bpo-36301: Error if decoding pybuilddir.txt fails (GH-12422)

vstinner · web-flow · commit 5f9cf23502fe · 2019-03-19T01:46:25.000+01:00
Python initialization now fails if decoding pybuilddir.txt
configuration file fails at startup.

_PyPathConfig_Calculate() now reports memory allocation failure and
decoding error on decoding pybuilddir.txt content from
UTF-8/surrogateescape.
diff --git a/Include/internal/pycore_fileutils.h b/Include/internal/pycore_fileutils.h
@@ -30,7 +30,8 @@ PyAPI_FUNC(int) _Py_EncodeUTF8Ex(
 
 PyAPI_FUNC(wchar_t*) _Py_DecodeUTF8_surrogateescape(
     const char *arg,
-    Py_ssize_t arglen);
+    Py_ssize_t arglen,
+    size_t *wlen);
 
 PyAPI_FUNC(int) _Py_GetForceASCII(void);
 
diff --git a/Misc/NEWS.d/next/Core and Builtins/2019-03-19-00-54-31.bpo-36301.xvOCJb.rst b/Misc/NEWS.d/next/Core and Builtins/2019-03-19-00-54-31.bpo-36301.xvOCJb.rst
@@ -0,0 +1,2 @@
+Python initialization now fails if decoding ``pybuilddir.txt`` configuration
+file fails at startup.
diff --git a/Modules/getpath.c b/Modules/getpath.c
@@ -563,23 +563,27 @@ search_for_exec_prefix(const _PyCoreConfig *core_config,
         }
         else {
             char buf[MAXPATHLEN+1];
-            wchar_t *rel_builddir_path;
             n = fread(buf, 1, MAXPATHLEN, f);
             buf[n] = '\0';
             fclose(f);
-            rel_builddir_path = _Py_DecodeUTF8_surrogateescape(buf, n);
-            if (rel_builddir_path) {
-                wcsncpy(exec_prefix, calculate->argv0_path, MAXPATHLEN);
-                exec_prefix[MAXPATHLEN] = L'\0';
-                err = joinpath(exec_prefix, rel_builddir_path);
-                PyMem_RawFree(rel_builddir_path );
-                if (_Py_INIT_FAILED(err)) {
-                    return err;
-                }
 
-                *found = -1;
-                return _Py_INIT_OK();
+            size_t dec_len;
+            wchar_t *pybuilddir;
+            pybuilddir = _Py_DecodeUTF8_surrogateescape(buf, n, &dec_len);
+            if (!pybuilddir) {
+                return DECODE_LOCALE_ERR("pybuilddir.txt", dec_len);
             }
+
+            wcsncpy(exec_prefix, calculate->argv0_path, MAXPATHLEN);
+            exec_prefix[MAXPATHLEN] = L'\0';
+            err = joinpath(exec_prefix, pybuilddir);
+            PyMem_RawFree(pybuilddir );
+            if (_Py_INIT_FAILED(err)) {
+                return err;
+            }
+
+            *found = -1;
+            return _Py_INIT_OK();
         }
     }
 
diff --git a/Objects/unicodeobject.c b/Objects/unicodeobject.c
@@ -5064,12 +5064,21 @@ _Py_DecodeUTF8Ex(const char *s, Py_ssize_t size, wchar_t **wstr, size_t *wlen,
     return 0;
 }
 
+
 wchar_t*
-_Py_DecodeUTF8_surrogateescape(const char *arg, Py_ssize_t arglen)
+_Py_DecodeUTF8_surrogateescape(const char *arg, Py_ssize_t arglen,
+                               size_t *wlen)
 {
     wchar_t *wstr;
-    int res = _Py_DecodeUTF8Ex(arg, arglen, &wstr, NULL, NULL, 1);
+    int res = _Py_DecodeUTF8Ex(arg, arglen,
+                               &wstr, wlen,
+                               NULL, _Py_ERROR_SURROGATEESCAPE);
     if (res != 0) {
+        /* _Py_DecodeUTF8Ex() must support _Py_ERROR_SURROGATEESCAPE */
+        assert(res != -3);
+        if (wlen) {
+            *wlen = (size_t)res;
+        }
         return NULL;
     }
     return wstr;
diff --git a/Python/pathconfig.c b/Python/pathconfig.c
@@ -712,7 +712,7 @@ _Py_FindEnvConfigValue(FILE *env_file, const wchar_t *key,
             continue;
         }
 
-        wchar_t *tmpbuffer = _Py_DecodeUTF8_surrogateescape(buffer, n);
+        wchar_t *tmpbuffer = _Py_DecodeUTF8_surrogateescape(buffer, n, NULL);
         if (tmpbuffer) {
             wchar_t * state;
             wchar_t * tok = WCSTOK(tmpbuffer, L" \t\r\n", &state);

Original file line number	Diff line number	Diff line change
`@@ -0,0 +1,2 @@`
	`1`	+Python initialization now fails if decoding ``pybuilddir.txt`` configuration
	`2`	`+file fails at startup.`
Original file line number	Diff line number	Diff line change
`@@ -712,7 +712,7 @@ _Py_FindEnvConfigValue(FILE env_file, const wchar_t key,`
`712`	`712`	`continue;`
`713`	`713`	`}`
`714`	`714`
`715`		`- wchar_t *tmpbuffer = _Py_DecodeUTF8_surrogateescape(buffer, n);`
	`715`	`+ wchar_t *tmpbuffer = _Py_DecodeUTF8_surrogateescape(buffer, n, NULL);`
`716`	`716`	`if (tmpbuffer) {`
`717`	`717`	`wchar_t * state;`
`718`	`718`	`wchar_t * tok = WCSTOK(tmpbuffer, L" \t\r\n", &state);`