Version check when setting up lxml

- Expose both the compiled version and the linked version of libxml2

- Do a check that the versions match when initializing the module.
  Otherwise raise an exception. This seems like a better result then
  a difficult to diagnose segfault.

Author: jonathangreen

src/lxml.c

@@ -9,6 +9,7 @@
 
 #include "common.h"
 #include "lxml.h"
+#include "exception.h"
 
 #include <etree_defs.h>
 #include <etree_api.h>
@@ -17,8 +18,102 @@
 #include <libxml/parser.h>
 #include <libxml/dict.h>
 
+#define XMLSEC_EXTRACT_VERSION(x, y) ((x / (y)) % 100)
+
+#define XMLSEC_EXTRACT_MAJOR(x) XMLSEC_EXTRACT_VERSION(x, 100 * 100)
+#define XMLSEC_EXTRACT_MINOR(x) XMLSEC_EXTRACT_VERSION(x, 100)
+#define XMLSEC_EXTRACT_PATCH(x) XMLSEC_EXTRACT_VERSION(x, 1)
+
+static long PyXmlSec_GetLibXmlVersionLong() {
+    return PyOS_strtol(xmlParserVersion, NULL, 10);
+}
+long PyXmlSec_GetLibXmlVersionMajor() {
+    return XMLSEC_EXTRACT_MAJOR(PyXmlSec_GetLibXmlVersionLong());
+}
+long PyXmlSec_GetLibXmlVersionMinor() {
+    return XMLSEC_EXTRACT_MINOR(PyXmlSec_GetLibXmlVersionLong());
+}
+long PyXmlSec_GetLibXmlVersionPatch() {
+    return XMLSEC_EXTRACT_PATCH(PyXmlSec_GetLibXmlVersionLong());
+}
+
+long PyXmlSec_GetLibXmlCompiledVersionMajor() {
+    return XMLSEC_EXTRACT_MAJOR(LIBXML_VERSION);
+}
+long PyXmlSec_GetLibXmlCompiledVersionMinor() {
+    return XMLSEC_EXTRACT_MINOR(LIBXML_VERSION);
+}
+long PyXmlSec_GetLibXmlCompiledVersionPatch() {
+    return XMLSEC_EXTRACT_PATCH(LIBXML_VERSION);
+}
+
+static int PyXmlSec_CheckLibXmlLibraryVersion(void) {
+    // Make sure that the version of libxml2 that we were compiled against is the same as the one
+    // that is loaded. If there is a version mismatch, we could run into segfaults.
+
+    if (PyXmlSec_GetLibXmlVersionMajor() != PyXmlSec_GetLibXmlCompiledVersionMajor() ||
+        PyXmlSec_GetLibXmlVersionMinor() != PyXmlSec_GetLibXmlCompiledVersionMinor()) {
+        return -1;
+    }
+
+    return 0;
+}
+
+static int PyXmlSec_CheckLxmlLibraryVersion(void) {
+    // Make sure that the version of libxml2 lxml is using is the same as the one we are using. Because
+    // we pass trees between the two libraries, we need to make sure that they are using the same version
+    // of libxml2, or we could run into difficult to debug segfaults.
+    // See: https://github.com/xmlsec/python-xmlsec/issues/283
+
+    PyObject* lxml = NULL;
+    PyObject* version = NULL;
+
+    // Default to failure
+    int result = -1;
+
+    lxml = PyImport_ImportModule("lxml.etree");
+    if (lxml == NULL) {
+        goto FINALIZE;
+    }
+    version = PyObject_GetAttrString(lxml, "LIBXML_VERSION");
+    if (version == NULL) {
+        goto FINALIZE;
+    }
+    if (!PyTuple_Check(version) || PyTuple_Size(version) != 3) {
+        goto FINALIZE;
+    }
+
+    PyObject* major = PyTuple_GetItem(version, 0);
+    PyObject* minor = PyTuple_GetItem(version, 1);
+
+    if (!PyLong_Check(major) || !PyLong_Check(minor)) {
+        goto FINALIZE;
+    }
+
+    if (PyLong_AsLong(major) != PyXmlSec_GetLibXmlVersionMajor() || PyLong_AsLong(minor) != PyXmlSec_GetLibXmlVersionMinor()) {
+        goto FINALIZE;
+    }
+
+    result = 0;
+
+FINALIZE:
+    // Cleanup our references, and return the result
+    Py_XDECREF(lxml);
+    Py_XDECREF(version);
+    return result;
+}
 
 int PyXmlSec_InitLxmlModule(void) {
+    if (PyXmlSec_CheckLibXmlLibraryVersion() < 0) {
+        PyXmlSec_SetLastError("xmlsec libxml2 library compiled version vs runtime version mismatch");
+        return -1;
+    }
+
+    if (PyXmlSec_CheckLxmlLibraryVersion() < 0) {
+        PyXmlSec_SetLastError("lxml & xmlsec libxml2 library version mismatch");
+        return -1;
+    }
+
     return import_lxml__etree();
 }
 

src/lxml.h

@@ -29,4 +29,13 @@ PyXmlSec_LxmlElementPtr PyXmlSec_elementFactory(PyXmlSec_LxmlDocumentPtr doc, xm
 // converts o to PyObject, None object is not allowed, does not increment ref_counts
 int PyXmlSec_LxmlElementConverter(PyObject* o, PyXmlSec_LxmlElementPtr* p);
 
+// get version numbers for libxml2 both compiled and loaded
+long PyXmlSec_GetLibXmlVersionMajor();
+long PyXmlSec_GetLibXmlVersionMinor();
+long PyXmlSec_GetLibXmlVersionPatch();
+
+long PyXmlSec_GetLibXmlCompiledVersionMajor();
+long PyXmlSec_GetLibXmlCompiledVersionMinor();
+long PyXmlSec_GetLibXmlCompiledVersionPatch();
+
 #endif // __PYXMLSEC_LXML_H__

src/main.c

@@ -10,6 +10,7 @@
 #include "common.h"
 #include "platform.h"
 #include "exception.h"
+#include "lxml.h"
 
 #include <xmlsec/xmlsec.h>
 #include <xmlsec/crypto.h>
@@ -127,10 +128,27 @@ static PyObject* PyXmlSec_GetLibXmlSecVersion() {
 }
 
 static char PyXmlSec_GetLibXmlVersion__doc__[] = \
-    "get_libxml_version() -> tuple\n"
-    "Returns Version tuple of wrapped libxml library.";
+    "get_libxml_version() -> tuple[int, int, int]\n"
+    "Returns version tuple of libxml2 library xmlsec is using.";
 static PyObject* PyXmlSec_GetLibXmlVersion() {
-    return Py_BuildValue("(iii)", XMLSEC_LIBXML_VERSION_MAJOR, XMLSEC_LIBXML_VERSION_MINOR, XMLSEC_LIBXML_VERSION_PATCH);
+    return Py_BuildValue(
+        "(iii)",
+        PyXmlSec_GetLibXmlVersionMajor(),
+        PyXmlSec_GetLibXmlVersionMinor(),
+        PyXmlSec_GetLibXmlVersionPatch()
+    );
+}
+
+static char PyXmlSec_GetLibXmlCompiledVersion__doc__[] = \
+    "get_libxml_compiled_version() -> tuple[int, int, int]\n"
+    "Returns version tuple of libxml2 library xmlsec was compiled with.";
+static PyObject* PyXmlSec_GetLibXmlCompiledVersion() {
+    return Py_BuildValue(
+        "(iii)",
+        PyXmlSec_GetLibXmlCompiledVersionMajor(),
+        PyXmlSec_GetLibXmlCompiledVersionMinor(),
+        PyXmlSec_GetLibXmlCompiledVersionPatch()
+    );
 }
 
 static char PyXmlSec_PyEnableDebugOutput__doc__[] = \
@@ -412,6 +430,12 @@ static PyMethodDef PyXmlSec_MainMethods[] = {
         METH_NOARGS,
         PyXmlSec_GetLibXmlVersion__doc__
     },
+    {
+        "get_libxml_compiled_version",
+        (PyCFunction)PyXmlSec_GetLibXmlCompiledVersion,
+        METH_NOARGS,
+        PyXmlSec_GetLibXmlCompiledVersion__doc__
+    },
     {
         "enable_debug_trace",
         (PyCFunction)PyXmlSec_PyEnableDebugOutput,

src/platform.h

@@ -12,20 +12,13 @@
 
 #define PY_SSIZE_T_CLEAN 1
 
-#include <libxml/xmlversion.h>
 #include <xmlsec/version.h>
 #include <Python.h>
 
 #ifdef MS_WIN32
 #include <windows.h>
 #endif /* MS_WIN32 */
 
-#define XMLSEC_EXTRACT_VERSION(x, y) ((x / (y)) % 100)
-
-#define XMLSEC_LIBXML_VERSION_MAJOR XMLSEC_EXTRACT_VERSION(LIBXML_VERSION, 100 * 100)
-#define XMLSEC_LIBXML_VERSION_MINOR XMLSEC_EXTRACT_VERSION(LIBXML_VERSION, 100)
-#define XMLSEC_LIBXML_VERSION_PATCH XMLSEC_EXTRACT_VERSION(LIBXML_VERSION, 1)
-
 #define XMLSEC_VERSION_HEX ((XMLSEC_VERSION_MAJOR << 16) | (XMLSEC_VERSION_MINOR << 8) | (XMLSEC_VERSION_SUBMINOR))
 
 // XKMS support was removed in version 1.2.21