LLVM objmode compilation error: Instruction does not dominate all uses

[Azmisov]

Reproducer:

from numba import float64, njit, objmode
import numpy as np

def cbk(arr):
	print(arr)

@njit
def test(flag):
	arr = np.zeros((10,), float64)
	if flag:
		with objmode():
			print("test")
	with objmode():
		cbk(arr)

test(1)

With output:

RuntimeError: Instruction does not dominate all uses!
  %.69 = bitcast i8* %.68 to { i8*, i8* }*
  %.141 = getelementptr inbounds { i8*, i8* }, { i8*, i8* }* %.69, i32 0, i32 1

[guilhermeleobas]

Thanks for the report, can reproduce this bug on Numba main.