Handle zero-length sublist correctly in Python -> SQL array conversion.

If PLySequence_ToArray came across a zero-length sublist, it'd compute
the overall array size as zero, possibly leading to a memory clobber.
(This would likely qualify as a security bug, were it not that plpython
is an untrusted language already.)

I think there are other corner-case issues in this code as well, notably
that the error messages don't match the core code and for some ranges
of array sizes you'd get "invalid memory alloc request size" rather than
the intended message about array size.

Really this code has no business doing its own array size calculation
at all, so remove the faulty code in favor of using ArrayGetNItems().

Per bug #17912 from Alexander Lakhin.  Bug seems to have come in with
commit 94aceed31, so back-patch to all supported branches.

Discussion: https://postgr.es/m/17912-82ceed78731d9cdc@postgresql.org

diff --git a/src/pl/plpython/expected/plpython_types.out b/src/pl/plpython/expected/plpython_types.out
index 98b89b7d5c11d1045fc21cdb2e04983a8b07c307..8e5a73ceedfb42b4f6e989596fc9637a50ba0d4e 100644 (file)
--- a/src/pl/plpython/expected/plpython_types.out
+++ b/src/pl/plpython/expected/plpython_types.out
@@ -687,6 +687,15 @@ SELECT * FROM test_type_conversion_array_mixed2();
 ERROR:  invalid input syntax for type integer: "abc"
 CONTEXT:  while creating return value
 PL/Python function "test_type_conversion_array_mixed2"
+CREATE FUNCTION test_type_conversion_array_mixed3() RETURNS text[] AS $$
+return [[], 'a']
+$$ LANGUAGE plpythonu;
+SELECT * FROM test_type_conversion_array_mixed3();
+ test_type_conversion_array_mixed3 
+-----------------------------------
+ {[],a}
+(1 row)
+
 CREATE FUNCTION test_type_conversion_mdarray_malformed() RETURNS int[] AS $$
 return [[1,2,3],[4,5]]
 $$ LANGUAGE plpythonu;

diff --git a/src/pl/plpython/expected/plpython_types_3.out b/src/pl/plpython/expected/plpython_types_3.out
index a6ec10d5e1821f00381e719ca811c3bd9dac454c..4f72a626e181d49f513f2a674335b643fbf61e18 100644 (file)
--- a/src/pl/plpython/expected/plpython_types_3.out
+++ b/src/pl/plpython/expected/plpython_types_3.out
@@ -687,6 +687,15 @@ SELECT * FROM test_type_conversion_array_mixed2();
 ERROR:  invalid input syntax for type integer: "abc"
 CONTEXT:  while creating return value
 PL/Python function "test_type_conversion_array_mixed2"
+CREATE FUNCTION test_type_conversion_array_mixed3() RETURNS text[] AS $$
+return [[], 'a']
+$$ LANGUAGE plpython3u;
+SELECT * FROM test_type_conversion_array_mixed3();
+ test_type_conversion_array_mixed3 
+-----------------------------------
+ {[],a}
+(1 row)
+
 CREATE FUNCTION test_type_conversion_mdarray_malformed() RETURNS int[] AS $$
 return [[1,2,3],[4,5]]
 $$ LANGUAGE plpython3u;

diff --git a/src/pl/plpython/plpy_typeio.c b/src/pl/plpython/plpy_typeio.c
index 47785735c304e01f96e3600d6f79df35ee53e4ca..dfe24d303ae1693f93ce819227f70c9684a45648 100644 (file)
--- a/src/pl/plpython/plpy_typeio.c
+++ b/src/pl/plpython/plpy_typeio.c
@@ -1151,7 +1151,7 @@ PLySequence_ToArray(PLyObToDatum *arg, PyObject *plrv,
        int                     i;
        Datum      *elems;
        bool       *nulls;
-       int64           len;
+       int                     len;
        int                     ndim;
        int                     dims[MAXDIM];
        int                     lbs[MAXDIM];
@@ -1170,7 +1170,6 @@ PLySequence_ToArray(PLyObToDatum *arg, PyObject *plrv,
         * Determine the number of dimensions, and their sizes.
         */
        ndim = 0;
-       len = 1;
 
        Py_INCREF(plrv);
 
@@ -1186,13 +1185,6 @@ PLySequence_ToArray(PLyObToDatum *arg, PyObject *plrv,
                if (dims[ndim] < 0)
                        PLy_elog(ERROR, "could not determine sequence length for function return value");
 
-               if (dims[ndim] > MaxAllocSize)
-                       PLy_elog(ERROR, "array size exceeds the maximum allowed");
-
-               len *= dims[ndim];
-               if (len > MaxAllocSize)
-                       PLy_elog(ERROR, "array size exceeds the maximum allowed");
-
                if (dims[ndim] == 0)
                {
                        /* empty sequence */
@@ -1220,15 +1212,18 @@ PLySequence_ToArray(PLyObToDatum *arg, PyObject *plrv,
                        PLy_elog(ERROR, "return value of function with array return type is not a Python sequence");
 
                ndim = 1;
-               len = dims[0] = PySequence_Length(plrv);
+               dims[0] = PySequence_Length(plrv);
        }
 
+       /* Allocate space for work arrays, after detecting array size overflow */
+       len = ArrayGetNItems(ndim, dims);
+       elems = palloc(sizeof(Datum) * len);
+       nulls = palloc(sizeof(bool) * len);
+
        /*
         * Traverse the Python lists, in depth-first order, and collect all the
         * elements at the bottom level into 'elems'/'nulls' arrays.
         */
-       elems = palloc(sizeof(Datum) * len);
-       nulls = palloc(sizeof(bool) * len);
        currelem = 0;
        PLySequence_ToArray_recurse(arg->u.array.elm, plrv,
                                                                dims, ndim, 0,

diff --git a/src/pl/plpython/sql/plpython_types.sql b/src/pl/plpython/sql/plpython_types.sql
index cc0524ee806bddcc94d031d6c39ac7a132728060..8fa8f6bee7f3b700d73be3f973beb11eb5a6615c 100644 (file)
--- a/src/pl/plpython/sql/plpython_types.sql
+++ b/src/pl/plpython/sql/plpython_types.sql
@@ -328,6 +328,13 @@ $$ LANGUAGE plpythonu;
 
 SELECT * FROM test_type_conversion_array_mixed2();
 
+CREATE FUNCTION test_type_conversion_array_mixed3() RETURNS text[] AS $$
+return [[], 'a']
+$$ LANGUAGE plpythonu;
+
+SELECT * FROM test_type_conversion_array_mixed3();
+
+
 CREATE FUNCTION test_type_conversion_mdarray_malformed() RETURNS int[] AS $$
 return [[1,2,3],[4,5]]
 $$ LANGUAGE plpythonu;