Bad access segmentation fault in AVX-512 stack as compiled with clang

[JustinTArthur]

There may be unaligned memory involved in the AVX-512 functions. clang tends to not work around this automatically. I reproduced this in both r9 and r10 with either 8-bit and 16-bit video samples. The crash happens within updateFrame_avx512 but the offending variable or register could have been created upstream of that call.

CPU: Intel Xeon W-3235
OS: Apple macOS 13.5.1 (22G90)
C++ Toolchain:

Apple clang version 14.0.3 (clang-1403.0.22.14.1)
Target: x86_64-apple-darwin22.6.0
Thread model: posix

Compilation extras: -march=cascadelake

Crash:

Crashed Thread:        59

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

Termination Reason:    Namespace SIGNAL, Code 11 Segmentation fault: 11
Terminating Process:   exc handler [3851]

VM Region Info: 0 is not in any region.  Bytes before following region: 4397551616
      REGION TYPE                    START - END         [ VSIZE] PRT/MAX SHRMOD  REGION DETAIL
      UNUSED SPACE AT START
--->  
      __TEXT                      1061d5000-1061d9000    [   16K] r-x/r-x SM=COW  .../MacOS/Python

Thread 59 Crashed:
0   libaddgrain.dylib             	       0x111bbc707 void updateFrame_avx512<unsigned char, signed char>(void const*, void*, int, int, long, int, int, AddGrainData const*) + 183
1   libaddgrain.dylib             	       0x111bbbc79 addgrainGetFrame(int, int, void*, void**, VSFrameContext*, VSCore*, VSAPI const*) + 393
2   libvapoursynth.dylib          	       0x1070c27ca VSNode::getFrameInternal(int, int, VSFrameContext*) + 144
3   libvapoursynth.dylib          	       0x1070d40d2 VSThreadPool::runTasks(std::__1::atomic<bool>&) + 870
4   libvapoursynth.dylib          	       0x1070d5ba1 void* std::__1::__thread_proxy<std::__1::tuple<std::__1::unique_ptr<std::__1::__thread_struct, std::__1::default_delete<std::__1::__thread_struct>>, void (*)(VSThreadPool*, std::__1::atomic<bool>&), VSThreadPool*, std::__1::reference_wrapper<std::__1::atomic<bool>>>>(void*) + 44
5   libsystem_pthread.dylib       	    0x7ff8131221d3 _pthread_start + 125
6   libsystem_pthread.dylib       	    0x7ff81311dbd3 thread_start + 15

Thread 59 crashed with X86 Thread State (64-bit):
  rax: 0x00007fc372723448  rbx: 0x0000000000000000  rcx: 0x0000000000000120  rdx: 0x00000000000002d0
  rdi: 0x00007fc372788320  rsi: 0x00007fc3727c0320  rbp: 0x000070000764ec40  rsp: 0x000070000764ec30
   r8: 0x00000000000002e0   r9: 0x0000000000000000  r10: 0x0000600002208080  r11: 0x0000000000000120
  r12: 0x00007fc372788040  r13: 0x0000600002208080  r14: 0x0000000000000000  r15: 0x0000600003e31200
  rip: 0x0000000111bbc707  rfl: 0x0000000000010246  cr2: 0x00000001070aa7af
  
Logical CPU:     12
Error Code:      0x00000000 
Trap Number:     13

Thread 59 instruction stream:
  2e 0f 1f 84 00 00 00 00-00 48 63 db 62 f1 fd 48  .........Hc.b..H
  ef 14 1f 62 f1 6d 48 ec-14 18 62 f1 ed 48 ef d1  ...b.mH...b..H..
  62 f1 7d 48 e7 14 1e 41-03 5a 40 39 d3 7c da 4c  b.}H...A.Z@9.|.L
  01 c7 4c 01 c6 4b 63 5c-8a 34 48 01 d8 31 db 66  ..L..Kc\.4H..1.f
  2e 0f 1f 84 00 00 00 00-00 48 63 db 62 f1 fd 48  .........Hc.b..H
  ef 14 1f 62 f1 6d 48 ec-14 18 62 f1 ed 48 ef d1  ...b.mH...b..H..
 [62]f1 7d 48 e7 14 1e 41-03 5a 40 39 d3 7c da 4c  b.}H...A.Z@9.|.L	<==
  01 c7 4c 01 c6 4b 63 5c-8a 34 48 01 d8 41 83 c6  ..L..Kc\.4H..A..
  02 45 39 de 0f 85 6f ff-ff ff f6 c1 01 74 40 31  .E9...o......t@1
  c9 62 f1 fd 48 6f 05 3e-05 00 00 62 f2 fd 48 59  .b..Ho.>...b..HY
  0d a4 04 00 00 0f 1f 40-00 48 63 c9 62 f1 fd 48  [email protected]
  ef 14 0f 62 f1 6d 48 ec-14 08 62 f1 ed 48 ef d1  ...b.mH...b..H..