`node --help' crashes inside qemu-user in 17.2.0

[hexchain]

Nodejs crashes when executing node --help, probably after nodejs/node#40488.

Version:
qemu-user: 6.2.0
node: 17.2.0 (but bisected to that PR)

GDB stack trace doesn't seem very useful with QEMU_GDB:

>>> thr apply all bt

Thread 7 (Thread 1.129 (CPU#6 [running])):
#0  0x00000040066abc06 in __futex_abstimed_wait_common64 () from /mnt/rv64/usr/lib/libpthread.so.0
#1  0x00000040066a9128 in __new_sem_wait_slow64.constprop.0 () from /mnt/rv64/usr/lib/libpthread.so.0
#2  0x00000040040f1f70 in uv_sem_wait () from /mnt/rv64/usr/lib/libuv.so.1
Backtrace stopped: frame did not save the PC

Thread 6 (Thread 1.128 (CPU#5 [running])):
#0  0x00000040066abc06 in __futex_abstimed_wait_common64 () from /mnt/rv64/usr/lib/libpthread.so.0
#1  0x00000040066a7024 in pthread_cond_wait@@GLIBC_2.27 () from /mnt/rv64/usr/lib/libpthread.so.0
#2  0x00000040040f1f38 in uv_cond_wait () from /mnt/rv64/usr/lib/libuv.so.1
Backtrace stopped: frame did not save the PC

Thread 5 (Thread 1.127 (CPU#4 [running])):
#0  0x00000040066abc06 in __futex_abstimed_wait_common64 () from /mnt/rv64/usr/lib/libpthread.so.0
#1  0x00000040066a7024 in pthread_cond_wait@@GLIBC_2.27 () from /mnt/rv64/usr/lib/libpthread.so.0
#2  0x00000040040f1f38 in uv_cond_wait () from /mnt/rv64/usr/lib/libuv.so.1
Backtrace stopped: frame did not save the PC

Thread 4 (Thread 1.126 (CPU#3 [running])):
#0  0x00000040066abc06 in __futex_abstimed_wait_common64 () from /mnt/rv64/usr/lib/libpthread.so.0
#1  0x00000040066a7024 in pthread_cond_wait@@GLIBC_2.27 () from /mnt/rv64/usr/lib/libpthread.so.0
#2  0x00000040040f1f38 in uv_cond_wait () from /mnt/rv64/usr/lib/libuv.so.1
Backtrace stopped: frame did not save the PC

Thread 3 (Thread 1.125 (CPU#2 [running])):
#0  0x00000040066abc06 in __futex_abstimed_wait_common64 () from /mnt/rv64/usr/lib/libpthread.so.0
#1  0x00000040066a7024 in pthread_cond_wait@@GLIBC_2.27 () from /mnt/rv64/usr/lib/libpthread.so.0
#2  0x00000040040f1f38 in uv_cond_wait () from /mnt/rv64/usr/lib/libuv.so.1
Backtrace stopped: frame did not save the PC

Thread 2 (Thread 1.124 (CPU#1 [running])):
#0  0x000000400675b322 in epoll_pwait () from /mnt/rv64/usr/lib/libc.so.6
Backtrace stopped: previous frame identical to this frame (corrupt stack?)

Thread 1 (Thread 1.121 (CPU#0 [running])):
#0  0x00007f8e54046b54 in ?? ()

Instructions around $pc:

>>> x/40i 0x00007f8e54046b00
   0x7f8e54046b00:      bnez    a0,0x7f8e54046c4c
   0x7f8e54046b04:      addi    sp,sp,-56
   0x7f8e54046b08:      ld      t2,-40(s0)
   0x7f8e54046b0c:      ld      a0,-48(s0)
   0x7f8e54046b10:      sub     a6,a0,t2
   0x7f8e54046b14:      ld      a1,-56(s0)
   0x7f8e54046b18:      addi    a0,a6,-1
   0x7f8e54046b1c:      slli    t1,a1,0x0
   0x7f8e54046b20:      sub     a0,a0,t1
   0x7f8e54046b24:      sd      a0,-80(s0)
   0x7f8e54046b28:      lui     a5,0x3fc73
   0x7f8e54046b2c:      addi    a5,a5,-1534
   0x7f8e54046b30:      slli    a5,a5,0xb
   0x7f8e54046b34:      ori     a5,a5,409
   0x7f8e54046b38:      slli    a5,a5,0x6
   0x7f8e54046b3c:      ori     a5,a5,33
   0x7f8e54046b40:      bnez    a1,0x7f8e54046b4c
   0x7f8e54046b44:      li      a7,10
   0x7f8e54046b48:      j       0x7f8e54046b58
   0x7f8e54046b4c:      addi    s3,a6,-1
   0x7f8e54046b50:      add     t1,t2,s3
=> 0x7f8e54046b54:      lbu     a7,0(t1)
   0x7f8e54046b58:      sd      a0,-104(s0)
   0x7f8e54046b5c:      sd      a0,-112(s0)
   0x7f8e54046b60:      sd      a0,-120(s0)
   0x7f8e54046b64:      sd      a0,-128(s0)
   0x7f8e54046b68:      sd      a0,-136(s0)
   0x7f8e54046b6c:      sd      a0,-144(s0)
   0x7f8e54046b70:      lui     t0,0x200
   0x7f8e54046b74:      addi    t0,t0,329
   0x7f8e54046b78:      slli    t0,t0,0xb
   0x7f8e54046b7c:      ori     t0,t0,700
   0x7f8e54046b80:      slli    t0,t0,0x6
   0x7f8e54046b84:      ori     t0,t0,24
   0x7f8e54046b88:      ld      t0,0(t0)
   0x7f8e54046b8c:      j       0x7f8e540466cc
   0x7f8e54046b90:      ld      a1,-48(s0)
   0x7f8e54046b94:      ld      a0,-32(s0)
   0x7f8e54046b98:      ld      a2,-56(s0)
   0x7f8e54046b9c:      sub     a1,t2,a1

[luyahan]

V17.2 node use v8 9.6. This v8version is unstable. Please use V17.1 or wait node checkout v8 to V9.7.

[qjivy]

@hexchain Any update？ Does the problem still happen after the version change?

[hexchain]

It seems that they are still in the progress of updating v8 to 9.7. It is blocked for now due to some problems in v8 itself: nodejs/node#40907

[qjivy]

OK, I just close this issue now.
After the updating to 9.7, if there still crashes on RISCV64 target, pls reopen this issue.
Thx!