These APIs started out as private APIs in qnumeric_p.h, but have since
been made pseudo-public in qnumeric.h. The qnumeric_p.h versions just
forward to the qnumeric.h ones, so just use the latter.
This is in preparation of removing the {add,sub,mul}_overflow
versions, which, despite being defined in the unnamed namespace, don't
sport the q prefix, so potentially clash with global symbols.
The change is a simple textual search and replace, manually excluding
qnumeric_p.h.
Picking to 6.5 to avoid cherry-pick conflicts going forward.
Pick-to: 6.6 6.5
Change-Id: Ic0f7c92f7c47923317109e8a9dc06fa66bdff2c2
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
CMakeLists.txt and .cmake files of significant size
(more than 2 lines according to our check in tst_license.pl)
now have the copyright and license header.
Existing copyright statements remain intact
Task-number: QTBUG-88621
Change-Id: I3b98cdc55ead806ec81ce09af9271f9b95af97fa
Reviewed-by: Jörg Bornemann <joerg.bornemann@qt.io>
Replace the current license disclaimer in files by
a SPDX-License-Identifier.
Files that have to be modified by hand are modified.
License files are organized under LICENSES directory.
Task-number: QTBUG-67283
Change-Id: Id880c92784c40f3bbde861c0d93f58151c18b9f1
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Jörg Bornemann <joerg.bornemann@qt.io>
The testing of infinity and NaN somewhat duplicated existing tests in
tst_QNumeric and, in any case, belongs there.
Change-Id: I6b5d1ff9767daf8e4bbe0025d3efab3d74ed35de
Reviewed-by: Miikka Heikkinen <miikka.heikkinen@qt.io>
Reviewed-by: Sona Kurazyan <sona.kurazyan@qt.io>
Replaces 0, 0.0, 1, 1.0 and 2.0 with zero, one and two constants of
the templated floating-point type, to avoid type-conversions which
might have meant the test only checked doubles, or similar.
Change-Id: Id9d4488f0cc7226b5b8958d2a204a76fe5ae469d
Reviewed-by: Sona Kurazyan <sona.kurazyan@qt.io>
- current INTEGRITY development pack don't support denormals for float and double.
All values are rounded to 0.
Task-number: QTBUG-99123
Pick-to: 6.2 6.3
Change-Id: Iaaacdc4210c7ac2ec3ec337c61164a1ade0efb01
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
This hasn't worked for some time. It's not in our CI and I don't think
it was working at all. When I tried to build it, I ran into several
problems with C++17 and an Internal Compiler Error I did not have any
interest in working around.
After discussing with the Intel compiler team, it was decided that
fixing those issues in the old compiler is not going to happen. Instead,
their recommendation is to adopt the new LLVM-based compiler, which
the last commit added support for.
This commit does not remove qmake support for the old ICC. It's possible
someone is using qmake with a non-Qt6 project and ICC.
Change-Id: Icb2516126f674e7b8bb3fffd16ad6350ddbd49e5
Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
They were missing.
I also wanted to verify that it's true for sub-normal values.
At the same time, relocate qfloat16's implementation of qFuzzyIsNull()
to between those of qFuzzyCompare() and qIsNull(), since its apparent
absence initially confused me.
Change-Id: I9637c0070e754d16744c76fc9f846596257c6a63
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Remove the qmake project files for most of Qt.
Leave the qmake project files for examples, because we still test those
in the CI to ensure qmake does not regress.
Also leave the qmake project files for utils and other minor parts that
lack CMake project files.
Task-number: QTBUG-88742
Change-Id: I6cdf059e6204816f617f9624f3ea9822703f73cc
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Kai Koehne <kai.koehne@qt.io>
Complete search and replace of QtTest and QtTest/QtTest with QTest, as
QtTest includes the whole module. Replace all such instances with
correct header includes. See Jira task for more discussion.
Fixes: QTBUG-88831
Change-Id: I981cfae18a1cabcabcabee376016b086d9d01f44
Pick-to: 6.0
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
Four code-paths that weren't tested are simply those with the
parameters swapped from code-paths we did test. In any case, the
float-distance between values should be symmetric, so test that.
Task-number: QTBUG-88183
Change-Id: I2060eb77b1abada5b0fd5f4557dbb1761c5cfd02
Reviewed-by: Andrei Golubev <andrei.golubev@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
We were missing 64-bit signed mul_overflow on 32-bit platforms and in
those where we did have it, the detection was awful (both for signed and
for unsigned). So if one of the parameters is a constant, we can
simplify the code generated.
Change-Id: Ia99afccf0c474e20b3ddfffd162a60d269eb1892
Reviewed-by: Andrei Golubev <andrei.golubev@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Modify special case locations to use the new API as well.
Clean up some stale .prev files that are not needed anymore.
Clean up some project files that are not used anymore.
Task-number: QTBUG-86815
Change-Id: I9947da921f98686023c6bb053dfcc101851276b5
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Use pro2cmake with '--api-version 2' to force regenerate
projects to use the new prefixed qt_foo APIs.
Change-Id: I055c4837860319e93aaa6b09d646dda4fc2a4069
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
There is no reason for keep using our macro now that we have C++17.
The macro itself is left in for the moment being, as well as its
detection logic, because it's needed for C code (not everything
supports C11 yet). A few more cleanups will arrive in the next few
patches.
Note that this is a mere search/replace; some places were using
double braces to work around the presence of commas in a macro, no
attempt has been done to fix those.
tst_qglobal had just some minor changes to keep testing the macro.
Change-Id: I1c1c397d9f3e63db3338842bf350c9069ea57639
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Do this by templating the floating-point tests, which removes some
existing duplication as well as avoiding new duplication. Did some
renaming in the process. Added some tests of fuzzyCompare that come
closer to its boundary. Increased number of tests from 69 to 97. Use
std::numeric_limits to replace assorted hard-coded constants and old
C-library boundary-value macros.
It turns out MSVC's float conflates quiet and signaling NaN (although
MinGW's doesn't); and WebAssembly's old fastcomp compiler conflates
NaNs for both float and double; so XFAIL the test for distinct NaNs in
those cases.
Change-Id: I0a1c0d2f68f75d51b8cda9e3ddfe7fa9c190a3e2
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Erik Verbruggen <erik.verbruggen@me.com>
One of our compilers for emscripten coerces all signaling NaNs to
quiet ones, so won't do any actual signaling. Anyone relying on them
to do so shall be disappointed, so it's better that they know about it
at compile-time - or, at least, have the ability to find it out.
Put the signaling NaN producers (and remaining (test) code using them)
under the control of a feature that's disabled when numeric_limits
claims double has no signaling NaN. Assume the bootstrap library
doesn't need signaling NaNs. Sadly, until C++20 <bit>, there's no
contexpr way to test that alleged signalling and quiet NaNs are
actually distinct.
Added some auto-tests for signaling NaN, including that it's distinct
from quiet NaN. Any platform on which the last fails should disable
this feature.
Task-number: QTBUG-77967
Change-Id: I57e9d14bfe276732cd313887adc9acc354d88f08
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Separate quiet NaN from infinity and expand the nan-with-payload test
to a general test that bits outside the exponent don't break qIsNan().
Generally test more thoroughly and systematically.
Tests for signalling NaN shall follow.
Change-Id: Ib35dabacc8ebcc9a0761df38f6f419f0398d0e20
Reviewed-by: Erik Verbruggen <erik.verbruggen@me.com>
The code had min and max of an integral type and tested min + min / 2,
which naturally overflowed, provoking a compiler warning. The test
was meant to be testing min - min / 2; but min is even, so this is
just min / 2; and doubling that won't overflow (which is what the test
is about). As it happens, min + min / 2 is in fact max - max / 2,
which *would* be a good value to test, since max is odd. So add a
test for that and remove the broken test.
Change-Id: Iec34acbf0d5d7993d41ff844875dc10480b8eb1f
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The rules of std don't permit us to add an overload for
fpclassify(qfloat16), so we need our own equivalent that we *can*
overload. Deploy it in the few places we use fpclassify().
Extended qnumeric's testing to cover qFpClassify().
Change-Id: Ie5a0a5cc24599d1571404c573d33c682b0d305a5
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Paul Wicking <paul.wicking@qt.io>
Renamed the test, since it covers both, verified slightly more and
added checks that QCOMPARE() copes as intended. Fixed some minor
coding-style defects in the process.
Change-Id: I49c2ffa0568a29e9e4b7f7395d4cacdeb0401da0
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Commit 29bc68cf16 added support for
unsigned and commit 5ff7a3d96e later added
support for int. This commit adds support for qsizetype, which isn't int
on 64-bit platforms.
We do this by reorganizing the code and using the generic version of
__builtin_{add,sub,mul}_overflow from GCC 5 and Clang 3.8, which ICC 18
seems to support now too on Linux. That leaves older versions of GCC and
Clang, as well as MSVC, ICC on Windows, and the GHS compiler, to use the
generic implementations, as I've removed the assembly code those
versions of GCC and Clang on x86 are now uncommon.
Note: any older version of ICC probably breaks. We only support the
latest.
Change-Id: I9e2892cb6c374e93bcb7fffd14fc11bcd5f067a7
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
I wrote a script to help find the files, but I reviewed the
contributions manually to be sure I wasn't claiming copyright for search
& replace, adding Q_DECL_NOTHROW or adding "We mean it" headers.
Change-Id: I7a9e11d7b64a4cc78e24ffff142b506368fc8842
Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com>
From Qt 5.7 -> tools & applications are lisenced under GPL v3 with some
exceptions, see
http://blog.qt.io/blog/2016/01/13/new-agreement-with-the-kde-free-qt-foundation/
Updated license headers to use new GPL-EXCEPT header instead of LGPL21 one
(in those files which will be under GPL 3 with exceptions)
Change-Id: I42a473ddc97101492a60b9287d90979d9eb35ae1
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com>
In C++, signed overflow math is Undefined Behavior. However, many CPUs
do implement some way to check for overflow. Some compilers expose
intrinsics to use this functionality. If the no intrinsic is exposed,
overflow checking can be done by widening the result type and "manually"
checking for overflow. Or, for X86, by using inline assembly to use the
CPU features.
Used in QtQml.
Change-Id: I2ef2523ccaa98f6757a45e24862a2fa730a26bb0
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The previous implementations did not check the full mantissa. The result
was that certain NaN values were seen as +/-Infinity.
A nice benefit is that the generated code for this implementation is also
faster.
Task-number: QTBUG-47692
Change-Id: I1507ec579ccd9a2ab97da8cf83dabbc5d6e28597
Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com>
The code still produces Internal Compiler Errors in release mode.
Task-number: QTBUG-46344
Change-Id: I86d3608b13a197a0b65b83829d1512203e1578f8
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@theqtcompany.com>
Most processors have carry flags which they set on addition overflow, so
it's a good idea to access them whenever possible. Most of them also
have widening multiply instructions that can be used to detect overflow
of the non-widening version.
Tested to compile on:
Architecture Compiler
x86 GCC 4.9, GCC 5*, Clang 3.6*, ICC 16 beta
x86-64 GCC 4.9, GCC 5*, Clang 3.6*, ICC 16 beta
x86-64 ILP32 GCC 4.9, GCC 5*, Clang 3.6*
IA-64 LP64 GCC 4.8
ARMv7-A GCC 4.9, Clang 3.6*
AArch64 Clang 3.6*
MIPS GCC 4.9, Clang 3.6*
MIPS64 GCC 4.9, Clang 3.6*
PowerPC GCC 4.9, Clang 3.6*
PowerPC 64 GCC 4.9, Clang 3.6*
SPARC Clang 3.6*
SPARCv9 Clang 3.6*
[*] supports the intrinsics
If the compiler does not offer a way to detect an overflow, we do it by
hand. For unsigned additions, that's easy, since the C++ language
specifies the behavior of the overflow. That's also the reason why this
code is implemented only for unsigned integers.
For the multiplication, if the compiler does not support widening
multiplications, we do it with a division instead. This is necessary for
GCC < 4.5 and compilers not compatible with GCC or MSVC.
Change-Id: I049a653beeb5454c9539ffff13e637de0f1338c1
Reviewed-by: Ulf Hermann <ulf.hermann@theqtcompany.com>
Reviewed-by: Alex Trotsenko <alex1973tr@gmail.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The keyword no longer has a meaning for the new CI.
Change-Id: Ibcea4c7a82fb7f982cf4569fdff19f82066543d1
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
The Intel compiler defaults to "fast math" mode, which is why those
tests had been failing. So for the test that is trying to check whether
we conform to IEEE strict requirements, turn on strict requirements.
Change-Id: I02f8426b1c8e4241ac10ffff13e8efa224f313b2
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
Qt copyrights are now in The Qt Company, so we could update the source
code headers accordingly. In the same go we should also fix the links to
point to qt.io.
Outdated header.LGPL removed (use header.LGPL21 instead)
Old header.LGPL3 renamed to header.LGPL3-COMM to match actual licensing
combination. New header.LGPL-COMM taken in the use file which were
using old header.LGPL3 (src/plugins/platforms/android/extract.cpp)
Added new header.LGPL3 containing Commercial + LGPLv3 + GPLv2 license
combination
Change-Id: I6f49b819a8a20cc4f88b794a8f6726d975e8ffbe
Reviewed-by: Matti Paaso <matti.paaso@theqtcompany.com>
Extending this to stock QNX as well since it is not
BlackBerry 10 specific.
- tst_QNumeric::floatDistance()
- tst_QNumeric::floatDistance_double()
- tst_QtJson::testNumbers_2()
- tst_QtJson::toJsonLargeNumericValues()
- tst_QtJson::parseNumbers()
Task-number: QTBUG-37066
Change-Id: If0e5d4fbefac5e8a0efed8ef8b1b7655ff6e7766
Reviewed-by: Fabian Bumberger <fbumberger@rim.com>
These tests seem to fail because denormalized numbers are not
supported on QNX yet, so marking them as expected failures.
- floatDistance(denormal)
- floatDistance_double(denormal)
Task-number: QTBUG-37094
Change-Id: I79dbc78da6e9bef8466264fd2cab4af0ee8b868f
Reviewed-by: Fabian Bumberger <fbumberger@rim.com>
This function is useful if a floating point comparison requires a
certain precision. The return value can be considered as the precision.
For instance, if you want to compare two 32-bit floating point values
and all you need is a 24-bit precision, you can use this function like
this:
if (qFloatDistance(a,b) < (1 << 7)) { // The last 7 bits are not
// significant
// precise enough
}
Task-number: QTBUG-32632
Change-Id: I020a58d2f9f9452ac3c510b4bb560dc806f0d93c
Reviewed-by: Shawn Rutledge <shawn.rutledge@digia.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
ICC apparently optimises 0 * anything directly to zero, even when it
should be doing a multiplication to conform to IEEE requirements. GCC in
fast-math mode does the same, but that also makes the rest of the
function unreliable, so we try to turn off fast-math mode if we can.
Task-number: QTBUG-22340
Change-Id: I0e3c5f4927b0a6bcb3189bb156c18843fc4b29b9
Reviewed-by: Caroline Chao <caroline.chao@digia.com>
