lk/linux-kernel-test
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

net: avoid limits overflow

Browse Source 
Robin Holt tried to boot a 16TB machine and found some limits were
reached : sysctl_tcp_mem[2], sysctl_udp_mem[2]

We can switch infrastructure to use long "instead" of "int", now
atomic_long_t primitives are available for free.

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Reported-by: Robin Holt <holt@sgi.com>
Reviewed-by: Robin Holt <holt@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit is contained in:
Eric Dumazet
2010-11-09 23:24:26 +00:00
committed by David S. Miller
parent 67286640f6
commit 8d987e5c75
15 changed files with 40 additions and 38 deletions
Download Patch File Download Diff File Expand all files Collapse all files

11
net/ipv4/tcp_input.c
View File

@@ -259,8 +259,11 @@ static void tcp_fixup_sndbuf(struct sock *sk)
int sndmem = tcp_sk(sk)->rx_opt.mss_clamp + MAX_TCP_HEADER + 16 +
sizeof(struct sk_buff);
if (sk->sk_sndbuf < 3 * sndmem)
sk->sk_sndbuf = min(3 * sndmem, sysctl_tcp_wmem[2]);
if (sk->sk_sndbuf < 3 * sndmem) {
sk->sk_sndbuf = 3 * sndmem;
if (sk->sk_sndbuf > sysctl_tcp_wmem[2])
sk->sk_sndbuf = sysctl_tcp_wmem[2];
}
}
/* 2. Tuning advertised window (window_clamp, rcv_ssthresh)
@@ -396,7 +399,7 @@ static void tcp_clamp_window(struct sock *sk)
if (sk->sk_rcvbuf < sysctl_tcp_rmem[2] &&
!(sk->sk_userlocks & SOCK_RCVBUF_LOCK) &&
!tcp_memory_pressure &&
atomic_read(&tcp_memory_allocated) < sysctl_tcp_mem[0]) {
atomic_long_read(&tcp_memory_allocated) < sysctl_tcp_mem[0]) {
sk->sk_rcvbuf = min(atomic_read(&sk->sk_rmem_alloc),
sysctl_tcp_rmem[2]);
}
@@ -4861,7 +4864,7 @@ static int tcp_should_expand_sndbuf(struct sock *sk)
return 0;
/* If we are under soft global TCP memory pressure, do not expand. */
if (atomic_read(&tcp_memory_allocated) >= sysctl_tcp_mem[0])
if (atomic_long_read(&tcp_memory_allocated) >= sysctl_tcp_mem[0])
return 0;
/* If we filled the congestion window, do not expand. */