Neither the Intel compiler nor Visual C++ have a dedicated switch to
enable F16C support, like GCC and Clang do. So we used the AVX switch
for that in commit 8241d51f70, as it was
the closest, lowest denominator. That was incorrect and insufficient.
The Intel compiler silently miscompiles the intrinsics with -xAVX,
making calls to out-of-line functions like _mm_cvtps_ph, which don't
exist. So we actually have to use AVX2 support to generate correct code.
That might be a problem later, since Ivy Bridge supports F16C but not
AVX2.
Visual C++ is able to generate F16C code with just -arch:AVX.
Either way, since there's no dedicated command-line switch, there's also
no dedicated preprocessor macro. We're using __AVX2__ for both
compilers, as that's a sufficient condition to indicate a processor that
supports F16C.
Change-Id: I27b55fdf514247549455fffd14b205b8d8b86da7
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
MSVC, Apple's Clang and Clang prior to 3.9 do not recognize _cvtss_sh
and _cvtsh_ss. So expand the operation to use directly the packed
intrinsics.
Change-Id: I27b55fdf514247549455fffd14b2046fd638593d
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
GCC is wrong. Type-punning is when you read something of a given
type as something else. We're not doing that, as it's only read
as integer.
qnativesocketengine_unix.cpp:1011:79: error: dereferencing type-punned pointer will break strict-aliasing rules [-Werror=strict-aliasing]
Too bad my plan for a good C++ solution was foiled by glibc
developers.
Change-Id: I27b55fdf514247549455fffd14b1a27667745e94
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
qsysinfo.h(235): error #1786: enum "QSysInfo::MacVersion" (declared at line 156) was declared deprecated ("Use QOperatingSystemVersion")
Take this opportunity to merge the two groups.
Change-Id: I27b55fdf514247549455fffd14b1c2a1d8eab869
Reviewed-by: Jake Petroules <jake.petroules@qt.io>
c0157a9f03 was incomplete.
Change-Id: I27b55fdf514247549455fffd14b158f54403663d
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Aligns with EPERM (Operation not permitted) and ENOENT (No such file
or directory), and is what errno is set to on macOS when opening the
/dev/tty device when running inside e.g. Xcode, where isatty() will
return true.
Change-Id: I09b88eaa3ff611d95ab37f0ff4df9aaaca52747d
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
If we do not the fontDef of the multifont will be the default 0.
Task-number: QTBUG-59443
Change-Id: Ib223517975b2a57b2371e309d12cd8f918d30825
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@qt.io>
Set stretch to always have no transform on raw fonts.
Task-number: QTBUG-59799
Change-Id: Ibfacc5c247e0b4a8410572e207f09f6e67b74f9d
Reviewed-by: Konstantin Tokarev <annulen@yandex.ru>
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@qt.io>
On some ARM devices the font glyph generation is broken
Add an environment variable to enable workaround_brokenFBOReadBack
in QOpenGLContext, to fix font rendering on such devices as
Mali and Adreno
Change-Id: I9cc99ecb8b71a35bc369ec9dd11b877016b1179e
Reviewed-by: Laszlo Agocs <laszlo.agocs@qt.io>
Test failed on QNX 7, even though alloca is available. On QNX7,
it's a macro that expands to a line with NULL, but without define
for it.
alloca.cpp:44:5: error: 'NULL' was not declared in this scope
Task-number: QTBUG-59700
Change-Id: I3631d139990020a3adbab8b72e49929b6e721e80
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
If QNAM's general policy is 'Manual' and QNetworkRequest has neither
policy set not FollowRedirectAttribute - do NOT set this 'Manual' policy -
its implicitly implied. This fixes previously unnoticed auto test failure
(was blacklisted) and also makes QNetworkRequest::operator == work correctly.
Change-Id: If17c9af4baf8a470659f82d1a40488078ea8ede0
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
Changed the file position and offset types so they can properly handle
files larger than the 32bit limit
Task-number: QTBUG-59493
Change-Id: I00e1741c7682c4c79f35fef808fe1ea26e67c8b5
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
Reviewed-by: Eirik Aavitsland <eirik.aavitsland@qt.io>
Reviewed-by: Hendrick Melo
When the last QSqlDatabase object is destroyed the destructor implicitly
calls close() to release the database connection.
Task-number: QTBUG-59919
Change-Id: I04c15c4999cdaaa8800a44a1a1006f977a90d8a6
Reviewed-by: Samuel Gaist <samuel.gaist@edeltech.ch>
Reviewed-by: Andy Shaw <andy.shaw@qt.io>
The QCoreTextFontEngine constructor already sets the member (like it should).
Change-Id: I6e59ff36c439857418a636de4056c20d2f65e8d4
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
The round trip via a CGFontRef to a CTFontRef so that we could pull out
the kCTFontURLAttribute of the font makes no sense, as the input is just
raw font data. None of the tst_QRawFont tests are able to pull out a
URL, and none of the other platforms where FreeType is available resolve
a filename from raw fonts.
Without this change the tst_QRawFont test has 13 failed test cases and
spits out 1250 QWARNs. With the patch all tests pass without warnings.
Change-Id: I17965e7405d7161958ff0870825ce8190598288f
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
So that the logic can be used without subclassing the FreeType font database.
Change-Id: Ib1ac0e1948ce18003b8eda0c1e8fbb3c7e6c8360
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
The Core Text font database can produce both Core Text and FreeType font
engines. Refactor the code a bit so that the actual factory methods that
differ between the two stand out, and do not require a granular runtime
check in each method.
Change-Id: Ib70f76f4a9001a8108d87c1101a50699a6ea8f55
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
More fallthrough-are-errors fixed.
Change-Id: I9a6cb6efe988400ed3f9cb95d1e426dac317e6c4
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
QOpenGLWidget is not supported on all platforms. Skip tests on those.
Change-Id: I0f9500553427903f20d248acaa20803276e3ab00
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
files is not a test function.
Change-Id: I6d23dac5d1c87bd35961406dd62cc3cf37d652e3
Reviewed-by: Oliver Wolff <oliver.wolff@qt.io>
Reviewed-by: Leena Miettinen <riitta-leena.miettinen@qt.io>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Fix QT_SCALE_FACTOR usage on macOS. Follow-up to
2d2d9078
QRasterBackingStore should account for native scaling
only. Any Qt scaling will have already been factored
into the size argument.
Change-Id: I26a67addfcbec3d45f4ed87f03b8dd79fd99cb62
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
Cherry picked:
this change was first merged into dev, but now we also need it in 5.9
to enable SSL socket tests on the new CI VMs (macOS 10.11, 10.12).
As we do not merge dev->5.9, we need this cherry-pick.
Since day one Secure Transport socket has two annoying problems on macOS:
when we call SecPKCS12Import, we indeed import certs and keys into the default keychain
and also (which is more serious) later a dialog can pop up, asking for permission
to use a private key (this is especially annoying if you're running SSL autotests or
have a server application). Apparently, it's possible to work around those problems
if we create our own (temporary) keychain and pass it in the 'options' parameter
to SecPKCS12Import.
[ChangeLog][QtNetwork] Allow QSslSocket to use a temporary keychain on macOS.
Task-number: QTBUG-56102
Change-Id: Ic3a56c905100dc80d907a25fe6ebfa232dcf5b9e
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
(cherry picked from commit 17927392cf)
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
The platform backingstore might need access to the window that the
backingstore was created for, e.g. for makingCurrent to release OpenGL
resources. In that case leaving it to the QRasterWindowPrivate destructor
would be too late, as the QWindow was gone.
This was seen on iOS, where the backingstore inherits QRasterBackingStore,
and uses composeAndFlush to composit via GL. The raster backingstore
cleans up these GL resources in its destructor, so the QIOSBackingStore
destructor makes sure that the GL context is current for the window,
resulting in a crash since the window is long gone by then.
Change-Id: I5a22597842819f0fe3b580856b9e75e4fab32ae5
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
The FreeType font engine setup is spread out between factory functions,
constructors, and init() functions, so let's at least try to share as
much as possible of it to make it easier to reason about and possibly
refactor in the future.
Change-Id: Ic39353d2b3111024e0589a70211bac80feb8498e
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@qt.io>
Previously, only bool isDefault was used, which did not discriminate
changes in ButtonColor. Use the color name instead.
Task-number: QTBUG-59850
Change-Id: I1e006f98371a5f2039dcca1207addc0396e7c1e5
Reviewed-by: J-P Nurmi <jpnurmi@qt.io>
qCInfo must never be used. For qt.* logging categories Debug is disabled
while everything else is enabled by default. This means that doing qCInfo
is equivalent to putting a qDebug which is not acceptable.
Amends 3e0355014e
Change-Id: I428e620f12afa324cf6af8dbe3912a55189a38a9
Reviewed-by: Andy Shaw <andy.shaw@qt.io>
Targets were always exported as .lib, for all windows compilers
which is correct for msvc, but not for mingw. Hence use
QMAKE_EXTENSION_STATICLIB to switch between .lib and .a
Task-number: QTBUG-59906
Change-Id: I948f5dd96e0fb46d679c474b7beececc379ad436
Reviewed-by: Volker Krause <volker.krause@kdab.com>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Kevin Funk <kevin.funk@kdab.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
When we call realloc, the alignment of the new block may be different
from the old one. When that happens, we need to memmove the data to the
new position, before we start overwriting things.
Task-number: QTBUG-59804
Change-Id: I27b55fdf514247549455fffd14b07ea78918a3d0
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
shouldShowFilename method has been deprecated since macOS 10.6.
In 10.11 the behavior of this method has been broken, causing
files containing metadata (e.g. audio) to be incorrectly filtered out,
displayed them as disabled in file dialog even though they shouldn’t be.
This erratic behavior applies also to NSOpenPanel setAllowedFileTypes
if set to anything but nil. This has been confirmed to be a known bug
in Cocoa.
Using shouldEnableURL solves this problem and also removes risk
of breaking compatibility with future SDKs.
Renamed and simplified private method isHiddenFile to
isHiddenFileAtURL.
Renamed to be consistent with other Cocoa file query methods.
Simplified to return true only if the file is hidden as the name of the method
implies. Previously it might have returned true also if the file has
not existed which was in fact very metaphysical answer.
Check for presence of the file is done by other method before calling
this one and the scope of the method is limited to one source file.
Task-number: QTBUG-57527
Change-Id: I2fded712d4e7098eb444331d92e38cee71655100
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
GCC 7 defines __has_cpp_attribute even when invoked as "gcc" (possibly,
Clang does the same, according to a comment in the code, did not test
myself).
Hence, define the fallthrough declaration (as C++11 attributes)
only when compiling as C++, otherwise we pick them up even in C mode,
and they cause build failures.
Change-Id: I3f13205e014bb1dea59ee3664b29111521a7eae3
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
Remove Xutil.h include from qxcbmime.cpp as it does
not use any Xlib APIs. Using API from Xutil.h requires
Xlib as noted in Xutil.h:
/* You must include <X11/Xlib.h> before including this file */
Everywhere else we do check for presence of Xlib, before
including Xutil.h
And remove some useless #undef(s)
Task-number: QTBUG-39665
Change-Id: Ibfd2341338fe7e902b47eae2df6b9dafe4ab962d
Reviewed-by: Alexander Volkov <a.volkov@rusbitech.ru>
Reviewed-by: Laszlo Agocs <laszlo.agocs@qt.io>
