oggdropXPd/GT3b2: Random crash bug Options

[p0s]

Hi. I was told that this is the right place to find the Vorbis developers, so I'm posting my bug report here:

- Ripped an original album using EAC V0.95 prebeta 5 as multiple WAV files
- Downloaded "oggdropXPd V.1.7.11 using GT3b2 2004-03-05" from rarewares.org
- Started encoding

=> OggdropXPd crashes at random track numbers, at the point where it's finished with encoding one file and about to switch to the next one. It always crashes, I am unable to encode the complete 25 tracks of the album.

Then I tried the several encoders to find out which ones are affected:

- OggdropXPd with GT3b2 crashes, both the standard and the P4 compile.
- OggdropXPd with LibVorbis POST v1.0.1 CVS (P4 compile) does NOT crash.
- Oggenc2.3 using GT3b2 (P4 compile) does NOT crash.

So it looks like this issue only appears when I use OggdropXPd in combination with GT3b2, and it's compiler independent there. *edit* I thought about this again. Compiler "independent" might be wrong, I mean all of them are ICL. But at least it's not the P4 compile.

Short summary of my system specs:
- OS: WinXP Professional SP1
- CPU: Intel P4 2.8C HT 800MHz FSB
- MB: Intel D875PBZLK
- RAM: 2x Corsair CL2 256MB in dual-channel mode

I doubt very much that the reason for this might be instability of my system. It has been perfectly stable since I built it three months ago, I never experienced any random crashes (and there is some heavy usage on it). Furthermore, I am as arrogant to blame that I do not run any weird software in background (not even a background virus scanner etc.).

Alright, please feel free to request any information and further testing you need for tracking this problem. If you want me to run a debug build, just tell me where I can get it. I also have a debugger installed (VS.NET) which I could use myself in the case that this is only reproduceable on my system - just didn't bother to dig into the Vorbis code yet.

*edit* Oups, I forgot that I wanted to look at the offset where it crashes. I will post it soon...
Alright:

Crash offset for "oggdropXPd V.1.7.11 using GT3b2 2004-03-05" (P4 compile): 0x77F58E22

I have tried (only?) twice and both times it was at this offset.


...Thanks for your help!

This post has been edited by p0s: May 9 2004, 12:52

[phoolgobi]

i can confirm this. oggdropXPd versions >= 1.7.6 have had this problem.

just when oggdropXPd is about to finish encoding one file and is about to start with the next one it crashes. the fish keeps spinning and a msgbox pops up saying that oggdropXPd has crashed and will be terminated. if i tried again, it would encode the file successfully.

as a work around, i switched to using oggifier + oggenc and tOGGer

This post has been edited by phoolgobi: May 8 2004, 20:08

[QuantumKnot]

John33 would be the best person to answer this.

[john33]

So, only the GT3b2 build crashes, and not the Post 1.0.1 CVS? If this is the case, there should be some fundamental difference in the build parameters of the libs. I'll check it out and see what I can find.

Thanks for the notification. I'll get back on this.

[p0s]

So, only the GT3b2 build crashes, and not the Post 1.0.1 CVS?

Yes.

[john33]

I've uploaded a compile, the P4 version, for testing. This is identical, except for the P4 optimisations, to the standard version that runs perfectly on my AthlonXP system. Can you give this a 'quick whirl' and tell me if you still have the problem?

http://homepage.ntlworld.com/jfe1205/testXPd.zip

[p0s]

I've uploaded a compile, the P4 version, for testing. This is identical, except for the P4 optimisations, to the standard version that runs perfectly on my AthlonXP system. Can you give this a 'quick whirl' and tell me if you still have the problem?

http://homepage.ntlworld.com/jfe1205/testXPd.zip

Tried it, crashes often at 0x77F82E87.
And once it was at 0x00402015.
So the offsets seem to vary depending on the input :\

This post has been edited by p0s: May 10 2004, 14:58

[john33]

Tried it, crashes often at 0x77F82E87.
And once it was at 0x00402015.
So the offsets seem to vary depending on the input  :\

Can you tell me what encoder settings you're using and, in particular, if you're applying tagging and what the settings are for that? TIA.

[p0s]

Tried it, crashes often at 0x77F82E87.
And once it was at 0x00402015.
So the offsets seem to vary depending on the input  :\

Can you tell me what encoder settings you're using and, in particular, if you're applying tagging and what the settings are for that? TIA.

Here's my INI. And btw, the filenames are like that: "04 - The Coasters - Yakety Yak.wav"

[OggDropXPd]
Window_X=499
Window_Y=657
Always_on_top=8
Logerr=8
Shownbr=8
Mode=1013
Tagging=8
Quality=600
MinBitrate=0
NomBitrate=128
MaxBitrate=0
CBRBitrate=0
DelInputWav=0
DelInputOgg=0
Title=%3
Artist=%2
Album=the Sixties - 25 Classic Hits Volume 1
EncodedBy=
Separator= -
TrackNo=%1
Year=
Genre=Oldies
Volume=0
Limiter=0
Dither=0
DecTagMode=0
DecodeMode=1
WaveFormat=1
ReSampleRate=44100
ReSampleMode=0
ScaleFactor=100,000000
LowPassMode=0
LowPass=0
RenameOnTag=0
RenameFormat=
OutputDirMode=0
OutputDir=
TempDir=
PriorityClass=3
PlaylistOnTag=0
PlaylistFormat=
CopyFLACTag=0
UseFLACTag=0
FLACScaleMode=0
RenameFormatFLAC=
DelReplayGainTags=0

This post has been edited by p0s: May 10 2004, 15:20

[john33]

Thanks.

[john33]

New P4 compile on the same link (http://homepage.ntlworld.com/jfe1205/testXPd.zip). Hopefully some improvement(??).

[p0s]

New P4 compile on the same link (http://homepage.ntlworld.com/jfe1205/testXPd.zip). Hopefully some improvement(??). :unsure:

Negative, sorry :|
Still crashes at the same point. (I figured out that the track numbers where it crashes are not really 100% random. Sometimes the same. They depend somehow on the order of the input or so.)

[john33]

OK, thanks - back to the drawing board!!

I'll try to get some testing time on a friend's system (P4-3.2EE! ).

[john33]

OK, new compile (P4 only) available on same link: http://homepage.ntlworld.com/jfe1205/testXPd.zip

This version has been tested on my own system (AthlonXP 2700+, 1GB RAM, WinXPPro SP1 + all critical updates) using the standard compile and also on a P4 system (P4 3.2 Extreme Edition, 1GB RAM, same OS) using the P4 compile.

I used a similar naming scheme to the one quoted above and 48 files were encoded/tagged without any problem. Further, both sets of encoded files were bit identical.

Hopefully this is now OK?

[p0s]

OK, new compile (P4 only) available on same link: http://homepage.ntlworld.com/jfe1205/testXPd.zip

This version has been tested on my own system (AthlonXP 2700+, 1GB RAM, WinXPPro SP1 + all critical updates) using the standard compile and also on a P4 system (P4 3.2 Extreme Edition, 1GB RAM, same OS) using the P4 compile.

I used a similar naming scheme to the one quoted above and 48 files were encoded/tagged without any problem. Further, both sets of encoded files were bit identical.

Hopefully this is now OK?

I'm sorry but I have to tell you that it still crashes :((
(0x77F58E22,0x77F82E87)
And it seems like the tracknumbers at which it crashes have become "more random" (just an impression).

If you feel like doing faster debugging, you can add me in ICQ. I've sent you a PM with my UIN.

This post has been edited by p0s: May 12 2004, 15:48

[p0s]

Any news????

This post has been edited by p0s: Jun 13 2004, 15:59