[pypy-commit] cffi default: Issue #440

arigo pypy.commits at gmail.com
Sun Jan 19 04:23:44 EST 2020


Author: Armin Rigo <arigo at tunes.org>
Branch: 
Changeset: r3330:13004e3998dc
Date: 2020-01-19 10:23 +0100
http://bitbucket.org/cffi/cffi/changeset/13004e3998dc/

Log:	Issue #440

	Limit the amount of memory that is requested from alloca() for
	temporary conversion of arguments. Non-small requests are instead
	handled with PyObject_Malloc() and PyObject_Free().

diff --git a/c/_cffi_backend.c b/c/_cffi_backend.c
--- a/c/_cffi_backend.c
+++ b/c/_cffi_backend.c
@@ -2991,6 +2991,10 @@
     CTypeDescrObject *fresult;
     char *resultdata;
     char *errormsg;
+    struct freeme_s {
+        struct freeme_s *next;
+        union_alignment alignment;
+    } *freeme = NULL;
 
     if (!(cd->c_type->ct_flags & CT_FUNCTIONPTR)) {
         PyErr_Format(PyExc_TypeError, "cdata '%s' is not callable",
@@ -3111,7 +3115,21 @@
             else if (datasize < 0)
                 goto error;
             else {
-                tmpbuf = alloca(datasize);
+                if (datasize <= 512) {
+                    tmpbuf = alloca(datasize);
+                }
+                else {
+                    struct freeme_s *fp = (struct freeme_s *)PyObject_Malloc(
+                        offsetof(struct freeme_s, alignment) +
+                        (size_t)datasize);
+                    if (fp == NULL) {
+                        PyErr_NoMemory();
+                        goto error;
+                    }
+                    fp->next = freeme;
+                    freeme = fp;
+                    tmpbuf = (char *)&fp->alignment;
+                }
                 memset(tmpbuf, 0, datasize);
                 *(char **)data = tmpbuf;
                 if (convert_array_from_object(tmpbuf, argtype, obj) < 0)
@@ -3156,6 +3174,11 @@
     /* fall-through */
 
  error:
+    while (freeme != NULL) {
+        void *p = (void *)freeme;
+        freeme = freeme->next;
+        PyObject_Free(p);
+    }
     if (buffer)
         PyObject_Free(buffer);
     if (fvarargs != NULL) {
diff --git a/cffi/_cffi_include.h b/cffi/_cffi_include.h
--- a/cffi/_cffi_include.h
+++ b/cffi/_cffi_include.h
@@ -269,6 +269,54 @@
         return _cffi_from_c_wchar3216_t((int)x);
 }
 
+union _cffi_union_alignment_u {
+    unsigned char m_char;
+    unsigned short m_short;
+    unsigned int m_int;
+    unsigned long m_long;
+    unsigned long long m_longlong;
+    float m_float;
+    double m_double;
+    long double m_longdouble;
+};
+
+struct _cffi_freeme_s {
+    struct _cffi_freeme_s *next;
+    union _cffi_union_alignment_u alignment;
+};
+
+_CFFI_UNUSED_FN static int
+_cffi_convert_array_argument(struct _cffi_ctypedescr *ctptr, PyObject *arg,
+                             char **output_data, Py_ssize_t datasize,
+                             struct _cffi_freeme_s **freeme)
+{
+    char *p;
+    if (datasize < 0)
+        return -1;
+
+    p = *output_data;
+    if (p == NULL) {
+        struct _cffi_freeme_s *fp = (struct _cffi_freeme_s *)PyObject_Malloc(
+            offsetof(struct _cffi_freeme_s, alignment) + (size_t)datasize);
+        if (fp == NULL)
+            return -1;
+        fp->next = *freeme;
+        *freeme = fp;
+        p = *output_data = (char *)&fp->alignment;
+    }
+    memset((void *)p, 0, (size_t)datasize);
+    return _cffi_convert_array_from_object(p, ctptr, arg);
+}
+
+_CFFI_UNUSED_FN static void
+_cffi_free_array_arguments(struct _cffi_freeme_s *freeme)
+{
+    do {
+        void *p = (void *)freeme;
+        freeme = freeme->next;
+        PyObject_Free(p);
+    } while (freeme != NULL);
+}
 
 /**********  end CPython-specific section  **********/
 #else
diff --git a/cffi/recompiler.py b/cffi/recompiler.py
--- a/cffi/recompiler.py
+++ b/cffi/recompiler.py
@@ -560,23 +560,24 @@
             tovar, tp.get_c_name(''), errvalue))
         self._prnt('    %s;' % errcode)
 
-    def _extra_local_variables(self, tp, localvars):
+    def _extra_local_variables(self, tp, localvars, freelines):
         if isinstance(tp, model.PointerType):
             localvars.add('Py_ssize_t datasize')
+            localvars.add('struct _cffi_freeme_s *large_args_free = NULL')
+            freelines.add('if (large_args_free != NULL)'
+                          ' _cffi_free_array_arguments(large_args_free);')
 
     def _convert_funcarg_to_c_ptr_or_array(self, tp, fromvar, tovar, errcode):
         self._prnt('  datasize = _cffi_prepare_pointer_call_argument(')
         self._prnt('      _cffi_type(%d), %s, (char **)&%s);' % (
             self._gettypenum(tp), fromvar, tovar))
         self._prnt('  if (datasize != 0) {')
-        self._prnt('    if (datasize < 0)')
-        self._prnt('      %s;' % errcode)
-        self._prnt('    %s = (%s)alloca((size_t)datasize);' % (
+        self._prnt('    %s = ((size_t)datasize) <= 640 ? '
+                   '(%s)alloca((size_t)datasize) : NULL;' % (
             tovar, tp.get_c_name('')))
-        self._prnt('    memset((void *)%s, 0, (size_t)datasize);' % (tovar,))
-        self._prnt('    if (_cffi_convert_array_from_object('
-                   '(char *)%s, _cffi_type(%d), %s) < 0)' % (
-            tovar, self._gettypenum(tp), fromvar))
+        self._prnt('    if (_cffi_convert_array_argument(_cffi_type(%d), %s, '
+                   '(char **)&%s,' % (self._gettypenum(tp), fromvar, tovar))
+        self._prnt('            datasize, &large_args_free) < 0)')
         self._prnt('      %s;' % errcode)
         self._prnt('  }')
 
@@ -699,9 +700,10 @@
             prnt('  %s;' % arg)
         #
         localvars = set()
+        freelines = set()
         for type in tp.args:
-            self._extra_local_variables(type, localvars)
-        for decl in localvars:
+            self._extra_local_variables(type, localvars, freelines)
+        for decl in sorted(localvars):
             prnt('  %s;' % (decl,))
         #
         if not isinstance(tp.result, model.VoidType):
@@ -709,6 +711,7 @@
             context = 'result of %s' % name
             result_decl = '  %s;' % tp.result.get_c_name(' result', context)
             prnt(result_decl)
+            prnt('  PyObject *pyresult;')
         else:
             result_decl = None
             result_code = ''
@@ -742,9 +745,14 @@
         if numargs == 0:
             prnt('  (void)noarg; /* unused */')
         if result_code:
-            prnt('  return %s;' %
+            prnt('  pyresult = %s;' %
                  self._convert_expr_from_c(tp.result, 'result', 'result type'))
+            for freeline in freelines:
+                prnt('  ' + freeline)
+            prnt('  return pyresult;')
         else:
+            for freeline in freelines:
+                prnt('  ' + freeline)
             prnt('  Py_INCREF(Py_None);')
             prnt('  return Py_None;')
         prnt('}')
diff --git a/cffi/vengine_cpy.py b/cffi/vengine_cpy.py
--- a/cffi/vengine_cpy.py
+++ b/cffi/vengine_cpy.py
@@ -275,22 +275,23 @@
             tovar, tp.get_c_name(''), errvalue))
         self._prnt('    %s;' % errcode)
 
-    def _extra_local_variables(self, tp, localvars):
+    def _extra_local_variables(self, tp, localvars, freelines):
         if isinstance(tp, model.PointerType):
             localvars.add('Py_ssize_t datasize')
+            localvars.add('struct _cffi_freeme_s *large_args_free = NULL')
+            freelines.add('if (large_args_free != NULL)'
+                          ' _cffi_free_array_arguments(large_args_free);')
 
     def _convert_funcarg_to_c_ptr_or_array(self, tp, fromvar, tovar, errcode):
         self._prnt('  datasize = _cffi_prepare_pointer_call_argument(')
         self._prnt('      _cffi_type(%d), %s, (char **)&%s);' % (
             self._gettypenum(tp), fromvar, tovar))
         self._prnt('  if (datasize != 0) {')
-        self._prnt('    if (datasize < 0)')
-        self._prnt('      %s;' % errcode)
-        self._prnt('    %s = alloca((size_t)datasize);' % (tovar,))
-        self._prnt('    memset((void *)%s, 0, (size_t)datasize);' % (tovar,))
-        self._prnt('    if (_cffi_convert_array_from_object('
-                   '(char *)%s, _cffi_type(%d), %s) < 0)' % (
-            tovar, self._gettypenum(tp), fromvar))
+        self._prnt('    %s = ((size_t)datasize) <= 640 ? '
+                   'alloca((size_t)datasize) : NULL;' % (tovar,))
+        self._prnt('    if (_cffi_convert_array_argument(_cffi_type(%d), %s, '
+                   '(char **)&%s,' % (self._gettypenum(tp), fromvar, tovar))
+        self._prnt('            datasize, &large_args_free) < 0)')
         self._prnt('      %s;' % errcode)
         self._prnt('  }')
 
@@ -369,15 +370,17 @@
             prnt('  %s;' % type.get_c_name(' x%d' % i, context))
         #
         localvars = set()
+        freelines = set()
         for type in tp.args:
-            self._extra_local_variables(type, localvars)
-        for decl in localvars:
+            self._extra_local_variables(type, localvars, freelines)
+        for decl in sorted(localvars):
             prnt('  %s;' % (decl,))
         #
         if not isinstance(tp.result, model.VoidType):
             result_code = 'result = '
             context = 'result of %s' % name
             prnt('  %s;' % tp.result.get_c_name(' result', context))
+            prnt('  PyObject *pyresult;')
         else:
             result_code = ''
         #
@@ -409,9 +412,14 @@
         if numargs == 0:
             prnt('  (void)noarg; /* unused */')
         if result_code:
-            prnt('  return %s;' %
+            prnt('  pyresult = %s;' %
                  self._convert_expr_from_c(tp.result, 'result', 'result type'))
+            for freeline in freelines:
+                prnt('  ' + freeline)
+            prnt('  return pyresult;')
         else:
+            for freeline in freelines:
+                prnt('  ' + freeline)
             prnt('  Py_INCREF(Py_None);')
             prnt('  return Py_None;')
         prnt('}')
@@ -981,6 +989,59 @@
     return PyBool_FromLong(was_alive);
 }
 
+union _cffi_union_alignment_u {
+    unsigned char m_char;
+    unsigned short m_short;
+    unsigned int m_int;
+    unsigned long m_long;
+    unsigned long long m_longlong;
+    float m_float;
+    double m_double;
+    long double m_longdouble;
+};
+
+struct _cffi_freeme_s {
+    struct _cffi_freeme_s *next;
+    union _cffi_union_alignment_u alignment;
+};
+
+#ifdef __GNUC__
+  __attribute__((unused))
+#endif
+static int _cffi_convert_array_argument(CTypeDescrObject *ctptr, PyObject *arg,
+                                        char **output_data, Py_ssize_t datasize,
+                                        struct _cffi_freeme_s **freeme)
+{
+    char *p;
+    if (datasize < 0)
+        return -1;
+
+    p = *output_data;
+    if (p == NULL) {
+        struct _cffi_freeme_s *fp = (struct _cffi_freeme_s *)PyObject_Malloc(
+            offsetof(struct _cffi_freeme_s, alignment) + (size_t)datasize);
+        if (fp == NULL)
+            return -1;
+        fp->next = *freeme;
+        *freeme = fp;
+        p = *output_data = (char *)&fp->alignment;
+    }
+    memset((void *)p, 0, (size_t)datasize);
+    return _cffi_convert_array_from_object(p, ctptr, arg);
+}
+
+#ifdef __GNUC__
+  __attribute__((unused))
+#endif
+static void _cffi_free_array_arguments(struct _cffi_freeme_s *freeme)
+{
+    do {
+        void *p = (void *)freeme;
+        freeme = freeme->next;
+        PyObject_Free(p);
+    } while (freeme != NULL);
+}
+
 static int _cffi_init(void)
 {
     PyObject *module, *c_api_object = NULL;
diff --git a/testing/cffi0/test_function.py b/testing/cffi0/test_function.py
--- a/testing/cffi0/test_function.py
+++ b/testing/cffi0/test_function.py
@@ -520,3 +520,16 @@
         assert str(e.value).startswith("library '")
         assert str(e.value).endswith("' has already been closed")
         ffi.dlclose(lib)    # does not raise
+
+    def test_passing_large_list(self):
+        if self.Backend is CTypesBackend:
+            py.test.skip("the ctypes backend doesn't support this")
+        ffi = FFI(backend=self.Backend())
+        ffi.cdef("""
+            void getenv(char *);
+        """)
+        needs_dlopen_none()
+        m = ffi.dlopen(None)
+        arg = [b"F", b"O", b"O"] + [b"\x00"] * 20000000
+        x = m.getenv(arg)
+        assert x is None
diff --git a/testing/cffi0/test_verify.py b/testing/cffi0/test_verify.py
--- a/testing/cffi0/test_verify.py
+++ b/testing/cffi0/test_verify.py
@@ -2550,3 +2550,13 @@
         """.replace('?', str(a)))
         # the verify() crashes if the values in the enum are different from
         # the values we computed ourselves from the cdef()
+
+def test_passing_large_list():
+    ffi = FFI()
+    ffi.cdef("""void passing_large_list(long[]);""")
+    lib = ffi.verify("""
+        static void passing_large_list(long a[]) { }
+    """)
+    arg = list(range(20000000))
+    lib.passing_large_list(arg)
+    # assert did not segfault
diff --git a/testing/cffi1/test_recompiler.py b/testing/cffi1/test_recompiler.py
--- a/testing/cffi1/test_recompiler.py
+++ b/testing/cffi1/test_recompiler.py
@@ -2445,3 +2445,13 @@
         };
     """)
     py.test.raises(RuntimeError, ffi.new, "struct BinaryTree *")
+
+def test_passing_large_list():
+    ffi = FFI()
+    ffi.cdef("""void passing_large_list(long[]);""")
+    lib = verify(ffi, "test_passing_large_list", """
+        static void passing_large_list(long a[]) { }
+    """)
+    arg = list(range(20000000))
+    lib.passing_large_list(arg)
+    # assert did not segfault


More information about the pypy-commit mailing list