python_tutorial_lesson_10_aot_compilation_run incompatible with Numpy 2 #8380
Description
Numpy no longer automatically truncates Python integers to the destination type. So the following code causes an overflow error now:
input = np.empty((640, 480), dtype=np.uint8, order='F')
for y in range(480):
for x in range(640):
input[x, y] = x ^ (y + 1)
We can recover the old behavior by manually masking.
input = np.empty((640, 480), dtype=np.uint8, order='F')
for y in range(480):
for x in range(640):
input[x, y] = (x ^ (y + 1)) & 0xFF