* Linux: only disable WebKit compositing for WebKitGTK older than 2.42
The default Wayland path set WEBKIT_DISABLE_COMPOSITING_MODE=1 whenever the
session had both DISPLAY and WAYLAND_DISPLAY, i.e. any Wayland session with
XWayland. That forces WebKitGTK accelerated compositing off and also disables
threaded scrolling, so the Device tab web view (Mainsail/Fluidd) lagged unless
the app was launched with GDK_BACKEND=x11, the only path that skipped the
variable. Disabling compositing also turns off threaded scrolling.
The XWayland compositing bug the fallback worked around is fixed in WebKitGTK
>= 2.42, so apply it only for older WebKit. WebView::NeedsCompositingWorkaround()
reads the runtime WebKitGTK version; the X11 opt-in path is unchanged.
Behavior is unchanged on WebKit < 2.42, on pure X11, and on native Wayland
without XWayland. The wxWidgets skill reference is updated to match.
* Linux: use WebKit's shared-memory transport for the NVIDIA DMABUF renderer
The Device tab web view is slow on NVIDIA proprietary drivers even with
accelerated compositing left on: since the DMABUF renderer became the
default in WebKitGTK 2.42, the driver cannot back it, and web views fall
back to a slow path. Gating the old WEBKIT_DISABLE_COMPOSITING_MODE
fallback on WebKit < 2.42 is the wrong axis, because the failure is
version-independent from 2.42 on.
Force the shared-memory transport with WEBKIT_DMABUF_RENDERER_FORCE_SHM=1.
That keeps compositing and threaded scrolling and skips only the hardware
path. Also set WEBKIT_FORCE_DMABUF_RENDERER=1, which clears the isNVIDIA()
early return that Debian/Ubuntu/openSUSE patch into
rendererBufferTransportMode() and which otherwise returns before the
FORCE_SHM check is reached.
Both are set before the X11-vs-Wayland split so the workaround also covers
XWayland, and only when WebKitGTK is new enough to honour FORCE_SHM
(>= 2.44, WebView::SupportsForceSharedMemory()).
The AppImage launcher exported the superseded
WEBKIT_DISABLE_DMABUF_RENDERER=1, which is read first, leaves the transport
set empty and so disables compositing outright; replace it with the same
pair. The wxWidgets skill reference is updated to match.
* Linux: drop the AppImage WebKit DMABUF exports; correct the skill notes
The launcher cannot gate on the runtime WebKitGTK version, so exporting
WEBKIT_DMABUF_RENDERER_FORCE_SHM/WEBKIT_FORCE_DMABUF_RENDERER there would
clear a distro isNVIDIA() guard on hosts older than 2.44, where FORCE_SHM
is ignored, re-exposing the broken hardware path. The binary already sets
both with the correct gate, so remove them and say who owns the workaround.
Also correct two claims in the skill reference. An empty transport set only
means "no accelerated compositing" from 2.44 on; 2.42/2.43 still fall back
to the Wayland/X11 renderers, removed in the same release that added
FORCE_SHM. And threaded scrolling is off by default in 2.44/2.45 (on from
2.46), so FORCE_SHM keeps compositing rather than scrolling. Document the
module-level detection cost and the WEBKIT_DMABUF_RENDERER_FORCE_SHM=0
opt-out.
* Flatpak: force WebKit's shared-memory transport instead of disabling DMABUF
The entrypoint exported WEBKIT_DISABLE_DMABUF_RENDERER=1 for NVIDIA. WebKit
reads that variable first and leaves the transport set empty, which turns
accelerated compositing off, so the Device tab still lagged for NVIDIA
Flatpak users, and the binary could not override it because its setenv is
non-replacing.
The runtime's WebKitGTK is vanilla, so there is no distro isNVIDIA() guard
to clear and WEBKIT_DMABUF_RENDERER_FORCE_SHM=1 alone is the right switch:
compositing stays on and only the hardware DMABUF path is skipped.
* Flatpak: drop the WebKit NVIDIA export; the binary owns the workaround
/proc/driver/nvidia/version is readable inside the Flatpak sandbox, so the
binary's NVIDIA branch already sets WEBKIT_DMABUF_RENDERER_FORCE_SHM with
the correct runtime gate and the entrypoint's own export is redundant.
Verified with:
flatpak run --command=cat com.orcaslicer.OrcaSlicer /proc/driver/nvidia/version
which prints the NVRM version (driver 615.71.09) from inside the sandbox.
* Linux: say what the XWayland compositing gate actually tests
DISPLAY and WAYLAND_DISPLAY are both set in a native Wayland session as
well, because GNOME and KDE run Xwayland as part of the session and export
DISPLAY for it whether or not this process is an X client. The pair means
XWayland is available in the session, not that GTK picked X11 for this
process, so "(an XWayland session)" reads as more than the condition
establishes.
Say what the pair means, and name the case the fallback actually covers:
the X11 backend inside a Wayland session, which is what the forced
GDK_BACKEND=x11 above produces in a build without EGL.
Comment only; behaviour is unchanged.
* Linux: apply the XWayland compositing fallback only where GTK picks X11
The default path set WEBKIT_DISABLE_COMPOSITING_MODE whenever DISPLAY and
WAYLAND_DISPLAY were both set, but a native Wayland app inherits both, so a
build with EGL disabled accelerated compositing — and with it threaded
scrolling — while running on Wayland, where the XWayland bug does not apply.
The only default-path case that really runs through XWayland is the forced
GDK_BACKEND=x11 used when wx lacks EGL, so set the variable there. The
backend cannot be re-checked later: that branch exports GDK_BACKEND itself,
so the force and a user's GDK_BACKEND=x11 opt-in, which deliberately keeps
compositing, would be indistinguishable.
Behaviour is unchanged for WebKitGTK >= 2.42, where the gate already skipped
the variable, and for builds without EGL. A build with EGL on WebKitGTK <
2.42 in a Wayland session no longer loses compositing and threaded scrolling.
The wxWidgets skill reference is updated to match.
---------
Co-authored-by: SoftFever <103989404+SoftFever@users.noreply.github.com>
## Speeds up OrcaSlicer incremental rebuild on Linux
Profiled `build_linux_image.sh`: 96 s, of which 51 s in the dependency
audit.
**`appimage_is_elf_file()`** ran `file` and `grep` per candidate. An
AppDir holds ~9.6k of them, 4.8k being the bundled Python runtime and
none of them ELF: ~19k processes, 14 s. Reads the four-byte magic
instead. Checked against the old result on 4000 files, no disagreement.
**The dependency walk** popped its queue with `"${queue[@]:1}"`, which
rebuilds the whole array each time. At ~4.8k entries that was 22 s of
copying an array around. Uses a read index.
Audit still passes. `shellcheck` v0.11.0, the version CI uses, is clean.
## Notes
The 96 s -> 12.7 s. Measured on a 32-core / 48 GB machine, but the audit
is a serial bash loop, so cores and RAM is not the bottleneck. On slower
hardware the saving should be larger
## Images
<img width="1987" height="782" alt="Screenshot_20261007_092516"
src="https://github.com/user-attachments/assets/96fcb917-38e5-49e7-8cbe-b37be1a2f23a"
/>
<img width="1807" height="742" alt="Screenshot_20261007_092621"
src="https://github.com/user-attachments/assets/4a65ffae-4d59-41f0-a1c7-ee5b49f3c4be"
/>
The Bambu network plug-in's code protector rewrites one page of its own signed __TEXT after loading. The hardened runtime tolerates that until the page is evicted; the next read of it then kills OrcaSlicer with CODESIGNING Invalid Page. Bambu Studio signs with allow-unsigned-executable-memory for this reason; with it added, the same build survives critical memory pressure that killed it in 30 s without.
Profiling build_linux_image.sh: 96 s, of which 51 s in the dependency audit.
appimage_is_elf_file() ran file(1) and grep per candidate. An AppDir holds ~9.6k
of them, 4.8k being the bundled Python runtime, none of them ELF: ~19k processes
for 14 s. Read the four-byte magic instead; checked against the old result on
4000 files, no disagreement.
The dependency walk popped its queue with "${queue[@]:1}", which rebuilds the
whole array each time. At ~4.8k entries that was 22 s of copying. Use a read
index.
96 s -> 12.7 s. The audit still passes.
The AppImage dependency closure resolves each bundled ELF's DT_NEEDED
entries with plain ldd, which cannot resolve the deps-built FFmpeg stack
(libavcodec/libavutil/libswscale) once it is copied into the bundle:
those libs are not installed in any standard loader path and carry no
RUNPATH of their own, so ldd reports the siblings as missing and the
build aborts. Extend the loader path with the bundle directory plus the
source directories of already-bundled files (mirroring
scripts/check_appimage_libs.sh), and key the dedup set on the bundled
file path instead of the source path so dependencies resolved from the
bundle directory are not copied onto themselves.
Co-Authored-By: Claude <noreply@anthropic.com>
* Add runtime display backend detection for Wayland support
Add LinuxDisplayBackend utility to detect X11 vs Wayland at runtime
using GDK_IS_X11_DISPLAY / GDK_IS_WAYLAND_DISPLAY macros. This is
the foundation for removing the forced GDK_BACKEND=x11 and enabling
native Wayland support.
- New files: LinuxDisplayBackend.hpp/.cpp with get_linux_display_backend(),
is_running_on_wayland(), and is_running_on_x11()
- Propagate wxHAVE_GDK_X11 / wxHAVE_GDK_WAYLAND from FindGTK3.cmake
as compile definitions to libslic3r_gui
- No-op on non-Linux platforms (returns Unknown / false)
* Fix Phase 1 code quality: pragma once, source ordering, static cache
* Make X11 initialization conditional for Wayland support
Remove the unconditional GDK_BACKEND=x11 force that blocked native
Wayland. Replace with conditional logic:
- EGL safety fallback: re-force X11 only when wxUSE_GLCANVAS_EGL is
off and WAYLAND_DISPLAY is set, with a warning log
- XInitThreads() only called when DISPLAY is set (X11 in use)
- __GLX_VENDOR_LIBRARY_NAME only set when DISPLAY is present (GLX-specific)
- WEBKIT_DISABLE_COMPOSITING_MODE only set under XWayland (both
DISPLAY and WAYLAND_DISPLAY present)
- Guard X11/Xlib.h include with __has_include for robustness
- Restore display validation to accept either DISPLAY or WAYLAND_DISPLAY
This is Phase 2 of the Wayland support plan.
* Fix Phase 2: safer EGL macro check, add clarifying comments
* Add GLAD2 library and replace GLEW linkage in build system
Set up GLAD2 as a static library to replace GLEW for OpenGL loading.
GLAD2 supports both GLX and EGL, which is required for Wayland support.
- Create src/glad/ with pre-generated GLAD2 sources (GL 4.6 compat)
- Add src/glad/CMakeLists.txt building glad as a static library
- Wire glad into src/CMakeLists.txt before libvgcode
- Modify libvgcode to use shared glad for GL path (keeps local copy
only for GLES2/Emscripten) to avoid duplicate symbol conflicts
- Replace GLEW::GLEW with glad in libslic3r_gui link libraries
Note: GLEW is kept in deps for OpenCSG. Code migration from GL/glew.h
to glad/gl.h headers will follow in Phase 3B+3C.
* Fix Phase 3A+3D: libvgcode GLAD include, dead files, dlopen dep, OpenGL link var
* Migrate from GLEW to GLAD: replace headers and API calls across codebase
Replace all #include <GL/glew.h> with <glad/gl.h> across 49 source files.
Migrate GLEW API calls to GLAD equivalents:
- glewInit/glewExperimental -> gladLoaderLoadGL()
- GLEW_EXT_* / GLEW_ARB_* extension checks -> GLAD_GL_EXT_* / GLAD_GL_ARB_*
- Remove GLEW-specific EGL/GLX mismatch #error guards (not needed with GLAD)
- Replace unavailable EXT symbols with core GL equivalents in
GLCanvas3D.cpp (GL_MAX_SAMPLES, glRenderbufferStorageMultisample,
glBlitFramebuffer, GL_READ/DRAW_FRAMEBUFFER)
- Update log messages from glewInit to gladLoadGL
* Fix Phase 3B+3C: remove GLEW find, clean EXT symbols, update attribution
- Remove find_package(GLEW) block from root CMakeLists.txt since GLEW
is no longer linked by any main application code
- Remove "glew" from SLIC3R_STATIC option description
- Replace all remaining EXT framebuffer symbols with core equivalents
in render_thumbnail_framebuffer_ext and _rectangular_selection_picking_pass
- Update AboutDialog credits from GLEW to GLAD
* Enable EGL in wxWidgets and add runtime GLX/EGL selection for Wayland
- Set wxUSE_GLCANVAS_EGL=ON in wxWidgets build and Flatpak manifest
- Add PreferGLX() call on X11 sessions for driver compatibility
- Remove Phase 2 safety fallback (EGL is now always compiled in)
- Guard SwapBuffers against hidden canvases to prevent Wayland stalls
* Fix Phase 4: move PreferGLX to app startup, fix FPS counter guard
Move wxGLCanvas::PreferGLX() from OpenGLManager::create_wxglcanvas()
(static initializer) to GUI_App::on_init_inner() before any wxGLCanvas
is constructed. This prevents a race where SkipPartCanvas could trigger
wxGLBackend::Init() before the GLX preference is set. The new location
also adds explicit is_running_on_wayland() detection with a warning for
unknown backends.
Move increment_fps_counter() inside the IsShownOnScreen() guard so FPS
is only counted when a frame is actually swapped.
* Update GLFW from 3.3.7 to 3.4 for runtime Wayland/X11 backend selection
Replace the compile-time GLFW_USE_WAYLAND flag (which locked to a single
backend) with GLFW 3.4's GLFW_BUILD_WAYLAND + GLFW_BUILD_X11 flags that
build both backends and auto-select at runtime based on the available
display server. This enables the CLI thumbnail renderer to work on both
Wayland and X11 sessions without separate builds.
* wayland: Fix UI call sites that rely on global screen coordinates
On Wayland, wxGetMousePosition() returns (0,0) and SetPosition() is a
no-op for top-level windows. Fix the highest-impact call sites:
- GLCanvas3D: Use cached m_mouse.position from event handlers instead
of wxGetMousePosition() + ScreenToClient() in get_local_mouse_position()
- Plater: Use event-relative coords via ClientToScreen(e.GetPosition())
instead of wxGetMousePosition() in 3 leave-window handlers
- BBLTopbar: Use event.GetPosition() and FindToolByPosition() directly
in mouse handlers instead of wxGetMousePosition()/FindToolByCurrentPosition()
- Search: Use focus-based dismiss logic on Wayland instead of
wxGetMousePosition()-based rect checks in SearchDialog and
SearchObjectDialog
- GUI_App: Skip SetPosition() in window_pos_restore() on Wayland where
it is a no-op; still restore size and maximize state
- Button: Position tooltip relative to button widget via ClientToScreen
instead of wxGetMousePosition()
* Fix SearchDialog Wayland dismiss: guard against search_line focus
* flatpak: Add Wayland socket permission for native Wayland support
* spec
* Fix crash on Wayland when wxWidgets lacks EGL support
Restore the safety fallback that forces GDK_BACKEND=x11 when wxWidgets
was not built with wxUSE_GLCANVAS_EGL=ON. Without this, the GLX backend
tries to access a non-existent X11 display on native Wayland, crashing
in wxGLCanvas::IsDisplaySupported() with SIGSEGV at offset 0xe4.
Also add a defense-in-depth guard in detect_multisample() that skips
the IsDisplaySupported call entirely on Wayland without EGL.
Root cause: deps/wxWidgets must be rebuilt after enabling EGL. The
compile-time check in OrcaSlicer.cpp detects the mismatch and falls
back safely.
* Fix EGL detection: use wxHAS_EGL instead of wxUSE_GLCANVAS_EGL
wxUSE_GLCANVAS_EGL is a CMake build option, NOT a C++ preprocessor
macro. The actual macro defined in wxWidgets setup.h is wxHAS_EGL.
All compile-time EGL checks were using the wrong macro, causing
the safety fallback to always trigger even with a properly built
EGL-enabled wxWidgets.
* Fix GL function pointers invalidated on Wayland/EGL
gladLoaderLoadGL() dlopen's libGL.so.1 to resolve GL function pointers
via dlsym, then immediately dlclose's the handle. On X11/GLX this is
fine because the GLX context keeps libGL.so mapped. On Wayland/EGL,
nothing else holds libGL.so open, so dlclose unmaps it and all function
pointers become dangling — causing SIGSEGV on the first GL call.
Fix: on Wayland, use gladLoadGL(eglGetProcAddress) which resolves
function pointers through the EGL loader without opening/closing
libGL.so.
* fix crash on start and various rendering issues
* fix crash on close
* small refactor
* move GPU selection to desktop file
* clean up a bit
* clean up more
* fix appimage error
- add a new mode to build in docker
- Improve AppImage
1. fix libbz2 soname differeence issue on appimage
2. Downgrade to 22.04 for better compatibility
3. improve appimage overall
* Actually build tests on Linux and allow RelWithDebInfo
They weren't being built.
Also cleaned up --config flags which enables RelWithDebInfo on Linux,
now that Ninja Multi-Config is used, it's quite trivial.
* Remove obsolete Slic3r Perl tests
The directory doesn't exist, they're already gone.
* Add GH job for running unit tests
* Move unit test execution to script and upload test results
* Don't run scheduled builds on forks
* Only deploy from SoftFever/OrcaSlicer
Will stop failures on forks
* Use artifact instead of cache
* Tweak archive and checkout paths
Keep getting error:
```
/home/runner/work/_temp/902d0a0a-6d23-4fe0-a643-8b5cc4efd25b.sh: line 1: scripts/run_unit_tests.sh: Permission denied
```
That seems to be because I didn't use actions/checkout, the working
directory is never setup correctly? So using checkout to get scripts
directory. Unsure if archive will preserve the `build/tests/` prefix;
will find out soon.
* Use tar to package directory and write results to correct directory
Tar preserves filenames and directory structure
* Use tar -xvf not -xzf
Muscle memory failed me
* Add testing wiki page
* Save test logs on failure and choose correct directory for junit
* Consolidate apt install steps, use for unit tests too, disable non-Linux builds
Temporarily disable non-Linux builds to save time while developing
this.
Cache the apt packages to save some time searching apt and downloading
them again (though I realize this is also downloading, but hopefully
by something closer and faster).
Remove all the redundant packages listed in the workflow and debian
distribution lists.
* Remove apt install steps from workflow
`./build-linux.sh -u` is supposed to install all needed packages, so
it should build without needing anything besides that. If I'm wrong
this commit will be dropped.
* Need composite action checked out locally
* Re-enable non-Linux builds now that it's working
* Skip a deploy and a notarize in forks
They only succeed in the main repo.
* Fix multi-build for non-Release builds: share CONFIG
* Correct build errors in unit tests
Indeterminate method signatures resolved. Updated script to build all
the tests.
* Fix -g vs -e for RelWithDebInfo
* Change CONFIG->BUILD_CONFIG
Missed one in prior commits
* Reduce wasteful redundant build artifact copies
1. Don't copy the artifacts and leave them; make a hard link first;
only make a copy only while creating AppImage.
2. Don't tar up the `package` directory; nothing uses this tar AFAICT
* Fix directory name
* Change jigsaw auth test URLs to httpbin.org
No idea why the basic auth doesn't work, but it doesn't work for
`curl` CLI either. This does.
* Remove force-build
It got reverted at
https://github.com/SoftFever/OrcaSlicer/commit/e3f049829bff3ac4de9fd271768948efc150e4c3
for unknown reasons.
* Add timeout for unit tests in GitHub Actions workflow (#11146)
---------
Co-authored-by: SoftFever <softfeverever@gmail.com>
Move many third-party components' source codes from the src folder to a new folder called deps_src. The goal is to make the code structure clearer and easier to navigate.
