- Use local MTU detection
- i2np.udp.mtu now sets max MTU, not initial MTU
- Put local MTU in netDb
- Fix receive MTU calculations
- Track remote MTU based on actual received packet size
- Display local MTU on peers page
GMP 4.3.2
Compiled on Raspberry Pi with gcc version: gcc (Debian 4.6.3-8+rpi1) 4.6.3
java version "1.6.0_24"
OpenJDK Runtime Environment (IcedTea6 1.11.3) (6b24-1.11.3-2+rpi1)
OpenJDK Zero VM (build 20.0-b12, mixed mode)
Stripped.
Had a report that the ARMv5 jbigi worked on the RPi but it didn't for me.
See NativeBigInteger for more info.
/proc/cpuinfo:
Processor : ARMv6-compatible processor rev 7 (v6l)
BogoMIPS : 697.95
Features : swp half thumb fastmult vfp edsp java tls
CPU implementer : 0x41
CPU architecture: 7
CPU variant : 0x0
CPU part : 0xb76
CPU revision : 7
Hardware : BCM2708
Revision : 0002
- Limit PacketHandler threads to 1 (ticket #660)
- Limit queue sizes between UDPReceiver and PacketHandler,
and between PacketHandler and MessageReceiver, to prevent OOMs
and/or excessive queue delays
- Increase UDPPacket cache size based on max mem
- Remove more stats
At the same time I'm making more Debian-specific changes, such as
- not checking for Gentoo (this is obviously never going to be true :P)
- not mentioning IzPack replacing variables (also not applicable)
- inform about dpkg-reconfigure i2p in the event someone tries to remove or
install an initscript with "i2prouter install|remove|uninstall"
- remove the manual wrapper instructions
- Catch some cases where UDPPackets were not returned to the cache (ticket #675)
- Fix inverted logging logic (ticket #663)
- Remove check in UDPPacket.getLifetime() (ticket #664)
- RemoteHostID cleanup, cache hashcode
- Remove udp.fetchRemoteSlow stat
- Remove some time stamping in UDPPacket
- Other cleanups - see http://zzz.i2p/topics/1198