Performance tested on simulator: Before optimization: -25% bandwidth @2k entries, -64% @10k entries hashtable size= 2(1bit): -24% bandwidth @2k entries, -65% @10k entries hashtable size= 4(2bits):-15% bandwidth @2k entries, -51% @10k entries hashtable size= 32(5bits): -3% bandwidth @2k entries, -14% @10k entries hashtable size=256(8bits): -1% bandwidth @2k entries, -3% @10k entries Note: Tested on worst performance, the earliest entry will be the worst. Signed-off-by: Zhe Weng <wengzhe@xiaomi.com> |
||
|---|---|---|
| .. | ||
| ipv4_nat.c | ||
| ipv4_nat_entry.c | ||
| Kconfig | ||
| Make.defs | ||
| nat.h | ||