Incorrct matrix multiplication : same value different type get different result

[thematteroftime]

Describe the bug same matrix multiply same value with different type, one is int the other is float, but get different result

To Reproduce

import taichi as ti

arch = ti.cpu
ti.init(arch=arch)
n = 4
K = ti.linalg.SparseMatrixBuilder(n, n, max_num_triplets=100)

@ti.kernel
def fill(A: ti.types.sparse_matrix_builder()):
    for i in range(n):
        A[i, i] += 1

fill(K)
K.print_triplets()
A = K.build()

print("----------------")
print(A * 3)
print("----------------")
print(A * 3.0)

Log/Screenshots

$ python my_sample_code.py
[Taichi] version 1.7.0, llvm 15.0.1, commit 2fd24490, win, python 3.9.18
[Taichi] Starting on arch=x64
n=4, m=4, num_triplets=4 (max=100)
[0, 0] = 1.0
[1, 1] = 1.0
[2, 2] = 1.0
[3, 3] = 1.0
----------------
None
----------------
[3, 0, 0, 0]
[0, 3, 0, 0]
[0, 0, 3, 0]
[0, 0, 0, 3]
...

hope this issue will be fixed well Thank you