Black Screen While Traveling
available in future release
Salbug
On the current Open Beta cycle, I've been running into the black screen and getting soft locked by it at a much more regular frequency than I've ever experienced prior to this.
I currently have no repro steps for this but wanted to at least get it on the radar again.
Log In
Riria Nee-chan
It seems like this asynchronous deadlock issue is highly sensitive to CPU thread timing. High-end PCs with aggressive optimization might trigger this bug more easily, which explains why I've been experiencing this since the live release, while others are only starting to notice it now in the open beta as the code became more unstable.
(Specs: Ryzen 9 7950X3D / RTX 4090) with launch option --affinity=0000FFFF(for CCD0 lock)
Riria Nee-chan
also, I tried with default loading screen only and classic warp tunnel only and both loading screen cause same issue.
So I guess this issue is not releated with random effect
even with live release build
StormRel
Merged in a post:
[Critical] world loading issue on high-end PC
Riria Nee-chan
- System Specs:
CPU: AMD Ryzen 9 7950X3D
GPU: NVIDIA GeForce RTX 4090
Launch Options: --affinity=0000FFFF (Tested with and without)
- Issue Summary:
After the recent update focused on loading screen, the client consistently blackout(deadlock) trying to start the loading process(not freezing but cant display anything), nothing displayed but still loading at background(checked with log). This issue seems to be a "race condition" or "deadlock" caused by the high-speed parallel processing of assets on high-end hardware.
- Technical Observations (from Output Log):
The deadlock occurs immediately after [Behaviour] Avatar is Ready, Initializing.
Logs show frequent Cysharp.Threading.Tasks.UniTaskExtensions+ToCoroutineEnumerator.MoveNext() exceptions right at the moment of the deadlock.
It appears that the main thread becomes unresponsive while trying to resolve asset logic, while the parallel loading tasks wait indefinitely for a response, leading to a complete system hang.
Paradox: This issue is more prevalent on high-speed hardware where the "density" of concurrent operations is much higher than on lower-spec machines.
- Steps to Reproduce:
Use a high-end system (7950X3D/4090).
Attempt to join a world and theres no loading screen and loading sound, just a blackout
not always happened but frequently
- Suggested Request:
Please investigate if the current parallel loading pipeline can be optimized to handle "logical errors" in assets without locking the main thread. A "serial loading" or "safe mode" launch option to throttle the asset initialization speed would be a great temporary workaround for high-end users.
This post was marked as
available in future release
StormRel
marked this post as
tracked
RuneFox237
Also gonna leave a link to some logs here:
Almost all of these were triggered when hopping from my homeworld to another world. Was only able to get into 2 other instances on the beta, one I made myself, and one I joined immediately after that. I did get a black screen after leaving that world to go to another tho.
RuneFox237
I think I found the issue that causes the Black Screen.
I think it's the random loading screen bug again.
Noticed some errors in the logs where I was getting the black screen that mentioned the random loading screen.
Turning off Random let me load into multiple worlds in a row. As soon as I turned on Random again I got a black screen.
Raingram
I was never soft locked by the black screen before this open beta, and while I was still on it I'd estimate that it happened (and softlocked me) 33% of the time I traveled to worlds. The 3rd time it happened I swapped out of the beta.
I also got a BSoD from something in the kernel attempting to write to read-only memory. I've saved the memory dump if it could be useful, but I won't be posting it publicly.