QVector2D/QVector3D/QVector4D: ensure well-defined behavior

The purpose of QVectorND classes is to store N floats packed
together.  One of their usecases is to build arrays of them, then
using them as an array of floats (e.g. when uploading data to
OpenGL).

The design of the class however has a major problem: using separate
members does not guarantee that the compiler does not insert padding
between them (although that could be static-asserted). What's worse,
the implementation of operator[] just does pointer arithmetic on the
first member of the class; that's undefined behavior, and will
trigger ASAN warnings in the future [1].

Solve both problems by using an array of floats instead of
individual x/y/z/w members. Now the compiler is not allowed to
insert hidden padding any more, and makes operator[] well-defined.
However this might be a BIC (IF the compiler added paddings in the
past): hence, add static_asserts checking that the memory layout of
the classes hasn't changed.

For good measure, also add static_asserts checking

1) that the class is standard_layout, so it's safe to
reinterpret_cast it to its first (and only) member;

2) that there's no padding at the _end_ of the class.

Note: an alternative solution to the operator[] problem could've
been leaving the class untouched, and reimplementing the operator
via a switch. Unfortunately none between GCC, clang and MSVC
compile away the switch, dramatically pessimizing the code.

[1] https://github.com/google/sanitizers/wiki/AddressSanitizerIntraObjectOverflow

Change-Id: Iec00ffb6044c58cf728de1754a780068f88152cb
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>

1 files changed, 37 insertions, 37 deletions

diff --git a/src/gui/math3d/qvector2d.h b/src/gui/math3d/qvector2d.h
index 2af5132665..88d8bc199e 100644
--- a/src/gui/math3d/qvector2d.h
+++ b/src/gui/math3d/qvector2d.h
@@ -123,7 +123,7 @@ public:
     operator QVariant() const;
 
 private:
-    float xp, yp;
+    float v[2];
 
     friend class QVector3D;
     friend class QVector4D;
@@ -131,76 +131,76 @@ private:
 
 Q_DECLARE_TYPEINFO(QVector2D, Q_PRIMITIVE_TYPE);
 
-Q_DECL_CONSTEXPR inline QVector2D::QVector2D() : xp(0.0f), yp(0.0f) {}
+Q_DECL_CONSTEXPR inline QVector2D::QVector2D() : v{0.0f, 0.0f} {}
 
-Q_DECL_CONSTEXPR inline QVector2D::QVector2D(float xpos, float ypos) : xp(xpos), yp(ypos) {}
+Q_DECL_CONSTEXPR inline QVector2D::QVector2D(float xpos, float ypos) : v{xpos, ypos} {}
 
-Q_DECL_CONSTEXPR inline QVector2D::QVector2D(const QPoint& point) : xp(point.x()), yp(point.y()) {}
+Q_DECL_CONSTEXPR inline QVector2D::QVector2D(const QPoint& point) : v{float(point.x()), float(point.y())} {}
 
-Q_DECL_CONSTEXPR inline QVector2D::QVector2D(const QPointF& point) : xp(float(point.x())), yp(float(point.y())) {}
+Q_DECL_CONSTEXPR inline QVector2D::QVector2D(const QPointF& point) : v{float(point.x()), float(point.y())} {}
 
 inline bool QVector2D::isNull() const
 {
-    return qIsNull(xp) && qIsNull(yp);
+    return qIsNull(v[0]) && qIsNull(v[1]);
 }
 
-Q_DECL_CONSTEXPR inline float QVector2D::x() const { return xp; }
-Q_DECL_CONSTEXPR inline float QVector2D::y() const { return yp; }
+Q_DECL_CONSTEXPR inline float QVector2D::x() const { return v[0]; }
+Q_DECL_CONSTEXPR inline float QVector2D::y() const { return v[1]; }
 
-inline void QVector2D::setX(float aX) { xp = aX; }
-inline void QVector2D::setY(float aY) { yp = aY; }
+inline void QVector2D::setX(float aX) { v[0] = aX; }
+inline void QVector2D::setY(float aY) { v[1] = aY; }
 
 inline float &QVector2D::operator[](int i)
 {
     Q_ASSERT(uint(i) < 2u);
-    return *(&xp + i);
+    return v[i];
 }
 
 inline float QVector2D::operator[](int i) const
 {
     Q_ASSERT(uint(i) < 2u);
-    return *(&xp + i);
+    return v[i];
 }
 
 inline QVector2D &QVector2D::operator+=(const QVector2D &vector)
 {
-    xp += vector.xp;
-    yp += vector.yp;
+    v[0] += vector.v[0];
+    v[1] += vector.v[1];
     return *this;
 }
 
 inline QVector2D &QVector2D::operator-=(const QVector2D &vector)
 {
-    xp -= vector.xp;
-    yp -= vector.yp;
+    v[0] -= vector.v[0];
+    v[1] -= vector.v[1];
     return *this;
 }
 
 inline QVector2D &QVector2D::operator*=(float factor)
 {
-    xp *= factor;
-    yp *= factor;
+    v[0] *= factor;
+    v[1] *= factor;
     return *this;
 }
 
 inline QVector2D &QVector2D::operator*=(const QVector2D &vector)
 {
-    xp *= vector.xp;
-    yp *= vector.yp;
+    v[0] *= vector.v[0];
+    v[1] *= vector.v[1];
     return *this;
 }
 
 inline QVector2D &QVector2D::operator/=(float divisor)
 {
-    xp /= divisor;
-    yp /= divisor;
+    v[0] /= divisor;
+    v[1] /= divisor;
     return *this;
 }
 
 inline QVector2D &QVector2D::operator/=(const QVector2D &vector)
 {
-    xp /= vector.xp;
-    yp /= vector.yp;
+    v[0] /= vector.v[0];
+    v[1] /= vector.v[1];
     return *this;
 }
 
@@ -209,68 +209,68 @@ QT_WARNING_DISABLE_CLANG("-Wfloat-equal")
 QT_WARNING_DISABLE_GCC("-Wfloat-equal")
 Q_DECL_CONSTEXPR inline bool operator==(const QVector2D &v1, const QVector2D &v2)
 {
-    return v1.xp == v2.xp && v1.yp == v2.yp;
+    return v1.v[0] == v2.v[0] && v1.v[1] == v2.v[1];
 }
 
 Q_DECL_CONSTEXPR inline bool operator!=(const QVector2D &v1, const QVector2D &v2)
 {
-    return v1.xp != v2.xp || v1.yp != v2.yp;
+    return v1.v[0] != v2.v[0] || v1.v[1] != v2.v[1];
 }
 QT_WARNING_POP
 
 Q_DECL_CONSTEXPR inline const QVector2D operator+(const QVector2D &v1, const QVector2D &v2)
 {
-    return QVector2D(v1.xp + v2.xp, v1.yp + v2.yp);
+    return QVector2D(v1.v[0] + v2.v[0], v1.v[1] + v2.v[1]);
 }
 
 Q_DECL_CONSTEXPR inline const QVector2D operator-(const QVector2D &v1, const QVector2D &v2)
 {
-    return QVector2D(v1.xp - v2.xp, v1.yp - v2.yp);
+    return QVector2D(v1.v[0] - v2.v[0], v1.v[1] - v2.v[1]);
 }
 
 Q_DECL_CONSTEXPR inline const QVector2D operator*(float factor, const QVector2D &vector)
 {
-    return QVector2D(vector.xp * factor, vector.yp * factor);
+    return QVector2D(vector.v[0] * factor, vector.v[1] * factor);
 }
 
 Q_DECL_CONSTEXPR inline const QVector2D operator*(const QVector2D &vector, float factor)
 {
-    return QVector2D(vector.xp * factor, vector.yp * factor);
+    return QVector2D(vector.v[0] * factor, vector.v[1] * factor);
 }
 
 Q_DECL_CONSTEXPR inline const QVector2D operator*(const QVector2D &v1, const QVector2D &v2)
 {
-    return QVector2D(v1.xp * v2.xp, v1.yp * v2.yp);
+    return QVector2D(v1.v[0] * v2.v[0], v1.v[1] * v2.v[1]);
 }
 
 Q_DECL_CONSTEXPR inline const QVector2D operator-(const QVector2D &vector)
 {
-    return QVector2D(-vector.xp, -vector.yp);
+    return QVector2D(-vector.v[0], -vector.v[1]);
 }
 
 Q_DECL_CONSTEXPR inline const QVector2D operator/(const QVector2D &vector, float divisor)
 {
-    return QVector2D(vector.xp / divisor, vector.yp / divisor);
+    return QVector2D(vector.v[0] / divisor, vector.v[1] / divisor);
 }
 
 Q_DECL_CONSTEXPR inline const QVector2D operator/(const QVector2D &vector, const QVector2D &divisor)
 {
-    return QVector2D(vector.xp / divisor.xp, vector.yp / divisor.yp);
+    return QVector2D(vector.v[0] / divisor.v[0], vector.v[1] / divisor.v[1]);
 }
 
 Q_DECL_CONSTEXPR inline bool qFuzzyCompare(const QVector2D& v1, const QVector2D& v2)
 {
-    return qFuzzyCompare(v1.xp, v2.xp) && qFuzzyCompare(v1.yp, v2.yp);
+    return qFuzzyCompare(v1.v[0], v2.v[0]) && qFuzzyCompare(v1.v[1], v2.v[1]);
 }
 
 Q_DECL_CONSTEXPR inline QPoint QVector2D::toPoint() const
 {
-    return QPoint(qRound(xp), qRound(yp));
+    return QPoint(qRound(v[0]), qRound(v[1]));
 }
 
 Q_DECL_CONSTEXPR inline QPointF QVector2D::toPointF() const
 {
-    return QPointF(qreal(xp), qreal(yp));
+    return QPointF(qreal(v[0]), qreal(v[1]));
 }
 
 #ifndef QT_NO_DEBUG_STREAM