This is a Windows only feature that uses policies set in the registry to control settings in KeePassXC. Overrides are set in HKLM\Software\Policies\KeePassXC and uses the same structure as the ini settings file. Overridden settings are disabled in the user interface and their values are not written to the user's own settings file.
Under QT_QPA_PLATFORM=offscreen the X11 native interface exists but has no
display, and DefaultRootWindow() dereferenced the null pointer at startup.
The global shortcut code already checks for a null display.
DatabaseWidget::lock() returned early when its autosave failed and a reload
began, without clearing m_attemptingLock. Every later lock() then returned at
the m_attemptingLock check, so the database could not be locked again until
KeePassXC was restarted.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
QDBusAbstractInterface::isValid() can latch an empty service owner
cached while xdg-desktop-portal is still activating, permanently
disabling auto-type and the global shortcut. Use the interfaces'
version() property reads instead: real blocking calls that trigger
activation if needed and return 0 only on failure.
* Fixes a specific warning that spams the console due to a deprecated signal that was used to indicate checkboxes state changed.
* Converted all QCheckBox::stateChanged(int) to QCheckBox::toggled(bool). All functionality still works as intended.
* Fix snap build to use latest format
- add x11-related libs to make the program functional.
- additional metadata
- remove desktop-launch
* Move to core24
* Fix Secret Service message being null in some cases when unlocking a database
* Update changelog
---------
Co-authored-by: Janek Bevendorff <[email protected]>
* Add minimal native Windows Arm64 support
Extend the Qt-via-vcpkg work from #13473 with only the source changes required for Windows Arm64 triplet selection, deployqt resolution, ordered runtime dependency deployment, and architecture-correct MSI metadata.
Co-authored-by: Janek Bevendorff <[email protected]>
Co-authored-by: Copilot <[email protected]>
---------
Co-authored-by: Gordon Lam (SH) (from Dev Box) <[email protected]>
Co-authored-by: Janek Bevendorff <[email protected]>
Co-authored-by: Copilot <[email protected]>
parseGroupTreeState() retrieves the number of record tree states from a 32-bit integer. It then calculates the number of bytes to read as num * 5, which is checked against the actual buffer length. The multiplication occurs as quint32, which wraps at 2^32. An attacker could craft a 32-bit value that wraps to match the observed buffer length. The following loop, however, uses num directly and may thus loop millions of times, reading beyond the end of the buffer.
No bytes are written and the minimum number that wraps is 2^32 / 5, which is well beyond any reasonable buffer size. Hence this is exploitable only as a denial of service. Given that this requires a crafted KDB file, there are easier ways to make the contents of a file unreadable.
Previously KDE Plasma implicitly bound them on session start but this
behavior was recently changed so it is now always required to bind them
which is correct according to spec. Our workaround was only needed for
fairly old KDE releases.
Because depending on the version and type of the desktop portal the
behavior might be all over the place we will always list the shortcuts
first as they might be active on session start and calling BindShortcuts
might fail and not give us the list back making our UI stale.
Docs have been updated to indicate version support for both KDE and
GNOME.
* Use C++ allocation in zlib
Marginally reduces the chances of decrypted database fragments remaining in memory.
Addresses CVE-2024-33901, but does not fix the main issue. For a proper solution, we need to replace QXmlStreamReader with something else.
* Fix Linux build always falling back to std::free() branch due to undefined HAVE_MALLOC_USABLE_SIZE
The search parser now passes the CASE_SENSITIVE flag to the term
when the '+' prefix modifier is used, so searches like '+p:test1'
correctly perform case-sensitive matching.
Fixes#13000
[skip ci]
The item access confirmation dialog used to hold raw Entry pointers in
its model, its deny buttons, and its result signal. When the database
locks or closes while the dialog is open, DatabaseWidget replaces the
Database and destroys every Entry, while the dialog stays up: painting
the model, or answering the dialog, then dereferenced freed entries
(crash in UnlockPrompt::itemUnlockFinished via Entry::uuid, see the
backtrace in #9104).
Make the dialog self-contained: snapshot each entry's uuid and display
data (title, username, icon) at construction and report decisions as a
QHash<QUuid, AuthDecision>, so the dialog never touches an Entry again
after it is shown.
On the UnlockPrompt side, restore the invariant that a non-null item
reference implies a usable backend: track items with QPointer and clear
all references as soon as an item retires (Item::itemAboutToDelete),
mirroring how collection retirement is handled. A decision for an item
that has retired counts as rejected, so the prompt completes as
dismissed instead of reporting success for nothing.
Finally, match the browser behavior: the dialog now withdraws itself
(rejects) when a collection it is asking about locks or goes away,
instead of keeping a stale question on screen.
Fixes#9104
Co-Authored-By: Claude Fable 5 <[email protected]>
Claude-Session: https://claude.ai/code/session_01AqHu6RpvLpaoH2aUGSYQzR