's a pile of files... sigh.
Note this spuriously duplicates list.h @ src/libvmon/list.h, the old Makefile
shared list.h between vwm and libvmon, but I'm letting them have their own
instances with autotools. Libvmon was always an independent project I just
pulled in for vwm's use, and will likely continue to be developed independent
of vwm with occasional syncs.
|
|
Long overdue house cleaning.
The addition of compositing/monitoring overlays in vwm3 pushed vwm well past
what is a reasonable size for a simple thousand line file. This is a first
step towards restoring sanity in the code, but no behavioral differences are
intended, this is mostly just shuffling around and organizing code.
I expect some performance regressions initially, follow-on commits will make
more improvements to that end as the dust settles.
|
|
I've warmed up to space after control flow keywords
|
|
Approximately:
s/if(/if (/g
s/switch(/switch (/g
s/do(/do (/g
s/while(/while (/g
s/for(/for (/g
Macro use continues to be spaceless, even when used as control flow primitives
(i.e. list_for_each()), as do function calls.
|
|
Fix handling of non-visible focus switches
|
|
It's inappropriate to assume the window being focused is in the visible
context. This caused the input focus to spuriously vanish if a window on
another desktop went away causing the next one, also invisible, to get focused.
I'm surprised how long I went without ever noticing this bug! My workflow
basically never has windows vanishing that aren't in the focused context.
|
|
This can just be derived from the state of the basis window, I think originally
I had intended to implement context switching via this function.
Also changes the implementation to stop comparing against
focused_(shelf,desktop) which prevented doing transitions on invisible
contexts. It's not usually an issue, but windows can go away at any time, and
we must be able to focus next when they do, even if their context isn't the
visible one.
|
|
Renaming to better describe what this actually checks.
Initially I was leaning towards adding a data structure for actual visibility
checks rather than just assuming all vwm-mapped windows are visible. But
there's a need for knowing just if windows are currently mapped as well, so
this function can stay implemented this way with the appropriate name.
|
|
I didn't experience any bug in particular which brought this to my attention,
but it's an obvious omission, though apparently quite ignorable.
|
|
|
|
|
|
|
|
|
|
See changes to README, but basically if you're holding down Alt and doing
some window management operations, while still pressing Alt hit the other
Alt key to return focus to the starting window+desktop. Nothing is undone,
focus is simply returned to the starting window+desktop.
|
|
|
|
Automatically "autoconf" screen-sized new windows as "allscreened"
|
|
This enables automagic mplayer -fs borderless fullscreened playback windows
with minor additions to the code. Previously one had to focus the playback
window and mod1-triple-k it to lose the borders, now it "just works".
|
|
|
|
Using bash was a bit heavy-handed when all I really want is a /bin/sh -c
just like popen()/system(). On some systems this won't make any
difference, but on Debian /bin/sh -> dash.
|
|
Major changes from vwm2 include:
- Introduction of integrated X client process and descendants
monitoring in the form of per-window composited overlays.
The rendering is done using the Composite, Damage, and Render
extensions. When the monitors are visible, vwm3 is a compositing
manager. When invisible, vwm3 continues to be an immediate-mode
classic X11 minimalist window manager.
Monitors are toggled via Mod1-;, Mod1-LeftArrow and Mod1-RightArrow
may be used to decrease and increase the sampling frequency,
respectively. Mod1-' clears the monitored tasks history for the
focused window when monitoring is visible.
This feature depends on the CONFIG_CHECKPOINT_RESTORE kernel
configuration option for the discovery of descendant processes.
Without this kernel option enabled, you'll only get a single process
monitored per window; the X client process as indicated by the
_NET_WM_PID atom of the window.
A library called libvmon has been introduced for the abstraction of
lightweight system and process statistics sampling.
Since vwm2 received backported features unrelated to monitoring or
compositing while vwm3 was developed, there isn't really much
difference between the two outside the monitoring context.
This isn't to say there isn't much activity in the code, the addition
of compositing and monitoring requires a substantial amount of code
relative to the scale of vwm[12].
|
|
Major changes from vwm1:
- GNU screen-based "console" integration for monitored launching of X
clients via screen remote commands, replacing the simple double fork
approach used in vwm1. Clients exiting with non-zero status retain
their screen window in the console for 86400 seconds, facilitating
easier debugging and troubleshooting. The console xterm is accessed in
the shelf and has a red border by default.
- Xinerama/multihead support backported from vwm3, including the "screen
fencing" implementation facilitating screen-oriented window focus
cycling.
Shifting the Mod1-Tab window cycling focuses the next most recently
used window on another display. Unshifted stays confined to the
current display.
- SYNC extension integration for prioritizing the WM over other X clients
- setpriority() integration for "nicing" X client processes relative to
the WM process
- "autoconf" windows, horizontal/vertical halfscreen windows,
quarterscreen windows in addition to the full/all screen functions.
Mod1-[ and Mod1-] resize the focused window vertically with left and
right justification to half the screen in width. Shifting these does
the same thing just horizontally. Repeating the operation with a
second ] or [ press quarters the window in the respective screen
corner, extending upon the repeater pattern established in vwm1 for
full/allscreen windows with Mod1-k[k[k]]
- Exit now requires 3 consecutive strikes of Mod1-Esc
- Introduction of a README file
|
|
- Basic window management, window resizing, virtual desktops and shelf
implementation
- Simple double fork execution of launched clients
|
