Hi Matthias, your change looks good. Please note that I'm not a reviewer! I have a suggestion you might want to think about: Why don't you move the free-loop outside the if-else? Except for the NULL check, it's the same code in both branches. Thanks for fixing! Lutz On 26.09.18, 16:30, "core-libs-dev on behalf of Baesken, Matthias" <core-libs-dev-bounces@openjdk.java.net on behalf of matthias.baesken@sap.com> wrote: Hello, could you please review this small change (windows only) ? Currently, the function "getJavaIDFromLangID" (located in windows java_props_md.c) only does proper deallocations after a successful call to the function SetupI18nProps. See if (SetupI18nProps(MAKELCID(langID, SORT_DEFAULT), &(elems[0]), &(elems[1]), &(elems[2]), &(elems[3]), &(elems[4]))) { ...... for (index = 0; index < 5; index++) { free(elems[index]); } However a failing call (SetupI18nProps returning false) might still need deallocations, because the function SetupI18nProps can malloc memory in the failing case as well . The change initializes the pointers in char * elems[5]; And later frees them in case they are not NULL . Webrev and bug : http://cr.openjdk.java.net/~mbaesken/webrevs/8211149.0/8211149.0/webrev/ https://bugs.openjdk.java.net/browse/JDK-8211149 Best regards, Matthias