panda/tests/development/register_hashmap_spread.py

51 lines
1.3 KiB
Python
Executable File

#!/usr/bin/env python3
import matplotlib.pyplot as plt # pylint: disable=import-error
HASHING_PRIME = 23
REGISTER_MAP_SIZE = 0x3FF
BYTES_PER_REG = 4
# From ST32F413 datasheet
REGISTER_ADDRESS_REGIONS = [
(0x40000000, 0x40007FFF),
(0x40010000, 0x400107FF),
(0x40011000, 0x400123FF),
(0x40012C00, 0x40014BFF),
(0x40015000, 0x400153FF),
(0x40015800, 0x40015BFF),
(0x40016000, 0x400167FF),
(0x40020000, 0x40021FFF),
(0x40023000, 0x400233FF),
(0x40023800, 0x40023FFF),
(0x40026000, 0x400267FF),
(0x50000000, 0x5003FFFF),
(0x50060000, 0x500603FF),
(0x50060800, 0x50060BFF),
(0x50060800, 0x50060BFF),
(0xE0000000, 0xE00FFFFF)
]
def _hash(reg_addr):
return (((reg_addr >> 16) ^ ((((reg_addr + 1) & 0xFFFF) * HASHING_PRIME) & 0xFFFF)) & REGISTER_MAP_SIZE)
# Calculate hash for each address
hashes = []
double_hashes = []
for (start_addr, stop_addr) in REGISTER_ADDRESS_REGIONS:
for addr in range(start_addr, stop_addr + 1, BYTES_PER_REG):
h = _hash(addr)
hashes.append(h)
double_hashes.append(_hash(h))
# Make histograms
plt.subplot(2, 1, 1)
plt.hist(hashes, bins=REGISTER_MAP_SIZE)
plt.title("Number of collisions per _hash")
plt.xlabel("Address")
plt.subplot(2, 1, 2)
plt.hist(double_hashes, bins=REGISTER_MAP_SIZE)
plt.title("Number of collisions per double _hash")
plt.xlabel("Address")
plt.show()