[v2.0.6][Windows/RTX 3060] Repeated pipeline_cache_save crash shortly after loading Hyrule Castle #1022

Closed

opened 2026-11-09 06:24:53 +00:00 by Yannise-A · 2 comments

Yannise-A commented

2026-11-09 06:24:53 +00:00

(Migrated from github.com)

Hookshot version

v2.0.6 — log: hookshot v2.0.6 (source 9b541c9c), official Windows launcher update

What kind of problem

Crash, freeze or black screen

What happened

After updating to v2.0.6, the game starts and loads my existing save in the Hyrule Castle area, but then closes unexpectedly while I walk around. On the first attempt it crashed after roughly 10 seconds of gameplay; on the next attempt I could stay a little longer before it crashed again. There is heavy initial stuttering while I explore. Expected: remain playable while new shaders/pipelines are being prepared.

Two v2.0.6 sessions with the same crash stack are attached. Similar crashes also occurred on v2.0.5, so I am not claiming this began in v2.0.6. The earlier v2.0.4 test did not show these crashes.

Steps observed

Update with the official Windows launcher to v2.0.6.
Start the game and allow the initial shader preparation to finish.
Load the existing save in the Hyrule Castle area.
Walk around the area during the initial stutters; the game closes. Restarting and repeating produced another crash.

Evidence / tentative lead

Both logs contain:

CRASH: exception 0x20474343 at KERNELBASE.dll+0xc408a
... hookshot_vk.dll+0x2a810d hookshot_vk.dll+0x2a65ea
    hookshot_vk.dll+0x2a60fe hookshot_vk.dll+0x14205b
    hookshot_vk.dll+0x14e2e9 ...
CRASH: exception 0xc00000ff at ntdll.dll+0x1d3df

Symbolizing the installed DLL with the bundled llvm-symbolizer resolves +0x14205b to mkwvk::pipeline_cache_save(bool)::$_0::operator()() const and +0x14e2e9 to its std::thread proxy.

The first log also reports mesh program f93c1d64658c850b not translated: std::bad_alloc. Its preparation completed with 38,809 programs (97 failed). The second log loads a pipeline cache of 1,372,241 KB and a program cache of 6,693,541 KB. Its last memory sample is working set 4,169 MB, private 26,427 MB; VRAM is 8,726 MB used of 11,167 MB allowed. These are observations, not proof of the underlying cause or of VRAM exhaustion.

I searched for the version, allocation errors, exception codes and pipeline_cache_save and did not find an existing report matching this Windows crash. This is separate from my fluorescent-flower evidence in #866.

Report prepared with AI assistance from my gameplay observations and locally preserved logs.

Where in the game

Hyrule Castle area: loading my existing save, then walking around during the first seconds of gameplay.

How often

Every time

Graphics card and driver version

NVIDIA GeForce RTX 3060 12 GB; Vulkan 1.4; driver reported in both logs: 0x9b404000

CPU, RAM and operating system

AMD Ryzen 7 5800X3D, 16 GB RAM, Windows 11 Pro (10.0.26300)

Changes from the defaults

Recorded settings: fullscreen=1, frame_generation=0, frame_pacing=0, resolution=0 (Auto), upscaling=0, sharpness=0, game_speed=1, handheld_profile=0, ultrawide=0, fps_counter=1. No ReShade, MSI Afterburner/RTSS, OptiScaler or similar tools were used during these crashes. Existing caches were kept through the update; I have not deleted them or tested a fresh cache.

Other software hooked into the game

No, nothing hooked in

hookshot_log.txt

Two complete v2.0.6 crash logs from consecutive attempts. Only local installation/user paths have been redacted.

hookshot-v2.0.6-crash-1-redacted.txt
hookshot-v2.0.6-crash-2-redacted.txt

Screenshots or video

No response

Before you send it

I'm on the newest release, or the problem is still there on it.
I searched the open issues for this problem.
I'm not attaching game files, keys, hookshot.exe, the .bin caches or saves.

### Hookshot version v2.0.6 — log: hookshot v2.0.6 (source 9b541c9c), official Windows launcher update ### What kind of problem Crash, freeze or black screen ### What happened After updating to v2.0.6, the game starts and loads my existing save in the Hyrule Castle area, but then closes unexpectedly while I walk around. On the first attempt it crashed after roughly 10 seconds of gameplay; on the next attempt I could stay a little longer before it crashed again. There is heavy initial stuttering while I explore. Expected: remain playable while new shaders/pipelines are being prepared. Two v2.0.6 sessions with the same crash stack are attached. Similar crashes also occurred on v2.0.5, so I am not claiming this began in v2.0.6. The earlier v2.0.4 test did not show these crashes. ### Steps observed 1. Update with the official Windows launcher to v2.0.6. 2. Start the game and allow the initial shader preparation to finish. 3. Load the existing save in the Hyrule Castle area. 4. Walk around the area during the initial stutters; the game closes. Restarting and repeating produced another crash. ### Evidence / tentative lead Both logs contain: ```text CRASH: exception 0x20474343 at KERNELBASE.dll+0xc408a ... hookshot_vk.dll+0x2a810d hookshot_vk.dll+0x2a65ea hookshot_vk.dll+0x2a60fe hookshot_vk.dll+0x14205b hookshot_vk.dll+0x14e2e9 ... CRASH: exception 0xc00000ff at ntdll.dll+0x1d3df ``` Symbolizing the installed DLL with the bundled llvm-symbolizer resolves +0x14205b to `mkwvk::pipeline_cache_save(bool)::$_0::operator()() const` and +0x14e2e9 to its std::thread proxy. The first log also reports `mesh program f93c1d64658c850b not translated: std::bad_alloc`. Its preparation completed with 38,809 programs (97 failed). The second log loads a pipeline cache of 1,372,241 KB and a program cache of 6,693,541 KB. Its last memory sample is working set 4,169 MB, private 26,427 MB; VRAM is 8,726 MB used of 11,167 MB allowed. These are observations, not proof of the underlying cause or of VRAM exhaustion. I searched for the version, allocation errors, exception codes and pipeline_cache_save and did not find an existing report matching this Windows crash. This is separate from my fluorescent-flower evidence in #866. Report prepared with AI assistance from my gameplay observations and locally preserved logs. ### Where in the game Hyrule Castle area: loading my existing save, then walking around during the first seconds of gameplay. ### How often Every time ### Graphics card and driver version NVIDIA GeForce RTX 3060 12 GB; Vulkan 1.4; driver reported in both logs: 0x9b404000 ### CPU, RAM and operating system AMD Ryzen 7 5800X3D, 16 GB RAM, Windows 11 Pro (10.0.26300) ### Changes from the defaults Recorded settings: fullscreen=1, frame_generation=0, frame_pacing=0, resolution=0 (Auto), upscaling=0, sharpness=0, game_speed=1, handheld_profile=0, ultrawide=0, fps_counter=1. No ReShade, MSI Afterburner/RTSS, OptiScaler or similar tools were used during these crashes. Existing caches were kept through the update; I have not deleted them or tested a fresh cache. ### Other software hooked into the game No, nothing hooked in ### hookshot_log.txt Two complete v2.0.6 crash logs from consecutive attempts. Only local installation/user paths have been redacted. [hookshot-v2.0.6-crash-1-redacted.txt](https://github.com/user-attachments/files/34566265/hookshot-v2.0.6-crash-1-redacted.txt) [hookshot-v2.0.6-crash-2-redacted.txt](https://github.com/user-attachments/files/34566266/hookshot-v2.0.6-crash-2-redacted.txt) ### Screenshots or video _No response_ ### Before you send it - [x] I'm on the newest release, or the problem is still there on it. - [x] I searched the open issues for this problem. - [x] I'm not attaching game files, keys, hookshot.exe, the .bin caches or saves.

NotRexed commented

2026-11-10 01:59:46 +00:00

(Migrated from github.com)

Fixed in the v2.0.12 branch, closing when it ships. Allocation failures in the pipeline cache save are now contained instead of ending the game.

NotRexed commented

2026-11-10 10:07:43 +00:00

(Migrated from github.com)

Fixed in v2.0.12 (https://github.com/NotRexed/Hookshot/releases/tag/v2.0.12): pipeline cache save allocation failures are contained (#1046) and a cache over 2 GB is read on Windows (#1226). Update with the Hookshot Builder (or the game's update offer) and rebuild. If it still happens on v2.0.12, reopen this with your hookshot_log.txt.

Fixed in **v2.0.12** (https://github.com/NotRexed/Hookshot/releases/tag/v2.0.12): pipeline cache save allocation failures are contained (#1046) and a cache over 2 GB is read on Windows (#1226). Update with the Hookshot Builder (or the game's update offer) and rebuild. If it still happens on v2.0.12, reopen this with your `hookshot_log.txt`.

Commenting is not possible because the repository is archived.