Skip to content

Commit

Permalink
tcp: change tcp_adv_win_scale and tcp_rmem[2]
Browse files Browse the repository at this point in the history
tcp_adv_win_scale default value is 2, meaning we expect a good citizen
skb to have skb->len / skb->truesize ratio of 75% (3/4)

In 2.6 kernels we (mis)accounted for typical MSS=1460 frame :
1536 + 64 + 256 = 1856 'estimated truesize', and 1856 * 3/4 = 1392.
So these skbs were considered as not bloated.

With recent truesize fixes, a typical MSS=1460 frame truesize is now the
more precise :
2048 + 256 = 2304. But 2304 * 3/4 = 1728.
So these skb are not good citizen anymore, because 1460 < 1728

(GRO can escape this problem because it build skbs with a too low
truesize.)

This also means tcp advertises a too optimistic window for a given
allocated rcvspace : When receiving frames, sk_rmem_alloc can hit
sk_rcvbuf limit and we call tcp_prune_queue()/tcp_collapse() too often,
especially when application is slow to drain its receive queue or in
case of losses (netperf is fast, scp is slow). This is a major latency
source.

We should adjust the len/truesize ratio to 50% instead of 75%

This patch :

1) changes tcp_adv_win_scale default to 1 instead of 2

2) increase tcp_rmem[2] limit from 4MB to 6MB to take into account
better truesize tracking and to allow autotuning tcp receive window to
reach same value than before. Note that same amount of kernel memory is
consumed compared to 2.6 kernels.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Tom Herbert <therbert@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
  • Loading branch information
Eric Dumazet authored and David S. Miller committed May 3, 2012
1 parent 84768ed commit b49960a
Show file tree
Hide file tree
Showing 3 changed files with 8 additions and 7 deletions.
4 changes: 2 additions & 2 deletions Documentation/networking/ip-sysctl.txt
Original file line number Diff line number Diff line change
Expand Up @@ -147,7 +147,7 @@ tcp_adv_win_scale - INTEGER
(if tcp_adv_win_scale > 0) or bytes-bytes/2^(-tcp_adv_win_scale),
if it is <= 0.
Possible values are [-31, 31], inclusive.
Default: 2
Default: 1

tcp_allowed_congestion_control - STRING
Show/set the congestion control choices available to non-privileged
Expand Down Expand Up @@ -410,7 +410,7 @@ tcp_rmem - vector of 3 INTEGERs: min, default, max
net.core.rmem_max. Calling setsockopt() with SO_RCVBUF disables
automatic tuning of that socket's receive buffer size, in which
case this value is ignored.
Default: between 87380B and 4MB, depending on RAM size.
Default: between 87380B and 6MB, depending on RAM size.

tcp_sack - BOOLEAN
Enable select acknowledgments (SACKS).
Expand Down
9 changes: 5 additions & 4 deletions net/ipv4/tcp.c
Original file line number Diff line number Diff line change
Expand Up @@ -3243,7 +3243,7 @@ void __init tcp_init(void)
{
struct sk_buff *skb = NULL;
unsigned long limit;
int max_share, cnt;
int max_rshare, max_wshare, cnt;
unsigned int i;
unsigned long jiffy = jiffies;

Expand Down Expand Up @@ -3303,15 +3303,16 @@ void __init tcp_init(void)
tcp_init_mem(&init_net);
/* Set per-socket limits to no more than 1/128 the pressure threshold */
limit = nr_free_buffer_pages() << (PAGE_SHIFT - 7);
max_share = min(4UL*1024*1024, limit);
max_wshare = min(4UL*1024*1024, limit);
max_rshare = min(6UL*1024*1024, limit);

sysctl_tcp_wmem[0] = SK_MEM_QUANTUM;
sysctl_tcp_wmem[1] = 16*1024;
sysctl_tcp_wmem[2] = max(64*1024, max_share);
sysctl_tcp_wmem[2] = max(64*1024, max_wshare);

sysctl_tcp_rmem[0] = SK_MEM_QUANTUM;
sysctl_tcp_rmem[1] = 87380;
sysctl_tcp_rmem[2] = max(87380, max_share);
sysctl_tcp_rmem[2] = max(87380, max_rshare);

pr_info("Hash tables configured (established %u bind %u)\n",
tcp_hashinfo.ehash_mask + 1, tcp_hashinfo.bhash_size);
Expand Down
2 changes: 1 addition & 1 deletion net/ipv4/tcp_input.c
Original file line number Diff line number Diff line change
Expand Up @@ -85,7 +85,7 @@ int sysctl_tcp_ecn __read_mostly = 2;
EXPORT_SYMBOL(sysctl_tcp_ecn);
int sysctl_tcp_dsack __read_mostly = 1;
int sysctl_tcp_app_win __read_mostly = 31;
int sysctl_tcp_adv_win_scale __read_mostly = 2;
int sysctl_tcp_adv_win_scale __read_mostly = 1;
EXPORT_SYMBOL(sysctl_tcp_adv_win_scale);

int sysctl_tcp_stdurg __read_mostly;
Expand Down

0 comments on commit b49960a

Please sign in to comment.