On Tue, 17 Sep 2013 11:58:05 +0700 "Andrey Shernyukov" <andreysh@xxxxxxxxxxxx> wrote: > On Mon, 16 Sep 2013 19:16:53 +0700, Jeff Layton <jlayton@xxxxxxxxx> wrote: > > > No, sorry -- that doesn't tell me much. I had pretty much inferred that > > from the info before. What would be more interesting is the capture of > > the FIND_FIRST/NEXT calls, in binary format so I can open them with > > wireshark and see what it's sending. > > > > О.К. Capture packets are in attachment (except first packets with > authorization information) as "libpcap" wireshark format. > Ok, based on this and the previous text dumps of the QUERY_PATH_INFO packets, it looks like FIND_FIRST/NEXT is able to fetch this info without issue, but the QUERY_PATH_INFO call gets back an error. There's not much we can do about that without some major reworking of the lookup code. I think we really ought to convert the lookup code to use a FIND_FIRST call with a specific name instead of QUERY_PATH_INFO. It seems to "perturb" things less, plus it also gives us the index number which we use as an inode number. That allows us to do lookups on unknown dentries without a second call to fetch the UniqueID. Additionally, I think we ought to make that error not print by default. I'll send a patch to tone that down in a bit. -- Jeff Layton <jlayton@xxxxxxxxx> -- To unsubscribe from this list: send the line "unsubscribe linux-cifs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
