]> granicus.if.org Git - postgresql/commitdiff
Fix low-probability memory leak in regex execution.
authorTom Lane <tgl@sss.pgh.pa.us>
Fri, 18 Sep 2015 17:55:17 +0000 (13:55 -0400)
committerTom Lane <tgl@sss.pgh.pa.us>
Fri, 18 Sep 2015 17:55:17 +0000 (13:55 -0400)
After an internal failure in shortest() or longest() while pinning down the
exact location of a match, find() forgot to free the DFA structure before
returning.  This is pretty unlikely to occur, since we just successfully
ran the "search" variant of the DFA; but it could happen, and it would
result in a session-lifespan memory leak since this code uses malloc()
directly.  Problem seems to have been aboriginal in Spencer's library,
so back-patch all the way.

In passing, correct a thinko in a comment I added awhile back about the
meaning of the "ntree" field.

I happened across these issues while comparing our code to Tcl's version
of the library.

src/backend/regex/regexec.c

index 6135f59f22b110db90665a7c6ac10dbe8ed2be2a..6e2adf5cac07158f2fc8f95cd8f6e97979ffbcff 100644 (file)
@@ -323,7 +323,11 @@ find(struct vars * v,
                                                   (chr **) NULL, &hitend);
                else
                        end = longest(v, d, begin, v->stop, &hitend);
-               NOERR();
+               if (ISERR())
+               {
+                       freedfa(d);
+                       return v->err;
+               }
                if (hitend && cold == NULL)
                        cold = begin;
                if (end != NULL)