Short summary of your issue
recurring late-game freeze due to inability to allocate memory
Game Version
1.1.2
What OS are you playing on?
Mac
Do you have mods enabled?
No
Have you tried verifying your game files?
Yes
How much "pain" is this causing you?
7
Please explain the issue you experienced in the most condensed way possible
In the late (-ish) game, it starts hanging (indefinitely until killed), with increasing frequency as the game progresses. I believe the cause is a memory limit, even though there is sufficient memory on my machine (more details below)
Please explain how to reproduce the issue
Play for a while on a late-game save. It appears inconsistently, so no reliable way to reproduce.
Is there anything else you think could help us identify/replicate the issue?
The game freezes indefinitely. During that freeze, one of the threads' stack is in `CPdxNewDeleteAllocator::Allocate`, which calls `operator new`, which fails. Reason unclear but the only one I can imagine is that it runs into some usage limit. The signal is intercepted by `::AbortHandler`, which invokes the CPdxCrashReporter, but when the reporter tries to `CreateDirectories`, it invokes `NPdxStringUtils::Split` which calls `CPdxNewDeleteAllocator::Allocate` again, presumably running into the same issue again. The last frame on the stack is `_os_unfair_lock_recursive_abort`, which sounds like it's waiting for the previous error to be handled (alas, inside the error handler itself). There is no crash log attached because the crash log fails to generate. The save file is a save taken manually immediately after loading the last autosave before a crash (I cannot find the save location of the (non-cloud) autosave itself). The crash happened to me 1 month later, on the 30th of november (in-game time), but crashes are not consistent - reloading the save will allow progress to be made beyond the previous crash date, until it reoccurs. Other crashes have occurred in the middle of the month, so the issue is not connected to the autosave. I have attached the profile in which I found the thread mentioned above too. The trace of that thread starts at line 141.
Starting from the attached save, it happened to me again on april 2nd. Similar stacktrace, but the allocation that fails occurred in `BPdxParticle2Internal::SContext:: PrepareWorldCamerasSetup()` and the second allocation attempt was made in `CPdxCrashReporter::LaunchCrashReporterApp` while forking. The pattern of "failed allocation causes crash, followed by another failed allocation while handling the crash" repeats itself.
I have attached a save game
Yes
Upload Attachment
File(s) attached
recurring late-game freeze due to inability to allocate memory
Game Version
1.1.2
What OS are you playing on?
Mac
Do you have mods enabled?
No
Have you tried verifying your game files?
Yes
How much "pain" is this causing you?
7
Please explain the issue you experienced in the most condensed way possible
In the late (-ish) game, it starts hanging (indefinitely until killed), with increasing frequency as the game progresses. I believe the cause is a memory limit, even though there is sufficient memory on my machine (more details below)
Please explain how to reproduce the issue
Play for a while on a late-game save. It appears inconsistently, so no reliable way to reproduce.
Is there anything else you think could help us identify/replicate the issue?
The game freezes indefinitely. During that freeze, one of the threads' stack is in `CPdxNewDeleteAllocator::Allocate`, which calls `operator new`, which fails. Reason unclear but the only one I can imagine is that it runs into some usage limit. The signal is intercepted by `::AbortHandler`, which invokes the CPdxCrashReporter, but when the reporter tries to `CreateDirectories`, it invokes `NPdxStringUtils::Split` which calls `CPdxNewDeleteAllocator::Allocate` again, presumably running into the same issue again. The last frame on the stack is `_os_unfair_lock_recursive_abort`, which sounds like it's waiting for the previous error to be handled (alas, inside the error handler itself). There is no crash log attached because the crash log fails to generate. The save file is a save taken manually immediately after loading the last autosave before a crash (I cannot find the save location of the (non-cloud) autosave itself). The crash happened to me 1 month later, on the 30th of november (in-game time), but crashes are not consistent - reloading the save will allow progress to be made beyond the previous crash date, until it reoccurs. Other crashes have occurred in the middle of the month, so the issue is not connected to the autosave. I have attached the profile in which I found the thread mentioned above too. The trace of that thread starts at line 141.
Starting from the attached save, it happened to me again on april 2nd. Similar stacktrace, but the allocation that fails occurred in `BPdxParticle2Internal::SContext:: PrepareWorldCamerasSetup()` and the second allocation attempt was made in `CPdxCrashReporter::LaunchCrashReporterApp` while forking. The pattern of "failed allocation causes crash, followed by another failed allocation while handling the crash" repeats itself.
I have attached a save game
Yes
Upload Attachment
File(s) attached
Last edited: