Closed sapertuz closed 1 year ago
I solved it... the code should call the attribute udmabuf.array
instead.
There were also some code typos that I fixed here:
from udmabuf import Udmabuf
import numpy as np
import time
def test_1(a):
for i in range (0,9):
a[i] *= 0
a[i] += 0x31
if __name__ == '__main__':
udmabuf = Udmabuf('udmabuf0')
test_dtype = np.uint8
test_size = udmabuf.buf_size//(np.dtype(test_dtype).itemsize)
comparison = np.zeros(test_size, dtype=test_dtype)
udmabuf.memmap(dtype=test_dtype, shape=(test_size))
print ("test_size : %d" % test_size)
start = time.time()
test_1(udmabuf.array)
elapsed_time = time.time() - start
print ("udmabuf-0 : elapsed_time: {0}".format(elapsed_time) + " sec" )
start = time.time()
test_1(comparison)
elapsed_time = time.time() - start
print ("comparison : elapsed_time: {0}".format(elapsed_time) + " sec")
if np.array_equal(udmabuf.array, comparison):
print ("udmabuf-0 == comparison : OK")
else:
print ("udmabuf-0 != comparison : NG")
I wonder however if this is correct... in the test_1`` it didn't increment in the array. I added it, however... I am also curious about why the range of
iis only from 0 to 9... should it not be 0 to
test_size-1``` ?
Btw, feel free to use the code to be included in your readme file :)
Thank you for the issue.
Readme.md was corrected.
I wonder however if this is correct... in the
test_1
it didn't increment in the array. I added it, however... I am also curious about why the range of iis only from 0 to 9... should it not be 0totest_size-1
?
The argument a
of test_1
is a numpy array.
Thus, a *= 0
clears the entire contents of the array, and a += 0x31
adds 0x31 to each element of the array.
This is repeated 10 times and the time it takes is measured.
in the python test code:
in
the attribute
udmabuf.mem_map
does not exist and the code doesn't work. am I doing something wrong?