gnrc_sixlowpan_iphc: add fragment forwarding stubs#12629
gnrc_sixlowpan_iphc: add fragment forwarding stubs#12629miri64 merged 4 commits intoRIOT-OS:masterfrom
Conversation
8f99811 to
4fd7dc9
Compare
|
I also added a subroutine to do the actual forwarding in now. This code will be common between #11068 and SFR, so it makes sense to have a common base PR. Currently the function doesn't do anything except for returning |
|
I started a test application. Up until now, only the success case is tested, I plan however to add some failure cases. When setting |
eac9232 to
799261d
Compare
|
Finished tests and squashd. No longer WIP. |
|
Updated the testing procedures |
|
Not really sure why Murdock is failing on Arch Linux Operating System Environment
-----------------------------
Operating System: "Arch Linux"
Kernel: Linux 5.3.10-arch1-1 x86_64 unknown
Ubuntu 19.10Operating System Environment
-----------------------------
Operating System: "Ubuntu" "19.10 (Eoan Ermine)"
Kernel: Linux 5.3.0-23-generic x86_64 x86_64
|
|
Mhhhhh... they do work on my machine :-/ |
|
Ah, there might be something going wrong with current master rebased. Will check! |
b0aa053 to
d964e69
Compare
|
First of all: squashed! |
|
(and rebased, though this wasn't my intention ^^") |
d964e69 to
ddcc6ea
Compare
|
Fixed. The reason was that the arrival time of the VRB entries that filled the VRB was not set for the VRB-full test, so the garbage collection just through them out when the new entry was created. |
|
Not sure why this worked with the old version of this PR... But now its fixed so 🤷♀️ |
ddcc6ea to
b31d0db
Compare
|
Another update to |
benpicco
left a comment
There was a problem hiding this comment.
Looks good overall, most of it is just adding stub code that doesn't do anything on it's own yet.
b1d4cb3 is a nice small cleanup.
Only found some small things - comments are always good to not get lost in the sea of acronyms.
sys/net/gnrc/network_layer/sixlowpan/iphc/gnrc_sixlowpan_iphc.c
Outdated
Show resolved
Hide resolved
b31d0db to
3273602
Compare
|
Addressed comments and squashed directly as otherwise I wouldn't have been able to adapt the comment that was removed in a later commit. |
benpicco
left a comment
There was a problem hiding this comment.
Code looks good and tests are still passing - changes outside of MODULE_GNRC_SIXLOWPAN_FRAG_VRB are minimal anyway.
|
Thanks for the review! |
Contribution description
... when VRB is available
This is once again one of these tiny PRs that belong to a bigger PR, @benpicco is always talking about. However, in this case it was already introduced in an earlier version to #11068 (edc9b42) but is now needed also for SFR. So instead of having this piece of code copy-pasted in both, I decided to split it out (even though it is not functional without it).
Testing procedure
There is a test application
tests/gnrc_sixlowpan_iphc_w_vrbin this PR that should work on any board that it fits on:Confirm that a fragmented and unfragmented UDP packet can still be sent between two 6LoWPAN nodes.
Additionally, you can check-out #11068 and
gnrc_sixlowpan_iphc.cmake -C tests/gnrc_sixlowpan_frag_minfwd flash teston a board they fit onIssues/PRs references
Taken out of #11068's edc9b42 (as it is also required for other 6LoWPAN fragment forwarding schemes like SFR). It is however completely independent of that.