Re: sk_lock: inconsistent {RECLAIM_FS-ON-W} -> {IN-RECLAIM_FS-W} usage

On Mon, Jul 06, 2009 at 06:52:16PM +0800, Herbert Xu wrote:
Good suggestion! NFS already sets sk_allocation to GFP_ATOMIC in
linux/net/sunrpc/xprtsock.c <<xs_tcp_finish_connecting>>.

To fix this warning and possible recursions, I converted some
GPF_KERNEL cases to sk_allocation in the tcp/ipv4 code:

---
tcp: replace hard coded GFP_KERNEL with sk_allocation

This fixed a lockdep warning which appeared when doing stress
memory tests over NFS:

        inconsistent {RECLAIM_FS-ON-W} -> {IN-RECLAIM_FS-W} usage.

        mount_root => nfs_root_data => tcp_close => lock sk_lock =>
                        tcp_send_fin => alloc_skb_fclone => page reclaim

        page reclaim => nfs_writepage => tcp_sendmsg => lock sk_lock

CC: Arnaldo Carvalho de Melo <acme@...>
CC: David S. Miller <davem@...>
CC: Herbert Xu <herbert@...>
Signed-off-by: Wu Fengguang <fengguang.wu@...>
---
 net/ipv4/tcp.c        |    4 ++--
 net/ipv4/tcp_ipv4.c   |    5 +++--
 net/ipv4/tcp_output.c |    5 +++--
 3 files changed, 8 insertions(+), 6 deletions(-)

--- linux.orig/net/ipv4/tcp_ipv4.c
+++ linux/net/ipv4/tcp_ipv4.c
@@ -970,8 +970,9 @@ static int tcp_v4_parse_md5_keys(struct
 
  if (!tcp_sk(sk)->md5sig_info) {
  struct tcp_sock *tp = tcp_sk(sk);
- struct tcp_md5sig_info *p = kzalloc(sizeof(*p), GFP_KERNEL);
+ struct tcp_md5sig_info *p;
 
+ p = kzalloc(sizeof(*p), sk->sk_allocation);
  if (!p)
  return -EINVAL;
 
@@ -979,7 +980,7 @@ static int tcp_v4_parse_md5_keys(struct
  sk->sk_route_caps &= ~NETIF_F_GSO_MASK;
  }
 
- newkey = kmemdup(cmd.tcpm_key, cmd.tcpm_keylen, GFP_KERNEL);
+ newkey = kmemdup(cmd.tcpm_key, cmd.tcpm_keylen, sk->sk_allocation);
  if (!newkey)
  return -ENOMEM;
  return tcp_v4_md5_do_add(sk, sin->sin_addr.s_addr,
--- linux.orig/net/ipv4/tcp_output.c
+++ linux/net/ipv4/tcp_output.c
@@ -2100,7 +2100,8 @@ void tcp_send_fin(struct sock *sk)
  } else {
  /* Socket is locked, keep trying until memory is available. */
  for (;;) {
- skb = alloc_skb_fclone(MAX_TCP_HEADER, GFP_KERNEL);
+ skb = alloc_skb_fclone(MAX_TCP_HEADER,
+       sk->sk_allocation);
  if (skb)
  break;
  yield();
@@ -2358,7 +2359,7 @@ int tcp_connect(struct sock *sk)
  sk->sk_wmem_queued += buff->truesize;
  sk_mem_charge(sk, buff->truesize);
  tp->packets_out += tcp_skb_pcount(buff);
- tcp_transmit_skb(sk, buff, 1, GFP_KERNEL);
+ tcp_transmit_skb(sk, buff, 1, sk->sk_allocation);
 
  /* We change tp->snd_nxt after the tcp_transmit_skb() call
  * in order to make this packet get counted in tcpOutSegs.
--- linux.orig/net/ipv4/tcp.c
+++ linux/net/ipv4/tcp.c
@@ -1834,7 +1834,7 @@ void tcp_close(struct sock *sk, long tim
  /* Unread data was tossed, zap the connection. */
  NET_INC_STATS_USER(sock_net(sk), LINUX_MIB_TCPABORTONCLOSE);
  tcp_set_state(sk, TCP_CLOSE);
- tcp_send_active_reset(sk, GFP_KERNEL);
+ tcp_send_active_reset(sk, sk->sk_allocation);
  } else if (sock_flag(sk, SOCK_LINGER) && !sk->sk_lingertime) {
  /* Check zero linger _after_ checking for unread data. */
  sk->sk_prot->disconnect(sk, 0);
@@ -2666,7 +2666,7 @@ static struct tcp_md5sig_pool **__tcp_al
  struct tcp_md5sig_pool *p;
  struct crypto_hash *hash;
 
- p = kzalloc(sizeof(*p), GFP_KERNEL);
+ p = kzalloc(sizeof(*p), sk->sk_allocation);
  if (!p)
  goto out_free;
  *per_cpu_ptr(pool, cpu) = p;
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/