Repository navigation
Closed
Description
What were you doing when it happened?
Diagnostics: please drag crash-2026-11-02_14-58-00.zip (the launcher showed it in Finder) onto this issue: it has the full session log and macOS's crash reports.
Bloodborne for Mac 0.5.12: crash report
- Mac: Mac17,7 (Apple M5 Max), 48 GB, macOS 27.0.1 (Build 26A434)
- Settings: output 1920x1080, upscaler off, cap 60 FPS, fixed timing, full screen
- Session: started 2026-11-02 14:58:00 05:30, ran 121 min 20 s
- Exit status: 76
- Cause: the game stopped responding (no new frame for 30 seconds, 429814 frames before; BB_HANG_SECONDS) (at 16:59:19.922)
- Log: game-2026-11-02_14-58-00.log, 2395 lines: 1 critical, 189 errors, 38 warnings
Warnings and errors before the cause (5 most recent)
16:55:30.198 ERROR GPU [Render.Vulkan] <Error> vk_pipeline_common.cpp:58 Fail: Graphics pipeline fs_0x000000001c3953c6,vs_0x000000009da6aacb not built (the Metal pipeline state failed): its draws are skipped (×2)
16:58:49.640 ERROR GPU [Render] <Error> mtl_pso_cache.mm:699 GetRenderPipeline: Metal render pipeline (vs 0x328b17b2101 fs 0x10990d20b69): vertex attribute 0 (location 0) format Undefined has no Metal vertex format
16:58:49.640 ERROR GPU [Render.Vulkan] <Error> vk_pipeline_common.cpp:58 Fail: Graphics pipeline fs_0x00000000108e11bc,vs_0x00000000327e1a5e not built (the Metal pipeline state failed): its draws are skipped (×2)
16:58:49.649 ERROR GPU [Render] <Error> mtl_pso_cache.mm:699 GetRenderPipeline: Metal render pipeline (vs 0xce85eda9768 fs 0x75edbeda9): vertex attribute 0 (location 0) format Undefined has no Metal vertex format
16:58:49.649 ERROR GPU [Render.Vulkan] <Error> vk_pipeline_common.cpp:58 Fail: Graphics pipeline fs_0x00000000006c079b,vs_0x00000000ce7e5857 not built (the Metal pipeline state failed): its draws are skipped
game-2026-11-02_14-58-00.log (12 lines around the cause)
16:59:19.921 INFO EzWorkPool_TypeB1 blocked in condvar (0x71091009b0) for 30278 ms; __psynch_cvwait 0xa<cond_wait 0xcc<2083ecd<207e9e8<208bfe5<20340ab<263b246<host_start 0xa9<_pthread_start 0x73<thread_start 0xf
16:59:19.921 INFO EzWorkPool_TypeA3 blocked in condvar (0x71091007b0) for 30957 ms; __psynch_cvwait 0xa<cond_wait 0xcc<2083ecd<207e9e8<208bfe5<20340ab<263b246<host_start 0xa9<_pthread_start 0x73<thread_start 0xf
16:59:19.922 INFO EzWorkPool_TypeA2 blocked in condvar (0x7109100610) for 30957 ms; __psynch_cvwait 0xa<cond_wait 0xcc<2083ecd<207e9e8<208bfe5<20340ab<263b246<host_start 0xa9<_pthread_start 0x73<thread_start 0xf
16:59:19.922 INFO EzWorkPool_TypeA1 blocked in condvar (0x7109100470) for 30957 ms; __psynch_cvwait 0xa<cond_wait 0xcc<2083ecd<207e9e8<208bfe5<20340ab<263b246<host_start 0xa9<_pthread_start 0x73<thread_start 0xf
16:59:19.922 INFO EzWorkPool_TypeA0 blocked in condvar (0x71091002d0) for 30957 ms; __psynch_cvwait 0xa<cond_wait 0xcc<2083ecd<207e9e8<208bfe5<20340ab<263b246<host_start 0xa9<_pthread_start 0x73<thread_start 0xf
16:59:19.922 INFO FD4JobWorker:0 blocked in semaphore (0x13) for 7275648 ms; __psynch_cvwait 0xa<sem_wait 0x247<263ae6a<106747f<263b246<host_start 0xa9<_pthread_start 0x73<thread_start 0xf
16:59:19.922 INFO Core.Logging.BackgroundStrategy blocked in semaphore (0x12) for 7275660 ms; __psynch_cvwait 0xa<sem_wait 0x247<263ae6a<26471c5<263b246<host_start 0xa9<_pthread_start 0x73<thread_start 0xf
16:59:19.922 INFO main running; swtch_pri 0xa<thread_yield 0x9<1067fc5<75b25e<76a8e5<420965<fd55d7<fd35a1<17f9e98<103d438
16:59:19.922 CRIT Crash: the game stopped responding (no new frame for 30 seconds, 429814 frames before; BB_HANG_SECONDS)
16:59:19.922 INFO bbgpu-host: exiting: SIGTERM
16:59:19.922 INFO Exit: pipeline archive flushed in 0 ms
16:59:20.394 ERROR launcher: the game stopped with status 76 after 121 min 20 s
Metadata
Metadata
Assignees
Labels
No labels
Activity
Findings: 2026-11-02 triage (#128 reproduces, #117 cost, #109, #125-#127
Wooloomooloo2 commentedon Nov 2, 2026
Thanks @globbo-glitch. Please drag the
crash-2026-11-02_14-58-00.zipthe launcher showed onto this issue. The report needs the full thread list, because the lines above only show the workers that were waiting, not the one that was stuck. Also: what were you doing when it froze (an area, a loading screen, online play)?globbo-glitch commentedon Nov 3, 2026
No idea where that zip would be located. I was on a loading screen.
khhan0703 commentedon Nov 4, 2026
Hi, I reported a crash just now and found the crash report in the following folder.
~/Library/Logs/Bloodborne/Reports
globbo-glitch commentedon Nov 5, 2026
Thank you so much
Here is the crash report. It was loading into the loading screen when it crashed.
crash-2026-11-02_14-58-00.zip
Youne97 commentedon Nov 5, 2026
Hello! Same problem here on M4 / 16 GB at 40 FPS (twice, at 102 and 128 min). At 30 FPS cap(fixed steps) I did not experience this problem.
crash-2026-11-04_17-23-41.zip
crash-2026-11-05_11-33-44.zip
Wooloomooloo2 commentedon Nov 5, 2026
Thanks @globbo-glitch and @Youne97, the zips were exactly what we needed. All three show the same sequence, and so does #139.
During play, the game keeps a list of occlusion queries (GPU checks of whether something is visible). On the port almost none of them ever report a result, so the list keeps growing. The game copies it into a small 256 KiB per-frame memory area every frame, and after about 1.5 to 2 hours of play that area is full. Shortly after, its draw records start going missing and the game stops. Standing still doesn't grow the list; playing does. That's also why 30 FPS seemed safe, @Youne97: fewer frames per hour, so it fills more slowly. It would still fill eventually.
The details are in our notes (FINDINGS 2026-11-05). We've found where the list grows, and the next step is finding why the GPU results for those queries never arrive. Until it's fixed: quit and restart the game about every 90 minutes (rest at a lamp first). That resets the list.
Backlog: long-session hang from occlusion queries that never finish (#…
README: 0.5.13 (#76 Mac Settings, #117, #79, #121); known #127, #129
Findings: query begin/end are the #99 guard sites; COND_EXEC not invo…
GPU: write occlusion query dumps in order with the draws (#127)
README, findings, backlog: 0.5.14 (#127 fixed, #129)
Wooloomooloo2 commentedon Nov 7, 2026
Fixed in 0.5.14. The game's occlusion queries never finished on the port: each result was written and then wiped by the game's own clear of the same memory, which ran later. The list of waiting queries grew every frame until the game's per-frame memory filled, and then it froze. The results now arrive in order. A 2.5-hour test session stayed at 60 FPS with that memory at 3% full. Thanks for the report and the log. If you see a freeze on 0.5.14, please reopen with the new log.
GPU: write occlusion query dumps in order with the draws (Wooloomoolo…
Findings and backlog: Wooloomooloo2#127 occlusion dumps fixed on this…