Socket cleanup #182
Merged
Socket cleanup #182
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Also rename TX bind function to match
Don't list loopback in nicselect dialog Work offline always selects loopback Listener socket binds to all, but only accept destinations that are addresses on that interface (and multicast sACN!)
|
A couple of things I noticed on a quick play. 1: QNetworkDatagram didn't exist in Qt 5.6, which we need for XP support.. so : probably need to fix that, then I can do some extended testing on at least windows and mac.. you want to eliminate the QNetworkDatagram reference? or I can.. 2: There's something funky in the sequence number logic, I keep getting the message Source coming up, not processing packet on a simple test of the application sending to itself. Needs some more digging to find out why. |
|
|
Windows (10, 64bit). I actually see problems both with and without the Listen all Interfaces option checked; 2 interfaces plus loopback, I'm listening on a wifi interface. I get a ton of sequence errors, in a simple case (run application, fire up a single source, double click on a universe to view it). Debug output below.. |
ip_multicast_loop seems to be not respected on Windows. Windows specific define added
|
ok, lets try this... Unfortunately I need the destination IP from the datagram, I don't see how to do this with out QNetworkDatagram. The older method only returns source... |
Use IFDEFs to support Qt5.6; disable multiple nic selection on windows
|
Take a quick look at branch marcusbirkin-SocketCleanup, in particular sacnlistener.cpp. I IFDEF'd out the QNetworkDatagram part, you can still get the sender, and I assume the destination IP is the IP of the socket. I think it's fine to disable multi-NIC reception for XP, so I added that. Works on a quick test on Win10, I'll have to find an XP machine to try it on.. |
Ensure multicast frame is sent to correct listener/thread Cross platform tweaks PCap - Restore send to self, but send to correct listener WinXP - Unable to listen to unicast
|
Commit: #3d324ab OS: Ubuntu 18.04.1 LTS Interface selection: Only physical interfaces shown. OK Interface: Physical interface
Interface: Physical interface
Interface: Work offline (Localhost)
|
|
Looking good now : tested on win10 and mac today. Results below. Haven't yet tested pcap support but good enough for a merge. OS: Windows 10 Enterprise 17134.523 Interface selection: Only physical interfaces shown. Interface: Physical interface Interface: Physical interface Interface: Work offline (Localhost) OS: macOS Mojave 10.14 Interface selection: Only physical interfaces shown. Interface: Physical interface Interface: Physical interface Interface: Work offline (Localhost) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Resolves #179
Possible and likely cause of #167
The socket backend could receive frames multiple times, while this wouldn't have affected the resultant DMX data, it would affect the source statistics.
The socket arrangement has been altered, in particular multicast frames for display locally are passed internally and not re-received over the wire (Because Windows and Unix handle ip_multicast_loop slightly differently with multi homed machines, so this is simply disabled). Additionally the number of open sockets reduced.
This will require a good amount of testing to prove it's good. I'm unable to test on MacOS!