diff options
| author | Eric Dumazet <eric.dumazet@gmail.com> | 2011-06-09 02:31:27 -0400 |
|---|---|---|
| committer | David S. Miller <davem@davemloft.net> | 2011-06-09 02:31:27 -0400 |
| commit | 2b77bdde97ae8241dcc23110a4e837acfbc83438 (patch) | |
| tree | a9af81a3ed7b6ee04e7c1b38c3bc77ec295562c8 /include/net/inetpeer.h | |
| parent | 1a7a10325d370e2cbed0c5bb7313904545f6dac8 (diff) | |
inetpeer: lower false sharing effect
Profiles show false sharing in addr_compare() because refcnt/dtime
changes dirty the first inet_peer cache line, where are lying the keys
used at lookup time. If many cpus are calling inet_getpeer() and
inet_putpeer(), or need frag ids, addr_compare() is in 2nd position in
"perf top".
Before patch, my udpflood bench (16 threads) on my 2x4x2 machine :
5784.00 9.7% csum_partial_copy_generic [kernel]
3356.00 5.6% addr_compare [kernel]
2638.00 4.4% fib_table_lookup [kernel]
2625.00 4.4% ip_fragment [kernel]
1934.00 3.2% neigh_lookup [kernel]
1617.00 2.7% udp_sendmsg [kernel]
1608.00 2.7% __ip_route_output_key [kernel]
1480.00 2.5% __ip_append_data [kernel]
1396.00 2.3% kfree [kernel]
1195.00 2.0% kmem_cache_free [kernel]
1157.00 1.9% inet_getpeer [kernel]
1121.00 1.9% neigh_resolve_output [kernel]
1012.00 1.7% dev_queue_xmit [kernel]
# time ./udpflood.sh
real 0m44.511s
user 0m20.020s
sys 11m22.780s
# time ./udpflood.sh
real 0m44.099s
user 0m20.140s
sys 11m15.870s
After patch, no more addr_compare() in profiles :
4171.00 10.7% csum_partial_copy_generic [kernel]
1787.00 4.6% fib_table_lookup [kernel]
1756.00 4.5% ip_fragment [kernel]
1234.00 3.2% udp_sendmsg [kernel]
1191.00 3.0% neigh_lookup [kernel]
1118.00 2.9% __ip_append_data [kernel]
1022.00 2.6% kfree [kernel]
993.00 2.5% __ip_route_output_key [kernel]
841.00 2.2% neigh_resolve_output [kernel]
816.00 2.1% kmem_cache_free [kernel]
658.00 1.7% ia32_sysenter_target [kernel]
632.00 1.6% kmem_cache_alloc_node [kernel]
# time ./udpflood.sh
real 0m41.587s
user 0m19.190s
sys 10m36.370s
# time ./udpflood.sh
real 0m41.486s
user 0m19.290s
sys 10m33.650s
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'include/net/inetpeer.h')
| -rw-r--r-- | include/net/inetpeer.h | 24 |
1 files changed, 13 insertions, 11 deletions
diff --git a/include/net/inetpeer.h b/include/net/inetpeer.h index 1f0966f9176..39d123081e7 100644 --- a/include/net/inetpeer.h +++ b/include/net/inetpeer.h | |||
| @@ -32,12 +32,17 @@ struct inet_peer { | |||
| 32 | struct inet_peer __rcu *avl_left, *avl_right; | 32 | struct inet_peer __rcu *avl_left, *avl_right; |
| 33 | struct inetpeer_addr daddr; | 33 | struct inetpeer_addr daddr; |
| 34 | __u32 avl_height; | 34 | __u32 avl_height; |
| 35 | __u32 dtime; /* the time of last use of not | 35 | |
| 36 | * referenced entries */ | 36 | u32 metrics[RTAX_MAX]; |
| 37 | atomic_t refcnt; | 37 | u32 rate_tokens; /* rate limiting for ICMP */ |
| 38 | unsigned long rate_last; | ||
| 39 | unsigned long pmtu_expires; | ||
| 40 | u32 pmtu_orig; | ||
| 41 | u32 pmtu_learned; | ||
| 42 | struct inetpeer_addr_base redirect_learned; | ||
| 38 | /* | 43 | /* |
| 39 | * Once inet_peer is queued for deletion (refcnt == -1), following fields | 44 | * Once inet_peer is queued for deletion (refcnt == -1), following fields |
| 40 | * are not available: rid, ip_id_count, tcp_ts, tcp_ts_stamp, metrics | 45 | * are not available: rid, ip_id_count, tcp_ts, tcp_ts_stamp |
| 41 | * We can share memory with rcu_head to help keep inet_peer small. | 46 | * We can share memory with rcu_head to help keep inet_peer small. |
| 42 | */ | 47 | */ |
| 43 | union { | 48 | union { |
| @@ -46,17 +51,14 @@ struct inet_peer { | |||
| 46 | atomic_t ip_id_count; /* IP ID for the next packet */ | 51 | atomic_t ip_id_count; /* IP ID for the next packet */ |
| 47 | __u32 tcp_ts; | 52 | __u32 tcp_ts; |
| 48 | __u32 tcp_ts_stamp; | 53 | __u32 tcp_ts_stamp; |
| 49 | u32 metrics[RTAX_MAX]; | ||
| 50 | u32 rate_tokens; /* rate limiting for ICMP */ | ||
| 51 | unsigned long rate_last; | ||
| 52 | unsigned long pmtu_expires; | ||
| 53 | u32 pmtu_orig; | ||
| 54 | u32 pmtu_learned; | ||
| 55 | struct inetpeer_addr_base redirect_learned; | ||
| 56 | }; | 54 | }; |
| 57 | struct rcu_head rcu; | 55 | struct rcu_head rcu; |
| 58 | struct inet_peer *gc_next; | 56 | struct inet_peer *gc_next; |
| 59 | }; | 57 | }; |
| 58 | |||
| 59 | /* following fields might be frequently dirtied */ | ||
| 60 | __u32 dtime; /* the time of last use of not referenced entries */ | ||
| 61 | atomic_t refcnt; | ||
| 60 | }; | 62 | }; |
| 61 | 63 | ||
| 62 | void inet_initpeers(void) __init; | 64 | void inet_initpeers(void) __init; |