Messages in this thread | | | Date | Mon, 6 Jul 2009 18:52:16 +0800 | From | Herbert Xu <> | Subject | Re: sk_lock: inconsistent {RECLAIM_FS-ON-W} -> {IN-RECLAIM_FS-W} usage |
| |
Wu Fengguang <fengguang.wu@intel.com> wrote: > > This lockdep warning appears when doing stress memory tests over NFS. > > page reclaim => nfs_writepage => tcp_sendmsg => lock sk_lock > > tcp_close => lock sk_lock => tcp_send_fin => alloc_skb_fclone => page reclaim
Well perhaps not this particular path, but it is certainly possible if an existing NFS socket dies and NFS tries to reestablish it.
I suggest that NFS should utilise the sk_allocation field and set an appropriate value. Note that you may have to patch TCP so that it uses sk_allocation everywhere necessary, e.g., in tcp_send_fin.
Cheers, -- Visit Openswan at http://www.openswan.org/ Email: Herbert Xu ~{PmV>HI~} <herbert@gondor.apana.org.au> Home Page: http://gondor.apana.org.au/~herbert/ PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
| |