#gemrb@irc.freenode.net logs for 3 Oct 2014 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[02:20:38] <-- brada has left IRC (Quit: brada)
[02:21:31] --> brada has joined #gemrb
[03:11:43] --> Eli2_ has joined #gemrb
[03:14:52] <-- Eli2 has left IRC (Ping timeout: 260 seconds)
[03:30:06] <-- brada has left IRC (Quit: brada)
[03:51:52] <-- raevol has left IRC (Ping timeout: 240 seconds)
[04:05:45] --> raevol has joined #gemrb
[06:40:40] <-- Lightkey has left IRC (Ping timeout: 260 seconds)
[06:53:10] --> Lightkey has joined #gemrb
[07:37:38] --> Yoshimo has joined #gemrb
[07:37:38] <-- Yoshimo has left IRC (Changing host)
[07:37:38] --> Yoshimo has joined #gemrb
[07:43:07] <-- Yoshimo has left IRC (Read error: Connection reset by peer)
[08:40:03] --> fizzle has joined #gemrb
[08:53:32] --> Yoshimo has joined #gemrb
[08:53:32] <-- Yoshimo has left IRC (Changing host)
[08:53:32] --> Yoshimo has joined #gemrb
[09:33:09] <-- fizzle has left #gemrb
[09:55:10] <-- Yoshimo has left IRC (Excess Flood)
[10:10:58] --> Yoshimo has joined #gemrb
[10:10:58] <-- Yoshimo has left IRC (Changing host)
[10:10:58] --> Yoshimo has joined #gemrb
[12:24:45] <-- raevol has left IRC (Quit: Leaving)
[13:41:22] --> brada has joined #gemrb
[16:53:00] <-- Yoshimo has left IRC (Quit: Yoshimo)
[17:51:31] --> Yoshimo has joined #gemrb
[18:21:50] <-- Yoshimo has left IRC (Ping timeout: 272 seconds)
[21:29:35] <brada> sooo many sprite leaks in gemrb :(
[21:30:04] <fuzzie> in your shiny new branch?
[21:30:11] <brada> in all of gemrb
[21:30:20] <fuzzie> well, define leaks..?
[21:30:20] <brada> lots of calls to GetFrame that dont release
[21:30:53] <brada> this is being pointed out to me bay a valgrind esque profiler btw
[21:31:20] <fuzzie> well
[21:31:32] <fuzzie> the reason I ask is, I think that would've shown up in my valgrind runs
[21:32:36] <brada> its possible this is lying to me i suppose
[21:32:48] <fuzzie> well
[21:32:50] <brada> one was certainly valid, but my fault and isolated to this branch
[21:32:53] <fuzzie> I'm wondering what's going wrong
[21:33:02] <fuzzie> I don't make any judgement :)
[21:33:26] <brada> its flagging GemRB_AddNewArea
[21:33:46] <brada> but doesnt give any more info other than SDLSurfaceSprite2D is being leaked there
[21:34:28] <brada> we have some confusing use of reference counting too :/
[21:34:34] <fuzzie> yes
[21:35:03] <brada> i dont like that GetXXX does ref++ then we pass that to something and call it good
[21:35:05] <fuzzie> I mean, we *do* leak some stuff, and we should complain about it on exit
[21:35:10] <brada> when really that something should bedoing ref++
[21:35:40] <fuzzie> which is where this complaining on exit about refcounts comes from
[21:36:03] <fuzzie> but that is all BAMmery
[21:37:06] <fuzzie> quite possible that the !SupportsBAMSprites() case is broken too.
[21:37:15] <fuzzie> and most likely explanation is that it's all broke n;)
[21:37:18] <fuzzie> but I have to sleep
[21:37:38] <brada> i dont get a complaint on exit
[21:37:43] <brada> do you have to have some flag on?
[21:37:56] <fuzzie> no, but it does depend on you supporting BAM sprites I expect.
[21:38:03] <brada> well im not on opengl
[21:38:07] <brada> on sdl 1
[21:38:26] <brada> anyway we will talk when you are rested :)
[21:38:27] <fuzzie> but the BAMSprite2D system keeps internal refcounts which will whine on exit if it's not zero
[21:38:28] <brada> good night
[21:38:44] <fuzzie> sure, after tomorrow I'm hopefully back to it being a bit quieter.. :) ninight
[22:12:56] <brada> horray i figured out how to get more useful information!
[22:13:41] --> Yoshimo has joined #gemrb
[22:13:45] <brada> and it does seem to be my fault
[22:13:50] <Lightkey> with a $5 wrench?
[22:13:52] <brada> but i cant see how this is leaking :/
[22:14:14] <brada> randomly clicking buttons is like a $5 wrench
[22:15:07] <brada> anyway its flagging the GetAnimationFactoryCall in the BAMFontManager. but i delete that at the end…
[22:22:00] <brada> yeah all back to that odd/irritating GetFrame thing
[22:22:03] <brada> so thats fixed
[22:22:14] <brada> plenty more leaks that arent my fault… i think
[22:22:32] <brada> still that Add_newArea one
[22:22:48] <brada> something in openal
[22:23:03] <brada> and something in initactortables
[22:34:11] <-- Yoshimo has left IRC (Quit: Yoshimo)
[22:46:21] <brada> ah
[22:46:31] <brada> indeed some of these arent my fault! horray!
[22:47:19] <Lightkey> hehe
[23:19:52] <-- Mechanimal has left IRC (Ping timeout: 240 seconds)
[23:28:01] <brada> only leak left is this openal one… but it doesnt seem to come from gemrb
[23:29:03] <brada> call stack all the way back to pthread_start doesnt have any gemrb calls