Change meaning of offset in QStringData

It used to be an index into the first element in 'd' that came after
'offset'. It is now the byte offset from the beginning of the
QStringData structure.

By no longer using an actual array to access characters, we also steer
clear of GCC bug #43247:

    http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43247

This aligns this data structure with QArrayData. The intention is to
have QVector, QString and QByteArray share the same memory layout and
possibly code.

Change-Id: I4850813e1bd47c3cb670c50c9a8ccc1bff2e8597
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>

2 files changed, 25 insertions, 24 deletions

diff --git a/src/corelib/tools/qstring.cpp b/src/corelib/tools/qstring.cpp
index 57210901ee..be6f48808c 100644
--- a/src/corelib/tools/qstring.cpp
+++ b/src/corelib/tools/qstring.cpp
@@ -94,6 +94,8 @@
 #define ULLONG_MAX quint64_C(18446744073709551615)
 #endif
 
+#define IS_RAW_DATA(d) ((d)->offset != sizeof(QStringData))
+
 QT_BEGIN_NAMESPACE
 
 #ifndef QT_NO_TEXTCODEC
@@ -800,8 +802,8 @@ const QString::Null QString::null = { };
     \sa split()
 */
 
-const QStaticStringData<1> QString::shared_null = { { Q_REFCOUNT_INITIALIZE_STATIC, 0, 0, false, { 0 } }, { 0 } };
-const QStaticStringData<1> QString::shared_empty = { { Q_REFCOUNT_INITIALIZE_STATIC, 0, 0, false, { 0 } }, { 0 } };
+const QStaticStringData<1> QString::shared_null = { { Q_REFCOUNT_INITIALIZE_STATIC, 0, 0, false, sizeof(QStringData) }, { 0 } };
+const QStaticStringData<1> QString::shared_empty = { { Q_REFCOUNT_INITIALIZE_STATIC, 0, 0, false, sizeof(QStringData) }, { 0 } };
 
 int QString::grow(int size)
 {
@@ -1052,7 +1054,7 @@ QString::QString(const QChar *unicode, int size)
             d->size = size;
             d->alloc = (uint) size;
             d->capacityReserved = false;
-            d->offset = 0;
+            d->offset = sizeof(QStringData);
             memcpy(d->data(), unicode, size * sizeof(QChar));
             d->data()[size] = '\0';
         }
@@ -1076,7 +1078,7 @@ QString::QString(int size, QChar ch)
         d->size = size;
         d->alloc = (uint) size;
         d->capacityReserved = false;
-        d->offset = 0;
+        d->offset = sizeof(QStringData);
         d->data()[size] = '\0';
         ushort *i = d->data() + size;
         ushort *b = d->data();
@@ -1100,7 +1102,7 @@ QString::QString(int size, Qt::Initialization)
     d->size = size;
     d->alloc = (uint) size;
     d->capacityReserved = false;
-    d->offset = 0;
+    d->offset = sizeof(QStringData);
     d->data()[size] = '\0';
 }
 
@@ -1122,7 +1124,7 @@ QString::QString(QChar ch)
     d->size = 1;
     d->alloc = 1;
     d->capacityReserved = false;
-    d->offset = 0;
+    d->offset = sizeof(QStringData);
     d->data()[0] = ch.unicode();
     d->data()[1] = '\0';
 }
@@ -1220,7 +1222,7 @@ void QString::resize(int size)
     if (size < 0)
         size = 0;
 
-    if (d->offset && !d->ref.isShared() && size < d->size) {
+    if (IS_RAW_DATA(d) && !d->ref.isShared() && size < d->size) {
         d->size = size;
         return;
     }
@@ -1294,14 +1296,14 @@ void QString::resize(int size)
 // ### Qt 5: rename reallocData() to avoid confusion. 197625
 void QString::realloc(int alloc)
 {
-    if (d->ref.isShared() || d->offset) {
+    if (d->ref.isShared() || IS_RAW_DATA(d)) {
         Data *x = static_cast<Data *>(::malloc(sizeof(Data) + (alloc+1) * sizeof(QChar)));
         Q_CHECK_PTR(x);
         x->ref.initializeOwned();
         x->size = qMin(alloc, d->size);
         x->alloc = (uint) alloc;
         x->capacityReserved = d->capacityReserved;
-        x->offset =0;
+        x->offset = sizeof(QStringData);
         ::memcpy(x->data(), d->data(), x->size * sizeof(QChar));
         x->data()[x->size] = 0;
         if (!d->ref.deref())
@@ -1312,7 +1314,7 @@ void QString::realloc(int alloc)
         Q_CHECK_PTR(p);
         d = p;
         d->alloc = alloc;
-        d->offset = 0;
+        d->offset = sizeof(QStringData);
     }
 }
 
@@ -3741,7 +3743,7 @@ QString::Data *QString::fromLatin1_helper(const char *str, int size)
         d->size = size;
         d->alloc = (uint) size;
         d->capacityReserved = false;
-        d->offset = 0;
+        d->offset = sizeof(QStringData);
         d->data()[size] = '\0';
         ushort *dst = d->data();
         /* SIMD:
@@ -4769,7 +4771,7 @@ int QString::localeAwareCompare_helper(const QChar *data1, int length1,
 
 const ushort *QString::utf16() const
 {
-    if (d->offset)
+    if (IS_RAW_DATA(d))
         const_cast<QString*>(this)->realloc();   // ensure '\\0'-termination for ::fromRawData strings
     return d->data();
 }
@@ -7050,7 +7052,7 @@ QString QString::fromRawData(const QChar *unicode, int size)
         x->size = size;
         x->alloc = 0;
         x->capacityReserved = false;
-        x->offset = (const ushort *)unicode - (x->d + sizeof(qptrdiff)/sizeof(ushort));
+        x->offset = reinterpret_cast<const char *>(unicode) - reinterpret_cast<char *>(x);
     }
     return QString(x, 0);
 }
@@ -7076,9 +7078,9 @@ QString &QString::setRawData(const QChar *unicode, int size)
     } else {
         if (unicode) {
             d->size = size;
-            d->offset = (const ushort *)unicode - (d->d + sizeof(qptrdiff)/sizeof(ushort));
+            d->offset = reinterpret_cast<const char *>(unicode) - reinterpret_cast<char *>(d);
         } else {
-            d->offset = 0;
+            d->offset = sizeof(QStringData);
             d->size = 0;
         }
     }
diff --git a/src/corelib/tools/qstring.h b/src/corelib/tools/qstring.h
index df0ce73b3a..90fed78a15 100644
--- a/src/corelib/tools/qstring.h
+++ b/src/corelib/tools/qstring.h
@@ -75,12 +75,11 @@ struct QStringData {
     int size;
     uint alloc : 31;
     uint capacityReserved : 1;
-    union {
-        qptrdiff offset; // will always work as we add/subtract from a ushort ptr
-        ushort d[sizeof(qptrdiff)/sizeof(ushort)];
-    };
-    inline ushort *data() { return d + sizeof(qptrdiff)/sizeof(ushort) + offset; }
-    inline const ushort *data() const { return d + sizeof(qptrdiff)/sizeof(ushort) + offset; }
+
+    qptrdiff offset;
+
+    inline ushort *data() { return reinterpret_cast<ushort *>(reinterpret_cast<char *>(this) + offset); }
+    inline const ushort *data() const { return reinterpret_cast<const ushort *>(reinterpret_cast<const char *>(this) + offset); }
 };
 
 template<int N> struct QStaticStringData;
@@ -126,7 +125,7 @@ template<int N> struct QStaticStringData
 #  define QStringLiteral(str) ([]() -> QStaticStringDataPtr<sizeof(QT_UNICODE_LITERAL(str))/2 - 1> { \
         enum { Size = sizeof(QT_UNICODE_LITERAL(str))/2 - 1 }; \
         static const QStaticStringData<Size> qstring_literal = \
-        { { Q_REFCOUNT_INITIALIZE_STATIC, Size, 0, 0, { 0 } }, QT_UNICODE_LITERAL(str) }; \
+        { { Q_REFCOUNT_INITIALIZE_STATIC, Size, 0, 0, sizeof(QStringData) }, QT_UNICODE_LITERAL(str) }; \
         QStaticStringDataPtr<Size> holder = { &qstring_literal }; \
     return holder; }())
 
@@ -139,7 +138,7 @@ template<int N> struct QStaticStringData
     __extension__ ({ \
         enum { Size = sizeof(QT_UNICODE_LITERAL(str))/2 - 1 }; \
         static const QStaticStringData<Size> qstring_literal = \
-        { { Q_REFCOUNT_INITIALIZE_STATIC, Size, 0, 0, { 0 } }, QT_UNICODE_LITERAL(str) }; \
+        { { Q_REFCOUNT_INITIALIZE_STATIC, Size, 0, 0, sizeof(QStringData) }, QT_UNICODE_LITERAL(str) }; \
         QStaticStringDataPtr<Size> holder = { &qstring_literal }; \
         holder; })
 # endif
@@ -705,7 +704,7 @@ inline QChar *QString::data()
 inline const QChar *QString::constData() const
 { return reinterpret_cast<const QChar*>(d->data()); }
 inline void QString::detach()
-{ if (d->ref.isShared() || d->offset) realloc(); }
+{ if (d->ref.isShared() || (d->offset != sizeof(QStringData))) realloc(); }
 inline bool QString::isDetached() const
 { return !d->ref.isShared(); }
 inline QString &QString::operator=(const QLatin1String &s)