The new loop does 32 bytes (16 code units) at a time
Change-Id: I8f261579aad648fdb4f0fffd155412a4d77428e9
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
In some cases the DEVMODE structure is not updated with the selected
paper size, whereas the ptPaperSize structure is always set to the right
paper size. Therefore we set the page size on the printer to this after
the page dialog is shown.
Change-Id: Ieafd486232aca6e930f73a8131b7196ddecea305
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
In order to get a correctly not truncated size we need the hint to
be calculated based on the screen it is about to be shown on.
This patch places some code in QWidgetPrivate that is used by
QMenu and QToolTip + can be used to solve similar problems in the
future.
Task-number: QTBUG-72306
Change-Id: I58c058761f71b4a7675b6a078be62aa813ead752
Reviewed-by: Morten Kristensen <msk@nullpointer.dk>
Reviewed-by: Oliver Wolff <oliver.wolff@qt.io>
The keyboard navigation with MovePageUp/Down and MoveEnd did not honor
disabled cells in all cases which lead to inconsistencies in the
navigation (esp. since MoveHome does honor them correctly).
Therefore make sure that all four move operations work consistent by
refactoring the code to use common functions.
Fixes: QTBUG-72400
Change-Id: I63fa3b626510d21c66f4f9b2b1bfb3261728ecaf
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Luca Beldi <v.ronin@yahoo.it>
Reviewed-by: Samuel Gaist <samuel.gaist@idiap.ch>
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@qt.io>
these were new on dev while the original migration happened on 5.9, or
came from new changes which hadn't adapted yet.
Change-Id: I5e48437061a97e6df6e93881c98471455e177631
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
there is no particular reason to exclude static builds from the default.
misses are cheap now, so it's fine if nothing is found.
this affects only the legacy pre-1.1 library names under windows.
Change-Id: I998b9f7bfcce42ec990a236bb44372c4d6b3f631
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
both the mingw and msvc build have a "lib" prefix on the libraries. this
makes the msvc build unconventional, so it needs an extra source. for
the mingw build, otoh, this is the expected setup, so the source used
for unix will work just fine.
this doesn't fix any actual bug, because mingw will apparently resolve
-llibfoo to libfoo.a even though only liblibfoo.a and libfoo.lib are
documented (on mingw.org). however, this mix of conventions is ugly and
should be avoided.
Change-Id: I32b1621e4ac15db1f071c08ced738bfdafdcc11b
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
don't run the openssl_headers test pointlessly when openssl-linked is
selected but its test fails. implementing this cleanly required creating
a separate openssl-runtime feature, including 'redirecting' the -openssl
option (which is just an alias for -openssl-runtime) to it.
simplify the openssl-linked conditions: while "anything but the value
that enables it" in the 'disable' field effectively means "don't
auto-detect it", it's better to be explicit about that.
Change-Id: I6b117cc50711bb64d090fcfdb89ff009c60ed86c
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
this allows compile-testing the specified headers with no further tests.
Change-Id: I268ff328deee221d9b92386fe2bd133b19a6f8e2
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
as it is now cheap to test for individual libraries and headers, split
the xcb_syslibs monster-library into its parts.
the compile test remains a single blob, though, as that didn't get any
cheaper. whether it's worth keeping it in the first place is debatable.
Change-Id: Id7cae7925bb4d77069437512abecf14feea749f2
Reviewed-by: Gatis Paeglis <gatis.paeglis@qt.io>
When building ANGLE, we need the shader compiler (fxc.exe), which is not
shipped with MinGW. Previously, we required an installed DirectX SDK.
For Windows versions >= 8, the DX SDK is also part of the Windows Kit,
so we also allow the user to specify the location of the Windows Kit.
We also detect fxc on 64-bit hosts now, and in newer SDK versions which
version the binary directory.
The detected binary is now exported by configure, so the ANGLE project
file does not need to duplicate the logic anymore.
Task-number: QTBUG-52487
Change-Id: I41a17992909041dd84291b69498195cc8b8fab8a
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
properly atomize the libraries and express their dependencies, and
adjust the project files accordingly.
note that we don't try to use any additional paths, as all SDKs we
currently support have built-in directx 11 support:
- msvc2013 comes with win sdk 8.1; that is also used for win7 targets
- mingw-64 5.3 (though this one is missing fxc, which is why the code
path for using an external sdk for that remains)
Change-Id: Ib44e389ef46567308293c2bbcad20a96e8ef70c7
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Reviewed-by: Oliver Wolff <oliver.wolff@qt.io>
like all qt_*.prf files (exceptions prove the rule), the configure
system counts as private api.
Change-Id: Iea3057445e430029ecd8449e604e2d5c499ae10a
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
This was added as a platform-specific implementation of the multi engine
in 2005, before this code was generalized, and it currently only has the purpose
of special handling loadEngine().
The way this was special handled was by creating a new QFontEngine for every
fallback family *every* time, never checking if the font engine already exists
in the cache (like the superclass implementation does). The result of this was
that if you had 500 fonts and each of them had 500 fallback fonts, and made
a loop that would load all of them, then you would get 250000 font engines.
At some point before this, we would run out of available handles and crash.
There shouldn't be any need to have special handling of fallback font loading
on Windows (i.e. all the platform specific parts should go through the normal
mechanisms in QPA), so lets just go through the superclass implementation instead.
[ChangeLog][Windows][Text] Reduced the number of font engines that are
created when loading new fonts, fixing crashes in some special cases where
a large number of fonts are created during a short period of time.
Fixes: QTBUG-70032
Change-Id: I05040dd458e820510685e8c6df8f31876d9bdb89
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
I think this may have been for some POST method form queries, but
obviously was missing something.
Task-number: QTBUG-72382
Change-Id: I59016776aeedf4b5599b3b44af70610babb0a61e
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Morten Johan Sørvig <morten.sorvig@qt.io>
Reviewed-by: Ryan Chu <ryan.chu@qt.io>
When the mouse is released, it gets the current window to send the
event to from the pointer position.
If the release happens out of the browser viewport, that window is null
and the release event essentially gets ignored. This change keeps
the last known QWindow object to send the event to until the release.
Also replace hardcoded emscripten events with enum names.
Task-number: QTBUG-71948
Change-Id: I354d14479a43489f210cca31d6b9e0f65d083bb0
Fixes: QTBUG-71948
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Mikhail Svetkin <mikhail.svetkin@qt.io>
Reviewed-by: Morten Johan Sørvig <morten.sorvig@qt.io>
Otherwise HarfBuzz will fall back to using default font sizes, e.g. 12.0
in the CoreText backend.
hb_coretext_font_create() already does the right thing, but we're not using
that function, we create the hb_font_t ourselves in _hb_qt_font_create().
The mismatch didn't matter for vector-based fonts as the metrics (presumably)
scale linearly, but for bitmap-based font such as Emojis they don't, and
we ended up with advances that were way too large (increasingly so the
larger the font point size).
Change-Id: I4781dda965c745853732026da91590d8506ec3f5
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
To enable LTCG, clang-cl.exe need "-flto" instead of "-GL". But if you
enabled LTO, the lib tool will have to change to "llvm-lib.exe" and the
link tool also need to change to "lld-link.exe", because msvc linker
will see these lib files as corrupted files. Leave QMAKE_LFLAGS_LTCG
empty because lld-link doesn't need any additional parameters.
Fixes: QTBUG-72402
Change-Id: Ia34d98bd1ad919d542ea3eac83944bf8c40ce0cd
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
C++11 added the new enum class key as well as enum struct. While the
former is likely the most known and used, the later can be used in the
same contexts and with the same effects.
Currently moc doesn't parse enum struct while it does for enum class.
This patch fixes this.
[ChangeLog][moc] moc now parses enum struct the same way as enum class
therefore that keyword can be used with the Q_ENUM macro as well as
Q_FLAG and Q_DECLARE_FLAGS.
Change-Id: Iaac3814ad63a15ee4d91b281d451e786b510449c
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
The documentation is not explicit enough about the content of the
returned list of capturedList nor the element 0 when calling captured or
its friends.
The first element (aka element 0) is the whole string captured when one
or more groups are used.
Change-Id: I3c59ebfc9f6d762dd4d8aaf8f5c0de24359f53d7
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Sze Howe Koh <szehowe.koh@gmail.com>
The current implementation of wildcardToRegularExpression doesn't
anchor the pattern which makes it not narrow enough for globbing
patterns. This patch fixes that by applying anchoredPattern before
returning the wildcard pattern.
[ChangeLog][QtCore][QRegularExpression] The wildcardToRegularExpression
method now returns a properly anchored pattern.
Change-Id: I7bee73389d408cf42499652e4fb854517a8125b5
Fixes: QTBUG-72539
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
in addition to the actual library resolution, also resolve the headers
belonging to the library, to validate the include path, and possibly
ensure that the right version of the library is present.
the "include" entries were moved out of the "test" objects, and renamed
to "headers". this cleanly permits libraries without compile tests.
the headers were not put into the sources, because the variance among
the includes is generally orthogonal to the variance among the
libraries.
note that this - like the library resolution - provides no support for
darwin frameworks. consequently, the opengl libraries are excluded from
the conversion on darwin.
similarly, wasm is excluded (centrally), because emcc is magic and would
need advanced wizardry to be dealt with.
Change-Id: Ib390c75371efa2badcfec9b74274047ce67c3e5a
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Reviewed-by: Paul Wicking <paul.wicking@qt.io>
Silence warnings about signed constants being shifted out of int range.
Change-Id: I5dc397de71f4de09e54ce3cbc0f8e3a1cf977b03
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
When these docs were written, the context.file and context.function
pointers were never null. But in commit d78fb442d7
(Qt 5.4), we made the logging pass null pointers in release builds. The
C standard does not say that passing null pointers is permitted. In
fact, it says "If no l length modifier is present, the argument shall be
a pointer to the initial element of an array of character type." and
that's pretty explicit that it needs to point to the initial element of
a string.
Fixes: QTBUG-72478
Change-Id: I4ac1156702324f0fb814fffd156f624ffefa1445
Reviewed-by: Topi Reiniö <topi.reinio@qt.io>
Change ebd3a13b80 introduced a new
QPaletter::PlaceholderText color role which causes the uic-generated
code not to compile when using Qt Designer embedded in Qt Creator with
older (5.9 LTS) kits. Generate a version check macro to fix this.
Change-Id: I6d9f7edb0c6047c2f64ef3357b29f91655c52aac
Fixes: QTBUG-72555
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Christian Ehrlicher <ch.ehrlicher@gmx.de>
Reviewed-by: André Hartmann <aha_1980@gmx.de>
fbbe8aba9d introduced a check for
MSVC_VER to qmake, which is not set in win32-clang-msvc,
causing the build to fail:
Mkspec does not specify MSVC_VER. Cannot continue.
Unable to generate output for: .../config.tests/verifyspec/Makefile
Extract a minimal msvc-based-version.conf which determines
MSVC_VER from QMAKE_MSC_VER for win32-clang-msvc and win32-icc.
Task-number: QTBUG-63512
Change-Id: Ia6de8c4b1aae2ae1962cf4e60e3e6d51fdbbbabe
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
... when QMAKE_DEFAULT_{INC,LIB}DIRS cannot be determined.
it would have been nicer to actually persist empty results, but cache()
won't do that, and fixing it doesn't seem worth the effort now.
Change-Id: I95d5645e40a0da572f0def16462703373eaeb804
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
these cannot be possibly correct, and might mislead.
Change-Id: Ie10531807978def04768e2429304949415cafb2a
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
... which are specified by full filepath, by making the de-duplication
consistent with that applied to libs specified with -l, that is, last
one wins.
the problem existed "forever", but it became more visible after the
recent configure changes.
fwiw, Win32MakefileGenerator is not affected, because it has the
opposite problem: it de-duplicates everything (including object files)
in "last one wins mode". it might make sense to change that as well.
Change-Id: Id7ef1d394fcc9d444450672c06a6f11af2b19eab
Reviewed-by: Robert Griebl <robert.griebl@pelagicore.com>
There was a comment in the code that said:
// It seems we need to use invertedAppearance for Left and right, otherwise, things look weird.
It's not clear what that was referring to, but in its current state,
a slider with invertedControls set to true will not behave as expected:
pressing the left arrow key will decrease its value instead of increasing it,
and vice versa for the right arrow key.
As stated in the documentation (and by its name), invertedAppearance only
controls the appearance of the slider, and not the effect of key events.
Remove the comment and use invertedControls instead.
Change-Id: I13296cbda9244413978ef0d7f0856065f74fd0bf
Fixes: QTBUG-25988
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@qt.io>
Don't do several tests at once in the test function. Instead, move
the extra tests to the data function. This makes it possible to easily
add a self-contained test (i.e row) for an upcoming fix.
Task-number: QTBUG-25988
Change-Id: I65c8d7620f01107f8f59c96896b1a641d97f5fdc
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@qt.io>
uses of this function (or the "files" stanza in configure.json) which
don't explicitly target windows don't specify the .exe extension, so we
need to add it automatically if it's missing.
Task-number: QTBUG-57436
Change-Id: I1994378399bc3466c32ee065e752516f42652975
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
size is size_t, so it's never less than zero.
Fixes: QTBUG-72286
Change-Id: Idd0c85a4e7b64f9c9c7dfffd156d404d0de5ed8d
Reviewed-by: Oliver Wolff <oliver.wolff@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Insufficient memory was allocated when asking GetDIBits() to convert to 32bit.
Fix allocation size and use a QScopedArrayPointer.
Fixes: QTBUG-72343
Change-Id: I45f79c913a243316e01bc6efed08e50ccc7d25f4
Reviewed-by: Eirik Aavitsland <eirik.aavitsland@qt.io>
Changed 0 to nullptr, used more C++-style casts, simplified some
code for searching a button, and changed foreach to range-based for
loop.
Task-number: QTBUG-44131
Change-Id: I211b12751b0e2591d1d14294c31b51d52bb4e3f6
Reviewed-by: Christian Ehrlicher <ch.ehrlicher@gmx.de>
QFileDialogPrivate::init() sets the working directory derived
from the URL passed in, causing the lastVisitedDir to be set.
This in turn prevented the restoreState() logic from setting
the directory retrieved from the file. Clear lastVisitedDir
in init() in case the initial URL was invalid.
Fixes: QTBUG-70798
Change-Id: I19084e24eb6d469330c4dd8c50495b4996279189
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@qt.io>
The Qt and CoreText positioning is now in sync.
Change-Id: I0cbb5b150d1bef732674b8d42c64a040773a62ab
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@qt.io>
Previously, the C locale was treated as English because each back-end
takes the locale's bcp47Name(), which maps C to en. However, the C
locale has its own rules; which QString helpfully implements; so we
can delegate to it in this case. Extended this to sort keys, where
possible. Clean up existing implementations in the process.
Extended tst_QCollator::compare() with some cases to check this. That
required wrapping the test's calls to collator.compare() in a sign
canonicalizer, since it can return any -ve for < or +ve for >, not
just -1 and +1 for these cases (and it'd be rash to hard-code specific
negative and positive values, as they may vary between backends).
[ChangeLog][QtCore][QCollator] Added support for collation in the C
locale, albeit this is only well-defined for ASCII. Collation sort
keys remain unsupported on Darwin.
Fixes: QTBUG-58621
Change-Id: I327010d90f09bd1b1816f5590cb124e3d423e61d
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Without this applications using the Offscreen QPA don't have
access to any fonts on macOS and thus cannot render text correctly.
Task-number: QTBUG-72335
Change-Id: I8e58c066365d0231d0993ad3b480d957a32f7f7b
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
