Adds support for Unique Local Addresses #2856
Conversation
n-thumann
force-pushed
the
ipv6_ula_support
branch
from
2d4245b to
7cb4809
Compare
n-thumann
force-pushed
the
ipv6_ula_support
branch
from
7cb4809 to
9aa9ac4
Compare
Codecov Report
@@ Coverage Diff @@
## master #2856 +/- ##
==========================================
+ Coverage 88.15% 88.19% +0.03%
==========================================
Files 254 254
Lines 54240 54246 +6
==========================================
+ Hits 47817 47841 +24
+ Misses 6423 6405 -18
|
| @@ -783,12 +779,14 @@ def in6_getscope(addr): | |||
| """ | |||
| Returns the scope of the address. | |||
| """ | |||
| if in6_isgladdr(addr) or in6_isuladdr(addr): | |||
| if in6_isgladdr(addr): | |||
There was a problem hiding this comment.
According to https://tools.ietf.org/html/rfc4193#section-3.3 the scope is global. This function should not be modified.
| addrType = (IPV6_ADDR_UNICAST | IPV6_ADDR_GLOBAL) | ||
| if naddr[:2] == b' \x02': # Mark 6to4 @ | ||
| if in6_isaddr6to4(addr): # Mark 6to4 |
There was a problem hiding this comment.
Please replace @ with address instead of removing it.
| @@ -164,29 +167,27 @@ def rfc3484_cmp(source_a, source_b): | |||
| return candidate_set[0] | |||
|
|
|||
|
|
|||
| # Think before modify it : for instance, FE::1 does exist and is unicast | |||
| # there are many others like that. | |||
| # TODO : integrate Unique Local Addresses | |||
There was a problem hiding this comment.
Do not remove the other comments.
| addrType = (IPV6_ADDR_UNICAST | IPV6_ADDR_LINKLOCAL) | ||
| elif in6_isuladdr(addr): | ||
| addrType = (IPV6_ADDR_UNICAST | IPV6_ADDR_UNIQUE_LOCAL) |
There was a problem hiding this comment.
It should also be IPV6_ADDR_GLOBAL
| @@ -59,12 +60,14 @@ def cset_sort(x, y): | |||
| return -res | |||
|
|
|||
| cset = [] | |||
| if in6_isgladdr(addr) or in6_isuladdr(addr): | |||
There was a problem hiding this comment.
I am not convinced that this change is correct. Did you test it on Linux with unique local addresses? It seems that the Linux kernel does not specify IPV6_ADDR_UNIQUE_LOCAL. Therefore, this patch will likely break address selection.
|
I am closing this PR as we did not receive an answer, and that I doubt that it works correctly. |
As suggested in #399, this PR adds support for detecting Unique Local Addresses.