On 04/30/2015 02:14 AM, José Bollo wrote: > Le mercredi 29 avril 2015 à 13:58 -0500, Rob Landley a écrit : >> Of course I dunno if your security infrastructure is going to veto the >> open anyway. Still can't test it... > > The issue is maybe not related to security. The reason is that both > lgetxattr and open(READ|NOFOLLOW|NOATIME) will fail if security forbids > the read.
A) But will open(O_PATH) fail? That's open _without_ read, providing a filehandle to the file's metadata but not the contents. B) If it's not adding _new_ limitations to the command, then switching from lgetxattr to fgetxattr(openat(O_PATH)) sounds right to me. > The consistency of access time of the file can be managed with > NOATIME. I dunno if O_PATH updates atime anyway? (It probably shouldn't. If mv doesn't...) You may have noticed, I regularly hit design corners where "I wonder if they thought of X", and then sometimes have to push a kernel patch. (Rich has been outdoing me here lately, though.) Rob _______________________________________________ Toybox mailing list [email protected] http://lists.landley.net/listinfo.cgi/toybox-landley.net
