mirror of
https://github.com/hrydgard/ppsspp.git
synced 2026-04-24 21:56:10 +03:00
[GH-ISSUE #2451] Crazy Taxi: Fare Wars #945
Labels
No labels
Atrac3+
Audio
CPU emulation
D3D11
D3D9 (removed)
Depth / Z
Feature Request
Font Atlas
GE emulation
Guardband / Range Culling
HLE/Kernel
I/O
Input/Controller
MP3
Multithreading
Needs hardware testing
Networking/adhoc/infrastructure
No Feedback / Outdated?
OpenGL
PGF / sceFont
PSMF / MPEG
Platform-specific (Android)
Platform-specific (Windows)
Platform-specific (iOS)
PowerVR GPU
SDL2
Saving issue
User Interface
Vulkan
arm64jit
armjit
armv6
x86jit
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference
starred/ppsspp#945
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Originally created by @HelicopterP on GitHub (Jun 24, 2013).
Original GitHub issue: https://github.com/hrydgard/ppsspp/issues/2451
The game freezes at loading after selecting Crazi Taxi 1. The log is really huge but it goes something like this:
a few MB of this lines -
36:53:926 idle0 W[HLE]: HLE\sceKernelThread.cpp:2598 sceKernelSuspendThread(314)
36:53:926 user_main W[HLE]: HLE\sceKernelThread.cpp:1926 800201a4=sceKernelStartThread(thread=314, argSize=16, argPtr=0986be40): thread already running
36:53:926 user_main W[HLE]: HLE\sceKernelThread.cpp:2629 sceKernelResumeThread(314)
36:53:926 idle0 W[HLE]: HLE\sceKernelThread.cpp:2598 sceKernelSuspendThread(314)
36:53:926 user_main W[HLE]: HLE\sceKernelThread.cpp:1926 800201a4=sceKernelStartThread(thread=314, argSize=16, argPtr=0986be40): thread already running
36:53:926 user_main W[HLE]: HLE\sceKernelThread.cpp:2629 sceKernelResumeThread(314)
36:53:926 idle0 W[HLE]: HLE\sceKernelThread.cpp:2598 sceKernelSuspendThread(314)
than this:
36:53:943 user_main I[HLE]: HLE\sceKernel.cpp:461 KO 297: Alarm "[Alarm]": -
36:53:944 user_main I[HLE]: HLE\sceKernel.cpp:461 KO 298: Alarm "[Alarm]": -
36:53:944 user_main I[HLE]: HLE\sceKernel.cpp:461 KO 299: Thread "idle0":
than some stuff mentioning sceMp3 than a lot of Alarms again than crash.
Here is a full log: http://www.sendspace.com/file/fwuiss
@sum2012 commented on GitHub (Aug 3, 2013):
v0.8.1-1067-g0b502f1
Still freezes at loading after selecting Crazi Taxi 1
ppsspp debug log: https://docs.google.com/file/d/0B3OaSdeV0L8kOVcwX3hvZGZwS3M/edit?usp=sharing
jpcsp trace log: https://gist.github.com/sum2012/2a7780a71c641e04185a
@unknownbrackets commented on GitHub (Oct 28, 2013):
Callback changes and timing improvements could've improved this.
Does it still have the same issues?
-[Unknown]
@sum2012 commented on GitHub (Oct 28, 2013):
ppsspp-v0.9.1-2232-gc17dd02-windows-x86 same issues
info log
https://gist.github.com/sum2012/ba187ea3077d7107a902
@unknownbrackets commented on GitHub (Jan 11, 2014):
I see some mpeg stuff in there and a loadexec, both of which I think had fixes in the past 3 months. Any improvement in the latest git build?
-[Unknown]
@HelicopterP commented on GitHub (Jan 11, 2014):
Still hangs or crashes at loading CT1.
CT2 is ingame but slow ( i have an i5 2500k) and crashes after loading a game city.
The log is full of this:
40:31:732 user_main W[KERNEL]: HLE\sceKernelThread.cpp:2176 800201a4=sceKernelStartThread(thread=287, argSize=16, argPtr=0986be40): thread already running
40:31:732 idle0 E[KERNEL]: HLE\sceKernelThread.cpp:2848 sceKernelSuspendThread(287): thread already suspended
40:31:732 user_main W[KERNEL]: HLE\sceKernelThread.cpp:2176 800201a4=sceKernelStartThread(thread=287, argSize=16, argPtr=0986be40): thread already running
In case of CT2 it's
38:06:297 user_main W[KERNEL]: HLE\sceKernelThread.cpp:2176 800201a4=sceKernelStartThread(thread=287, argSize=16, argPtr=09776f30): thread already running
38:06:297 Decompress I[KERNEL]: HLE\sceKernelThread.cpp:2260 __KernelReturnFromThread: 13595
38:06:297 idle0 E[KERNEL]: HLE\sceKernelThread.cpp:2843 sceKernelSuspendThread(287): thread not running
38:06:315 user_main W[UTIL]: HLE\sceUtility.cpp:418 sceUtilityNetconfGetStatus(): wrong dialog type
38:06:315 user_main I[KERNEL]: HLE\sceKernelThread.cpp:2180 sceKernelStartThread(thread=287, argSize=16, argPtr=09776f30)
Full logs
http://www.sendspace.com/filegroup/MQ7fG1g6syu81XZAqLwCnA
@sum2012 commented on GitHub (Jan 12, 2014):
v0.9.6-406-g62eb701
CT1 debug log with stop on error,turn off fast memary
https://drive.google.com/file/d/0B3OaSdeV0L8kY3RyVzVMQU1hR0k/edit?usp=sharing
@dbz400 commented on GitHub (Jan 18, 2014):
CT2 is working but bit graphical issue and will crash after a while .Vertex cache/HW transform ON/OFF is not helping here.
@sum2012 commented on GitHub (Jan 18, 2014):
ppsspp-v0.9.6-453-gda97a67-windows-x86 debug log (CT1)
fast memory off,stop on first error
https://drive.google.com/file/d/0B3OaSdeV0L8kZ1NZZDFKVHF1bU0/edit?usp=sharing
@unknownbrackets commented on GitHub (Jun 28, 2014):
Has either game improved at all? The graphical glitches or the crash?
-[Unknown]
@sum2012 commented on GitHub (Jun 28, 2014):
CT2 crash strace (v0.9.8-1450-g63bcb65 )
@unknownbrackets commented on GitHub (Jun 28, 2014):
Hmm, trying to jump the list to 0x0E003F38. Is this affected at all by PSP-1000/PSP-2000?
-[Unknown]
@sum2012 commented on GitHub (Jun 28, 2014):
Not affected by PSP-1000/PSP-2000
same crash
@unknownbrackets commented on GitHub (Aug 25, 2014):
Was this improved at all by the duplicate stack address check or the duplicate interrupt handler thing?
-[Unknown]
@HelicopterP commented on GitHub (Aug 25, 2014):
Nope, CT2 still crashes, CT1 still hangs.
@sum2012 commented on GitHub (Aug 26, 2014):
CT2 ULUS10273
PPSSPP v0.9.9.1-135-gb4b8174
info log:
https://gist.github.com/sum2012/405bafaddd08cbd1bc8b
debug log:
https://drive.google.com/file/d/0B3OaSdeV0L8kZU1JaWZGVEl5MmM/edit?usp=sharing
@sum2012 commented on GitHub (Aug 26, 2014):
JPCSP emulator has this log "Texture mipmap with invalid dimension at level 1:"
full log: https://gist.github.com/sum2012/10cc917c7c1d26e54b5b
@unknownbrackets commented on GitHub (Sep 29, 2014):
I made a change not long ago to avoid the specific crash in the screenshots above, but it still means there's invalid ge data. Still happening, right?
-[Unknown]
@daniel229 commented on GitHub (Sep 29, 2014):
That crash is like the one happens in Resistance https://github.com/hrydgard/ppsspp/issues/4626
@sum2012 commented on GitHub (Sep 29, 2014):
CT1 (ULUS10273) ppsspp-v0.9.9.1-558-gfeada0e-windows-amd64 log
info log:
EDIT:invalid address log (Do sceuliity debug log):
https://gist.github.com/sum2012/461e7178751fffc10b06
EDIT:non-stop log ,do not get invalid address(Do not do sceuliity debug log):
https://drive.google.com/file/d/0B3OaSdeV0L8kUDVhVHVnREQ0X1k/edit?usp=sharing
debug log: (Just do non-stop log ,do not get invalid address)
https://drive.google.com/file/d/0B3OaSdeV0L8kWlZaYlF2bUtleU0/edit?usp=sharing
EDIT:
invalid addres debug log:
https://drive.google.com/file/d/0B3OaSdeV0L8kOWc5M0JxM2Vybkk/edit?usp=sharing
@sum2012 commented on GitHub (Sep 29, 2014):
@unknownbrackets some finding
PPSSPP do not return sceKernelCancelAlarm error
JPCSPtrace log:
05:24:09.755114 Decompress - -> sceKernelCancelAlarm 0x4394941 = 0x0
05:24:09.755204 Decompress - <- sceKernelCancelAlarm 0x4394941 = 0x8002019F
full JPCSPTrace log:
https://gist.github.com/sum2012/f4853b7851633e8e7ab4
JPCSP emulator also give this error code.
45926 [Decompress] WARN hle.ThreadManForUser - checkAlarmID unknown uid=0x6c
Full log:
https://gist.github.com/sum2012/bdf4ddc48bb3765317e3
@sum2012 commented on GitHub (Sep 29, 2014):
For CT2
Last good
v0.9.8-1397-g412acc7
First bad
v0.9.8-1405-gfea0a26
It mean this pull request
https://github.com/hrydgard/ppsspp/pull/6414
@sum2012 commented on GitHub (Sep 30, 2014):
@unknownbrackets For CT2
git bisect result
This commit
github.com/unknownbrackets/ppsspp@d8cff4509e@unknownbrackets commented on GitHub (Sep 30, 2014):
I guess it's a case of "accidentally working", then. Probably related to audio timing somewhere, though (e.g. sceAudio or sceAtrac or sceSas, all of which I'm pretty sure have fairly incorrect timing.)
-[Unknown]
@sum2012 commented on GitHub (Jan 16, 2015):
v0.9.9.1-1503-gc250074 same problem
ct1 debug log:
https://drive.google.com/file/d/0B3OaSdeV0L8kQWk2SE5zTHAwbDA/view?usp=sharing
ct2 debug log:
https://drive.google.com/file/d/0B3OaSdeV0L8kcXIxU1pzdzBnMEk/view?usp=sharing
@ppmeis commented on GitHub (Jul 25, 2015):
Tested with latest build. It crash in both CT1 and CT2 after a while. Maybe related to #3854?¿
@sum2012 commented on GitHub (Jul 31, 2015):
@ppmeis It doesn't work with interpreter mode,so that it is not related to #3854
@ppmeis commented on GitHub (Dec 6, 2015):
Tested with latest build. Same status.
@sum2012 commented on GitHub (Oct 30, 2016):
Update debug log to v1.3-135-g934aff4
https://drive.google.com/file/d/0B3OaSdeV0L8kV1Bmc21NQ3lGU0E/view?usp=sharing
@ppmeis commented on GitHub (Jan 10, 2017):
Tested with latest build. Instant crash in CT1 character select screen. CT2 goes ingame but also crash:
@Cade-H commented on GitHub (Jul 7, 2018):
v1.6.3-285-g1f60c85cb (default settings)
Crazy Taxi crashes before character select screen.
Crazy Taxi 2 graphics are now fixed, but crashes before you can drive:
@unknownbrackets commented on GitHub (Aug 25, 2018):
Do any of the I/O timing settings change whether this crashes?
-[Unknown]
@HelicopterP commented on GitHub (Aug 25, 2018):
ppsspp.zip
CT1 crashes at character select screen, CT2 crashes after character select screen
@unknownbrackets commented on GitHub (Aug 25, 2018):
Hmm, both logs call
sceUmdDeactivate(2, disc0:), then later enqueues an invalid display list. #11325 may help the crash (with fast memory disabled.)I wonder what
sceUmdDeactivate(2, "disc0:")does, actually...-[Unknown]
@Cade-H commented on GitHub (Aug 26, 2018):
#11325 does seem to help with fast memory disabled. PPSSPP no longer completely crashes and the game only freezes now. CT2 goes a bit further than before as well, but CT1 still crashes before character select.
@Cade-H commented on GitHub (Sep 1, 2018):
v1.6.3-404-ga004196d4 (default settings)
Crazy Taxi 2 is no longer crashing on Small Apple, but there are still some apparent issues. I am not sure which commit changed this unfortunately.
@unknownbrackets commented on GitHub (Sep 3, 2018):
That looks bad - but I can't tell how incorrect it is. From Google, the shadow and top screenshot look like they could be right. Any chance you could try to compare with a screenshot from a PSP (even if you have to drive to the same place or something?)
If so - in addition to the good and bad screenshot, could you try exporting a GE debugger dump on PC?
To do this, open the game and select Debug -> GE debugger..., then when it's displaying the scene, press Record in the top right. After a second or so, it'll finish and save a trace of the drawing activity.
After that, check the memstick/PSP/SYSTEM/DUMP folder and it'll have created a file named something like "ULES12345_0000.ppdmp". You can zip that and then drag and drop it into a reply here.
-[Unknown]
@Cade-H commented on GitHub (Sep 6, 2018):
@unknownbrackets Unfortunately I do not have access to a PSP, but I did as you said in PPSSPP. I attached it to this comment, hope that helps.
I used v1.6.3-426-gf42065ed3 (Default settings)
ULUS10273_0001.zip
@unknownbrackets commented on GitHub (Sep 9, 2018):
This game draws the scene to a 512x480 wide buffer, then downloads that framebuffer to RAM, then stretches that to 480x272.
There are a couple verts outside the guardband, but nothing significant. The texture it's trying to draw there is some sort of granite pattern or something, but it's far away and ends up looking ugly. Not sure.
https://www.neogaf.com/threads/crazy-taxi-fare-wars-looks-awful.145578/
https://i.neoseeker.com/p/Games/PSP/Racing/Car/crazy_taxi_fare_wars_image3.jpg
It's not a beautiful game, but I didn't find any screenshots that look as bad - I'm guessing somehow we have the wrong texture.
-[Unknown]
@unknownbrackets commented on GitHub (Feb 22, 2019):
The latest version of git has some changes that might help debug this better - any chance you could upload a new dump?
It works on more platforms now:
https://github.com/hrydgard/ppsspp/wiki/How-to-create-a-frame-dump
-[Unknown]
@ppmeis commented on GitHub (Feb 24, 2019):
I tested with latest build and CT1 still crashing the same. CT2 crash in both game modes so I can't create a frame dump. Here's the debuglog_
ppsspplog.zip
@ppmeis commented on GitHub (Feb 8, 2020):
Tested in latest Windows Build. Same status (offtopic last commit a year ago wow lol).
@unknownbrackets commented on GitHub (Apr 7, 2020):
An update GE dump would still help.
-[Unknown]
@Panderner commented on GitHub (Apr 7, 2020):
Here's a log that spam this error in crazy taxi 1's title screen:
And this one crashes after crazy taxi 1 character select screen:
And this one after selecting the game:
@Panderner commented on GitHub (Apr 7, 2020):
And this in crazy taxi 2 in-game:
@oskarangure commented on GitHub (Jun 27, 2020):
Here's a GE dump of the american version using v1.9.3:
ULUS10273_0001.zip
Also here's how it appears in ppsspp vs how it should be using an actual psp. The marble-like texture in the middle is similar (maybe the same) as the texture of the surface of the water/ocean in the game.
@sum2012 commented on GitHub (Aug 18, 2020):
CT1:
Remove spam of these Log:
23:873 user_main I[SCEKERNEL]: hle\scekernelthread.cpp:2102 -2147352156=sceKernelStartThread(287, 16, 0986be40)
00:23:873 user_main W[SCEKERNEL]: hle\scekernelthread.cpp:2094 800201a4=sceKernelStartThread(287, 16, 0986be40): thread already running
v1.10.3-449-g8c859ef0f Log:
https://gist.github.com/sum2012/b008d1153068232604115af4d9fea054
debug log:
ppsspplogv1.10.3-449-g8c859ef0f.txt.zip
We may need to do thing in sceKernelExitThread (// TODO: This should trigger ON the thread when it exits.)
and sceKernelTerminateDeleteThread do not same as jpcsp method https://github.com/jpcsp/jpcsp/blob/master/src/jpcsp/HLE/modules/ThreadManForUser.java#L3998
@hrydgard commented on GitHub (Aug 18, 2020):
Does the game work in JPCSP now?
@sum2012 commented on GitHub (Aug 18, 2020):
Yes, the game work at least in 2014 year
Newest log: https://gist.github.com/sum2012/6a6e7c5de2b4c1f7a882e3a3c0db9892
Some interested log:WARN hle.ThreadManForUser - Decompress - checkAlarmID unknown uid=0x73
@sum2012 commented on GitHub (Aug 18, 2020):
unlucky,in jpcsp ignore checkAlarmID check still in-game.
@sum2012 commented on GitHub (Aug 30, 2020):
I make some hack in sceatrac raise unusual error :/
@unknownbrackets commented on GitHub (Aug 30, 2020):
That seems like a broken memcpy?
// TODO: This should trigger ON the thread when it exits.is only for thread event handlers - is this game actually registering any? Most games don't. If it really does use them, it's very likely related - they are not well tested.I've been meaning for a while to try to properly simulate the delayed-delete behavior the actual firmware has. It is possible that could be a factor here too, but hard to say.
A JpcspTrace could verify if "thread already running" is a problem or just something the game normally does.
-[Unknown]
@sum2012 commented on GitHub (Aug 30, 2020):
JPCSPTrace log of @unknownbrackets https://gist.github.com/sum2012/066c74fbefff9a85d507c8e05e9d9c8e
@unknownbrackets why your jpcsptrace don't return
<- _sceKernelExitThread
or <- sceKernelExitThread ?
@sum2012 commented on GitHub (Sep 1, 2020):
Thanks @shenweip
confirm v1.10.3-543-ge8c2fce1b-windows-amd64 fixed CT1 and CT2
Closing
@shenweip commented on GitHub (Sep 1, 2020):
Good news!
@hrydgard commented on GitHub (Sep 1, 2020):
Wow! Fantastic work shenweip and sum2012. That such a little thing fixed it... Really happy to see this closed, it's bothered me for a long time.
@sum2012 commented on GitHub (Sep 1, 2020):
https://github.com/jpcsp/jpcsp/blob/master/src/jpcsp/HLE/modules/ThreadManForUser.java#L488
@shenweip @hrydgard
Just a bit worry that need for more hleReSchedule for thread function.
@unknownbrackets commented on GitHub (Sep 2, 2020):
Very nice. Shows the importance of each function and tests.
It just runs code before and after the function to do that. There's no after sceKernelExitThread(), so the code can't run.
As for other areas - we should be careful about assumptions, but we may definitely need more tests.
-[Unknown]
@SilentMRG commented on GitHub (Jul 5, 2021):
Hi...
Crazy Taxi 2 works perfectly in version 1.11.3, however Crazy Taxi 1 freezes the screen after selecting the game.
I am playing on Windows 7 Ultimate 64bit and with Direct3D 9 API.
Ahh, I'm using the game version: Crazy Taxi - Fare Wars (USA) (v2.01).
I apologize for posting this information in a thread that was closed, but it's just that I don't have any information about the bug, and I thought it would be unnecessary to open a thread to just cite an old bug,
@SilentMRG commented on GitHub (Jul 5, 2021):
I just did a quick test here using the OpenGL API, and the game froze 1 time when selecting Crazy Taxi 1 and froze again when selecting Crazy Taxi 2, that is, 1 freeze for each game when selecting them.
In Direct3D 9, the game froze several times and randomly, when selecting CT 1 and 2.
Unfortunately, my video card doesn't support Direct3D 11, so I can't test this API.
I think this old bug is still present, however, randomly, in the above two APIs. Perhaps, in Direct3D 11, this bug does not appear.
I leave this simple warning to those involved in the project.
@hrydgard commented on GitHub (Jul 6, 2021):
I don't think the hangs are graphics-API-related, though you never know, of course. It's clear that the game isn't completely fixed, so I'll re-open (although the situation is of course a lot better now than when this issue was originally reported).
@Panderner commented on GitHub (Jul 6, 2021):
@SilentMRG what specs for your PC are you using?
@SilentMRG commented on GitHub (Jul 6, 2021):
Laptop:
Windows 7 Ultimate 64 bits
i5 2.67 GHz
4GB DDR3
Intel HD Graphics 1000
Note: Obviously it's a low performance machine, however it manages to run both GOW at 60 fps at default settings. Anyway, I don't believe the bug is hardware related.
@Panderner commented on GitHub (Jul 7, 2021):
Works fine for me:
I tested OpenGL and D3D9 for Both CT games when i swap CT games it didn't freeze and Crazy Taxi 1 works fine in-game.
@Panderner commented on GitHub (Jul 7, 2021):
Here's my Specs:
AMD A8-7680
Nvidia GTX 1050 Ti
8GB RAM DDR3 1333 MHz
Windows 10 Pro
@SilentMRG commented on GitHub (Jul 7, 2021):
CT 1 and 2 working perfectly here too. The only problem is this random bug that freezes the screen when I select one of the games. Today I played CT 1 for 2 hours and didn't notice any bugs. Then I went to play CT 2 and the screen froze when selecting the game; I closed the emulation and went back to emulating CT 2, and it worked for 30 minutes without bugs.
Note: When the screen freezes the fps stays at 0/0 (100%) and sometimes drops to 90% or 80%, while the fps stays at 0/0, sometimes going up to 02/02.
I have 32 games here and this is the only one with a problem.
Because it's a random bug, it might not happen to everyone.
@unknownbrackets commented on GitHub (Jul 11, 2021):
If you can find a way or place to reproduce the issue more frequently, that may help. I don't personally have the game, but bugs that take hours to reproduce are typically very hard bugs to fix even when you do own the game in question.
Sometimes, a thing that takes a long time to reproduce just actually needs 2 uncommon things to both happen in a specific order. That's not always the case, but figuring out those kinds of things can help significantly in making bugs easier to fix.
If you see 0/0, that means the game has gone an extended period of time without attempting to draw to the screen and this most likely means an emulation issue. It can also be caused by cheats, disc issues (please double check your CRC), savedata, or certain settings like the CPU clock speed setting.
Such an issue should definitely not be graphics backend specific (the screen locking up certainly can, be showing 0/0 or 2/2 shouldn't be.)
-[Unknown]
@ghost commented on GitHub (Jul 27, 2021):
@SilentMRG did you also try this game on android and the issue can reproduce?
@SilentMRG commented on GitHub (Jul 27, 2021):
I understand. And I'll try as soon as I have time, because at the moment I have a lot of work.
@SilentMRG commented on GitHub (Jul 27, 2021):
Unfortunately, I don't have a good Android to test, and that makes me stuck on the PC.
@ghost commented on GitHub (Sep 9, 2021):
I cannot reproduce this on my android phone.
@iota97 why calculate crc is gone?
@hrydgard commented on GitHub (Sep 9, 2021):
Huh that is weird! CRC button should be there...
@ghost commented on GitHub (Sep 9, 2021):
Never mind about the calculate crc.
I forget to scroll down ✌️😂
@hrydgard commented on GitHub (Sep 9, 2021):
Ah yeah. we should probably make it more obvious that it's scrollable somehow...
@ghost commented on GitHub (Sep 11, 2021):
This should be added to the future-prio milestone.
@hrydgard commented on GitHub (Sep 11, 2021):
Thanks, missed that.
@ppmeis commented on GitHub (Dec 17, 2022):
Tested in latest windows build. It works just fine.
@sum2012 commented on GitHub (Oct 20, 2023):
@ANR2ME in https://github.com/hrydgard/ppsspp/issues/11062#issuecomment-1769770354 reported that Crazy Taxi Fare Wars can sometimes stuck/freezes after selecting either Crazy Taxi or Crazy Taxi 2 (still showing the game selection screen while it stucks
@ANR2ME Can you try Simulate UMD delays ?
@anr2me commented on GitHub (Oct 20, 2023):
Ah right, i was using the default I/O timing method (Fast) when i was testing the graphical glitches back then, haven't tried using Simulate UMD yet.
Even with Fast it was quite difficult to reproduce it.