Applied this, not sure if it's causal.

I get this failure for subthread-exit and bug-10127:

you are registering the same counter address twice: Discarded method code at 0x80432828
you are registering the same counter address twice: Time spent JITting discarded code at 0x80432830
you are registering the same counter address twice: Dynamic code allocs at 0x80484ee8
you are registering the same counter address twice: Dynamic code bytes at 0x80484ef0
you are registering the same counter address twice: Dynamic code frees at 0x80484ef8
you are registering the same counter address twice: Unwind info size at 0x80431728
you are registering the same counter address twice: Calls to trampolines at 0x80431498
you are registering the same counter address twice: JIT trampolines at 0x804314c8
you are registering the same counter address twice: Unbox trampolines at 0x80431494
you are registering the same counter address twice: Static rgctx trampolines at 0x80431490
you are registering the same counter address twice: RGCTX unmanaged lookups at 0x8043149c
you are registering the same counter address twice: RGCTX num lazy fetch trampolines at 0x804314cc
you are registering the same counter address twice: Async JIT info size at 0x80431288
* Assertion: should not be reached at domain.c:474

I also get failures in thread-exit and appdomain-threadpool-unload:

"Finalizer" tid=0x0x200029d7910 this=0x0x20000f74278 , thread handle : 0x80504d40, state : not waiting

@nealef for that assert at domain.c:474, can you get a native stack trace for that?

Seen the exact same thing with @lewurm two weeks ago: #5502 (comment)

@nealef how reproducible is it before and after applying this commit? The domain.c:474 assertion seems highly unlikely to be related to this change.

gdb shows the following thread attempt to Exit:

Thread 2 (Thread 0x3ff8567f910 (LWP 51756)):
#0 0x0000004536f783c6 in sem_wait@@GLIBC_2.2 () from /lib64/libpthread.so.0
#1 0x00000000802e6ca2 in mono_os_sem_wait () at ../../mono/utils/mono-os-semaphore.h:210
#2 mono_os_sem_timedwait () at ../../mono/utils/mono-os-semaphore.h:243
#3 mono_threads_wait_pending_operations () at mono-threads.c:246
#4 0x00000000802e8628 in mono_thread_info_safe_suspend_and_run (id=4396089853760,
interrupt_kernel=, callback=0x8020c064 <async_suspend_critical>,
user_data=0x3ff8567e1f8) at mono-threads.c:1042
#5 0x000000008020f276 in async_suspend_internal (thread=0x3ff831dc130, interrupt=)
at threads.c:4933
#6 0x000000008021092a in mono_thread_suspend_all_other_threads () at threads.c:3467
#7 0x000000008018e1ae in ves_icall_System_Environment_Exit (result=) at icall.c:6736
#8 0x000003ff856ad192 in ?? ()
#9 0x000003ff8b5ecdf4 in ?? ()
#10 0x000003ff8b5ecbba in ?? ()
#11 0x000000008025cb3a in mono_gc_run_finalize (obj=, data=)
at gc.c:312

The thread that appears to be doing the suspend handling shows:

Thread 1 (Thread 0x3ff8b5fc740 (LWP 51743)):
#0 0x0000004536f78762 in sem_post@@GLIBC_2.2 () from /lib64/libpthread.so.0
#1 0x00000000802e76b6 in mono_os_sem_post (info=)
at ../../mono/utils/mono-os-semaphore.h:283
#2 mono_threads_notify_initiator_of_suspend (info=) at mono-threads.c:106
#3 0x00000000802e9eda in suspend_signal_handler (_dummy=, info=,
context=0x3ffe637e160) at mono-threads-posix-signals.c:173
#4
#5 0x0000004536f735b6 in pthread_mutex_unlock () from /lib64/libpthread.so.0
#6 0x000000008028e3fa in mono_os_mutex_unlock () at ../../mono/utils/mono-os-mutex.h:121
#7 mono_coop_mutex_unlock () at ../../mono/utils/mono-coop-mutex.h:70
#8 sgen_gc_unlock () at sgen-gc.c:3703
#9 0x0000000080282f1c in sgen_alloc_obj (vtable=0xb337b3d0, size=24) at sgen-alloc.c:424
#10 0x00000000802675d0 in mono_gc_alloc_obj (vtable=, size=)
at sgen-mono.c:948
#11 0x00000000801eb39c in mono_object_new_alloc_specific_checked (vtable=0xb337b3d0, error=0x3ffe637e6b8)
at object.c:5420
#12 0x00000000801ec8d0 in mono_object_new_specific_checked (vtable=0xb337b3d0, error=0x3ffe637e6b8)
at object.c:5360
#13 0x00000000801ec960 in ves_icall_object_new_specific (vtable=) at object.c:5367
#14 0x000003ff831fe2f0 in ?? ()
#15 0x000003ff831fe62a in ?? ()