Improve headlessness detection for backend selection. #17396

anntzer · 2020-05-13T13:38:50Z

We currently check the $DISPLAY environment variable to autodetect
whether we should auto-pick a non-interactive backend on Linux, but that
variable can be set to an "invalid" value. A realistic use case is for
example a tmux session started interactively inheriting an initially
valid $DISPLAY, but to which one later reconnects e.g. via ssh, at which
point $DISPLAY becomes invalid.

Before this PR, something like

DISPLAY=:123 MPLBACKEND= MATPLOTLIBRC=/dev/null python -c 'import pylab'

(where we unset matplotlibrc to force backend autoselection) would crash
when we select qt and qt fails to initialize as $DISPLAY is invalid (qt
unconditionally abort()s via qFatal() in that case).
With this PR, we correctly autoselect a non-interactive backend.

Note that we were already relying on $DISPLAY being correctly set
before, and this PR also doesn't return "invalid display" if we can't
load X11, so this should not make anything worse on Wayland (at worst
we'll just fail to detect headlessness like before).

Also supporting Wayland was easy enough, so this also closes #18377.

PR Summary

PR Checklist

Has Pytest style unit tests
Code is Flake 8 compliant
New features are documented, with examples if plot related
Documentation is sphinx and numpydoc compliant
Added an entry to doc/users/next_whats_new/ if major new feature (follow instructions in README.rst there)
Documented in doc/api/api_changes.rst if API changed in a backward-incompatible way

tacaswell · 2020-05-13T17:05:03Z

lib/matplotlib/backends/backend_qt5.py

-                display = os.environ.get('DISPLAY')
-                if display is None or not re.search(r':\d', display):
-                    raise RuntimeError('Invalid DISPLAY variable')
+            if is_x11_build and matplotlib._c_internal_utils.invalid_display():


Does this check also need to land in wx/tk/gtk flavors of the backends?

Actually the important fix is in cbook._get_running_interactive_framework() because that's what'll determine fallback. The one in qt5 only affects people who directly instantiate qt canvases without going through the pyplot machinery (but then it's just unlikely that they're going through this path to create their QApplication; likely they're doing it themselves and this check doesn't help). tk and gtk throw proper exceptions when DISPLAY is invalid. wx apparently also abort()s, but as for qt I doubt that a check would really help.

mstoeckl · 2020-08-30T16:28:56Z

src/_c_internal_utils.c

+            dlsym(libwayland_client, "wl_display_connect");
+        int (* wl_display_disconnect)(struct wl_display*) =
+            dlsym(libwayland_client, "wl_display_disconnect");
+        if (wl_display_connect && wl_display_connect


Should this be wl_display_connect && wl_display_disconnect?

yup, thanks

mstoeckl · 2020-08-30T17:16:34Z

src/_c_internal_utils.c

 #ifdef _WIN32
 #include <Objbase.h>
 #include <Shobjidl.h>
 #include <Windows.h>
 #endif

+static PyObject* mpl_display_is_invalid(PyObject* module)
+{
+#ifdef __linux__


I think it would still make sense to check that environment variables DISPLAY, or WAYLAND_DISPLAY exist before hand. First, to avoid libwayland's behavior of defaulting to connect to 'wayland-0', even if no environment variables are set; second, because dlopen is very slow -- one one computer, calling dlopen followed by dlclose on libX11.so and libwayland-client.so costs between 1.4 msec (marginal cost), 2 msec (initial cost, hot cache), and 6 msec (uncached, initial cost). Importing matplotlib already costs me 240msec every script load. (Aside: It's not necessary to check env variable WAYLAND_SOCKET, because the socket named by it cannot be reused over two successive wl_display_connect calls.)

Can you check the speed of getenv()? (I genuinely do not know whether it is much faster than dlopen().)

getenv is very fast, since the environment variables are provided to a program at the start, just like the command line arguments. My benchmark of getenv uses around 0.0002 msec to call both getenv("WAYLAND_DISPLAY") and getenv("DISPLAY").

anntzer · 2020-08-31T07:42:23Z

I inverted the logic to display_is_valid as now we should(?) support all relevant cases (X and Wayland), and this behavior is much easier to document.

src/_c_internal_utils.c

We currently check the $DISPLAY environment variable to autodetect whether we should auto-pick a non-interactive backend on Linux, but that variable can be set to an "invalid" value. A realistic use case is for example a tmux session started interactively inheriting an initially valid $DISPLAY, but to which one later reconnects e.g. via ssh, at which point $DISPLAY becomes invalid. Before this PR, something like ``` DISPLAY=:123 MPLBACKEND= MATPLOTLIBRC=/dev/null python -c 'import pylab' ``` (where we unset matplotlibrc to force backend autoselection) would crash when we select qt and qt fails to initialize as $DISPLAY is invalid (qt unconditionally abort()s via qFatal() in that case). With this PR, we correctly autoselect a non-interactive backend.

anntzer · 2020-09-01T09:35:17Z

yes to all

timhoffm · 2020-10-11T10:20:22Z

src/_c_internal_utils.c

@@ -83,7 +148,7 @@ static PyMethodDef functions[] = {
     "always returns None."},
    {"Win32_SetForegroundWindow",
     (PyCFunction)mpl_SetForegroundWindow, METH_O,
-     "Win32_SetForegroundWindow(hwnd)\n--\n\n"
+     "Win32_SetForegroundWindow(hwnd, /)\n--\n\n"


What is the slash doing here?

This marks hwnd as being positional-only (https://www.python.org/dev/peps/pep-0570/), which it is (being implemented with METH_O).

Since matplotlib#17396, backends now check for Wayland settings as part of headless detection. However, the headless tests do not override those settings. When running on Wayland, they thus think the display exists when they are not supposed to.

rnhmjoj · 2021-05-25T22:35:49Z

I came here by bisecting: this commit breaks the automatic selection of an interactive backend for me.
When running matplotlib (built with either Tk, GTK3, QT5) in an X session and with a valid DISPLAY variable, the agg backend is chosen.

EDIT: It looks like the problem is in the mpl_display_is_valid function.
EDIT 2: It's the dlopen() failing because it can't find the shared object (NixOS here). I don't think there is any other way beside patching.

anntzer · 2021-07-18T17:18:59Z

Sorry for not noticing your message earlier (feel free to ping us if no one replies, sometimes no one watches old threads...). Your diagnosis seems reasonable, but is there any way to find the path to libX11.so/libwayland-client.so on nixos without linking them? attn @QuLogic, perhaps you may know a bit about this?

QuLogic · 2021-07-19T19:58:41Z

I think the point of NixOS is to hide all libraries between packages, unless they need them. So unless you can find a way mark X/Wayland as needed by Python/Matplotlib, I'm not sure what else can be done.

rnhmjoj · 2021-07-19T20:04:11Z

That's correct: there is no way to detect the library at runtime and it's the way is supposed to be.
There is something that could be done, though: if the python package were to fail at build time or signal in some way which shared libraries it needs at runtime, it could prevent this kind of problem in the future. I'm not too familiar with setuptools, so I'm not sure it can be done.

anntzer · 2021-07-28T18:07:35Z

Would it work for nixos to simply force _c_internal_utils to link with libX11.so/libwayland-client.so? (this could be done easily by patching setupext.py to add the right entry under libraries=...) I don't know think we want to do that by default (as we certainly don't want to force a dependency on either libraries on other OSes), though.

jicksaw · 2022-12-02T21:14:10Z

For any other timetravellers, the detection issue on NixOS was apparently fixed in NixOS/nixpkgs#124439. The fix requires matplotlib to be built with enableTk/Gtk/Qt.

I got here because you can't easily enable it when using poetry2nix nix-community/poetry2nix#514 (comment) and just using MPLBACKEND=tkagg had no effect. My workaround is to boldly mpl.use("tkagg")

I guess MPLBACKEND=tkagg would work if it was assumed the session is interactive when one of the interactive backends is set in it? Dunno what other effects that would have.

anntzer force-pushed the guillotine branch from 1a88ee6 to 2ad7160 Compare May 13, 2020 15:38

tacaswell added this to the v3.4.0 milestone May 13, 2020

tacaswell reviewed May 13, 2020

View reviewed changes

anntzer force-pushed the guillotine branch from 2ad7160 to 00dcbf6 Compare June 23, 2020 07:19

anntzer force-pushed the guillotine branch 2 times, most recently from 653f3d0 to a74a951 Compare August 22, 2020 09:57

anntzer mentioned this pull request Aug 30, 2020

Matplotlib picks a headless backend on Linux if Wayland is available but X11 isn't #18377

Closed

anntzer force-pushed the guillotine branch 6 times, most recently from 114975c to ab7f780 Compare August 30, 2020 14:12

mstoeckl reviewed Aug 30, 2020

View reviewed changes

anntzer force-pushed the guillotine branch 2 times, most recently from 9aa515f to 3e17318 Compare August 30, 2020 19:01

anntzer force-pushed the guillotine branch from 3e17318 to 45620ad Compare August 31, 2020 07:42

QuLogic approved these changes Sep 1, 2020

View reviewed changes

src/_c_internal_utils.c Outdated Show resolved Hide resolved

src/_c_internal_utils.c Outdated Show resolved Hide resolved

src/_c_internal_utils.c Outdated Show resolved Hide resolved

src/_c_internal_utils.c Outdated Show resolved Hide resolved

anntzer force-pushed the guillotine branch from 45620ad to 043c578 Compare September 1, 2020 09:35

timhoffm reviewed Oct 11, 2020

View reviewed changes

timhoffm approved these changes Oct 11, 2020

View reviewed changes

timhoffm merged commit 4c35d32 into matplotlib:master Oct 11, 2020

anntzer deleted the guillotine branch October 11, 2020 17:19

QuLogic mentioned this pull request Dec 9, 2020

Fix headless tests on Wayland. #19083

Merged

3 tasks

anntzer mentioned this pull request Jul 18, 2021

Reword docs about fallbacks on headless linux. #20664

Merged

7 tasks

jaicher mentioned this pull request Dec 13, 2021

Import hangs when matplotlib installed but no display available pydata/xarray#6062

Closed

hattom mentioned this pull request Jan 10, 2022

Matplotlib Agg backend easybuilders/easybuild#772

Open

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Improve headlessness detection for backend selection. #17396

Improve headlessness detection for backend selection. #17396

anntzer commented May 13, 2020 •

edited

Loading

tacaswell May 13, 2020

anntzer May 13, 2020

mstoeckl Aug 30, 2020

anntzer Aug 30, 2020

mstoeckl Aug 30, 2020 •

edited

Loading

anntzer Aug 30, 2020

mstoeckl Aug 30, 2020 •

edited

Loading

anntzer Aug 30, 2020

anntzer commented Aug 31, 2020

anntzer commented Sep 1, 2020

timhoffm Oct 11, 2020

anntzer Oct 11, 2020

rnhmjoj commented May 25, 2021 •

edited

Loading

anntzer commented Jul 18, 2021

QuLogic commented Jul 19, 2021

rnhmjoj commented Jul 19, 2021

anntzer commented Jul 28, 2021

jicksaw commented Dec 2, 2022

Improve headlessness detection for backend selection. #17396

Improve headlessness detection for backend selection. #17396

Conversation

anntzer commented May 13, 2020 • edited Loading

PR Summary

PR Checklist

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

mstoeckl Aug 30, 2020 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

mstoeckl Aug 30, 2020 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

anntzer commented Aug 31, 2020

anntzer commented Sep 1, 2020

Choose a reason for hiding this comment

Choose a reason for hiding this comment

rnhmjoj commented May 25, 2021 • edited Loading

anntzer commented Jul 18, 2021

QuLogic commented Jul 19, 2021

rnhmjoj commented Jul 19, 2021

anntzer commented Jul 28, 2021

jicksaw commented Dec 2, 2022

anntzer commented May 13, 2020 •

edited

Loading

mstoeckl Aug 30, 2020 •

edited

Loading

mstoeckl Aug 30, 2020 •

edited

Loading

rnhmjoj commented May 25, 2021 •

edited

Loading