-
Notifications
You must be signed in to change notification settings - Fork 17
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
sometimes ir compiler will deadloop (discovered by piglit) #38
Comments
shaders/glsl-fs-convolution-1 also fails in schedule_try_place_node() backtrace:
|
Current compiler is not in a complete state, some instructions are not implemented and control flow is not supported, so I'd expect many tests of piglit should fail. We can get a list of them to implement one by one. |
I will report all fails and crashes in a new issue when I get all (or all except these deadloop ones) tests executed. However, even the compiler is not complete it shouldn't deadloop. |
Yes, dead loop will block the piglit test run which causes much manual work. I see piglit has timeout function, maybe it needs some option to enable. You can add deadloop tests to the fail test list too. |
I was not able to reproduce a deadloop with those specific tests anymore, tried shaders/glsl-fs-convolution-1 and shaders/glsl-deadcode-varying now and they just failed with:
Does any other test still deadloop in the compiler? |
mesa-lima commit: 2a54f24
piglit commit: af1140c9d152e11380519ff79dde2cac110bb327
piglit test: shaders/glsl-deadcode-varying
The test hangs until the process is manually terminated.
stacktrace gained by sending SIGABRT to the shader_runner:
The text was updated successfully, but these errors were encountered: