Try starting to create the directory at the end, not at the front. This
is the same way as the Unix implementation is doing. This avoids problems
like us trying to enter directories we are not allowed to read, which
might be due to access rights or due to sandboxing.
Change-Id: I67c1ed4bdc20a15b1af9b33aa48d59fea359da22
Reviewed-by: Christoph Schleifenbaum <christoph.schleifenbaum@kdab.com>
When one specified a prefix "/usr/qt5" on Windows for a non-Windows
target (e.g. QNX) that prefix was translated to "C:/usr/qt5" and
passed on to the Makefiles, for example as
"-Wl,-rpath,C:/usr/qt5/lib".
The reason was that we called $$absolute_path on the user-specified
prefix. However, absolute_path operates according to the rules of the
host operating system.
When cross-building, the prefix is an on-device path. Therefore we
must not attempt to make it absolute to the build directory.
The check whether we're cross-compiling looks a bit arcane, but we
cannot use $$qtConfEvaluate(features.cross_compile) at this stage of
configure. Instead, we use XSPEC (set up by qtConfOutput_prepareSpec)
and $$[QMAKE_SPEC] which is the right host mkspec.
Fixes: QTBUG-79214
Change-Id: Id8664f8512cf1d9e178054a38e72323d7929547d
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
[ChangeLog][Third-Party Code] Updated DNS public suffix list
Task-number: QTBUG-79418
Change-Id: I02dbe2b1f5b5f3e4a1ed4fde60ee71f5b0a50cb5
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
The compiler didn't complain at a QLocale::FormatType values being
passed for the int year parameters of the month-name functions, which
all have a default for their final QLocale::FormatType parameters. So
we didn't notice that the year parameter was missing until the bug was
reported.
Removed some code duplication by giving QCalendarModel a monthName()
method. Reworked QCalendarMonthValidator::text() to avoid repeated
calendar calculations (and use fewer braces).
This commit amends commit 2dee006216
Fixes: QTBUG-79495
Change-Id: Iad48c3b648a0139ab43511e6fb4e6a8f63a0495f
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
The patch adds ANDROID_BUILD_ABI_<abi> CMake options, which when enabled
will determine CMake to build the current project with the enabled ABI
settings.
When building with CMake and the official Android CMake toolchain
one needs to specify the Qt base directory as an argument to
CMAKE_FIND_ROOT_PATH, which contains the Android NDK sysroot
set by the toolchain.
CMake will consider directories that contain this base path as
valid directories to search packages. In the developer build case we
have to append "lib/cmake" because the Qt base directory passed as
CMAKE_FIND_ROOT_PATH will be the same directory as the developer
build base, and will not be considered.
Change-Id: I180502032c8ea1105bde2456252b367497f511d6
Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
Reviewed-by: BogDan Vatra <bogdan@kdab.com>
[ChangeLog][Third-Party Code] libjpeg-turbo was updated to version 2.0.3
Task-number: QTBUG-79420
Change-Id: I9f9b8b3a913fd5843759c0610f43b22c5bee67dc
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
There are too many semicolons. Delete some.
Fixes two warnings from Clang 9.0.1.
Change-Id: I363a6a2de9c075c03da62c58ad46828c04a95440
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
This header file intentionally puts a 'using namespace' into the global
namespace, the artful cleverness of which Clang doesn't properly appreciate.
Teach Clang a lesson by disabling the warning.
Change-Id: I9754ac5fc9d4c53654854082e1145d8b5fef186d
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
Released on October 4th.
Adds Windows names for two time zones, Qyzylorda and Volgograd.
Added languages Chickasaw (cic), Muscogee (mus) and Silesian (szl).
Norwegian number formatting has flipped back to using colon rather
than dot as time separator; it's flipped back and forth over the last
several CLDR releases. The dot form is present as a variant, the
colon form was long given as the normal pattern, then went away; but
now it's back as a contributed draft and that's what we pick up.
The MS-Win time-zone ID script was iterating a dict, causing random
reshuffling when new entries are added. Fixed that by doing the
critical iteration in sorted order.
Omitted locales ccp_BD and ccp_IN due to QTBUG-69324.
Task-number: QTBUG-79418
Change-Id: I43869ee1810ecc1fe876523947ddcbcddf4e550a
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
The Unicode data tables moved with QString and friends.
So did the locale data generated from CLDR.
This amends commit a9aa206b7b.
Change-Id: If12f0420b559dcb78993adc00e9f39751bca684a
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
While I was looking into a bug related to problems building the test
without GUI, I noticed a lot of spurious #include lines so tidied up a
bit. Split some long lines, while I was about it.
Change-Id: Id87eb6f612c6b174f8240dfe9c00e0929244fb6c
Reviewed-by: Albert Astals Cid <albert.astals.cid@kdab.com>
Part of the 5.14.0 third-party component update.
[ChangeLog][Third-Party Code] Updated double-conversion code to
upstream version 3.1.5.
Task-number: QTBUG-79418
Change-Id: I70c3890fcfa0606c462cc0fe702d0f62fd9c7279
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Ulf Hermann <ulf.hermann@qt.io>
Thanks to Dmitriy Purgin for pointing out the serialization one.
Task-number: QTBUG-79353
Change-Id: Ia3d750b17ddd8fbb7a83a55df7e4546ca78c358b
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
The qmake config for tst_QString tried to impose QT_NO_CAST_TO_ASCII
on it, but the source file explicitly #undef-s this symbol and its
friends. Leave the define commented out in the .pro so that a comment
can explain why it's no good.
Change-Id: I7620f4e104f0cdab05fdc246b903c40026e63d76
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Test QT_CONFIG(icu) in the code instead of testing qtConfig(icu) in
the profile and setting an extra define just to shadow what's already
defined. Also remove the matching define from qcollator.pro, whose
test code didn't use it.
Noticed while reviewing the conversions to CMake.
Change-Id: I19d3b1026b2a8f50ec424c450614e721500fd38a
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@qt.io>
QSqlTableModelPrivate::ModifiedRow::rec returns a const &
so use it instead doing a copy. QSqlRecord is cheap to copy constructor
but not having to do it is faster
Change-Id: Iad6e79fcdcdf380ce681fe9426436f8cb98be553
Reviewed-by: Andy Shaw <andy.shaw@qt.io>
See the comment in the test for details.
Change-Id: Ie3d356e476ba0419d304bccd396fc18a831a30cd
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
Reviewed-by: Eirik Aavitsland <eirik.aavitsland@qt.io>
This reverts commit 631efee293.
Reason for revert: This causes a problem when running on iOS 11 devices.
Change-Id: If5194989b8d7a9f4cf2d72e770fdaad754d7e236
Reviewed-by: Jani Heikkinen <jani.heikkinen@qt.io>
As described in QTBUG-72404, the android-clang mkspec has the gcc pch
style hardcoded for no reason. This change removes the respective lines.
[ChangeLog][qmake][Android] Remove gcc-style PCH directives from the
android-clang mkspec.
Fixes: QTBUG-72404
Change-Id: Iad42651e25ecce08eda7aa5fa8bbf531c9497896
Reviewed-by: BogDan Vatra <bogdan@kdab.com>
Precompiled headers were put in a directory which had exactly the same
name as the binary to be generated. This resulted in a failure with any
mkspec that used clang_pch_style. The g++-mkspec avoid this problem by
extending the directory name. This change adopts this technique for
clang mkspecs.
[ChangeLog][qmake] Fixed precompiled headers for the Clang compiler.
Fixes: QTBUG-72404
Change-Id: I471462e2bcb1e33f19d277c21acde0c04b1ffcd6
Reviewed-by: BogDan Vatra <bogdan@kdab.com>
0.3.4 is the newest tagged release. This updates to upstream
0354e1ab5a453e9913dcd5f87c2cfe9a2510dfda which has a change that
affects behavior of the ".\n" case in QTBUG-78870 (it will be
seen as a paragraph rather than a list item).
Task-number: QTBUG-78870
Change-Id: Ib01f9c1d3f71a39782608da071c2f42512845382
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
We use md4c for parsing markdown. It provides flags to control the
feature set that will be supported when parsing particular documents.
QTextMarkdownImporter::Feature is a fine-grained set of flags that
exactly match the md4c feature flags that we support in Qt so far.
QTextMarkdownImporter is a private exported class (new in 5.14).
We don't expect the corresponding flags in md4c to change in
incompatible ways in the future: the md4c authors have as much respect
for avoiding compatibility issues as we do, and likely will only add
features, not remove them.
We now enforce QTextMarkdownImporter::Features compatibility with
QTextDocument::MarkdownFeatures by setting them directly. We check
QTextMarkdownImporter::Features compatibility with md4c's #define'd
feature flags using static asserts, so that any hypothetical
incompatibility would be detected at compile time.
The enum conversion from QTextDocument::MarkdownFeatures to
QTextMarkdownImporter::Features is moved to a new QTextMarkdownImporter
constructor; thus the conversions from QTextDocument::MarkdownFeatures
to QTextMarkdownImporter::Features, and then to unsigned (in
QTextMarkdownImporter::import()) are adjacent in the same private class
implementation. If incompatibility ever occurred, we would need to
replace one or both of those with another suitable conversion function.
Change-Id: I0bf8a21eb7559df1d38406b948ef657f9060c67b
Reviewed-by: Vitaly Fanaskov <vitaly.fanaskov@qt.io>
Make parent an optional parameter to be consistent with the rest of the
API.
Change-Id: I5942fee69ca33bd87d4d14b6919d2b8d5b60c0b4
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: David Faure <david.faure@kdab.com>
When the first column is hidden or not visible in the current viewport,
it is not possible to deselect the current row.
Fix it by passing the correct column to
QItemSelectionModel::selectedRows() when testing if the current index is
selected.
Fixes: QTBUG-79092
Change-Id: I9d8082d2b29ad2f799156aee910c6ff6e3217771
Reviewed-by: David Faure <david.faure@kdab.com>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
stat() and friends expect a null-terminated C string. There is no way to
generate anything useful from a string that has null bytes in the
middle. It's important to catch this early, as otherwise, for example, a
QDir::exists() on such a path can return true, as the path is silently
truncated.
Extend the checks for empty file names to windows and add checks for null
bytes.
Change-Id: Ie9794c3a7c4fd57f9a66bdbbab8b45a08b6f9170
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The Apple software renderer is perfectly capable of being used when
compositing CA layers.
Change-Id: I3b78ff61a79869ecdb7bd431388041f2c124472e
Reviewed-by: Morten Johan Sørvig <morten.sorvig@qt.io>
qtbase/src/corelib/serialization/qjson_p.h:230:38: error: ‘void* memcpy(void*, const void*, size_t)’ copying an object of type ‘QJsonPrivate::qle_ushort’ {aka ‘class QSpecialInteger<QLittleEndianStorageType<short unsigned int> >’} with ‘private’ member ‘QSpecialInteger<QLittleEndianStorageType<short unsigned int> >::val’ from an array of ‘const class QChar’; use assignment or copy-initialization instead [-Werror=class-memaccess]
230 | str.length()*sizeof(ushort));
| ^
Change-Id: Ie58e7fe4bae3003227364012ad56ab23bd560d8c
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
Some clients such as QOpenGLWidget will end up drawing and flushing
during the resize event, which for GL will result in an immediate update
on the screen. The problem is that the underlying Core Animation layer,
and the window's frame, has not been visually updated yet to the new
size, so we end up drawing "ahead" of what the window server is showing
the user.
Ideally we'd be able to present the GL drawing in a transaction, in sync
with the drawing of the window frame, but this API is only available for
CAMetalLayer and CAEAGLLayer.
As a workaround we detect when the exposed size is out of sync with the
window geometry, and skip the flush until the exposed size has caught
up. We know this will happen eventually as AppKit will always ask us
to display after a resize.
Change-Id: I1739ac8878b3fc6820a55dd017ddd170fd5f55d6
Fixes: QTBUG-79139
Reviewed-by: Morten Johan Sørvig <morten.sorvig@qt.io>
We were disabling wantsBestResolutionOpenGLSurface whenever we detected
the Apple software renderer, but this isn't needed when layer-backed,
and did in fact result in the exact same visual result as the bug the
code was working around -- only rendering to a quarter of the viewport.
We now apply the workaround only when software rendering is combined
with surface-backed views.
The logic has also been improved to not rely on string comparison to
look for the software renderer, but instead uses the renderer ID that
the context provides.
Since tweaking the wantsBestResolutionOpenGLSurface is only relevant
when using a window for GL rendering the logic has been moved into
QCocoaGLContext.
Change-Id: I021aaefbb7a9782bc8ee3c9703da246510326d50
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
This flag is needed by LLDB & simple perf tools to locate the right binary.
Change-Id: Iffa1b0678663cfb9d1d699da5ad6fe672863918c
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@qt.io>
Building examples inside a Qt source tree is considered unsupported.
Instead of checking, whether the build is done "in source" inside the
source directory, we also have to check for "shadowed example builds" as
they are done by Creator. An example would be:
qtbase/examples/widgets/widgets/build-wiggly as a shadow build for
qtbase/examples/widgets/widgets/wiggly.
Fixes: QTBUG-76237
Change-Id: Iceb88af006fad249a4c13fa0b0345cf3f7086252
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Pass the dialog as parent to the event filter class
in QColorDialogPrivate::_q_pickScreenColor().
Fixes: QTBUG-53469
Change-Id: I9110c19a8f49a545a0fbf7cfdb3ded70fea4dcce
Reviewed-by: Christian Ehrlicher <ch.ehrlicher@gmx.de>
Make it readable by using names instead of mere indices for the stages.
There is an important fix in there as well: when in a render pass, only
resource for VERTEX and FRAGMENT are taken into account, while in a compute
pass those are skipped. This ensures that we do not send messages to a nil or
invalid MTLRender/ComputeCommandEncoder. (nil would not be an error but the
other is fatal)
Task-number: QTBUG-79447
Change-Id: Ibef108cb7c82b5b0fdd2a299cd89fbebe8c3606a
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
