On Fri, 2022-10-28 at 10:48 -0400, Chuck Lever wrote:
> fh_match() is costly, especially when filehandles are large (as is
> the case for NFSv4). It needs to be used sparingly when searching
> data structures. Unfortunately, with common workloads, I see
> multiple thousands of objects stored in file_hashtbl[], which has
> just 256 buckets, making its bucket hash chains quite lengthy.
>
> Walking long hash chains with the state_lock held blocks other
> activity that needs that lock. Sizable hash chains are a common
> occurrance once the server has handed out some delegations, for
> example -- IIUC, each delegated file is held open on the server by
> an nfs4_file object.
>
> To help mitigate the cost of searching with fh_match(), replace the
> nfs4_file hash table with an rhashtable, which can dynamically
> resize its bucket array to minimize hash chain length.
>
> The result of this modification is an improvement in the latency of
> NFSv4 operations, and the reduction of nfsd CPU utilization due to
> eliminating the cost of multiple calls to fh_match() and reducing
> the CPU cache misses incurred while walking long hash chains in the
> nfs4_file hash table.
>
> Signed-off-by: Chuck Lever <chuck.lever@xxxxxxxxxx>
> Reviewed-by: NeilBrown <neilb@xxxxxxx>
> ---
> fs/nfsd/nfs4state.c | 77 ++++++++++++++++++++++++++-------------------------
> fs/nfsd/state.h | 4 ---
> 2 files changed, 40 insertions(+), 41 deletions(-)
>
> diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
> index c2ef2db9c84c..c78b66e678dd 100644
> --- a/fs/nfsd/nfs4state.c
> +++ b/fs/nfsd/nfs4state.c
> @@ -591,11 +591,8 @@ static void nfsd4_free_file_rcu(struct rcu_head *rcu)
> void
> put_nfs4_file(struct nfs4_file *fi)
> {
> - might_lock(&state_lock);
> -
> - if (refcount_dec_and_lock(&fi->fi_ref, &state_lock)) {
> + if (refcount_dec_and_test(&fi->fi_ref)) {
> nfsd4_file_hash_remove(fi);
> - spin_unlock(&state_lock);
> WARN_ON_ONCE(!list_empty(&fi->fi_clnt_odstate));
> WARN_ON_ONCE(!list_empty(&fi->fi_delegations));
> call_rcu(&fi->fi_rcu, nfsd4_free_file_rcu);
> @@ -709,20 +706,6 @@ static unsigned int ownerstr_hashval(struct xdr_netobj *ownername)
> return ret & OWNER_HASH_MASK;
> }
>
> -/* hash table for nfs4_file */
> -#define FILE_HASH_BITS 8
> -#define FILE_HASH_SIZE (1 << FILE_HASH_BITS)
> -
> -static unsigned int file_hashval(const struct svc_fh *fh)
> -{
> - struct inode *inode = d_inode(fh->fh_dentry);
> -
> - /* XXX: why not (here & in file cache) use inode? */
> - return (unsigned int)hash_long(inode->i_ino, FILE_HASH_BITS);
> -}
> -
> -static struct hlist_head file_hashtbl[FILE_HASH_SIZE];
> -
> static struct rhltable nfs4_file_rhltable ____cacheline_aligned_in_smp;
>
> static const struct rhashtable_params nfs4_file_rhash_params = {
> @@ -4687,12 +4670,14 @@ move_to_close_lru(struct nfs4_ol_stateid *s, struct net *net)
> static noinline_for_stack struct nfs4_file *
> nfsd4_file_hash_lookup(const struct svc_fh *fhp)
> {
> - unsigned int hashval = file_hashval(fhp);
> + struct inode *inode = d_inode(fhp->fh_dentry);
> + struct rhlist_head *tmp, *list;
> struct nfs4_file *fi;
>
> rcu_read_lock();
> - hlist_for_each_entry_rcu(fi, &file_hashtbl[hashval], fi_hash,
> - lockdep_is_held(&state_lock)) {
> + list = rhltable_lookup(&nfs4_file_rhltable, &inode,
> + nfs4_file_rhash_params);
> + rhl_for_each_entry_rcu(fi, tmp, list, fi_rlist) {
> if (fh_match(&fi->fi_fhandle, &fhp->fh_handle)) {
> if (refcount_inc_not_zero(&fi->fi_ref)) {
> rcu_read_unlock();
> @@ -4706,40 +4691,56 @@ nfsd4_file_hash_lookup(const struct svc_fh *fhp)
>
> /*
> * On hash insertion, identify entries with the same inode but
> - * distinct filehandles. They will all be in the same hash bucket
> - * because nfs4_file's are hashed by the address in the fi_inode
> - * field.
> + * distinct filehandles. They will all be on the list returned
> + * by rhltable_lookup().
> + *
> + * inode->i_lock prevents racing insertions from adding an entry
> + * for the same inode/fhp pair twice.
> */
> static noinline_for_stack struct nfs4_file *
> nfsd4_file_hash_insert(struct nfs4_file *new, const struct svc_fh *fhp)
> {
> - unsigned int hashval = file_hashval(fhp);
> + struct inode *inode = d_inode(fhp->fh_dentry);
> + struct rhlist_head *tmp, *list;
> struct nfs4_file *ret = NULL;
> bool alias_found = false;
> struct nfs4_file *fi;
> + int err;
>
> - spin_lock(&state_lock);
> - hlist_for_each_entry_rcu(fi, &file_hashtbl[hashval], fi_hash,
> - lockdep_is_held(&state_lock)) {
> + rcu_read_lock();
> + spin_lock(&inode->i_lock);
> +
> + list = rhltable_lookup(&nfs4_file_rhltable, &inode,
> + nfs4_file_rhash_params);
> + rhl_for_each_entry_rcu(fi, tmp, list, fi_rlist) {
> if (fh_match(&fi->fi_fhandle, &fhp->fh_handle)) {
> if (refcount_inc_not_zero(&fi->fi_ref))
> ret = fi;
> - } else if (d_inode(fhp->fh_dentry) == fi->fi_inode)
> + } else
> fi->fi_aliased = alias_found = true;
> }
> - if (likely(ret == NULL)) {
> - nfsd4_file_init(fhp, new);
> - hlist_add_head_rcu(&new->fi_hash, &file_hashtbl[hashval]);
> - new->fi_aliased = alias_found;
> - ret = new;
> - }
> - spin_unlock(&state_lock);
> + if (ret)
> + goto out_unlock;
> +
> + nfsd4_file_init(fhp, new);
> + err = rhltable_insert(&nfs4_file_rhltable, &new->fi_rlist,
> + nfs4_file_rhash_params);
> + if (err)
> + goto out_unlock;
> +
> + new->fi_aliased = alias_found;
> + ret = new;
> +
> +out_unlock:
> + spin_unlock(&inode->i_lock);
> + rcu_read_unlock();
> return ret;
> }
>
> static noinline_for_stack void nfsd4_file_hash_remove(struct nfs4_file *fi)
> {
> - hlist_del_rcu(&fi->fi_hash);
> + rhltable_remove(&nfs4_file_rhltable, &fi->fi_rlist,
> + nfs4_file_rhash_params);
> }
>
> /*
> @@ -5629,6 +5630,8 @@ nfsd4_process_open2(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nf
> * If not found, create the nfs4_file struct
> */
> fp = nfsd4_file_hash_insert(open->op_file, current_fh);
> + if (unlikely(!fp))
> + return nfserr_jukebox;
> if (fp != open->op_file) {
> status = nfs4_check_deleg(cl, open, &dp);
> if (status)
> diff --git a/fs/nfsd/state.h b/fs/nfsd/state.h
> index 190fc7e418a4..eadd7f465bf5 100644
> --- a/fs/nfsd/state.h
> +++ b/fs/nfsd/state.h
> @@ -536,16 +536,12 @@ struct nfs4_clnt_odstate {
> * inode can have multiple filehandles associated with it, so there is
> * (potentially) a many to one relationship between this struct and struct
> * inode.
> - *
> - * These are hashed by filehandle in the file_hashtbl, which is protected by
> - * the global state_lock spinlock.
> */
> struct nfs4_file {
> refcount_t fi_ref;
> struct inode * fi_inode;
> bool fi_aliased;
> spinlock_t fi_lock;
> - struct hlist_node fi_hash; /* hash on fi_fhandle */
> struct rhlist_head fi_rlist;
> struct list_head fi_stateids;
> union {
>
>
You can add this to 13 and 14:
Reviewed-by: Jeff Layton <jlayton@xxxxxxxxxx>
That said, I'd probably prefer to see the two squashed together, since
you're adding unused infrastructure in 13.
