On Sun, 6 Jun 2010 09:16:20 -0400, Matt Turner <[email protected]> wrote: > On Sun, Jun 6, 2010 at 9:14 AM, Mikhail Gusarov <[email protected]> > wrote: > > > > Twas brillig at 10:03:16 06.06.2010 UTC-03 when [email protected] > > did gyre and gimble: > > > > FC> By the way, the hunk below sounds rather fishy, right? > > > > >> @@ -109,14 +107,10 @@ RRCreateOutputProperty (Atom property) > > >> static void > > >> RRDestroyOutputProperty (RRPropertyPtr prop) > > >> { > > >> - if (prop->valid_values) > > >> - free(prop->valid_values); > > >> - if (prop->current.data) > > >> - free(prop->current.data); > > >> - if (prop->pending.data) > > >> - free(prop->pending.data); > > >> - if (prop->valid_values) > > >> - free(prop->valid_values); > > >> + free(prop->valid_values); > > >> + free(prop->current.data); > > >> + free(prop->pending.data); > > >> + free(prop->valid_values); > > >> free(prop); > > >> } > > > > Hmm, why? Prop is zero-initialized on allocation, so it's okay. > > > > -- > > http://fossarchy.blogspot.com/ > > > > _______________________________________________ > > [email protected]: X.Org development > > Archives: http://lists.x.org/archives/xorg-devel > > Info: http://lists.x.org/mailman/listinfo/xorg-devel > > > > I didn't see it at first, but prop->valid_values is in there twice > (first call to free and second-to-last). Nothing your patch did, it > was like that already. > > Matt
Yes, Matt is right. I wasn't clear in my first mail. Sorry. _______________________________________________ [email protected]: X.Org development Archives: http://lists.x.org/archives/xorg-devel Info: http://lists.x.org/mailman/listinfo/xorg-devel
