On Fri, Apr 21, 2023 at 10:58:42AM -0400, npiazza@xxxxxxxxxxx wrote: > taken from openbsd ports, and used by many linux distributions. > > https://github.com/openbsd/ports/blob/master/security/wpa_supplicant/patches/patch-src_utils_eloop_c > > does it make sense to include it in wpa_supplicant? Probably not unless someone can provide an explanation why that would be needed. > Don't try to access list members to free them unless already > initialised > > Index: src/utils/eloop.c > @@ -1254,6 +1254,9 @@ void eloop_destroy(void) > struct eloop_timeout *timeout, *prev; > struct os_reltime now; > > + if (eloop.timeout.prev == NULL) > + return; It is not valid to call eloop_destroy() without having first called eloop_init() successfully. As such, this condition should never be reached and should not be needed. If there are applications that do indeed have eloop.timeout.prev == NULL when eloop_destroy() is being called, those applications should be fixed instead. -- Jouni Malinen PGP id EFC895FA _______________________________________________ Hostap mailing list Hostap@xxxxxxxxxxxxxxxxxxx http://lists.infradead.org/mailman/listinfo/hostap
