Crash on exit after every long Sculpt session (Metal backend) #108836

New Issue

antonheg · 2023-06-10T07:18:53+02:00

antonheg commented

2023-06-10 07:18:53 +02:00

System Information
Operating system: macOS Monterey 12.6.2
Graphics card: Intel HD Graphics 530 / Radeon Pro 450

Blender Version
Broken: 4.0.0 Alpha (4.0.0 2023-06-09 and every previous release) and 3.6 Beta (the last few releases)
Worked: 3.5.1

Short description of error
Crashes on exit after every long (~30 minutes) sculpting session with a crash report:

Crashed Thread:        0  Dispatch queue: com.apple.main-thread

Exception Type:        EXC_BAD_ACCESS (SIGSEGV)
Exception Codes:       UNKNOWN_0xD at 0x0000000000000000
Exception Codes:       0x000000000000000d, 0x0000000000000000
Exception Note:        EXC_CORPSE_NOTIFY

Termination Reason:    Namespace SIGNAL, Code 11 Segmentation fault: 11

VM Region Info: 0 is not in any region.  Bytes before following region: 4397645824
      REGION TYPE                    START - END         [ VSIZE] PRT/MAX SHRMOD  REGION DETAIL
      UNUSED SPACE AT START
--->  
      __TEXT                      1061ec000-1124c8000    [194.9M] r-x/r-x SM=COW  ...MacOS/Blender

Thread 0 Crashed::  Dispatch queue: com.apple.main-thread
0   libobjc.A.dylib               	    0x7ff81768000f objc_release + 15
1   Blender                       	       0x10c5eb144 blender::gpu::MTLBufferPool::free() + 116
2   Blender                       	       0x10c5eafff blender::gpu::MTLBufferPool::~MTLBufferPool() + 15
3   Blender                       	       0x10c5defc3 blender::gpu::MTLContext::~MTLContext() + 931
4   Blender                       	       0x10c5dcc8e blender::gpu::MTLContext::~MTLContext() + 14
5   Blender                       	       0x10c438a3f GPU_context_discard + 15
6   Blender                       	       0x106cfdbcd DRW_gpu_context_destroy + 45
7   Blender                       	       0x106b53d80 WM_exit_ex + 944
8   Blender                       	       0x106b539bd WM_exit + 45
9   Blender                       	       0x106b5398b wm_exit_handler(bContext*, wmEvent const*, void*) + 11
10  Blender                       	       0x106b42c2e wm_handlers_do_intern(bContext*, wmWindow*, wmEvent*, ListBase*) + 974
11  Blender                       	       0x106b42220 wm_handlers_do(bContext*, wmEvent*, ListBase*) + 48
12  Blender                       	       0x106b415ba wm_event_do_handlers + 2554
13  Blender                       	       0x106b37420 WM_main + 32
14  Blender                       	       0x106386b67 main + 887
15  dyld                          	       0x120bef52e start + 462

Exact steps for others to reproduce the error

Add UV Sphere
Switch to Sculpt Mode
Remesh at 0.05
Sculpt for ~15 minutes with the Grab and Draw brushes
Remesh at 0.02
Sculpt again for ~15 minutes with the Grab and Draw brushes
Command-Q to Quit Blender
Crashes with the above report (only on long sessions ~30 minutes)

**System Information** Operating system: macOS Monterey 12.6.2 Graphics card: Intel HD Graphics 530 / Radeon Pro 450 **Blender Version** Broken: 4.0.0 Alpha (4.0.0 2023-06-09 and every previous release) and 3.6 Beta (the last few releases) Worked: 3.5.1 **Short description of error** Crashes on exit after every long (~30 minutes) sculpting session with a crash report: ``` Crashed Thread: 0 Dispatch queue: com.apple.main-thread Exception Type: EXC_BAD_ACCESS (SIGSEGV) Exception Codes: UNKNOWN_0xD at 0x0000000000000000 Exception Codes: 0x000000000000000d, 0x0000000000000000 Exception Note: EXC_CORPSE_NOTIFY Termination Reason: Namespace SIGNAL, Code 11 Segmentation fault: 11 VM Region Info: 0 is not in any region. Bytes before following region: 4397645824 REGION TYPE START - END [ VSIZE] PRT/MAX SHRMOD REGION DETAIL UNUSED SPACE AT START ---> __TEXT 1061ec000-1124c8000 [194.9M] r-x/r-x SM=COW ...MacOS/Blender Thread 0 Crashed:: Dispatch queue: com.apple.main-thread 0 libobjc.A.dylib 0x7ff81768000f objc_release + 15 1 Blender 0x10c5eb144 blender::gpu::MTLBufferPool::free() + 116 2 Blender 0x10c5eafff blender::gpu::MTLBufferPool::~MTLBufferPool() + 15 3 Blender 0x10c5defc3 blender::gpu::MTLContext::~MTLContext() + 931 4 Blender 0x10c5dcc8e blender::gpu::MTLContext::~MTLContext() + 14 5 Blender 0x10c438a3f GPU_context_discard + 15 6 Blender 0x106cfdbcd DRW_gpu_context_destroy + 45 7 Blender 0x106b53d80 WM_exit_ex + 944 8 Blender 0x106b539bd WM_exit + 45 9 Blender 0x106b5398b wm_exit_handler(bContext*, wmEvent const*, void*) + 11 10 Blender 0x106b42c2e wm_handlers_do_intern(bContext*, wmWindow*, wmEvent*, ListBase*) + 974 11 Blender 0x106b42220 wm_handlers_do(bContext*, wmEvent*, ListBase*) + 48 12 Blender 0x106b415ba wm_event_do_handlers + 2554 13 Blender 0x106b37420 WM_main + 32 14 Blender 0x106386b67 main + 887 15 dyld 0x120bef52e start + 462 ``` **Exact steps for others to reproduce the error** - Add UV Sphere - Switch to Sculpt Mode - Remesh at 0.05 - Sculpt for ~15 minutes with the Grab and Draw brushes - Remesh at 0.02 - Sculpt again for ~15 minutes with the Grab and Draw brushes - Command-Q to Quit Blender - Crashes with the above report (only on long sessions ~30 minutes)

antonheg added the

labels 2023-06-10 07:18:54 +02:00

Jesse Yurkovich commented

2023-06-10 07:22:24 +02:00

From the stack trace it looks like a duplicate of #108792 so I'll close this one. You can subscribe/watch the other issue for updates.

Jesse Yurkovich closed this issue

2023-06-10 07:22:26 +02:00

Blender Bot added

and removed

labels 2023-06-10 07:22:28 +02:00

Jesse Yurkovich added

and removed

labels 2023-06-10 07:22:35 +02:00

antonheg commented

2023-06-11 19:01:08 +02:00

cc @deadpin

I'm noticing this is not limited to the Sculpt mode. Blender 3.6 and 4.0 are crashing on exit after long work sessions whatever the mode. The crash report is identical.

cc @deadpin I'm noticing this is not limited to the Sculpt mode. Blender 3.6 and 4.0 are crashing on exit after long work sessions whatever the mode. The crash report is identical.

antonheg referenced this issue

2023-06-11 19:15:23 +02:00

Crash on exit with Metal backend #108792

antonheg referenced this issue

2023-06-16 18:55:44 +02:00

Fix #108792: Ensure Metal buffers correctly freed on exit #108940

Sign in to join this conversation.