Closed AlexTMjugador closed 4 years ago
game can show Replay other players, so i made TmForever.exe /silent /bench=X_BayA1.Replay.gbx
for "less size"
apitrace:https://drive.google.com/open?id=1Rtq8QOEGFc9P1okckZ50_q6eIuBonYkt
d3d9.log:https://drive.google.com/open?id=1sOepD9Wp7XhMBdslfgyOQXzn6DiNhj0M
the game freezes, so i used sigterm via tty2 to close it
err: DxvkSubmissionQueue: Failed to sync fence: VK_ERROR_DEVICE_LOST
err: DxvkDevice: waitForIdle: Operation failed
What exactly are we supposed to see here? The apitrace replays fine on my end.
Also, I guess it's fine to open a new issue for bugs but please don't spam our bug tracker with three issues for one single game.
yes, this is hopeless because I cannot record a complete trace the game just stops drawing new frames at the moment when the race should begin and it does not crash, so have to close it forcibly
similar problem appears in warriors orochi / orochi 2 / samurai warriors 2 if shader quality stay in low game entry correctly but with shader quality in medium or more game freeze
I tried next liquorix version and arch-LTS version of kernel and found (sudo journalctl -p 3 -xb
and dmesg
) that "freeze" actually is nvidia Xid error 8 - "GPU stopped processing"
I suppose this issue need to be reported to nvidia, not here
Can you get me an nvidia-bug-report
?
i made game app to window mode and run nvidia-script after "freeze"- all in one x-session
@Joshua-Ashton nvidia-bug-report.log
The game is using swvp and triggers this validation error on nv:
VUID-VkWriteDescriptorSet-descriptorType-00332(ERROR / SPEC): msgNum: 0 - vkUpdateDescriptorSetWithTemplate() failed write update validation for VkDescriptorSet 0x2dda0000002dda[] with error: Write update to VkDescriptorSet VkDescriptorSet 0x2dda0000002dda[] allocated with VkDescriptorSetLayout VkDescriptorSetLayout 0x2ca90000002ca9[] binding #0 failed with error message: Attempted write update to buffer descriptor failed due to: VkDescriptorBufferInfo range is 164096 which is greater than this device's maxUniformBufferRange (65536). The Vulkan spec states: If descriptorType is VK_DESCRIPTOR_TYPE_UNIFORM_BUFFER or VK_DESCRIPTOR_TYPE_UNIFORM_BUFFER_DYNAMIC, the range member of each element of pBufferInfo, or the effective range if range is VK_WHOLE_SIZE, must be less than or equal to VkPhysicalDeviceLimits::maxUniformBufferRange (https://www.khronos.org/registry/vulkan/specs/1.1-extensions/html/vkspec.html#VUID-VkWriteDescriptorSet-descriptorType-00332)
This issue was originally reported in #1363 by @Iglu47 and me, but as I like to keep things ordered I opened this issue specifically for it, as the original one has 41 comments and it's getting pretty messy.
However, if the shadows are turned down to PC2, @Iglu47 says that the issue is gone:
Software information
TrackMania United Forever, maximum settings.
System information
@Iglu47 system specs:
My system specs:
Apitrace file(s)
Log files