Hi, Replying to myself. Christophe's email made me realize that I use additional Link-Local addresses that are not obtained through SLAAC. This means that my test between the two Linux hosts are completely useless (as it might have not made any use of your new code). I'll perform some additional tests tonight and let you know the outcome. The test I performed with Contiki stands, because it verifies that a Linux host can understand a compressed header sent from a Contiki-enabled mote.
As a side note, I read RFC 4944 and RFC 6282 (this is what you implemented actually) and I'd say the current code pretty much follow the standard. I'm not too sure it would work well with short addresses, but that not part of your patch anyway. The whole address uncompression would need to be rewritten IMHO (and I believe Alexander Arig is on it). Regards, Tony Le 2013-04-11 15:57, Tony Cheneau a écrit : > Hi Ralph, > > First, thank you for your patches. I've tested them yesterday. > Because > I didn't had enough hardware to perform a proper packet capture, so I > just checked the connectivity. > > I could run all my regression tests (ICMP, TCP, UDP, varying packet > length) between two linux hosts with your patches applied. They all > passed and looked fine. > > I could also do basic interrop with Contiki (yeah!). Using one linux > host with your patches, I could ping a Contiki device on its > Link-Local > Address (I could not ping the global address though, but I guess it's > not supposed to be fixed by this patch, and I'll need to look more > into > it). I did not modify the Contiki example much, so I could not check > that UDP/TCP would work as well (those are done using the global > addresses by default). Still, I can also confirm that without the > patches, the ping echo requests are just not answered. > > From a functional perspective, your patches look fine. I didn't > checked > that your patches are good from a logical point of view (making sure > they indeed match the RFC better). I'll try to read both RFC 4944 and > RFC 6282 today, and provide additional comments. > > Regards, > Tony > > > Le 2013-04-11 13:51, Ralph Droms a écrit : >> These patches fix link-local address uncompression, for both source >> and destination addresses. >> >> - Ralph >> >> Ralph Droms (3): >> net/ieee802154: 6lowpan.c: Fixes for stateless address >> decompression >> to use MAC address >> net/ieee802154: 6lowpan.c: This patch fixes source address >> uncompression to send the MAC address to lowpan_uncompress_addr >> net/ieee802154: 6lowpan.c: This patch fixes destination address >> uncompression to send the MAC address to lowpan_uncompress_addr >> >> net/ieee802154/6lowpan.c | 53 >> +++++++++++++++++++++++++++------------------ >> 1 files changed, 32 insertions(+), 21 deletions(-) > > > ------------------------------------------------------------------------------ > Precog is a next-generation analytics platform capable of advanced > analytics on semi-structured data. The platform includes APIs for > building > apps and a phenomenal toolset for data science. Developers can use > our toolset for easy data analysis & visualization. Get a free > account! > http://www2.precog.com/precogplatform/slashdotnewsletter > _______________________________________________ > Linux-zigbee-devel mailing list > Linux-zigbee-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/linux-zigbee-devel ------------------------------------------------------------------------------ Precog is a next-generation analytics platform capable of advanced analytics on semi-structured data. The platform includes APIs for building apps and a phenomenal toolset for data science. Developers can use our toolset for easy data analysis & visualization. Get a free account! http://www2.precog.com/precogplatform/slashdotnewsletter _______________________________________________ Linux-zigbee-devel mailing list Linux-zigbee-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/linux-zigbee-devel
