patchWarzone 2100 Project - Patches: patch #978, updates netlog.c

 
 
Show feedback again

You are not allowed to post comments on this tracker with your current authentification level.

patch #978: updates netlog.c

Submitted by:  Bugs Buggy <buginator>
Submitted on:  Wed 13 Feb 2008 05:56:57 AM UTC  
 
Category: FixPriority: 5 - Normal
Status: NonePrivacy: Public
Assigned to: NoneOpen/Closed: Open
Planned Release: None

Sun 17 Feb 2008 05:06:45 AM UTC, comment #3:

another slight update

(file #3837)

Bugs Buggy <buginator>
Project Administrator
Wed 13 Feb 2008 08:49:47 PM UTC, comment #2:

This is just the netlog.c enhancement for all known packets.

(file #3824)

Bugs Buggy <buginator>
Project Administrator
Wed 13 Feb 2008 05:59:55 AM UTC, comment #1:

Looks like we got packet polution again. Or maybe to be more percise, we just moved the packet polution over.

This is not the fault of dpid being wrong, but when we are decoding packets, we never check if that is the correct packet to decode with! That is why it currently crashes.

Right now, in the lobby, with this patch applied, you will see a few things.
1) it will start spamming errors in the console about what the host/client is sending via the setMultiStats() call. Just shows dpid, #games played & #game wins.

2) if will also start spamming errors in the console about the main cause of the current crash issue we are having. And that is, we are hitting the recvMultiStats() routine with the wrong packet type. It tells you the type. In my tests, it showed
MSG_PLAYER_INFO (1 time) & NET_PING for all remaining iterations.

the actual packet type is NET_PING = 8 & MSG_PLAYER_INFO = 92.
In NETprocessSystemMessage(), pMsg->type = 12 though, for BOTH of them.

3) it will NOT crash now in the lobby, it will just keep spamming the above messages
over & over.

4)netplay.log is expanded to include all known packet types we send, and the output is a bit easier to read IMO.

5) this is NOT a final patch of anything except for netlog.c
This is just meant to show the problem more clearly.

The fix will take some more thinking on how we want to handle this, since I believe in r3759 they used the static NETMSG message; To handle the packets with the MSG* prefix.
Now, we are handling both MSG* & NET* which is screwing us up it looks like it.
But I could be wrong, and it is a simple fix. :)

Bugs Buggy <buginator>
Project Administrator
Wed 13 Feb 2008 05:56:57 AM UTC, original submission:

details to follow.

Bugs Buggy <buginator>
Project Administrator

 

Attached Files
file #3837:  netlog2b.patch added by buginator (5kB - text/x-diff)
file #3824:  netlog.patch added by buginator (9kB - text/x-diff)
file #3808:  netlogplus.patch added by buginator (13kB - text/x-diff)

 

Depends on the following items: None found

Items that depend on this one: None found

 

Carbon-Copy List
  • -unavailable- added by buginator (Submitted the item)
  •  

    Do you think this task is very important?
    If so, you can click here to add your encouragement to it.
    This task has 0 encouragements so far.

    Only logged-in users can vote.

     

    Please enter the title of George Orwell's famous dystopian book (it's a date):

     

     

    Follow 3 latest changes.

    Date Changed By Updated Field Previous Value => Replaced By
    Sun 17 Feb 2008 05:06:45 AM UTCbuginatorAttached File-=>Added netlog2b.patch, #3837
    Wed 13 Feb 2008 08:49:47 PM UTCbuginatorAttached File-=>Added netlog.patch, #3824
    Wed 13 Feb 2008 05:56:57 AM UTCbuginatorAttached File-=>Added netlogplus.patch, #3808
    Show feedback again

    Back to the top


    Powered by Savane 3.1-cleanup