forum

[resolved] [confirmed] Crash when spectating, playing, or retrying maps.

posted
Total Posts
217
show more
Toshi-Kun
This Nanahira map is as well crashed on me- oh wait, we stopped throwing maps... okay... :?
DarkKanaki
Well, if maps don't matter and it happens to every map, the only time it occurs is whenever it tries to load the song, or reload the song.
Perhaps it's something that happens while trying to load a specific beatmap?
Although, I am no programmer, but maybe it's having trouble loading the map? Perhaps new coding?
However, I really highly doubt it has something to do with programs that are installed onto your computer system and the graphical hardware we use, because it was working fine before this started to happen... Right?
TheVileOne
The crash has to do with sprites that are not being disposed before new sprites are loaded. The player is retrying and for whatever reason the sprites aren't unloading correctly and the game crashes when it tries to create new sprites. I think that's how it works.
ritsu101nightshade
I have a question. does it crash for you all if you play on multiplayer? Because I found beatmaps that usually crash doesn't when I play on multiplayer.
Blit-Z
Ive never tried before
I rarely play multiplayer so im not very sure about it in my case
I might try later tonight
Topic Starter
Niervaco
Okay, because your support, maybe you can provide us with what causes sprites to be disposed and created? that might help.
TheVileOne
I'm not a dev who would know exactly why. I'm going off what Millhiore said and it makes sense. There is a method called Begin starts a play and a method called End which ends a play. End is supposed to remove all the sprites before Begin is called. This leads to one of three possibilities.

1. End is not doing what it is supposed to be doing.

2. End is not being called or is not called quickly enough.

3. Something entirely unrelated to the sprite loading process is triggering this error message by mistake. This is unlikely.
Topic Starter
Niervaco

TheVileOne wrote:

I'm not a dev who would know exactly why. I'm going off what Millhiore said and it makes sense. There is a method called Begin starts a play and a method called End which ends a play. End is supposed to remove all the sprites before Begin is called. This leads to one of three possibilities.

1. End is not doing what it is supposed to be doing.

2. End is not being called or is not called quickly enough.

3. Something entirely unrelated to the sprite loading process is triggering this error message by mistake. This is unlikely.
So basically End is at fault here? If so what would cause End to become buggy? Or more likely, End not being called quickly enough.

I would love to see if I knew what sprite wasn't removed. Is there a way to decipher that?
[ Ayase ]
happening when i play with hr and hd only.
BM: http://osu.ppy.sh/s/28705
Skin: http://osu.ppy.sh/forum/t/159951

error:
System.InvalidOperationException: Begin cannot be called again until End has been successfully called.
at Microsoft.Xna.Framework.Graphics.SpriteBatch.Begin(SpriteBlendMode blendMode, SpriteSortMode sortMode, SaveStateMode stateMode, Matrix transformMatrix)
at Microsoft.Xna.Framework.Graphics.SpriteBatch.Begin(SpriteBlendMode blendMode, SpriteSortMode sortMode, SaveStateMode stateMode)
at #Lm.#sp.#8Fb(Boolean )
at #Lm.#sp.#Ms()
at #Vl.#Yo.#0tb()
at #5i.#4i.#Ms()
at #5i.#El.#Ms()
at #rb.#xb.#Ms()
at #1h.#0h.#Ms()
at #rb.#xb.#7s()
at #rb.#xb.#it(Object , EventArgs )
at #rb.#Bb.#Dt()
at #rb.#Rb.#Sv(Object , EventArgs )
at System.Windows.Forms.Application.ThreadContext.System.Windows.Forms.UnsafeNativeMethods.IMsoComponent.FDoIdle(Int32 grfidlef)
at System.Windows.Forms.Application.ComponentManager.System.Windows.Forms.UnsafeNativeMethods.IMsoComponentManager.FPushMessageLoop(Int32 dwComponentID, Int32 reason, Int32 pvLoopData)
at System.Windows.Forms.Application.ThreadContext.RunMessageLoopInner(Int32 reason, ApplicationContext context)
at System.Windows.Forms.Application.ThreadContext.RunMessageLoop(Int32 reason, ApplicationContext context)
at System.Windows.Forms.Application.Run(Form mainForm)
at #rb.#Rb.#ot()
at #rb.#xb.#ot()
at #1h.#np.#Iwb(String , Boolean )
FireFox2000000
Same issue with me. Running a high-end PC too-
8GB RAM
GTX GeForce 7700 Video Card
i5 processor (4th gen)

This is seriously frustrating and only seems to occur when restarting, not loading up maps itself.
Topic Starter
Niervaco
Just a question, is there a way to view the code of osu! when the game is running? (Like a command line interface)

EDIT: I have encountered this in Taiko just a few minutes ago. So it's still alive for me. I was able to get some audio before a crash. (a drum hit)
Franc[e]sco
Same problem here
details here: p/3027633

Graphics card: ATI HD 5770
Latest non-beta drivers
OS: Windows 8.1 x64

Also, I use raptr to track gameplay time which also provides an in-game overlay.

EDIT: it seems to happen more often on certain maps, mods don't seem to really matter. I'm, getting it very frequently on https://osu.ppy.sh/b/232721 with DT on insane
bawangcan
Same problem. Been happening for a while, but seems random to me.I didn't use any mods, and I never full screen.

Crash during spectating: even after the host had finished a song for a long time, osu crashed when he restarted, and gave me this report.

I only play osu and taiko mode, the random crash happened when I play taiko and I wasn't using any mods at all.

Crashing during spectating became really bad yesterday, osu crashed more than 10 times within 2 hours. I reinstalled osu last night, and it crashed on the 2nd map I played this morning which is in taiko mode.

I know this seems to be random but it doesn't mean the problem cannot be fixed right? Please fix this soon, it's really affecting me enjoying osu.

My pc:
Windows 8.1 64-bit (6.3, Build 9600) (9600.winblue_gdr.131030-1505)
Intel(R) Core(TM) i7-4700MQ CPU @ 2.40GHz, 2401 Mhz, 4 Core(s), 8 Logical Processor(s)
16GB RAM
Intel(R) HD Graphics 4600 1920X1080 (32bit) (60Hz)
NVIDIA GeForce GT 740M

System.InvalidOperationException: Begin cannot be called again until End has been successfully called.
at Microsoft.Xna.Framework.Graphics.SpriteBatch.Begin(SpriteBlendMode blendMode, SpriteSortMode sortMode, SaveStateMode stateMode, Matrix transformMatrix)
at Microsoft.Xna.Framework.Graphics.SpriteBatch.Begin(SpriteBlendMode blendMode, SpriteSortMode sortMode, SaveStateMode stateMode)
at #Lm.#sp.#8Fb(Boolean )
at #Lm.#sp.#Ms()
at #5i.#4i.#Ms()
at #5i.#Gl.#Ms()
at #rb.#xb.#Ms()
at #1h.#0h.#Ms()
at #rb.#xb.#7s()
at #rb.#xb.#it(Object , EventArgs )
at #rb.#Bb.#Dt()
at #rb.#Rb.#Sv(Object , EventArgs )
at System.Windows.Forms.Application.ThreadContext.System.Windows.Forms.UnsafeNativeMethods.IMsoComponent.FDoIdle(Int32 grfidlef)
at System.Windows.Forms.Application.ComponentManager.System.Windows.Forms.UnsafeNativeMethods.IMsoComponentManager.FPushMessageLoop(Int32 dwComponentID, Int32 reason, Int32 pvLoopData)
at System.Windows.Forms.Application.ThreadContext.RunMessageLoopInner(Int32 reason, ApplicationContext context)
at System.Windows.Forms.Application.ThreadContext.RunMessageLoop(Int32 reason, ApplicationContext context)
at System.Windows.Forms.Application.Run(Form mainForm)
at #rb.#Rb.#ot()
at #rb.#xb.#ot()
at #1h.#np.#Iwb(String , Boolean )
Franc[e]sco
Also, the issue does not appear on OpenGL, and this is probabilly because OpenGL's fullscreen really is a fullscreen borderless window.
Topic Starter
Niervaco
I'm thinking that this sprite issue is more than just songs.
t/202896
When MillhioreF mentioned sprite unloading problems. Oh boy. This is the background for the main menu.
Franc[e]sco

Niervaco wrote:

I'm thinking that this sprite issue is more than just songs.
t/202896
When MillhioreF mentioned sprite unloading problems. Oh boy. This is the background for the main menu.
I've had the result panel's sprite become invisible before. Not sure what was going on.
Topic Starter
Niervaco

Franc[e]sco wrote:

Niervaco wrote:

I'm thinking that this sprite issue is more than just songs.
t/202896
When MillhioreF mentioned sprite unloading problems. Oh boy. This is the background for the main menu.
I've had the result panel's sprite become invisible before. Not sure what was going on.
And the results page. Hm. I sould ask if anyone has been having sprite issues with other areas. (Main menu background for one) if so, please post here what area. (Main menu, song selection, ect.) and what sprite. (Background, skin element)

If we can confirm that there are sprite unload problems in other areas as well, then we could confirm that the bug is a broad sprite behavior bug.
Franc[e]sco
Since the crashes seem to be more frequent on specific maps, could it be related to storyboarding?
But yeah, I'm quite sure it originated from the recent alt-tab fix. Anyhow, I'm gonna keep playing on OpenGL until the crashes are fixed.
tsundoll

Franc[e]sco wrote:

Since the crashes seem to be more frequent on specific maps, could it be related to storyboarding?
But yeah, I'm quite sure it originated from the recent alt-tab fix. Anyhow, I'm gonna keep playing on OpenGL until the crashes are fixed.
Considering I play with storyboard off in options I'm not sure how. I wish I could play in OpenGL, it's really laggy on this card for any game using OpenGL.
Topic Starter
Niervaco

tsundoll wrote:

Franc[e]sco wrote:

Since the crashes seem to be more frequent on specific maps, could it be related to storyboarding?
But yeah, I'm quite sure it originated from the recent alt-tab fix. Anyhow, I'm gonna keep playing on OpenGL until the crashes are fixed.
Considering I play with storyboard off in options I'm not sure how. I wish I could play in OpenGL, it's really laggy on this card for any game using OpenGL.
Well that's probably because OpenGL renders differently (Duh, but its a double edge sword, no unlimited) At least we know what is causing the crash.
ThePooN
I have an Asus GTX 660 Ti and an i7-3770 and 8GB RAM DDR3. I have this error randomly too.
Toshi-Kun
Nearly 200 posts. :?
Blit-Z
bump
just keeping this thread up so it doesnt get lost
Copy Cat Master
This happen to me too when I try to play Taiko mode with hard rock mod.

PC:

Window 7
Intel(R) Pentium(R) CPU B940 @ 2.00GHz 2.00 GHz
4.00 GB RAM
64 bit OS

Pop up that appear when I use OSU! updater:



Error:

System.InvalidOperationException: Begin cannot be called again until End has been successfully called.
at Microsoft.Xna.Framework.Graphics.SpriteBatch.Begin(SpriteBlendMode blendMode, SpriteSortMode sortMode, SaveStateMode stateMode, Matrix transformMatrix)
at Microsoft.Xna.Framework.Graphics.SpriteBatch.Begin(SpriteBlendMode blendMode, SpriteSortMode sortMode, SaveStateMode stateMode)
at #Lm.#sp.#8Fb(Boolean )
at #Lm.#sp.#Ms()
at #Vl.#Yo.#0tb()
at #5i.#4i.#Ms()
at #rb.#xb.#Ms()
at #1h.#0h.#Ms()
at #rb.#xb.#7s()
at #rb.#xb.#it(Object , EventArgs )
at #rb.#Bb.#Dt()
at #rb.#Rb.#Sv(Object , EventArgs )
at System.Windows.Forms.Application.ThreadContext.System.Windows.Forms.UnsafeNativeMethods.IMsoComponent.FDoIdle(Int32 grfidlef)
at System.Windows.Forms.Application.ComponentManager.System.Windows.Forms.UnsafeNativeMethods.IMsoComponentManager.FPushMessageLoop(Int32 dwComponentID, Int32 reason, Int32 pvLoopData)
at System.Windows.Forms.Application.ThreadContext.RunMessageLoopInner(Int32 reason, ApplicationContext context)
at System.Windows.Forms.Application.ThreadContext.RunMessageLoop(Int32 reason, ApplicationContext context)
at System.Windows.Forms.Application.Run(Form mainForm)
at #rb.#Rb.#ot()
at #rb.#xb.#ot()
at #1h.#np.#Iwb(String , Boolean )

EDIT: Now I keep getting crash with or without mods on the Taiko mode. Anyway, I keep getting crash from this beatmap > https://osu.ppy.sh/s/161102
DragonSlayer96
After doing some testing, I found that if multiple people are spectating the same person, only one of the people will crash, not everyone. Could it be related more on the user end instead of the map itself not loading properly?
Topic Starter
Niervaco
Its definitely client-side. But imagine that if someone was playing OWC or CWC and then their map crash? That would be pretty bad. There is a decent chance of it happening too.

It doesn't take a retry or anything, but CtB is vulnerable. So I hope it doesn't happen.
TheVileOne
It's about time to confirm this thread.

It might happen in only certain songs, but it doesn't happen consistently on any song. There's no guarantee it will happen under any controlled environment. I can only note that it affects every person differently and usually after several retries. If there are more than one spectator, then an event that crashes me might not crash someone else who is spectating.

I have seen no evidence that mod usage, skin usage, whether or not the user misses or not makes any difference. Gamemode doesn't seem to matter either.

I have checked whether it was related to the userlog and it didn't crash because of the userlog.

I used the default skin, an old cookiezi skin, and Aesthetic 1.2 and can crash in all of them.

Some people have reported that it happens only in DirectX. I have not confirmed this.

All of my testing has been in public builds.
Topic Starter
Niervaco

TheVileOne wrote:

It's about time to confirm this thread.

It might happen in only certain songs, but it doesn't happen consistently on any song. There's no guarantee it will happen under any controlled environment. I can only note that it affects every person differently and usually after several retries. If there are more than one spectator, then an event that crashes me might not crash someone else who is spectating.

I have seen no evidence that mod usage, skin usage, whether or not the user misses or not makes any difference. Gamemode doesn't seem to matter either.

I have checked whether it was related to the userlog and it didn't crash because of the userlog.

I used the default skin, an old cookiezi skin, and Aesthetic 1.2 and can crash in all of them.

Some people have reported that it happens only in DirectX. I have not confirmed this.

All of my testing has been in public builds.
Thanks for the confirm :) I will try and crash on OpenGL for you, to see if it will or not.

EDIT: I cannot get OpenGL to crash. also, with OpenGL starting a song really lags when it loads.

Here that, finally a confirm tag!
Karuta-_old_1
Glad to see that this is confirmed

I keep getting random crashes from retrying map too much
Toshi-Kun
Yay! ><
Omio9999
Alright, reading a decent portion of the thread leads me to suspect a few things:

1) You all are running potentially buggy drivers for Direct3D, and you may all want to update your DirectX (won't walk through it here), as well as reporting this issue to DirectX (Tech crews of osu! may want to do the same, so a bit more advice can be pulled in). I base this on no crashes from OpenGL yet, and from the common link to graphics.
2) There's some stupid fancy setting (probably fps counter or something stupid) that keeps killing osu! somehow for you guys.

DISCLAIMER: I am nowhere close to osu! staff, these are my moderately-biased opinions/thoughts. Take it with a grain of salt, a hint of vinegar, and a splinter of wood. In all honesty, I am stabbing in the dark.
Topic Starter
Niervaco

Omio9999 wrote:

Alright, reading a decent portion of the thread leads me to suspect a few things:

1) You all are running potentially buggy drivers for Direct3D, and you may all want to update your DirectX (won't walk through it here), as well as reporting this issue to DirectX (Tech crews of osu! may want to do the same, so a bit more advice can be pulled in). I base this on no crashes from OpenGL yet, and from the common link to graphics.
2) There's some stupid fancy setting (probably fps counter or something stupid) that keeps killing osu! somehow for you guys.

DISCLAIMER: I am nowhere close to osu! staff, these are my moderately-biased opinions/thoughts. Take it with a grain of salt, a hint of vinegar, and a splinter of wood. In all honesty, I am stabbing in the dark.
We already went through the drivers phase. I am running the most updated drivers I can get for an optiplex 780, and even then, it's a modded driver that uses the same Direct3D as the latter i3 i5 i7 systems. I used to run the un modded driver, same result. I do use the FPS counter, this shouldn't have to do wth anything, since it's a rather small sprite they loads constantly, instead of at the beginning of a song. (Where the crash takes place.)

And with OpenGL, it handles sprites differently than DirectX does, so it's sensible that it wouldn't want to crash. So far for me, it doesn't. Also, other threads state some sprite disappearing (such as the main menu background, even without supporter, AKA defualt background) and song backgrounds while in song selection, and song selection sprites themselves. The hug might be a result of the recent .NET Framework update maybe? (As in from a Microsoft update, not osu!)
tsundoll

Omio9999 wrote:

2) There's some stupid fancy setting (probably fps counter or something stupid) that keeps killing osu! somehow for you guys.
I am running osu as bare as possible. No combo bursts, no video, no story board, no fps counter, etc.
peppy
How much video memory do people have who are experiencing this? Maybe it is getting saturated?
Blit-Z

peppy wrote:

How much video memory do people have who are experiencing this? Maybe it is getting saturated?
I think i have 4gb video memory if im thinking about the right one
Eizan Arizawa

peppy wrote:

How much video memory do people have who are experiencing this? Maybe it is getting saturated?
i have 16gb ram and 2gb video memory
DragonSlayer96

peppy wrote:

How much video memory do people have who are experiencing this? Maybe it is getting saturated?
I also have a 2 Gb video card on mine.
ritsu101nightshade

peppy wrote:

How much video memory do people have who are experiencing this? Maybe it is getting saturated?
I don't know if I'm looking at the right thing right now, but if it is correct I only have 384 mb . . .
tsundoll

peppy wrote:

How much video memory do people have who are experiencing this? Maybe it is getting saturated?
I have 358 MB of Graphics memory. That doesn't explain why some times I can play for hours and sometimes it crashes on the first map I play though. If it always happened after extended play, saturation would make sense.
TheVileOne
512 MB VRAM
show more
Please sign in to reply.

New reply