Panic when compiling python

[heatd]

Thread 4 (Thread 1.4 (CPU#3 [halted ])):
#0  sched_idle (ptr=0x0 <abi::abi_data>) at kernel/sched/scheduler.cpp:503
#1  0x0000000000000000 in ?? ()

Thread 3 (Thread 1.3 (CPU#2 [running])):
#0  spin_lock_slow_path (lock=0xffff807f15911400, what_to_insert=3) at kernel/spinlock.cpp:47
#1  0xffffffff81100488 in spin_lock_irqsave (lock=<optimized out>) at include/onyx/spinlock.h:67
#2  __sched_find_next (cpu=2) at kernel/sched/scheduler.cpp:224
#3  0xffffffff811009c0 in sched_find_next () at include/onyx/smp.h:85
#4  sched_find_runnable () at kernel/sched/scheduler.cpp:275
#5  sched_schedule (last_stack=last_stack@entry=0xffff807f27ca8d18) at kernel/sched/scheduler.cpp:459
#6  0xffffffff81100a9a in sched_schedule (last_stack=0xffff807f27ca8d18) at kernel/sched/scheduler.cpp:413
#7  0xffffffff81113051 in platform_yield () at arch/x86_64/interrupts.S:538
#8  0xffffffff81100f82 in sched_yield () at kernel/sched/scheduler.cpp:671
#9  0xffffffff810315a5 in futex::futex_queue::wait (this=0xffff807f27ca8eb0, s=0xffffffff811d644c <futex::futex_hashtable_locks+3084>) at kernel/futex.cpp:121
#10 futex::wait (uaddr=<optimized out>, val=<optimized out>, flags=<optimized out>, utimespec=<optimized out>) at kernel/futex.cpp:331
#11 0xffffffff81031fe1 in sys_futex (uaddr=<optimized out>, futex_op=<optimized out>, val=<optimized out>, timeout=<optimized out>, uaddr2=<optimized out>, val3=<optimized out>) at kernel/futex.cpp:522
#12 0xffffffff810446d9 in __sys_futex_thunk (arg0=<optimized out>, arg1=<optimized out>, arg2=<optimized out>, arg3=<optimized out>, arg4=<optimized out>, arg5=<optimized out>, arg6=68) at kernel/syscall_thunk.cpp:547
#13 0xffffffff81118452 in do_syscall64 (frame=0xffff807f27ca8f78) at arch/x86_64/syscall.cpp:44
#14 0xffffffff811120cf in syscall_ENTRY64 () at arch/x86_64/entry.S:130
#15 0x0000000000000033 in abi::abi_data ()
Backtrace stopped: previous frame inner to this frame (corrupt stack?)

Thread 2 (Thread 1.2 (CPU#1 [halted ])):
#0  sched_idle (ptr=0x0 <abi::abi_data>) at kernel/sched/scheduler.cpp:503
#1  0x0000000000000000 in ?? ()

Thread 1 (Thread 1.1 (CPU#0 [halted ])):
#0  halt () at arch/x86_64/debug.cpp:15
#1  0xffffffff81035936 in panic (msg=msg@entry=0xffffffff81161540 "Assertion %s failed in %s:%u, in function %s\n") at kernel/panic.cpp:95
#2  0xffffffff8103596b in __assert_fail (assertion=<optimized out>, file=<optimized out>, line=<optimized out>, function=<optimized out>) at kernel/panic.cpp:141
#3  0xffffffff8103411f in vprintf(const char * __restrict__, typedef __va_list_tag __va_list_tag *) (format=0xffffffff81170224 "panic: %s\n", va=va@entry=0xffff807f27ca2968) at kernel/kernlog.cpp:348
#4  0xffffffff810342cb in vprintk (format=<optimized out>, ap=0xffff807f27ca2968) at kernel/kernlog.cpp:311
#5  printk(const char * __restrict__, ...) (format=<optimized out>) at kernel/kernlog.cpp:319
#6  0xffffffff810358e1 in panic (msg=msg@entry=0xffffffff81161540 "Assertion %s failed in %s:%u, in function %s\n") at kernel/panic.cpp:125
#7  0xffffffff8103596b in __assert_fail (assertion=<optimized out>, file=<optimized out>, line=<optimized out>, function=<optimized out>) at kernel/panic.cpp:141
#8  0xffffffff8110183f in __thread_wake_up (thread=0xffff807f26c21800, cpu=0) at kernel/sched/scheduler.cpp:935
#9  0xffffffff811018bd in thread_wake_up (thread=0xffff807f26c21800) at kernel/sched/scheduler.cpp:987
#10 0xffffffff810418d8 in wait_queue_wake_all (queue=0xffff807f169eded0) at kernel/wait_queue.cpp:82
#11 0xffffffff81031928 in futex::futex_queue::wake (this=<optimized out>) at kernel/futex.cpp:131
#12 futex::wake (uaddr=<optimized out>, flags=<optimized out>, to_wake=1) at kernel/futex.cpp:381
#13 0xffffffff81031fa3 in sys_futex (uaddr=<optimized out>, futex_op=<optimized out>, val=<optimized out>, timeout=<optimized out>, uaddr2=<optimized out>, val3=<optimized out>) at kernel/futex.cpp:524
#14 0xffffffff810446d9 in __sys_futex_thunk (arg0=<optimized out>, arg1=<optimized out>, arg2=<optimized out>, arg3=<optimized out>, arg4=<optimized out>, arg5=<optimized out>, arg6=0) at kernel/syscall_thunk.cpp:547
#15 0xffffffff81118452 in do_syscall64 (frame=0xffff807f27ca2f78) at arch/x86_64/syscall.cpp:44
--Type <RET> for more, q to quit, c to continue without paging--
#16 0xffffffff811120cf in syscall_ENTRY64 () at arch/x86_64/entry.S:130
#17 0x0000000000000033 in abi::abi_data ()
Backtrace stopped: previous frame inner to this frame (corrupt stack?)

[heatd]

I think this is because thread migration is a little fucked when it comes to locking. sched_lock definitely needs a

for (;;)
{
    unsigned int cpu = READ_ONCE(thread->cpu);
    spin_lock(/*cpu->scheduler_lock*/);
    spin_lock(&thread->lock);
    if (thread->cpu == cpu)
       break;
   /* unlock locks, retry */
}

As to making sure a thread doesn't get migrated twice... Dunno. Is it an issue? Probably not? As long as we assume thread->cpu can change at any point in time (aka READ_ONCE)

[heatd]

or maybe ->cpu should get switched out while holding the thread's sched locks (needing a separate sched_unlock function for that)...