Skip to content

[Regression] MotoGP08 Graphical issue #13273

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
Triticum0 opened this issue Jan 22, 2023 · 7 comments
Closed

[Regression] MotoGP08 Graphical issue #13273

Triticum0 opened this issue Jan 22, 2023 · 7 comments
Assignees
Labels

Comments

@Triticum0
Copy link

Triticum0 commented Jan 22, 2023

Quick summary

The game has graphical issues and crashes past a certain build

Details

When testing this game I came across a shader cache issue which causes the game to have graphical issues and crash when decompiling shaders on the GPU.

The build that causes the graphical issue also crashes the game when first going ingame so not sure how I couldn't get an rsx file on the build that caused the issue. I should also mention that the crash also affects the latest build on both Vulkan and OpenGL.

Some build after v0.0.7-8619-7f99de36 do make it in-game without crashing but the graphical issue still persists. Was able to make it in-game twice with the latest build but didn't save the log there were two different graphical issues I got.

The first one was with default settings(This graphical issue does not appear on the original build with the regression)

screenshot-2023_01_22_00_51_37

The second one was with RDB+WDB+RCB+WCB which gives a different result(matches the effect of the build with the regression graphical issue).

2023-01-22.00-43-37.mp4

Crash log:·F 0:01:17.076559 {RSX [0x427b530]} SIG: Thread terminated due to fatal error: Unknown register name: unk
(in file C:\Users\ContainerAdministrator\AppData\Local\Temp\cirrus-ci-build\rpcs3\Emu\RSX\GL\GLCommonDecompiler.cpp:36[:23], in function get_varying_register_location)

Note: version 0.0.6-8608 regressed the game with a different issue causing the game to not display anything and was fixed on the regressed build so had to take a build before that build as a before log.

Build with regression

v0.0.7-8619-7f99de36

Attach two log files

v0.0.7-8619-7f99de36.gz
v0.0.6-8605-9672ad53.gz
v0.0.26-14600-53cc067c.gz
v0.0.14-11605-b96864c7 Ingame Log.gz

Attach capture files for visual issues

https://www.dropbox.com/s/v47o71hk72qi4rg/NPUB90181_20230122004412_capture.rrc.gz?dl=0
https://www.dropbox.com/s/x32bamum96khq3u/NPUB90181_20230122005503_capture.rrc.gz?dl=0

Note this is from build v0.0.14-11605-b96864c7 where I could get in-game
1.=RDB+WDB+RCB+WCB
2.=Default settings
3.=Opengl

System configuration

SYS: RPCS3 v0.0.26-14568-1852b370 Alpha | master
SYS: AMD Ryzen 5 2600 Six-Core Processor | 12 Threads | 15.95 GiB RAM | TSC: 3.400GHz | AVX+ | FMA3
SYS: Operating system: Windows, Major: 10, Minor: 0, Build: 19044, Service Pack: none, Compatibility mode: 0
SYS: Current Time: 2022-10-28T21:53:46
RSX: Found vulkan-compatible GPU: 'NVIDIA GeForce RTX 3060 Ti' running on driver 526.47.0.0
SYS: Firmware version: 4.89

Other details

This is properly GPU specific

@kd-11
Copy link
Contributor

kd-11 commented Jan 22, 2023

Have you tried using atomic FIFO?

@kd-11 kd-11 self-assigned this Jan 22, 2023
@kd-11 kd-11 added the RSX label Jan 22, 2023
@Virtua64
Copy link

Virtua64 commented Jan 22, 2023

This is the demo version. Got it to work with Atomic but it doesn't always work after reboot, even with the same exact settings. Trying to figure out why the inconsistency... video below is with Atomic enabled when it worked.

rpcs3mg8.mp4

@Triticum0
Copy link
Author

I tested it with atomic FIFO and it didn't fix the crashes

@kd-11
Copy link
Contributor

kd-11 commented Jan 26, 2023

Fixing the crash is easy by just ignoring the bad shader. Are the graphical issues still present with atomic FIFO is the question.

@kd-11
Copy link
Contributor

kd-11 commented Feb 6, 2023

The crash should be fixed by #13361

@Triticum0
Copy link
Author

@kd-11 The crashing has been fixed. You need to turn on the write depth buffer to fix the graphical issues. Turning on the Read Depth Buffer causes the blue screen and missing bikes. If you think the issue has been sufficiently fixed you can close the issue.
NPUB90181_screenshot_2023_02_07_20_13_07

@kd-11
Copy link
Contributor

kd-11 commented Feb 7, 2023

Thanks for the report. I'll close this and leave a comment in the buffer options ticket. Relates to #9133

@kd-11 kd-11 closed this as completed Feb 7, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Projects
None yet
Development

No branches or pull requests

3 participants