372 lines
13 KiB
Diff
372 lines
13 KiB
Diff
From 49a4cf8133eb8d76115fb7827bd76764e1b30d12 Mon Sep 17 00:00:00 2001
|
|
From: peijiankang <peijiankang@kylinos.cn>
|
|
Date: Wed, 29 Nov 2023 20:51:35 +0800
|
|
Subject: [PATCH] CVE-2023-38197
|
|
|
|
---
|
|
src/corelib/serialization/qxmlstream.cpp | 144 +++++++++++++++++-
|
|
src/corelib/serialization/qxmlstream_p.h | 11 ++
|
|
.../qxmlstream/tokenError/dtdInBody.xml | 21 +++
|
|
.../qxmlstream/tokenError/multipleDtd.xml | 21 +++
|
|
.../qxmlstream/tokenError/wellFormed.xml | 16 ++
|
|
.../qxmlstream/tst_qxmlstream.cpp | 38 +++++
|
|
6 files changed, 243 insertions(+), 8 deletions(-)
|
|
create mode 100644 tests/auto/corelib/serialization/qxmlstream/tokenError/dtdInBody.xml
|
|
create mode 100644 tests/auto/corelib/serialization/qxmlstream/tokenError/multipleDtd.xml
|
|
create mode 100644 tests/auto/corelib/serialization/qxmlstream/tokenError/wellFormed.xml
|
|
|
|
diff --git a/src/corelib/serialization/qxmlstream.cpp b/src/corelib/serialization/qxmlstream.cpp
|
|
index 3175517a..7dcf80c3 100644
|
|
--- a/src/corelib/serialization/qxmlstream.cpp
|
|
+++ b/src/corelib/serialization/qxmlstream.cpp
|
|
@@ -185,7 +185,7 @@ WRAP(indexOf, QLatin1StringView)
|
|
addData() or by waiting for it to arrive on the device().
|
|
|
|
\value UnexpectedElementError The parser encountered an element
|
|
- that was different to those it expected.
|
|
+ or token that was different to those it expected.
|
|
|
|
*/
|
|
|
|
@@ -322,13 +322,34 @@ QXmlStreamEntityResolver *QXmlStreamReader::entityResolver() const
|
|
|
|
QXmlStreamReader is a well-formed XML 1.0 parser that does \e not
|
|
include external parsed entities. As long as no error occurs, the
|
|
- application code can thus be assured that the data provided by the
|
|
- stream reader satisfies the W3C's criteria for well-formed XML. For
|
|
- example, you can be certain that all tags are indeed nested and
|
|
- closed properly, that references to internal entities have been
|
|
- replaced with the correct replacement text, and that attributes have
|
|
- been normalized or added according to the internal subset of the
|
|
- DTD.
|
|
+ application code can thus be assured, that
|
|
+ \list
|
|
+ \li the data provided by the stream reader satisfies the W3C's
|
|
+ criteria for well-formed XML,
|
|
+ \li tokens are provided in a valid order.
|
|
+ \endlist
|
|
+
|
|
+ Unless QXmlStreamReader raises an error, it guarantees the following:
|
|
+ \list
|
|
+ \li All tags are nested and closed properly.
|
|
+ \li References to internal entities have been replaced with the
|
|
+ correct replacement text.
|
|
+ \li Attributes have been normalized or added according to the
|
|
+ internal subset of the \l DTD.
|
|
+ \li Tokens of type \l StartDocument happen before all others,
|
|
+ aside from comments and processing instructions.
|
|
+ \li At most one DOCTYPE element (a token of type \l DTD) is present.
|
|
+ \li If present, the DOCTYPE appears before all other elements,
|
|
+ aside from StartDocument, comments and processing instructions.
|
|
+ \endlist
|
|
+
|
|
+ In particular, once any token of type \l StartElement, \l EndElement,
|
|
+ \l Characters, \l EntityReference or \l EndDocument is seen, no
|
|
+ tokens of type StartDocument or DTD will be seen. If one is present in
|
|
+ the input stream, out of order, an error is raised.
|
|
+
|
|
+ \note The token types \l Comment and \l ProcessingInstruction may appear
|
|
+ anywhere in the stream.
|
|
|
|
If an error occurs while parsing, atEnd() and hasError() return
|
|
true, and error() returns the error that occurred. The functions
|
|
@@ -659,6 +680,7 @@ QXmlStreamReader::TokenType QXmlStreamReader::readNext()
|
|
d->token = -1;
|
|
return readNext();
|
|
}
|
|
+ d->checkToken();
|
|
return d->type;
|
|
}
|
|
|
|
@@ -743,6 +765,11 @@ static constexpr auto QXmlStreamReader_tokenTypeString = qOffsetStringArray(
|
|
"ProcessingInstruction"
|
|
);
|
|
|
|
+static constexpr auto QXmlStreamReader_XmlContextString = qOffsetStringArray(
|
|
+ "Prolog",
|
|
+ "Body"
|
|
+);
|
|
+
|
|
/*!
|
|
\property QXmlStreamReader::namespaceProcessing
|
|
\brief the namespace-processing flag of the stream reader.
|
|
@@ -777,6 +804,15 @@ QString QXmlStreamReader::tokenString() const
|
|
return QLatin1StringView(QXmlStreamReader_tokenTypeString.at(d->type));
|
|
}
|
|
|
|
+/*!
|
|
+ \internal
|
|
+ \return \param loc (Prolog/Body) as a string.
|
|
+ */
|
|
+static constexpr QLatin1StringView contextString(QXmlStreamReaderPrivate::XmlContext ctxt)
|
|
+{
|
|
+ return QLatin1StringView(QXmlStreamReader_XmlContextString.at(static_cast<int>(ctxt)));
|
|
+}
|
|
+
|
|
#endif // QT_NO_XMLSTREAMREADER
|
|
|
|
QXmlStreamPrivateTagStack::QXmlStreamPrivateTagStack()
|
|
@@ -864,6 +900,8 @@ void QXmlStreamReaderPrivate::init()
|
|
|
|
type = QXmlStreamReader::NoToken;
|
|
error = QXmlStreamReader::NoError;
|
|
+ currentContext = XmlContext::Prolog;
|
|
+ foundDTD = false;
|
|
}
|
|
|
|
/*
|
|
@@ -3838,6 +3876,96 @@ void QXmlStreamWriter::writeCurrentToken(const QXmlStreamReader &reader)
|
|
break;
|
|
}
|
|
}
|
|
+static constexpr bool isTokenAllowedInContext(QXmlStreamReader::TokenType type,
|
|
+ QXmlStreamReaderPrivate::XmlContext loc)
|
|
+{
|
|
+ switch (type) {
|
|
+ case QXmlStreamReader::StartDocument:
|
|
+ case QXmlStreamReader::DTD:
|
|
+ return loc == QXmlStreamReaderPrivate::XmlContext::Prolog;
|
|
+
|
|
+ case QXmlStreamReader::StartElement:
|
|
+ case QXmlStreamReader::EndElement:
|
|
+ case QXmlStreamReader::Characters:
|
|
+ case QXmlStreamReader::EntityReference:
|
|
+ case QXmlStreamReader::EndDocument:
|
|
+ return loc == QXmlStreamReaderPrivate::XmlContext::Body;
|
|
+
|
|
+ case QXmlStreamReader::Comment:
|
|
+ case QXmlStreamReader::ProcessingInstruction:
|
|
+ return true;
|
|
+
|
|
+ case QXmlStreamReader::NoToken:
|
|
+ case QXmlStreamReader::Invalid:
|
|
+ return false;
|
|
+ }
|
|
+
|
|
+ // GCC 8.x does not treat __builtin_unreachable() as constexpr
|
|
+#if !defined(Q_CC_GNU_ONLY) || (Q_CC_GNU >= 900)
|
|
+ Q_UNREACHABLE_RETURN(false);
|
|
+#else
|
|
+ return false;
|
|
+#endif
|
|
+}
|
|
+
|
|
+/*!
|
|
+ \internal
|
|
+ \brief QXmlStreamReader::isValidToken
|
|
+ \return \c true if \param type is a valid token type.
|
|
+ \return \c false if \param type is an unexpected token,
|
|
+ which indicates a non-well-formed or invalid XML stream.
|
|
+ */
|
|
+bool QXmlStreamReaderPrivate::isValidToken(QXmlStreamReader::TokenType type)
|
|
+{
|
|
+ // Don't change currentContext, if Invalid or NoToken occur in the prolog
|
|
+ if (type == QXmlStreamReader::Invalid || type == QXmlStreamReader::NoToken)
|
|
+ return false;
|
|
+
|
|
+ // If a token type gets rejected in the body, there is no recovery
|
|
+ const bool result = isTokenAllowedInContext(type, currentContext);
|
|
+ if (result || currentContext == XmlContext::Body)
|
|
+ return result;
|
|
+
|
|
+ // First non-Prolog token observed => switch context to body and check again.
|
|
+ currentContext = XmlContext::Body;
|
|
+ return isTokenAllowedInContext(type, currentContext);
|
|
+}
|
|
+
|
|
+/*!
|
|
+ \internal
|
|
+ Checks token type and raises an error, if it is invalid
|
|
+ in the current context (prolog/body).
|
|
+ */
|
|
+void QXmlStreamReaderPrivate::checkToken()
|
|
+{
|
|
+ Q_Q(QXmlStreamReader);
|
|
+
|
|
+ // The token type must be consumed, to keep track if the body has been reached.
|
|
+ const XmlContext context = currentContext;
|
|
+ const bool ok = isValidToken(type);
|
|
+
|
|
+ // Do nothing if an error has been raised already (going along with an unexpected token)
|
|
+ if (error != QXmlStreamReader::Error::NoError)
|
|
+ return;
|
|
+
|
|
+ if (!ok) {
|
|
+ raiseError(QXmlStreamReader::UnexpectedElementError,
|
|
+ QObject::tr("Unexpected token type %1 in %2.")
|
|
+ .arg(q->tokenString(), contextString(context)));
|
|
+ return;
|
|
+ }
|
|
+
|
|
+ if (type != QXmlStreamReader::DTD)
|
|
+ return;
|
|
+
|
|
+ // Raise error on multiple DTD tokens
|
|
+ if (foundDTD) {
|
|
+ raiseError(QXmlStreamReader::UnexpectedElementError,
|
|
+ QObject::tr("Found second DTD token in %1.").arg(contextString(context)));
|
|
+ } else {
|
|
+ foundDTD = true;
|
|
+ }
|
|
+}
|
|
|
|
/*!
|
|
\fn bool QXmlStreamAttributes::hasAttribute(QAnyStringView qualifiedName) const
|
|
diff --git a/src/corelib/serialization/qxmlstream_p.h b/src/corelib/serialization/qxmlstream_p.h
|
|
index 7c46d187..f805cedb 100644
|
|
--- a/src/corelib/serialization/qxmlstream_p.h
|
|
+++ b/src/corelib/serialization/qxmlstream_p.h
|
|
@@ -296,6 +296,17 @@ public:
|
|
QStringDecoder decoder;
|
|
bool atEnd;
|
|
|
|
+ enum class XmlContext
|
|
+ {
|
|
+ Prolog,
|
|
+ Body,
|
|
+ };
|
|
+
|
|
+ XmlContext currentContext = XmlContext::Prolog;
|
|
+ bool foundDTD = false;
|
|
+ bool isValidToken(QXmlStreamReader::TokenType type);
|
|
+ void checkToken();
|
|
+
|
|
/*!
|
|
\sa setType()
|
|
*/
|
|
diff --git a/tests/auto/corelib/serialization/qxmlstream/tokenError/dtdInBody.xml b/tests/auto/corelib/serialization/qxmlstream/tokenError/dtdInBody.xml
|
|
new file mode 100644
|
|
index 00000000..68ef2962
|
|
--- /dev/null
|
|
+++ b/tests/auto/corelib/serialization/qxmlstream/tokenError/dtdInBody.xml
|
|
@@ -0,0 +1,21 @@
|
|
+<!DOCTYPE TEST [
|
|
+ <!ELEMENT TESTATTRIBUTE (CASE+)>
|
|
+ <!ELEMENT CASE (CLASS, FUNCTION)>
|
|
+ <!ELEMENT CLASS (#PCDATA)>
|
|
+
|
|
+ <!-- adding random ENTITY statement, as this is typical DTD content -->
|
|
+ <!ENTITY unite "∪">
|
|
+
|
|
+ <!ATTLIST CASE CLASS CDATA #REQUIRED>
|
|
+]>
|
|
+<TEST>
|
|
+ <CASE>
|
|
+ <CLASS>tst_QXmlStream</CLASS>
|
|
+ </CASE>
|
|
+ <!-- invalid DTD in XML body follows -->
|
|
+ <!DOCTYPE DTDTEST [
|
|
+ <!ELEMENT RESULT (CASE+)>
|
|
+ <!ATTLIST RESULT OUTPUT CDATA #REQUIRED>
|
|
+ ]>
|
|
+</TEST>
|
|
+
|
|
diff --git a/tests/auto/corelib/serialization/qxmlstream/tokenError/multipleDtd.xml b/tests/auto/corelib/serialization/qxmlstream/tokenError/multipleDtd.xml
|
|
new file mode 100644
|
|
index 00000000..1dbe75c4
|
|
--- /dev/null
|
|
+++ b/tests/auto/corelib/serialization/qxmlstream/tokenError/multipleDtd.xml
|
|
@@ -0,0 +1,21 @@
|
|
+<!DOCTYPE TEST [
|
|
+ <!ELEMENT TESTATTRIBUTE (CASE+)>
|
|
+ <!ELEMENT CASE (CLASS, FUNCTION, DATASET, COMMENTS)>
|
|
+ <!ELEMENT CLASS (#PCDATA)>
|
|
+
|
|
+ <!-- adding random ENTITY statements, as this is typical DTD content -->
|
|
+ <!ENTITY iff "⇔">
|
|
+
|
|
+ <!ATTLIST CASE CLASS CDATA #REQUIRED>
|
|
+]>
|
|
+<!-- invalid second DTD follows -->
|
|
+<!DOCTYPE SECOND [
|
|
+ <!ELEMENT SECONDATTRIBUTE (#PCDATA)>
|
|
+ <!ENTITY on "∘">
|
|
+]>
|
|
+<TEST>
|
|
+ <CASE>
|
|
+ <CLASS>tst_QXmlStream</CLASS>
|
|
+ </CASE>
|
|
+</TEST>
|
|
+
|
|
diff --git a/tests/auto/corelib/serialization/qxmlstream/tokenError/wellFormed.xml b/tests/auto/corelib/serialization/qxmlstream/tokenError/wellFormed.xml
|
|
new file mode 100644
|
|
index 00000000..9dfbc0f9
|
|
--- /dev/null
|
|
+++ b/tests/auto/corelib/serialization/qxmlstream/tokenError/wellFormed.xml
|
|
@@ -0,0 +1,16 @@
|
|
+<!DOCTYPE TEST [
|
|
+ <!ELEMENT TESTATTRIBUTE (CASE+)>
|
|
+ <!ELEMENT CASE (CLASS, FUNCTION, DATASET, COMMENTS)>
|
|
+ <!ELEMENT CLASS (#PCDATA)>
|
|
+
|
|
+ <!-- adding random ENTITY statements, as this is typical DTD content -->
|
|
+ <!ENTITY unite "∪">
|
|
+
|
|
+ <!ATTLIST CASE CLASS CDATA #REQUIRED>
|
|
+]>
|
|
+<TEST>
|
|
+ <CASE>
|
|
+ <CLASS>tst_QXmlStream</CLASS>
|
|
+ </CASE>
|
|
+</TEST>
|
|
+
|
|
diff --git a/tests/auto/corelib/serialization/qxmlstream/tst_qxmlstream.cpp b/tests/auto/corelib/serialization/qxmlstream/tst_qxmlstream.cpp
|
|
index b7f603c7..839d9edc 100644
|
|
--- a/tests/auto/corelib/serialization/qxmlstream/tst_qxmlstream.cpp
|
|
+++ b/tests/auto/corelib/serialization/qxmlstream/tst_qxmlstream.cpp
|
|
@@ -590,6 +590,8 @@ private slots:
|
|
|
|
void entityExpansionLimit() const;
|
|
|
|
+ void tokenErrorHandling_data() const;
|
|
+ void tokenErrorHandling() const;
|
|
private:
|
|
static QByteArray readFile(const QString &filename);
|
|
|
|
@@ -1818,6 +1820,42 @@ void tst_QXmlStream::roundTrip() const
|
|
QCOMPARE(out, in);
|
|
}
|
|
|
|
+void tst_QXmlStream::tokenErrorHandling_data() const
|
|
+{
|
|
+ QTest::addColumn<QString>("fileName");
|
|
+ QTest::addColumn<QXmlStreamReader::Error>("expectedError");
|
|
+ QTest::addColumn<QString>("errorKeyWord");
|
|
+
|
|
+ constexpr auto invalid = QXmlStreamReader::Error::UnexpectedElementError;
|
|
+ constexpr auto valid = QXmlStreamReader::Error::NoError;
|
|
+ QTest::newRow("DtdInBody") << "dtdInBody.xml" << invalid << "DTD";
|
|
+ QTest::newRow("multipleDTD") << "multipleDtd.xml" << invalid << "second DTD";
|
|
+ QTest::newRow("wellFormed") << "wellFormed.xml" << valid << "";
|
|
+}
|
|
+
|
|
+void tst_QXmlStream::tokenErrorHandling() const
|
|
+{
|
|
+ QFETCH(const QString, fileName);
|
|
+ QFETCH(const QXmlStreamReader::Error, expectedError);
|
|
+ QFETCH(const QString, errorKeyWord);
|
|
+
|
|
+ const QDir dir(QFINDTESTDATA("tokenError"));
|
|
+ QFile file(dir.absoluteFilePath(fileName));
|
|
+
|
|
+ // Cross-compiling: File will be on host only
|
|
+ if (!file.exists())
|
|
+ QSKIP("Testfile not found.");
|
|
+
|
|
+ file.open(QIODevice::ReadOnly);
|
|
+ QXmlStreamReader reader(&file);
|
|
+ while (!reader.atEnd())
|
|
+ reader.readNext();
|
|
+
|
|
+ QCOMPARE(reader.error(), expectedError);
|
|
+ if (expectedError != QXmlStreamReader::Error::NoError)
|
|
+ QVERIFY(reader.errorString().contains(errorKeyWord));
|
|
+}
|
|
+
|
|
void tst_QXmlStream::test_fastScanName_data() const
|
|
{
|
|
QTest::addColumn<QByteArray>("data");
|
|
--
|
|
2.41.0
|
|
|