nullptr can be used directly in the Qt code since Qt 5.7.
Use it in generated code for consistency.
Change-Id: I249aeaf0a39b46ce1106b29d3ea4569a399908b7
Reviewed-by: Jarek Kobus <jaroslaw.kobus@qt.io>
CID 158429 (#1 of 1): Dereference before null check (REVERSE_INULL)
check_after_deref: Null-checking ctx suggests that it may be null, but it has already been dereferenced on all paths leading to the check.
Coverity-Id: 158429
Change-Id: I6d202599e962dae5ea4a45401f34237dd496d38b
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
This amends patch f27d1ccbb2.
Change-Id: I4c7a390a5f2cdd3307007c7b6708692c36f861b4
Task-number: QTBUG-62396
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
Define the lib dependencies for corelib in corelib.pro, where they
belong.
Change-Id: I973d3b0c571782d869b27dea243e899db4dddc43
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
It is currently impossible to get output from autotests in Visual Studio
or Qt Creator when running under the debugger. Qt Creator's cdb
integration cannot distinguish between the inferior's console output and
cdb's console output. If the inferior's output came from
OutputDebugString we'd be able to catch and display it.
Pave a way to force QTestLib's logging facility use OutputDebugString.
Task-number: QTCREATORBUG-16161
Change-Id: Iccd69c283626266ee4384a6163a8b72bb0e7df27
Reviewed-by: hjk <hjk@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
This prevents unnecessarily removing and reinstalling the completer
as event filter on the same widget.
This does not prevent what's going on in QComboBox::focusInEvent(),
where we'd set the line edit as widget just to immediately override
it with combo box itself.
Change-Id: I70c081a920f4daf4d7560e5cd7158e4070042d42
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
With MSG_TRUNC on Linux, recv/recvfrom/recvmsg will return the full size
of the datagram, even if it won't fit the buffer you passed. On Darwin,
we have a getsockopt() option to get that value.
Change-Id: I6e9274c1e7444ad48c81fffd14da7d5b93815f90
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
I'm getting crashes in Akonadi processes due to libproxy. I don't have
direct evidence that this was caused by a threading condition, but it's
clear from the source code of libproxy that the plugins it runs for
expanding PAC scripts are not thread-safe. To overcome this problem, we
only run libproxy functions in one thread only.
#0 0x00007f745f0ac1d8 in JSC::HeapTimer::timerDidFire() () at /usr/lib64/libjavascriptcoregtk-4.0.so.18
#1 0x00007f745f0ac287 in () at /usr/lib64/libjavascriptcoregtk-4.0.so.18
#2 0x00007f748e5ae9c5 in g_main_context_dispatch () at /usr/lib64/libglib-2.0.so.0
#3 0x00007f748e5aed88 in () at /usr/lib64/libglib-2.0.so.0
#4 0x00007f748e5aee1c in g_main_context_iteration () at /usr/lib64/libglib-2.0.so.0
#5 0x00007f7494f4268f in QEventDispatcherGlib::processEvents(QFlags<QEventLoop::ProcessEventsFlag>) () at /usr/lib64/libQt5Core.so.5
#6 0x00007f7494eeb35a in QEventLoop::exec(QFlags<QEventLoop::ProcessEventsFlag>) () at /usr/lib64/libQt5Core.so.5
#7 0x00007f7494d1b31a in QThread::exec() () at /usr/lib64/libQt5Core.so.5
#8 0x00007f7494d1fd2e in () at /usr/lib64/libQt5Core.so.5
#9 0x00007f74913174e7 in start_thread () at /lib64/libpthread.so.0
The pacrunner implementation of libproxy uses libdbus-1 which
(officially) is thread-safe, but experience tells that it has
problems. Since it is not running a JS engine, we don't need a thread,
but we do need to lock around it.
Change-Id: I84e45059a888497fb55ffffd14d2f638f21e807d
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
This patch avoids intermediate copying of the data by using a
new QIODevice::skip() API.
Change-Id: I358a83b21e2a58b59481745162fce1894549c0ac
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Markus Goetz (Woboq GmbH) <markus@woboq.com>
Without TLS (and thus ALPN/NPN negotiation) HTTP/2 requires
a protocol upgrade procedure, as described in RFC 7540, 3.2.
We start as HTTP/1.1 (and thus we create QHttpProtocolHandler first),
augmenting the headers we send with 'Upgrade: h2c'. In case
we receive HTTP/1.1 response with status code 101 ('Switching
Protocols'), we continue as HTTP/2 session, creating QHttp2ProtocolHandler
and pretending the first request we sent was HTTP/2 request
on a real HTTP/2 stream. If the first response is something different
from 101, we continue as HTTP/1.1. This change also required
auto-test update: our toy-server now has to respond to
the initial HTTP/1.1 request on a platform without ALPN/NPN.
As a bonus a subtle flakyness in 'goaway' auto-test went
away (well, it was fixed).
[ChangeLog][QtNetwork][HTTP/2] In case of clear text HTTP/2 we
now initiate a required protocol upgrade procedure instead of
'H2Direct' connection.
Task-number: QTBUG-61397
Change-Id: I573fa304fdaf661490159037dc47775d97c8ea5b
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
Consider the following:
/root/target - a file
/root/path/link -> ../target
/root/path/other/exe - executable
Running from /root/path/other.
exe is:
#include <QDebug>
#include <QFileInfo>
int main()
{
qDebug() << QFileInfo("../link").symLinkTarget()
return 0;
}
The link references /root/target, but the current output is
/root/path/target.
The link doesn't depend on the PWD. It depends on its own directory.
Change-Id: I61e95018154a75e0e0d795ee801068e18870a5df
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Also add import script.
The remaining diff to clean 1.6.32 is archived in the qtpatches.diff file.
[ChangeLog][Third-Party Code] libpng was updated to version 1.6.32
Change-Id: I1b4b78e39a6eb098d1b66c2528e47544bd9e6713
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The code was factored out in 0058f00b6 to be shared with Wayland, but
the Wayland platform plugin now lives in its own repository (as of
75a66c3d), and doesn't support running on macOS anymore.
Change-Id: Ied54f5f3e0a3e6fdaaedfcd456a140b46706f33c
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
QtWidgets in particular expects a show event before any resize or expose
events. QWindow::setVisible creates the platform window before sending
the show event, so we block sending any resize events during platform
window creation of QCocoaWindow, which means we miss out on the resize
event that indicates that the window is maximized.
We should probably teach QWindow::setVisible to send the show event
before creating the window, but until then we work around it by sending
an explicit resize event in QCocoaWindow::setVisible.
Task-number: QTBUG-62774
Change-Id: I6d512c9775986bcbcbf5c02a9e49984015fa9782
Reviewed-by: Morten Johan Sørvig <morten.sorvig@qt.io>
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
The definition of isTouchScreen() is protected with XCB_USE_XINPUT22 so
the implementation needs to have this too.
Task-number: QTBUG-62226
Change-Id: Icc3de01a6cb1299b43e56fc9f77833764131ca4b
Reviewed-by: Gatis Paeglis <gatis.paeglis@qt.io>
Originally when the QPrinter was created it would create the engine with
the default printer and then change it afterwards even though the
desired printer may already be known here. So by passing the printer
name we ensure that it is initialized with the desired one right away.
Task-number: QTBUG-62221
Change-Id: Iaa90243708b57bf89354a527a982ac45c991f603
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
By sharing the implementations of the methods between QNSWindow and
QNSPanel we don't need a helper, and can remove duplicated code. This
duplication would expand in the future, as for each method added to
the QNSWindowProtocol, we would have to add forwarding functions
in both QNSWindow and QNSPanel, forwarding to QNSWindowHelper, and
then two more functions in QNSWindow and QNSPanel in case we wanted
to call super from the helper, similar to [QNSWindow superSendEvent].
The only snag is that calls to super are hard-coded to a specific
superclass during complication, so we provide our wrapper for
objc_msgSendSuper that resolves the superclass at runtime.
The helper class QSendSuperHelper provides compile time implicit
instantiation of the right template without having to provide
the return type as a template argument, via operator T and
a fallback for the case of no return type via the destructor.
Change-Id: Iaf13f27675d90f884470f5005270ea0d9d0316f3
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
The default on iOS has been raster for two years now, as of 3e892e4a97,
and we haven't seen any major performance regressions that would warrant
keeping the OpenGL based code-path alive.
This includes the default surface format, which was ony set so that
QPainter clip regions would work when using the GL backed backing store.
Change-Id: I37b880a758b9c3fad1f23ae60268629ffbe9bc3e
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@qt.io>
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
Change-Id: I83cbbb47af8580fa67cbc75fee07bc1e123895eb
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
This #if has been there since 0248ec4054b (Oct 2005), presumably due to
bugs reported after the Qt 4.0 release (the first containing QHostInfo).
Any macOS issues with getnameinfo() have long since been gone. I've
confirmed that it works on 10.9, 10.10, 10.11 and 10.12, so I'm
dropping the #ifndef.
[ChangeLog][Deprecation Notice] Starting with Qt 5.10, IPv6 support will
be mandatory for all platforms. Systems without proper IPv6 support,
such as the getaddrinfo() function or the proper socket address
structures, will not be able to build QtNetwork anymore.
Change-Id: I320d9d2f42284a69a4cbfffd14dd5bf479e5f678
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
No need to make workarounds for Android in the .cpp source. Just let it
fail (if it still has to fail).
Change-Id: Iaf4157b7efa2416d898cfffd14d94ebcb4d979be
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Minimal adjustments to config.h are necessary.
[ChangeLog][Third-Party Code] PCRE2 has been updated to version 10.30.
Change-Id: Iaca6a5ceffe4f5029212411eca8e2965ca7d9410
Reviewed-by: Kai Koehne <kai.koehne@qt.io>
When calling resize() from showEvent(), we'd set the full geometry
on the widget's QWindow. This resulted in the top-level window
being moved to the top-left corner, even though no other call to
move() or setGeometry() had happened before.
The solution consists on calling the proper QWindow methods depending
on whether setGeometry_sys() is called for a move, a resize or both.
Furthermore, this needs QWindow::resize() to set its position policy
to frame-exclusive. The documentation states that is already the case
and we're setting the full geometry on the platform window, so we need
to convey that bit of information.
This also solves the age-old conundrum: "### why do we have isMove as
a parameter?"
Change-Id: I2e00fd632929ade14b35ae5e6495ed1ab176d32f
Task-number: QTBUG-56277
Reviewed-by: Gabriel de Dietrich <gabriel.dedietrich@qt.io>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
That means a file is never created, unless you ask for the name. There's
no chance of left-over temporary files being left behind. QSaveFile also
benefits from this, since the save file is not present on disk until
commit(). Unfortunately, QSaveFile must go through a temporary name
because linkat(2) cannot overwrite -- we need rename(2) for that (for
now).
[ChangeLog][Important Behavior Changes][QTemporaryFile] On Linux,
QTemporaryFile will attempt to create unnamed temporary files. If that
succeeds, open() will return true but exists() will be false. If you
call fileName() or any function that calls it, QTemporaryFile will give
the file a name, so most applications will not see a difference.
Change-Id: I1eba2b016de74620bfc8fffd14cc843e5b0919d0
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
These introduce an unwanted locale-dependency; clients should use
suitable QLocale methods instead.
Task-number: QTBUG-28581
Change-Id: Ie7dfe712c50b9f5da94e4b20af7b231d8963cbc9
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This function had accumulated a fair bit of accidental
complexity over the years.
- No early returns, make sure to preserve fullscreen
state for all windows.
- Use windowIsPopupType() directly to set borderless
for Qt::Popup (but not Qt::Tool).
- Handle Qt::Tool explicitly.
- Deduplicate Qt::CustomizeWindowMask handling.
- Remove case that used the absence of NSResizableWindowMask
to remove the maximize button. Maximize is now
disabled elsewhere (setWindowZoomButton). All windows
now get NSResizableWindowMask by default.
- Qt::ForeignWindow now gets a standard window style
mask instead of NSBorderlessWindowMask. The old
code did not handle this case and left the mask
value unmodified.
Change-Id: I56499e9f05c3f481b5a96e0507da2fb195f207fa
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
88a8feeacb fixed QCryptographicHash's
implementation of SHA-3 -- before, QCryptographicHash was
calculating Keccak, not SHA-3. The change however breaks downstreams
that were relying on the "wrong" values calculated by earlier
versions of Qt. To restore compatibility with such calculations,
add the support for Keccak back, as deprecated and to-be-removed
in Qt 6.
[ChangeLog][QtCore][QCryptographicHash] In order to preserve
compatibility with earlier versions of Qt, QCryptographicHash is
now able to calculate Keccak message digests.
Task-number: QTBUG-59770
Change-Id: I740684cb3d0774292fa837a06b444c73fb4eed52
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
FICLONE only works on (currently) btrfs and xfs, and then only if it's
the same mount, so it's of very limited use. There are a couple other
techniques we may try that do not involve I/O through user-space, though
not immediate:
- sendfile(2) can be used on regular files, even across mountpoints
- sendfile(2) can be used on block devices too, but we need to get the
device's size first
- splice(2) can be used on pipes (FIFOs)
We only implement the first technique (earlier iterations of this patch
implemented all).
Change-Id: I81480fdb578d4d43b3fcfffd14d4b47cd70495a3
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
For some reason, fcopyfile(3) works on directories, so we need to make
we aren't operating in one.
Change-Id: I81480fdb578d4d43b3fcfffd14d4f2f485348515
Reviewed-by: Jake Petroules <jake.petroules@qt.io>
A case of nested Qt::WindowStaysOnTopHint may occur when
context menus are created on windows with Qt::WindowStaysOnTopHint
set. Raise the popup in that case.
Amends 329a029c36.
Task-number: QTBUG-62004
Change-Id: Ifb761edbd42b1447bec30735810c006d02e1aa97
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
[ChangeLog][Binary Compatibility Note] The variable
QOperatingSystemVersion::AndroidOreo was added in this release.
Code that uses this variable will not run under Qt 5.9.1.
If backwards compatibility is desired, use instead
QOperatingSystemVersion(QOperatingSystemVersion::Android, 8)
Change-Id: I1da5a5577bf6b719e543a1ded1f9b912a83665c3
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
App Nap may cause stalls or timer delays during test runs.
Change-Id: I828282d12127918439a9a2a4f7d7be6cac457b42
Task-number: QTBUG-61499
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
NSOpenGLProfileVersion4_1Core is available starting
with macOS 10.10.
Task-number: QTBUG-62333
Change-Id: I75d8c3bc1093ff38c6ca1db2d2da50fd448a149f
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
Move deprecated and obsoleted enum values to the end, as there already
was a category for them.
Fix linking to 'Q(Gui)Application'.
Update the usage of the name macOS.
To make the table more readable in online style, add zero-width spaces
to long strings, allowing browsers to word-break them, thus avoiding
text overflow/horizontal scroll bar.
Task-number: QTWEBSITE-783
Change-Id: I0a96156d24cba4a0405c4edd8d3829def30c69bf
Reviewed-by: Venugopal Shivashankar <Venugopal.Shivashankar@qt.io>
After 871966 we now do drawing as a result of drawRect calls, but layer
backed mode was not taken into account. This restores support for both
pull and push-mode drawing in layer-backed mode.
Change-Id: I35039ee9eb4486206f9f92f8230df104473368c9
Reviewed-by: Morten Johan Sørvig <morten.sorvig@qt.io>
We have to exit composition mode before inserting text. Otherwise, the
state of the input method will be out of sync with the contents of the
text editor.
Task-number: QTBUG-61717
Change-Id: I58bf3988ae9e0acf3302e810e46bb0ebeda30d17
Reviewed-by: BogDan Vatra <bogdan@kdab.com>
Propage event to all sibling screens resulting in enter and leave
events being properly sent.
Change-Id: Ia89d53105f6303fae3f304ce0920b5a4a24f86ae
Reviewed-by: Laszlo Agocs <laszlo.agocs@qt.io>
Removes the need to initialize QDesktopWidget, just to look up which
screen a widget would map to, the geometry of a screen, number of
screens, etc.
Change-Id: Ieb153b9ff6d3fba645fc528d6e430e8392f990bf
Reviewed-by: Gabriel de Dietrich <gabriel.dedietrich@qt.io>
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
Flushing outside of the display cycle does not care about any ordering
between views, including the NSThemeFrame responsible for drawing the
rounded corners of the window.
Since Qt Widgets is doing a lot of synchronous flushing (for now, until
we plumb update() to requestUpdate(), or enable layer-backing), we add
a workaround that explicitly draws the corners after flushing, just like
the logic in [NSView displayIfNeeded]. This is the same workaround used
by WebKit: https://trac.webkit.org/changeset/85376/webkit
Change-Id: I884152cdb2685569704e577b64b5ae278ed82c21
Reviewed-by: Morten Johan Sørvig <morten.sorvig@qt.io>
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
It seems to break the ability to grab an offscreen QQuickView.
That in turn breaks the tst_qquickimage::hugeImages autotest.
This reverts commit 096b56f336.
Task-number: QTBUG-62548
Change-Id: I11ce452341bfc2cc3cbc832b613c7366049b31d5
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
The artefacts appeared for square-capped dashed pens when the end of
the line fell a tiny fraction into the start of a new dash. At that
point in the dashing algorithm, accumulated precision errors in the
'length' variable could make it slightly differ from the actual length
between the start and end points of the line fragment. Although both
values would be "almost zero", the rasterizeLine() function's square
capping would make the error very visible; see the bug report.
Fix by calculating the precise length of the last line fragment.
Task-number: QTBUG-56969
Change-Id: I7b69c0d465649be61fb87ac7b8348f0c299486ee
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Change-Id: Id9558fa1db6a7a8f29149e26c761450f58b74b81
Reviewed-by: Jesus Fernandez <Jesus.Fernandez@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Change c12072c685 fixed a problem
where tab labels would overlap with the icons if the tab became
too small to contain the text. But it did not properly account
for the full area occupied by the icon, because the horizontal
padding is hardcoded to 4 in the mac style, whereas in the
common style (where the icon is drawn) it uses the pixel metric
for this.
In addition, the change only allocated space on the left side,
causing the label to no longer be centered.
Task-number: QTBUG-61235
Change-Id: Ieec4f7044584361f92045addbc8bbd81bd5c9fc7
Reviewed-by: Gabriel de Dietrich <gabriel.dedietrich@qt.io>
Calling QCoreApplication::processEvents() from a slot connected to the
readyRead() signal might cause desynchronization in the waitForXXX()
loop, if the process has been finished during the event processing.
This results in unnecessary timeouts and causes waitForFinished() to
fail unexpectedly.
So, a proposed solution is to check the state on each iteration of the
loop, as Windows implementation does.
Given issue is tested by tst_QProcess::processEventsInAReadyReadSlot()
which was unstable in CI.
Task-number: QTBUG-62584
Change-Id: I7438cf67b0163bbf49314008a9dc660c0977fb7b
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Previously, a sequence of middle-click pastes produced just one undo
item to undo/redo all of the pastes at once.
The new code seems to be the intended way to paste the selection anyway.
Change-Id: Ifc2e9714628da9e394053ff8c34709578656f54d
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
Unified headers now defines _POSIX_THREAD_SAFE_FUNCTIONS but not all
libc functions are available in all Android API versions.
Change-Id: I01c94f0b89e7f8aa8575e7bbda28d9fe41a68ff1
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
QWidget::event() will turn an unhandled double click into a
mouse press. Therefore, we must avoid sending a separate mouse press
in that case.
This copies the logic from 9063edef79,
the previous fix for QTBUG-25831, which only applied to the non-popup
case.
Task-number: QTBUG-25831
Change-Id: I616f398a5ebe1f95d4b65b9f6ce2fe39a6fad83e
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
The only reason our code wants PKCS12 files is for a private key, but
a valid file needn't contain one; and reading a file without lead to a
crash in QSslKeyPrivate::fromEVP_PKEY(). So check for missing key and
fail the load, since the file is useless to us. Also ensure the
caller's pkey is initialized, as we aren't promised that
PKCS12_parse() will set it when there is no private key.
Add a test for this case (it crashes without the fix) and update the
instructions for how to generate test data to cover it also.
(Corrected the wording there, too; at the interactive prompt,
"providing no password" really provides an empty password.)
Task-number: QTBUG-62335
Change-Id: I617508b903f6d9dee40d539b7136b0be8bc2c747
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
Making system calls with empty file names is not a good idea. When you
run qmake $srcdir, you see this in strace:
stat("", 0x7ffed229e250) = -1 ENOENT (No such file or directory)
(twice)
I've also inlined the isEmpty() function for better code generation.
Some functions take QSystemError and some don't. That needs to be
corrected at some point, possibly with something like std::expected.
Change-Id: I1eba2b016de74620bfc8fffd14ccbfa162f93631
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
The function returns a const char * out of a QByteArray. We must
be sure that the QByteArray outlives the function, otherwise the
pointer returned would be dangling. Add an assertion for that.
Found by clazy.
Change-Id: I3416af4eb5ec79ddb3e4baf3bdcfe046b44d4225
Reviewed-by: Eirik Aavitsland <eirik.aavitsland@qt.io>
The function and type is only used in the tests and has no benefit in
applications so it should be marked as internal to avoid confusion for
those implementing their own drivers.
Task-number: QTBUG-56278
Change-Id: I0f2ae27d41b133c4f0d3b0d390688fd3307592ce
Reviewed-by: Andre Somers <andre.somers@kdab.com>
Reviewed-by: Venugopal Shivashankar <Venugopal.Shivashankar@qt.io>
Reviewed-by: Andy Shaw <andy.shaw@qt.io>
Although the window is refused input for the most part from the system,
it does not act like that it is blocked by the application modal dialog.
This ensures that it is the case and prevents things like being able to
double click on the title bar to maximize the window on Windows.
Task-number: QTBUG-49102
Change-Id: If1582819b90cb2ec9d891f664da24f13bfec7103
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
5ad2e1cea1 exposed a
latent bug when matching font weights that do not exactly
match any of the canonical weights in the system .
It seems that when we registered the font with FontManager,
it would modify the font weight to match a system weight,
while we now get raw values from the font instead.
So for Roboto Medium, for instance, we would now get
0.2 instead of 0.23, and since our conversion logic
would convert everything between Light (-0.4) to
Medium (0.23) to QFont::Normal, we would regard this
as normal weight instead of medium.
But on a linear scale, it makes more sense to match
to the closest canonical weight instead of requiring
exact matches.
Note that the definition says that the middle between
two weights should tend upwards, therefore we do the
comparisons in decreasing order.
[ChangeLog][QtGui][Text] Fixed matching of non-regular
font weights for application fonts on macOS.
Task-number: QTBUG-61520
Change-Id: Ieda4927c2c69ec72125257340cf06c683b031d05
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
On Windows and macOS, that area of the menu can become inaccessible
if the menu is tall enough. Since these are not popups but tool
windows, the test for UseFullScreenForPopupMenu should not apply
for torn-off menus.
Change-Id: Ife7836bef568896a5bb67d42a2af412f06a871d6
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
A loopback address warrants the same exception as local files.
Task-number: QTBUG-59219
Change-Id: Ie0a75faa558d6596455da38656c8749c994d0fd8
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Empty menus on a menubar are hidden by default. If the menu gets
added to the menubar before it contains any item, we need to get
the menubar to sync the menu, which will update its native menu
item hidden property.
Menurama manual test's 'Add Many Items' button should now work.
Change-Id: I8ce1df21031c171789318fdf28ae495819458d71
Task-number: QTBUG-62260
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
Calling -[NSMenu update] every time we add a new item can result in
a quadratic behavior since the function itself will iterate over all
the items in the menu. We solve this by using a 0-timer which will
trigger the call to update the next time the event loop spins.
Menurama manual test updated.
Change-Id: Ic155d364515cc93eb81b1c8085c8e44c93799954
Task-number: QTBUG-62396
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
Statics and threads don't mix well. There can be multiple threads
calling QEvdevTabletHandler::readData() simultaneously if you have
several tablet devices registered with the plugin, creating a race on
the static buffer array.
Make the buffer a simple local variable instead, the array is small
enough that we can afford the per-thread stack allocation.
Change-Id: I4487add8df50743b8178ca6faeb9be45231ccb78
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Laszlo Agocs <laszlo.agocs@qt.io>
If we reach the maximum reconnect attempts for an HTTP request with
pipelining enabled, those also need to be requeued during cleanup for
the connection channel. Otherwise future successful requests on that
same channel will incorrectly assign the data to replies from the old
pipelined requests, resulting in swapped data in the replies.
Task-number: QTBUG-62286
Change-Id: I804b8ac280957b518d63b2341e469a13315a8c27
Reviewed-by: Markus Goetz (Woboq GmbH) <markus@woboq.com>
QLineEdit with a mask does not return empty fields with the
ImSurroundingText query. This is a problem for the input
context that is not aware of the mask and relies on the
fact that the cursor position never exceeds the boundaries
of the surrounding text.
This change fixes the issue by returning unmasked text with
the ImSurroundingText query.
[ChangeLog][QtWidgets][QLineEdit] Fixed behavior of the
ImSurroundingText query. Previously, it returned a masked
text whose length may be less than the cursor position.
Now it returns unmasked text, so the text length is always
greater than or equal to the cursor position.
Task-number: QTBUG-60319
Change-Id: I1c8009164836a1baa2e3a45958bf1ea5fa9be38d
Reviewed-by: Mitch Curtis <mitch.curtis@qt.io>
The relevant changes was made for 32-bit, but not 64
Change-Id: I7dc1a299d72a742efca7c56717274c0c2ea0c579
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@qt.io>
This will be used by the Qt X11 Extras module.
Task-number: QTBUG-50358
Change-Id: Ie095cd211c393ea6d78660b4d53cac28b435a3b2
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Using int as the operation parameter was wrong. It meant any write or RW
ioctls (which have the MSB set) would be sign-extended to 64-bit when
calling the native API. Depending on the OS, the upper 32 bits were not
cleared prior to comparing with the call number, resulting in
unexpected errors.
Linux and FreeBSD operated properly; I only got the error on Darwin.
Change-Id: Iaf4157b7efa2416d898cfffd14d985af65c1ff89
Reviewed-by: Jake Petroules <jake.petroules@qt.io>
- To use categorized logging.
- Print also non-core event types.
- Added QLoggingCategory as one of the parameters because I will
need it in a subsequent patch.
- A call to this function in QXcbConnection::handleXcbEvent I left
disabled from builds by default as the "handled" variable is broken.
Change-Id: I15cd71c40b5772a4352a69cdc24a0ae026626060
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@qt.io>
Profiling a large model showed that QSqlRecord was being copied and
destroyed from a few places, due to use of QMap<int, ModifiedRow>::value()
where ModifiedRow has two QSqlRecord members. This can easily be avoided
(in the common case with no modified rows) by using constFind() instead.
My testcase (iterating over a model with 126936 rows) went from 1266 ms
to 896 ms.
Change-Id: I04e98b5573ef24165bf6cff19946e5bedd0fb0ba
Reviewed-by: Jesus Fernandez <Jesus.Fernandez@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Replaced dependency to libdl.a with libshm_client.a. Defined symbols
'shm_area_password' and 'shm_area_name' internally. The build for
INTEGRITY is static only so libdl.a is not needed.
Change-Id: I7e34528835132d79ea582a30cf9ff61cdda198da
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Rolland Dudemaine <rolland@ghs.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Calculates the correct offsets and coordinate transforms for the
intermediate buffer. This means we can conceptually simplify our
path switches instead of having downscale routines handling mirrored
upscaling.
Change-Id: I60efa7feaba80165672ca0ce064515fdf620869d
Reviewed-by: Eirik Aavitsland <eirik.aavitsland@qt.io>
53d289ec4c fixed the issue of not getting
touch events when grabbing via the plain xcb functions. And due to wanting
to support a setup where mouse events are delivered via core events it was
chosen to use mouse via XI2 only when really necessary. Thus starting only
from 2.2, the version from which the mouse+touch grabbing becomes the issue
(XI2 introduced touch support from 2.2).
The same patch states that using QT_XCB_NO_XI2_MOUSE restores to the old
behavior (grabbing via core) with broken touch grabbing. Broken only with
2.2+, not 2.0 and 2.1 since those versions of protocol do not know about touch.
All of this implies the following:
1) The user code that already depends on QT_XCB_NO_XI2_MOUSE, won't see any
behavioural difference if we will use XInput starting from 2.0 for mouse.
Mouse grabbing will continue to be done via core (xcb_grab_pointer) as
mentioned above and thus touch grabbing will continue to be broken with 2.2+.
2) The code that has never cared how we get the native events (core vs
xinput2), won't see any behavioural difference. In this case grabbing will
always be done via XI2 (XIGrabDevice) grab when XI2.0+ is available.
Since there is no difference in the outcome, we migth as well use XI2 for
mouse from 2.0, not 2.2. Extension events are always better choice than core.
Besides the broken touch grabbing issue with QT_XCB_NO_XI2_MOUSE, there are
other issues with that code path, for details see internal documentation of
xi2SelectDeviceEventsCompatibility(), where the conclusion is:
*** If your code relies on QT_XCB_NO_XI2_MOUSE, then your code needs fixing. ***
This patch also cleans up how we select XInput2 events, by separating
the QT_XCB_NO_XI2_MOUSE code path. This has two benefits - improved code
readability and will make the deprecation of QT_XCB_NO_XI2_MOUSE easier.
The patch removes some sparse comments as the behavior is now documented in
one place, see xi2SelectDeviceEventsCompatibility().
[ChangeLog][Platform Specific Changes][Linux] The QT_XCB_NO_XI2_MOUSE
environment variable is deprecated and will be removed in Qt 6. If your
application relies on behavior set by QT_XCB_NO_XI2_MOUSE, it should be
updated accordingly.
[ChangeLog][Platform Specific Changes][Linux] Pointer event delivery on
X11 is now done starting from XInput version 2.0 (when available) instead
of 2.2. XInput support can be disabled by setting QT_XCB_NO_XI2=1 environment
variable. Note that using QT_XCB_NO_XI2 would also disable tablet and touch
support.
Change-Id: I661b36d6710b9f6ec71fecc8287ba479432bff4c
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
Commit 3ea04c7d made it so that we always set the
ignoresMouseEvent property on the NSWindow, based
on the WindowTransparentForInput flag.
However, this overwrites the magical secret initial
state where click-trough is determined based on window
content transparency - setting the property to false
makes the window capture all events.
Restore 5.6 behavior by not modifying ignoresMouseEvent
if we can. Toggling WindowTransparentForInput on and
off again is still broken.
Task-number: QTBUG-54830
Change-Id: I5f44ce14d9a7dc1713f9efb1ef929e2969838d90
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
"gtk-long-press-time" was introduced in 3.14. Therefore, if the property
does not exist, we should fallback to QGnomeTheme::themeHint().
Task-number: QTBUG-61393
Change-Id: I898c0ddebdbabb300d7ad9dd275d51836ad9cf8c
Reviewed-by: Dmitry Shachnev <mitya57@gmail.com>
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
The x11 native painting experimental feature requires fontconfig to
build, so add that dependency to configure.
Task-number: QTBUG-62121
Change-Id: Iac79c17fdeaef06b2e075dc34fe554c0e7c56edf
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
QFile::copy was assuming that the target file was native and therefore
it could simply take the file descriptor to clone. While that was not
currently a problem, in theory it could be as we do have one writeable
file engine besides QFSFileEngine (QWinRTFileEngine).
By refactoring to take the parameter as a QAbstractFileEngine, we can
ensure that the target file is a native file.
Change-Id: Ib7a1737987bf4c4a8c51fffd14d0c048fd509025
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
That code was added by ae5f2a6672 and
later incompletely removed by c173a50719.
This patch amends c173a50719.
Change-Id: Ibfd39aaf9b49424fc54d878dc588454eb841ed97
Reviewed-by: Laszlo Agocs <laszlo.agocs@qt.io>
The stat::st_flags and fchflags(2) syscall are marked:
HISTORY
The chflags() and fchflags functions first appeared in 4.4BSD.
Change-Id: I81480fdb578d4d43b3fcfffd14d4fd23bd27e37e
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
Reviewed-by: Tobias C. Berner <tcberner@FreeBSD.org>
If a QFileInfo was constructed with an empty path, which could happen
with QFileInfo(QFile()) or via QDir, etc., then it would issue system
calls to empty paths and could even produce warnings. This commit makes
am empty path name be the same as a default-constructed QFileInfo and
corrects the use if 0 for ownerId and groupId to match the
documentation.
[ChangeLog][Important Behavior Changes] QFileInfo on empty strings now
behaves like the default-constructed QFileInfo. Notably, path() will now
be the empty string too, instead of ".", which means absoluteFilePath()
is no longer the current working directory.
Change-Id: I8d96dea9955d4c749b99fffd14ce34968b1d9bbf
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
This system call, new in Linux 4.11, gives us the file birth time. It's
also extensible, representing the fourth generation of stat(2) on Linux
(the original sys_stat(), sys_newstat(), sys_stat64() and now
sys_statx()), not to be confused with glibc's __xstat function, which
wraps a call to stat64. Anyway, the new one is designed to be extensible.
Now we get birth times on ext[34] on Linux too:
Name: .
Path: . (/home/tjmaciei/src/qt)
Size: 4096 Type: Directory
Attrs: readable writable executable hidden nativepath
Mode: drwxr-xr-x
Owner: tjmaciei (1000) Group: users (100)
Access: 2017-07-02T14:47:49.608
Birth: 2016-05-02T13:20:33.097
Change: 2017-07-01T13:37:08.737
Modified: 2017-07-01T13:37:08.737
It's not supported in any other filesystems I have (Linux sources show
xfs has the feature too). Even on ext4, it depends on whether the
filesystem was created with 256-byte inodes, which my /boot fs wasn't.
Change-Id: I8d96dea9955d4c749b99fffd14cda23ed60d5e72
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
This fixes an issue where iOS, watchOS, and tvOS versions would be
mis-detected as 100 times their version (i.e. iOS 8.0 as iOS 800.0).
Instead of protecting the branch in (version >= 100000) with a macOS
ifdef, we can simply remove it entirely since Qt cannot be run on OSes
where the old encoding was used (macOS < 10.10).
Amends 8418a6335b
Change-Id: I32b307163c815799cb46c008b93f3b53d27c48b9
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Conflicts:
src/widgets/kernel/qwidget.cpp
This merge also extends the expected output of the pairdiagnostics
teamcity output (added in dev in commit
c608ffc56a) after the recent addition of
the flowId attribute to the teamcity output (commit
8f03656211 in 5.9).
Change-Id: I3868166e5efc45538544fffd14d8aba438f9173c
There is already some case accounting for when this warning appears with
warning level 4 and Visual C++ on Windows. However it was not catching
all the places it was coming from, so this extends it to cover those
places too.
Change-Id: I69b21440716361fda1c1ae0be0d9c17ced7f0792
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The qoffscreensurface::setScreen crashes in create if application is
closing. This happens in multimonitor setup when one of the screens is
already destroyed and qoffscreensurface handles the screenDestroyed
signal. In this case the QGuiApplication::primaryScreen returns non-null
surface and selects the codepath that recreates the surface. The create
function calls the qGuiApp->thread function, which crashes since the
application instance has already been set to nullptr. Prevent this in
qoffscreensurface by checking the application instance in the setScreen
function.
Task-number: QTBUG-62222
Change-Id: Ibb11a270c23a39e66009b45ab83dd7d47d5ccb1f
Reviewed-by: Janne Koskinen <janne.p.koskinen@qt.io>
Reviewed-by: Laszlo Agocs <laszlo.agocs@qt.io>
It was working on Linux because _GNU_SOURCE gets us POSIX.1-2008
compatibility, but not on macOS or the BSDs. There, we were still stuck
to full second precision.
This commit uses the template trick introduced by the futimes code
(which itself was inspired by commit 2fb42eb4af
in QtNetwork). Also note how it adds support for birth time, if the
system's stat struct has that information.
Tested to work on MacOS and FreeBSD. The manual filetest produces:
Name: .
Path: . (/usr/home/tjmaciei/src/qt/qt5)
Size: 1536 Type: Directory
Attrs: readable writable executable hidden nativepath
Mode: drwxr-xr-x
Owner: tjmaciei (1001) Group: tjmaciei (1001)
Access: 2017-07-13T20:03:47.916
Birth: 2017-07-13T20:03:47.916
Change: 2017-07-13T20:04:41.648
Modified: 2017-07-13T20:04:41.648
Linux will require support for statx(2).
Change-Id: I8d96dea9955d4c749b99fffd14cd97d7a8c6d45d
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Took me a long while to understand what it did. Now that I do, I can
also answer the question left behind during the original implementation
in Qt 4.8 (commit 4fd2aced96d9095254d89f9da9c911bd88f15245 in the old
history): how to know if a file exists?
Change-Id: I8d96dea9955d4c749b99fffd14cdae135499a0d3
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
qimage.cpp:2127: warning: Can't link to 'isDetached()'
Change-Id: Ib8f056441bf08e315ed97caf55c510f50ebaf4ed
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@qt.io>
In Qt, we have QTextOption::tabStop, QTextEdit::tabStopWidth and
QPlainTextEdit::tabStopWidth.
Neither are very good names, since the tab stop is neither a
numerical value as in the former, nor does it have any dimensions
that can be measured, as in the latter. Vertical text advances
may also be supported by Qt at some point in the future, at
which point the name would make even less sense.
At the same time, we expose the actual type of the tab stop
distance as floating point in the QTextEdit and QPlainTextEdit
API instead of always rounding it to an int.
To avoid duplicating either of these APIs in Qt Quick, we
introduce tabStopDistance as the common term instead and deprecate
the old names.
[ChangeLog][Text] Introduced tabStopDistance property in
QTextOption, QTextEdit and QPlainTextEdit as replacement for
the inconsistently named tabStop and tabStopWidth properties.
QTextOption::tabStop, QTextEdit::tabStopWidth and
QPlainTextEdit::tabStopWidth have subsequently been deprecated.
Change-Id: Ib7e01387910cddb58adaaaadcd56c0e69edc4bc2
Reviewed-by: Paolo Angelelli <paolo.angelelli@qt.io>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
This patch updates the code from the unix implementation of
QStandardPaths to use QRegularExpression in place of the deprecated
QRegExp.
Change-Id: I51fa231dcd70ca55d1bfffb31d8f28f964ac44fe
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: David Faure <david.faure@kdab.com>
The full documentation explains it well but a more prominent warning
will help avoid the wrong use of that class and encourage user to read
further the documentation.
Change-Id: I3178749f2b1b0350040f81eef253fd85c7ba0a5f
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The use as in the code:
futimesat(fd, NULL, &tv)
is not documented to work. The file descriptor should be a directory's
one, not an open file (though the Linux source code seems to handle that
case). This call was done as a fallback to futimes, so it's very
unlikely a system would have futimesat and not futimes.
Both the Linux and the FreeBSD man pages say it's deprecated anyway.
Change-Id: I8d96dea9955d4c749b99fffd14cd94068dc7668a
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
The Unix stat fields "st_ctime" and "st_ctim" mean "change time", the
last time that the file/inode status fields were changed. It does not
mean "creation time". So this commit splits all of the internal API to
"birth" and "metadata change" instead of "creation" to avoid the
conflict.
Change-Id: I149e0540c00745fe8119fffd1463fe78b619649e
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
[ChangeLog][QtCore][QFileInfo] Deprecated created() because it could
return one of three different file times depending on the OS and
filesystem type, without the ability to determine which one is which. It
is replaced by metadataChangeTime() and birthTime().
[ChangeLog][QtCore][QFileInfo] Added QFileInfo::metadataChangeTime(),
which returns the time the file's metadata was last changed, if it is
known, and falling back to the same value as lastModified() otherwise.
On Unix systems, this corresponds to the file's ctime.
[ChangeLog][QtCore][QFileInfo] Added QFileInfo::birthTime(), which
returns the file's birth time if it is known, an invalid QDateTime
otherwise. This function is supported on Windows and on some Unix
systems.
Change-Id: I0031aa609e714ae983c3fffd1467bd8b3e3a593d
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Added flowId='name' to each message when using TeamCity logging format.
This is necessary to distinguish separate processes running in parallel.
[ChangeLog][QtTest] Added flowId to messages when logging in TeamCity
format. FlowId is used to distinguish logging from multiple processes
running in parallel.
Change-Id: I7f5046c1058ff02770404caa2c9b3a5398f97f6b
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Building iOS shared failed with link errors to the CoreGraphics framework
when building without harfbuzz-ng.
This patch adds the missing dependency.
Change-Id: If464417039172c2854e455563153057f0cea2a22
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Code was lagging a bit behind published standards.
Change-Id: I0fe07c0162a54d9eed8b539887496b9313b37aa4
Reviewed-by: Laszlo Agocs <laszlo.agocs@qt.io>
Reviewed-by: Jesus Fernandez <Jesus.Fernandez@qt.io>
Reviewed-by: Sean Harmer <sean.harmer@kdab.com>
The current code fetches currentOwner in the constructor, then sets up
a connection between a service watcher and _q_serviceOwnerChanged() in
initOwnerTracking(). But an owner change notification could arrive
after we fetch the current owner and before the connection is made.
In this case the owner change notification will be lost, and the
interface will keep reporting being invalid permanently.
The fix is to delay initializing currentOwner until after the connection
is made.
Task-number: QTBUG-62284
Change-Id: I92b9d61004e14fd2ee2543488740a542dc7a9b7a
Reviewed-by: Romain Pokrzywka <romain.pokrzywka@gmail.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The file was added to glibc 2.7 along with the functions we need (Added
2007-10-05). But they forgot to install the file until a month and a
half later (2007-11-17), which means it missed the 2.7 release
(2007-10-19).
Note that EFD_CLOEXEC wasn't added until glibc 2.9, so effectively glibc
2.9 is required.
Change-Id: I3868166e5efc45538544fffd14d773ba576fb793
Reviewed-by: Kai Koehne <kai.koehne@qt.io>
Reviewed-by: David Faure <david.faure@kdab.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Several of these variables/macros are no longer defined. We didn't
validate the preconditions on iOS, tvOS, or watchOS, so no
need to bother validating them on macOS either. Nor did we check the
OSStatus result on any platform anyways.
Task-number: QTBUG-62266
Change-Id: Id19ebead5d3a8a08a0a56d798f0173d0d893fc91
Reviewed-by: Tony Sarajärvi <tony.sarajarvi@qt.io>
The renameat2(2) Linux system call, new in 3.16, allows for the atomic
renaming of a file if and only if it won't clobber an existing
file. None of the Linux libcs have enabled this syscall as an API, so we
use syscall(3) to place the call.
If your libc has SYS_renameat2 but your kernel doesn't support it, we'll
keep issuing the unknown syscall, every time. Users in that situation
should upgrade (3.16 is from 2014).
On Darwin, there's a similar renameatx_np (guessing "np" stands for
"non-portable"). I haven't found anything similar on the other BSDs.
Change-Id: I1eba2b016de74620bfc8fffd14ccb4e455a3ec9e
Reviewed-by: David Faure <david.faure@kdab.com>
This is required so that one can use QSettings in situations that
temporary files or renaming may not work.
[ChangeLog][QtCore][QSettings] Added setAtomicSyncRequired(), which
allows one to use QSettings with config files in unwriteable directories
or in Alternate Data Streams on NTFS on Windows. This used to work
before Qt 5.4, but remains a non-default behavior due to the potential
of data corruption.
Task-number: QTBUG-47379
Change-Id: I81480fdb578d4d43b3fcfffd14d4f77112f0402f
Reviewed-by: Jesus Fernandez <Jesus.Fernandez@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Kai Koehne <kai.koehne@qt.io>
Reviewed-by: Leena Miettinen <riitta-leena.miettinen@qt.io>
We can't use MoveFile to do atomic commits on an ADS, so QSaveFile needs
to detect when the target name is ADS and then use the direct fallback
mode.
[ChangeLog][QtCore][QSaveFile] Saving to Alternate Data Streams on NTFS
on Windows is now possible, but requires setDirectWriteFallback(true).
Task-number: QTBUG-47379
Change-Id: I81480fdb578d4d43b3fcfffd14d4bc062ae1750d
Reviewed-by: Kai Koehne <kai.koehne@qt.io>
Reviewed-by: David Faure <david.faure@kdab.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
First, instead of getting 32 bits of random for every character, we can
get reasonable randomness with just 10 bits, so we can replace 3
characters per get32() call. A typical template with six X will only
need 64 bits of random data.
Second, using the PID is not random at all. The reason why mktemp used
to use it is irrelevant (probably because applications failed to seed
rand()). Either way, we don't need nor want the PID anymore. And because
the code is completely rewritten, the UCB copyright no longer applies.
Change-Id: I84e45059a888497fb55ffffd14d2d4cfe159abca
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This means the workaround for Linux's behavior now works even if
QTemporaryFile is disabled. We also avoid the creation of an otherwise
unused temporary file just so we can take its name.
Tested with tests/manual/filetest. Strace shows:
renameat2(AT_FDCWD, "a", AT_FDCWD, "/var/run/media/tjmaciei/B852-6088/a.EBG705", RENAME_NOREPLACE) = 0
renameat2(AT_FDCWD, "/var/run/media/tjmaciei/B852-6088/a.EBG705", AT_FDCWD, "A", RENAME_NOREPLACE) = 0
Before it was:
open("/var/run/media/tjmaciei/B852-6088/a.VuL412", O_RDWR|O_CREAT|O_EXCL|O_CLOEXEC, 0600) = 4
fcntl(4, F_SETFD, FD_CLOEXEC) = 0
lseek(4, 0, SEEK_SET) = 0
rename("a", "/var/run/media/tjmaciei/B852-6088/a.VuL412") = 0
stat("A", 0x7fff13260f00) = -1 ENOENT (No such file or directory)
close(4) = 0
renameat2(AT_FDCWD, "/var/run/media/tjmaciei/B852-6088/a.VuL412", AT_FDCWD, "A", RENAME_NOREPLACE) = 0
(the absolute path comes from fd61059d35
and it seems it was needed for Symbian)
Change-Id: I1eba2b016de74620bfc8fffd14ccc7c4ded009c8
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
So we can use it in QTemporaryFile, QTemporaryDir and QFile::rename()
[ChangeLog][QtCore][QTemporaryDir] The class now supports the "XXXXXX"
replacement token anywhere in the template, not just at the end. This
behavior is similar to what QTemporaryFile supports.
Change-Id: I1eba2b016de74620bfc8fffd14ccb645729de170
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
This makes the code more reusable by the unnamed file feature.
This commit removes the backwards compatibility in using sequential file
names if the first one failed. Since 5483b30868,
there are at least three random characters, so the chance of collision
is 1 in 52³ = 140608.
That commit also did not take a system failure into account. If we ended
up getting EEXIST for all attempts, we'd attempt on average 26³*53³ file
creations. For that reason, I've added an upper limit in the number of
attempts to create a file.
Change-Id: I1eba2b016de74620bfc8fffd14cc7e31c6e50558
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Instead of storing it in d->fileEntry. The QFSFileEngine methods may try
to use it before the file entry is filled with the actual file name and
that's no good. This change is using a reference to avoid keeping an
extra QString reference and to avoid going out of sync.
Change-Id: I1eba2b016de74620bfc8fffd14cca9e340e4b1e2
Reviewed-by: Kai Koehne <kai.koehne@qt.io>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Calling the parent version is still ok, but if you call the new one you
get a bit of benefit. Since we control the file name, we don't have to
worry about a case-changing renaming (by choice). We also know that the
file is a regular one, because we created it.
[ChangeLog][Important Behavior Changes][QTemporaryFile] rename() no
longer attempts to do block copying, as that usually indicates a mistake
in the user's code. Instead, either create the temporary file in the
same directory as the new name to be, or use QSaveFile.
Change-Id: I1eba2b016de74620bfc8fffd14ccaac0cdb9fe87
Reviewed-by: David Faure <david.faure@kdab.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
The rename(2) system call overwrites, so instead of using it, we try to
use the link/unlink pair. This works for regular cases, but can fail if
trying to change case in case-insensitive filesystems, if we're
operating on a non-Unix filesystem (FAT) or, on Linux, if the file
doesn't belong to the calling user (BSDs permit this). For those cases,
we fall back to rename(2).
That means there's a race condition if a new file is created there. But
we at least reduce the likelihood of that happening for regular files.
Change-Id: I1eba2b016de74620bfc8fffd14ccb38fd929e5aa
Reviewed-by: David Faure <david.faure@kdab.com>
In the Android style, we matched both the full CE_PushButton,
and the individual, decomposed CE_PushButtonBevel and
CE_PushButtonLabel (and similar for other controls) to the
same drawing code which draws both the label and the bevel.
When used together with the style sheet style, this caused
the label of the button to be drawn multiple times in
different locations.
For labels, we now skip the part which draws the actual control,
and for the non-textual parts of the controls, we skip drawing
the label.
[ChangeLog][QtWidgets][Android] Fixed label duplication for
buttons when using style sheets with the Android style.
Task-number: QTBUG-48639
Change-Id: I1a1cdb0a82870cebd0438c5c72199afc3192938c
Reviewed-by: BogDan Vatra <bogdan@kdab.com>
A nullptr QVariant should become a null QString or QByteArray,
since null strings have previous in our APIs represented the null value
in the absence of a dedicated null metatype.
Change-Id: I3b8f6386ece314d7c196959fbcf042c4fe0508a0
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Actually check that there's a T where ISO 8601 wants it (instead of
just skipping over whatever's there), with something after it; move
some declarations later; add some comments; and use the QStringRef API
more cleanly (so that it's easier to see what's going on). Simplify a
loop condition to avoid the need for a post-loop fix-up.
This incidentally prevents an assertion failure (which brought the
mess to my attention) parsing a short string as an ISO date-time; if
there's a T with nothing after it, we won't try to read at index -1 in
the following text. (The actual fail seen had a Z where the T should
have been, with nothing after it.)
Add tests for invalid ISOdate cases that triggered the assertion.
Change-Id: Ided9adf62a56d98f144bdf91b40f918e22bd82cd
Reviewed-by: Israel Lins Albuquerque <israelins85@yahoo.com.br>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
I was mistaken before, the "es" part for that version isn't optional, it
*must* be omitted.
Change-Id: I9e83d2317523fb0a905e40b95a56033cf693b93b
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
