Adam Ierymenko
da9a720c3f
Hash table bug fix, and add copy constructor and assignment operator for principle of least surprise.
2015-09-03 17:33:06 -07:00
Adam Ierymenko
b11ffc9635
Integrate Hashtable into Multicaster, where @mwarning found heaviest std::map() overhead.
2015-08-27 16:17:21 -07:00
Adam Ierymenko
3947807b1f
A simple and fast Hashtable, tested but not yet integrated with anything.
2015-08-27 15:36:13 -07:00
Adam Ierymenko
5fcef91278
Ignore old root-topology if present -- fix for upgrading on Linux and possibly elsewhere.
2015-08-25 09:42:37 -07:00
Adam Ierymenko
facb009a1d
Add security notice to auto-update info in -h output, and fix a missing paren.
2015-07-31 09:50:55 -07:00
Adam Ierymenko
8d09c37140
Remove a bit of redundant logic, and also announce MULTICAST_LIKEs to controllers (for future use).
2015-07-31 09:37:13 -07:00
Adam Ierymenko
f6ced547be
Dead code removal.
2015-07-31 08:56:31 -07:00
Adam Ierymenko
eea8d58afa
docs,cleanup
2015-07-28 12:39:03 -07:00
Adam Ierymenko
21e6850722
Cancel NAT-t attempts if peer is no longer "alive"
2015-07-28 12:18:59 -07:00
Adam Ierymenko
5986d83738
Kill more kittens.
2015-07-28 12:04:14 -07:00
Adam Ierymenko
4564dd95ff
Revert... no luck with any of that.
2015-07-28 12:00:50 -07:00
Adam Ierymenko
d2bfdfa6e7
Play with NAT-t tweaks some more.
2015-07-28 11:57:18 -07:00
Adam Ierymenko
b69afa010e
Disable type punning on ARM by ifdef.
2015-07-28 11:50:01 -07:00
Adam Ierymenko
708aac1ea7
Remove some left over debug code, and fix attempt to send to self if we are an active bridge.
2015-07-28 11:43:09 -07:00
Adam Ierymenko
17bfd4d55e
Add TRACE for NAT-t debugging.
2015-07-28 11:32:34 -07:00
Adam Ierymenko
b31071463c
Try another NAT traversal improvement.
2015-07-28 11:28:47 -07:00
Adam Ierymenko
dda376c9eb
Nuke some abandoned code.
2015-07-28 11:16:43 -07:00
Adam Ierymenko
821f1f366e
Fix to NAT escalation sequence.
2015-07-27 17:34:58 -07:00
Adam Ierymenko
e99eda4a4a
Fix IP scoping bug, and disable remotely reported surface push... not helping. :(
2015-07-27 17:28:13 -07:00
Adam Ierymenko
fadb291962
Fix infinite loop typo.
2015-07-27 17:14:49 -07:00
Adam Ierymenko
f0003ea922
Push remote surface as reported by peers along with known interface direct paths to assist with (some) NAT traversal. (trying this, may back out if not effective)
2015-07-27 17:02:43 -07:00
Adam Ierymenko
e30ba3e138
Eliminate some aggressive port scanning NAT-t behavior that has proven ineffective.
2015-07-27 16:43:27 -07:00
Adam Ierymenko
7a15d8a7e3
Fix leaving of networks to actually call Network::destroy().
2015-07-24 14:50:44 -07:00
Adam Ierymenko
d647a587a1
(1) Fix updating of network revision counter on member change.
...
(2) Go back to timestamp as certificate revision number. This is simpler
and more robust than using the network revision number for this and
forcing network revision fast-forward, which could cause some peers
to fall off the horizon when you don't want them to.
2015-07-23 17:18:20 -07:00
Adam Ierymenko
b3516c599b
Add a rate limiting circuit breaker to the network controller to prevent flooding attacks and race conditions.
2015-07-23 10:10:17 -07:00
Adam Ierymenko
3ba54c7e35
Eliminate some poorly thought out optimizations from the netconf/controller interaction,
...
and go ahead and bump version to 1.0.4.
For a while in 1.0.3 -dev I was trying to optimize out repeated network controller
requests by using a ratcheting mechanism. If the client received a network config
that was indeed different from the one it had, it would respond by instantlly
requesting it again.
Not sure what I was thinking. It's fundamentally unsafe to respond to a message
with another message of the same type -- it risks a race condition. In this case
that's exactly what could happen.
It just isn't worth the added complexity to avoid a tiny, tiny amount of network
overhead, so I've taken this whole path out.
A few extra bytes every two minutes isn't worth fretting about, but as I recall
the reason for this optimization was to save CPU on the controller. This can be
achieved by just caching responses in memory *there* and serving those same
responses back out if they haven't changed.
I think I developed that 'ratcheting' stuff before I went full time on this. It's
hard to develop stuff like this without hours of sustained focus.
2015-07-23 09:50:10 -07:00
Adam Ierymenko
8c18c60c00
sp.
2015-07-17 15:30:13 -07:00
Adam Ierymenko
88949a750f
Workaround for uclibc missing map::operator==()
2015-07-13 11:39:55 -07:00
Adam Ierymenko
547b1c6157
Add additional TRACE output in pushDirectPaths.
2015-07-13 10:35:33 -07:00
Adam Ierymenko
0b354803f3
Clean up some YAGNI issues with implementation of GitHub issue #180 , and make best path choice aware of path rank.
2015-07-13 10:03:04 -07:00
Adam Ierymenko
0b9524f23d
Merge branch 'adamierymenko-dev' of http://git.int.zerotier.com/zerotier/zerotierone into adamierymenko-dev
2015-07-13 09:30:02 -07:00
Adam Ierymenko
4bf3bcbd55
Fixes to PUSH_DIRECT_PATHS.
2015-07-13 09:29:51 -07:00
Adam Ierymenko
3f0eca72f7
ZT_TRACE build fix.
2015-07-13 08:36:22 -07:00
Adam Ierymenko
fe20f0d7cd
Put back legacy code to listen for LAN announcements to support same network location with pre-1.0.4 clients.
2015-07-13 08:33:02 -07:00
Adam Ierymenko
a297e4a5bf
Add build def ZT_NO_TYPE_PUNNING, which when defined disables type punning code that might cause unaligned access errors on architectures that care (e.g. Android/ARM)
2015-07-08 09:12:51 -07:00
Adam Ierymenko
412389ec75
Implement ERROR_UNWATNED_MULTICAST
2015-07-07 11:49:38 -07:00
Adam Ierymenko
cf6f30963c
Kill a potential source of type punning BUS errors on Android, and besides that hack probably did not improve performance at all given the short lenghts of things compared with secureEq()
2015-07-07 10:59:59 -07:00
Adam Ierymenko
3f567a07ca
Save a little bit of RAM by getting rid of overkill CMWC4096 non-crypto PRNG and replacing it with a simple non-crypto PRNG that just uses Salsa20.
2015-07-07 10:49:50 -07:00
Adam Ierymenko
41fc08b330
etherTypeName() is only used in Switch and only with ZT_TRACE
2015-07-07 10:06:05 -07:00
Adam Ierymenko
07ea4fd4f9
Fix potential bug in controller config request.
2015-07-07 10:02:48 -07:00
Adam Ierymenko
778c7e6e70
More cleanup to direct path push, comment fixes, etc.
2015-07-07 10:00:34 -07:00
Adam Ierymenko
c863ff3f02
A bunch of comments and cleanup, including some to yesterday's direct path pushing changes. Move path viability check to one place, and stop trying to use link-local addresses since they are not reliable.
2015-07-07 08:54:48 -07:00
Adam Ierymenko
f398952a6c
Revert some bad docs in Packet -- I think we will still use that. Also rename addMembershipCertificate to more security-descriptive validateAndAddMembershipCertificate, give it a return value, and drop unused force parameter.
2015-07-07 08:14:41 -07:00
Adam Ierymenko
6da9d2d36f
Remove debug printf().
2015-07-06 17:23:22 -07:00
Adam Ierymenko
f881cdd767
Add new .h file to VS build, and Windows side of local interface address enumeration.
2015-07-06 17:22:37 -07:00
Adam Ierymenko
84ba365c77
Fix bug in direct path push send.
2015-07-06 17:20:41 -07:00
Adam Ierymenko
cac55105c3
Fix a regression.
2015-07-06 16:40:23 -07:00
Adam Ierymenko
a87cd2d094
Unix side of local interface address awareness for GitHub issue #180 .
2015-07-06 16:32:34 -07:00
Adam Ierymenko
235f4762b7
Plumbing for local interface addresses -- GitHub issue #180
2015-07-06 15:51:04 -07:00
Adam Ierymenko
79e9a8bcc2
Almost everything for GitHub issue #180 except direct path map setup.
2015-07-06 15:28:48 -07:00
Adam Ierymenko
fad9dff2db
Almost all of GitHub issue #180
2015-07-06 15:05:04 -07:00
Adam Ierymenko
1632aec102
Check Network::isAllowed() always on multicast send.
2015-07-06 14:53:27 -07:00
Adam Ierymenko
255320e2a6
pushDirectPaths() implementation
2015-07-06 14:39:28 -07:00
Adam Ierymenko
93bb934d4e
Some cleanup, docs, and Path -> Path > RemotePath refactor.
2015-07-06 14:08:13 -07:00
Adam Ierymenko
feddd946f9
For curiosity add Salsa20/8 to benchmarks.
2015-07-06 13:51:25 -07:00
Adam Ierymenko
6bfbc43e3c
Include COM with EXT_FRAME in bridged case.
2015-07-06 12:46:27 -07:00
Adam Ierymenko
35b5dcf89d
Kill debug line.
2015-07-06 12:39:20 -07:00
Adam Ierymenko
9743db3538
docs
2015-07-06 12:37:37 -07:00
Adam Ierymenko
e5f7c55c54
Documentation in Packet, more work on path push, and clean up ancient legacy support code in Switch.
2015-07-06 12:34:35 -07:00
Adam Ierymenko
f19c3c51d3
Revert slow non-SSE Salsa20 modification since it did not fix Android/ARM issue. Also update Salsa20 comments and clean up a bit.
2015-07-02 09:00:00 -07:00
Adam Ierymenko
7c9949eea3
For @glimberg -- a *possible* fix to the alignment headaches on Android/ARM. If this works we should find a define that can be used to enable it there since it will slow things down on non-x86 other architectures.
2015-07-01 12:29:23 -07:00
Adam Ierymenko
0cbbcf2884
Rename VERB_CMA to the more descriptive VERB_PHYSICAL_ADDRESS_PUSH
2015-06-29 16:01:01 -07:00
Adam Ierymenko
dbee1b38b3
Fix semantics of std::unique() to actually remove duplicates (hidden memory leak?)
2015-06-29 10:21:28 -07:00
Adam Ierymenko
3f71afd0fb
Put multicast txQueue back to list.
2015-06-26 14:26:35 -07:00
Kees Bos
8a68624dae
Fix cert verification check for self signed signatures
2015-06-26 07:22:13 +02:00
Adam Ierymenko
57c7992c78
GitHub issue #191 - kill intra-network multicast rate limits (which were not well supported or easily configurable anyway) -- this is really left over from the old collaborative multicast propagation algorithm. New algorithm (in for a while) has been sender-side replication in which sender "pays" all bandwidth, which intrinsically limits multicast.
2015-06-26 12:36:45 -07:00
Adam Ierymenko
c287ae4d1d
Redo conversion of _networks to a vector. Just use a simple linear search and put the nwid in a pair with the pointer so linear search can be done without pointer chasing. This should be the fastest option for anything less than dozens of networks, and should save memory over the old map.
2015-06-26 11:38:31 -07:00
Moritz Warning
def9cf2a6a
replace vector::erase, was missed in a previous commit (ref #186 )
2015-06-23 00:09:15 +02:00
Adam Ierymenko
4aa8449fe1
Merge branch '_networks_vector' of https://github.com/mwarning/ZeroTierOne into adamierymenko-dev
2015-06-22 10:34:55 -07:00
Moritz Warning
787608b568
use _network function to reduce source code complexity
2015-06-20 16:34:19 +02:00
Moritz Warning
2b04ac5e9b
replace _networks map by vector
...
_networks is usually quite small, using binary search on a vector might be faster comapred to std::map.
This is especially true when using uClibc++, which uses a list.
2015-06-20 16:25:21 +02:00
Adam Ierymenko
7bae95836c
Root server terminology cleanup, and tighten up a security check by checking full identity of peers instead of just address.
2015-06-19 10:23:25 -07:00
Adam Ierymenko
be4f08a548
Merge pull request #186 from mwarning/txQueue_vector
...
replace txQueue list by vector for faster memory access and less allo…
2015-06-18 15:42:12 -07:00
Moritz Warning
0073d0f694
replace txQueue list by vector for faster memory access and less allocations
2015-06-19 00:28:51 +02:00
Adam Ierymenko
87bb0086de
Almost certain fix for GitHub issue #184 on -dev
2015-06-17 12:46:12 -07:00
Adam Ierymenko
a9a390a930
Apply @marning suggested SSE auto-detect to make builds easier on non-x86 platforms: aa2fd044cb
2015-06-01 19:43:06 -07:00
Adam Ierymenko
aab17d9f1f
Kill obsolete amSupernode method -- we now have no differing behaviors.
2015-06-01 19:18:26 -07:00
Adam Ierymenko
ecb6a1d84f
docs
2015-06-01 19:12:20 -07:00
Adam Ierymenko
5341e32729
Fix to GitHub issue #140 -- network preferred relays. Also go ahead and allow RENDEZVOUS from regular peers.
2015-06-01 19:05:27 -07:00
Adam Ierymenko
960ceb4791
Rest of GitHub issue #140 implementation.
2015-06-01 17:50:44 -07:00
Adam Ierymenko
318c2f025b
Call it _relays since there might be other kinds later.
2015-06-01 17:09:37 -07:00
Adam Ierymenko
5bc89034bc
Always ping / contact network preferred relays (if any).
2015-06-01 16:53:58 -07:00
Adam Ierymenko
b3b9af0dd8
Fix for GitHub issue #170
2015-06-01 11:56:15 -07:00
Adam Ierymenko
5e3c6d9e0d
Some nodeJS work, and apply fix from GitHub issue #166 plus a small optimization to avoid repeated calls to _allMulticastGroups().
2015-05-25 14:21:05 -07:00
Adam Ierymenko
af1d29cc6f
Remove obsolete method.
2015-05-25 11:54:32 -07:00
Adam Ierymenko
d8783b14eb
Build fix.
2015-05-22 15:46:06 -07:00
Adam Ierymenko
6867922d9e
typo
2015-05-22 15:33:33 -07:00
Adam Ierymenko
9774f789f2
TCP fallback tunneling is now working. That hurt more than expected.
2015-05-22 14:52:23 -07:00
Adam Ierymenko
cdec05af24
One second delay between NAT-t events.
2015-05-22 13:12:14 -07:00
Adam Ierymenko
196f27f1f0
Add delay to NAT-t escalation stuff to try to address GitHub issue #167
2015-05-22 13:11:55 -07:00
Adam Ierymenko
d9006712f6
Completely factor out "desperation" from the core. I thought of a significantly simpler way to move all of this logic entirely into the containing service, liberating the core from any concern over the nature of its pipe to the outside world.
2015-05-21 15:58:26 -07:00
Adam Ierymenko
1213073916
Apple auto-update stuff, now for Windows.
2015-05-20 19:38:49 -07:00
Adam Ierymenko
2810cd7c15
Build fixes for G++, building without SQLite3 present, and warning removal.
2015-05-17 23:56:47 +00:00
Adam Ierymenko
0bdd56ebd6
A few revisions to PFS design.
2015-05-15 09:04:39 -07:00
Adam Ierymenko
d0e0f5dd12
Basic OpenBSD compile fixes -- still need to update BSDEthernetTap, will do that later. Should be able to re-use FreeBSD port for OpenBSD, but we will see.
2015-05-15 08:48:53 -07:00
Adam Ierymenko
e94518590d
First stab of PFS design work with PKC security -- may not implement in 1.0.3 but stubbing out.
2015-05-14 17:41:05 -07:00
Adam Ierymenko
a8835cd8b3
Some prep work to make room for perfect forward security (PFS). Will not affect existing clients.
2015-05-13 18:53:37 -07:00
Adam Ierymenko
bdce679d84
Should fix deadlock issue in GitHub issue #166
2015-05-13 16:55:18 -07:00
Kees Bos
a425bbc673
Renamed supernode to rootserver
2015-05-06 12:05:20 +02:00
Adam Ierymenko
845955dea5
Add definition for VERB_CMA -- GitHub issue #180
2015-06-13 18:08:00 +02:00
