Age | Commit message (Collapse) | Author | |
---|---|---|---|
2011-09-04 | NEWS and version update. | Ondrej Zajicek | |
2011-07-08 | NEWS and version update.v1.3.2 | Ondrej Zajicek | |
2011-05-02 | NEWS and version update.v1.3.1 | Ondrej Zajicek | |
2011-03-30 | NEWS and version update. | Ondrej Zajicek | |
(and minor changes in documentation) | |||
2010-05-31 | NEWS and version update.v1.2.3 | Ondrej Zajicek | |
2010-04-09 | NEWS and version update.v1.2.2 | Ondrej Zajicek | |
2010-02-10 | NEWS and version update. | Ondrej Zajicek | |
2010-01-06 | NEWS, version and documentation update. | Ondrej Zajicek | |
2009-12-19 | NEWS and version update. | Ondrej Zajicek | |
2009-11-19 | New version. | Ondrej Zajicek | |
2009-11-02 | Version 1.1.5 | Ondrej 'Feela' Filip | |
2009-10-28 | New version. | Ondrej Zajicek | |
2009-09-30 | New version. | Ondrej Zajicek | |
2009-08-12 | NEWS and version update. | Ondrej Zajicek | |
2009-06-25 | NEWS and version update. | Ondrej Zajicek | |
2009-05-22 | Update versions. | Ondrej Zajicek | |
2004-06-07 | Build and run both IPv4 and IPv6. | Ondrej Filip | |
2004-06-06 | RPM is now able to build IPv4 and IPv6 bird. | Ondrej Filip | |
2001-06-09 | Yes, 1.0.5 released. | Ondrej Filip | |
2000-09-09 | Minor rpm changes to make RedHat happy. | Ondrej Filip | |
2000-09-03 | Serious bug in ext lsa origination found. Going for 1.0.4. | Ondrej Filip | |
2000-09-03 | Version 1.0.3. | Ondrej Filip | |
2000-08-25 | Minor bug. It does not work on non RH systems. Thanx to | Ondrej Filip | |
Craig Bourne <cbourne@cbourne.com> | |||
2000-08-24 | Added files for RPM building. | Ondrej Filip | |
1998-12-19 | Variance estimation fixed. | Martin Mares | |
1998-12-19 | Added several tools for fib hashing function analysis. It turned out | Martin Mares | |
we can use very simple function which is monotonic with respect to re-hashing: n ^= n >> 16; n ^= n << 10; h = (n >> (16 - o)) & ((1 << o) - 1); where o is table order. Statistical analysis for both backbone routing table and local OSPF routing tables gives values near theoretical optimum for uniform distribution (see ips.c for formulae). The trick is very simple: We always calculate a 16-bit hash value n and use o most significant bits (this gives us monotonity wrt. rehashing if we sort the chains by the value of n). The first shift/xor pair reduces the IP address to a 16-bit one, the second pair makes higher bits of the 16-bit value uniformly distributed even for tables containing lots of long prefixes (typical interior routing case with 24-bit or even longer prefixes). |