nfsd: allocate/free session-based DRC slots on demand
Subsystems (2)
v5 cover nfsd: allocate/free session-based DRC slots on demand
Differences from v4:
- added Jeff's Reviewed-by - thanks.
- fixed problem in final patch. There are two places where a session can be detached from lists and both need to detach from the new global list.
Thanks,
NeilBrown
[PATCH 1/6] nfsd: use an xarray to store v4.1 session slots [PATCH 2/6] nfsd: remove artificial limits on the session-based DRC [PATCH 3/6] nfsd: add session slot count to
[PATCH 4/6] nfsd: allocate new session-based DRC slots on demand. [PATCH 5/6] nfsd: add support for freeing unused session-DRC slots [PATCH 6/6] nfsd: add shrinker to reduce number of slots allocated
PATCH v5 1/6 nfsd: use an xarray to store v4.1 session slots
Using an xarray to store session slots will make it easier to change the number of active slots based on demand, and removes an unnecessary limit.
To achieve good throughput with a high-latency server it can be helpful to have hundreds of concurrent writes, which means hundreds of slots. So increase the limit to 2048 (twice what the Linux client will currently use). This limit is only a sanity check, not a hard limit.
--- fs/nfsd/nfs4state.c | 28 ++++++++++++++++++---------- fs/nfsd/state.h | 9 ++++++--- 2 files changed, 24 insertions(+), 13 deletions(-)
fs/nfsd/nfs4state.c +18 -10
@@ -1915,8 +1915,11 @@ free_session_slots(struct nfsd4_session *ses)int i;for (i =0 ; i < ses->se_fchannel.maxreqs; i++) { -free_svc_cred (&ses->se_slots[i]->sl_cred); -kfree (ses->se_slots[i]); +struct nfsd4_slot *slot =xa_load (&ses->se_slots, i); + +xa_erase (&ses->se_slots, i); +free_svc_cred (&slot->sl_cred); +kfree (slot); } }
@@ -1996,17 +1999,20 @@ static struct nfsd4_session *alloc_session(struct nfsd4_channel_attrs *fattrs,struct nfsd4_session *new;int i; -BUILD_BUG_ON (struct_size (new, se_slots,NFSD_MAX_SLOTS_PER_SESSION ) - >PAGE_SIZE ); - - new =kzalloc (struct_size (new, se_slots, numslots),GFP_KERNEL ); + new =kzalloc (sizeof (*new),GFP_KERNEL );if (!new)return NULL ; +xa_init (&new->se_slots);/* allocate each struct nfsd4_slot and data cache in one piece */ for (i =0 ; i < numslots; i++) { - new->se_slots[i] =kzalloc (slotsize,GFP_KERNEL ); -if (!new->se_slots[i]) +struct nfsd4_slot *slot; + slot =kzalloc (slotsize,GFP_KERNEL ); +if (!slot) goto out_free; +if (xa_is_err (xa_store (&new->se_slots, i, slot,GFP_KERNEL ))) { +kfree (slot); + goto out_free; + } }memcpy (&new->se_fchannel, fattrs,sizeof (struct nfsd4_channel_attrs ));
@@ -2017,7 +2023,8 @@ static struct nfsd4_session *alloc_session(struct nfsd4_channel_attrs *fattrs,return new; out_free:while (i--) -kfree (new->se_slots[i]); +kfree (xa_load (&new->se_slots, i)); +xa_destroy (&new->se_slots);kfree (new);return NULL ; }
@@ -2124,6 +2131,7 @@ static void nfsd4_del_conns(struct nfsd4_session *s)static void __free_session (struct nfsd4_session *ses) {free_session_slots (ses); +xa_destroy (&ses->se_slots);kfree (ses); }
@@ -4278,7 +4286,7 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,if (seq->slotid >= session->se_fchannel.maxreqs) goto out_put_session; - slot = session->se_slots[seq->slotid]; + slot =xa_load (&session->se_slots, seq->slotid);dprintk ("%s: slotid %d\n" , __func__, seq->slotid);/* We do not negotiate the number of slots yet, so set the
fs/nfsd/state.h +6 -3
@@ -227,8 +227,11 @@ static inline struct nfs4_delegation *delegstateid(struct nfs4_stid *s) return container_of(s, struct nfs4_delegation, dl_stid); } -/* Maximum number of slots per session. 160 is useful for long haul TCP */ -#define NFSD_MAX_SLOTS_PER_SESSION 160 +/* Maximum number of slots per session. This is for sanity-check only. +* It could be increased if we had a mechanism to shutdown misbehaving clients. +* A large number can be needed to get good throughput on high-latency servers. +*/ +#define NFSD_MAX_SLOTS_PER_SESSION 2048/* Maximum session per slot cache size */ #define NFSD_SLOT_CACHE_SIZE 2048/* Maximum number of NFSD_SLOT_CACHE_SIZE slots per session */
@@ -327,7 +330,7 @@ struct nfsd4_session {struct nfsd4_cb_sec se_cb_sec;struct list_head se_conns;u32 se_cb_seq_nr[NFSD_BC_SLOT_TABLE_SIZE ]; -struct nfsd4_slot *se_slots[];/* forward channel slots */ +struct xarray se_slots;/* forward channel slots */ };/* formatted contents of nfs4_sessionid */
Signature · 1 line
2.47.0
PATCH v5 2/6 nfsd: remove artificial limits on the session-based DRC
In reply to NeilBrownRather than guessing how much space it might be safe to use for the DRC, simply try allocating slots and be prepared to accept failure.
The first slot for each session is allocated with GFP_KERNEL which is unlikely to fail. Subsequent slots are allocated with the addition of __GFP_NORETRY which is expected to fail if there isn't much free memory.
This is probably too aggressive but clears the way for adding a shrinker interface to free extra slots when memory is tight.
--- fs/nfsd/nfs4state.c | 94 ++++++++------------------------------------- fs/nfsd/nfsd.h | 3 -- fs/nfsd/nfssvc.c | 32 --------------- 3 files changed, 16 insertions(+), 113 deletions(-)
fs/nfsd/nfs4state.c +16 -78
@@ -1938,65 +1938,13 @@ static inline u32 slot_bytes(struct nfsd4_channel_attrs *ca)return size +sizeof (struct nfsd4_slot ); } -/* -* XXX: If we run out of reserved DRC memory we could (up to a point) -* re-negotiate active sessions and reduce their slot usage to make -* room for new connections. For now we just fail the create session. -*/ -static u32 nfsd4_get_drc_mem (struct nfsd4_channel_attrs *ca,struct nfsd_net *nn) -{ -u32 slotsize =slot_bytes (ca); -u32 num = ca->maxreqs; -unsigned long avail, total_avail; -unsigned int scale_factor; - -spin_lock (&nfsd_drc_lock); -if (nfsd_drc_max_mem > nfsd_drc_mem_used) - total_avail = nfsd_drc_max_mem - nfsd_drc_mem_used; -else -/* We have handed out more space than we chose in -* set_max_drc() to allow. That isn't really a -* problem as long as that doesn't make us think we -* have lots more due to integer overflow. -*/ - total_avail =0 ; - avail =min ((unsigned long )NFSD_MAX_MEM_PER_SESSION , total_avail); -/* -* Never use more than a fraction of the remaining memory, -* unless it's the only way to give this client a slot. -* The chosen fraction is either 1/8 or 1/number of threads, -* whichever is smaller. This ensures there are adequate -* slots to support multiple clients per thread. -* Give the client one slot even if that would require -* over-allocation--it is better than failure. -*/ - scale_factor = max_t(unsigned int, 8, nn->nfsd_serv->sv_nrthreads); - - avail = clamp_t(unsigned long, avail, slotsize, - total_avail/scale_factor); - num =min_t (int, num, avail / slotsize); - num =max_t (int, num,1 ); - nfsd_drc_mem_used += num * slotsize; -spin_unlock (&nfsd_drc_lock); - -return num; -} - -static void nfsd4_put_drc_mem (struct nfsd4_channel_attrs *ca) -{ -int slotsize =slot_bytes (ca); - -spin_lock (&nfsd_drc_lock); - nfsd_drc_mem_used -= slotsize * ca->maxreqs; -spin_unlock (&nfsd_drc_lock); -} -static struct nfsd4_session *alloc_session (struct nfsd4_channel_attrs *fattrs,struct nfsd4_channel_attrs *battrs) {int numslots = fattrs->maxreqs;int slotsize =slot_bytes (fattrs);struct nfsd4_session *new; +struct nfsd4_slot *slot;int i; new =kzalloc (sizeof (*new),GFP_KERNEL );
@@ -2004,17 +1952,21 @@ static struct nfsd4_session *alloc_session(struct nfsd4_channel_attrs *fattrs,return NULL ;xa_init (&new->se_slots);/* allocate each struct nfsd4_slot and data cache in one piece */ -for (i =0 ; i < numslots; i++) { -struct nfsd4_slot *slot; - slot =kzalloc (slotsize,GFP_KERNEL ); + slot =kzalloc (slotsize,GFP_KERNEL ); +if (!slot ||xa_is_err (xa_store (&new->se_slots,0 , slot,GFP_KERNEL ))) + goto out_free; + +for (i =1 ; i < numslots; i++) { +const gfp_t gfp =GFP_KERNEL | __GFP_NORETRY | __GFP_NOWARN; + slot =kzalloc (slotsize, gfp);if (!slot) - goto out_free; -if (xa_is_err (xa_store (&new->se_slots, i, slot,GFP_KERNEL ))) { +break ; +if (xa_is_err (xa_store (&new->se_slots, i, slot, gfp))) {kfree (slot); - goto out_free; +break ; } } - + fattrs->maxreqs = i;memcpy (&new->se_fchannel, fattrs,sizeof (struct nfsd4_channel_attrs )); new->se_cb_slot_avail = ~0U ; new->se_cb_highest_slot = min(battrs->maxreqs - 1,
@@ -2022,8 +1974,7 @@ static struct nfsd4_session *alloc_session(struct nfsd4_channel_attrs *fattrs,spin_lock_init (&new->se_lock);return new; out_free: -while (i--) -kfree (xa_load (&new->se_slots, i)); +kfree (slot);xa_destroy (&new->se_slots);kfree (new);return NULL ;
@@ -2138,7 +2089,6 @@ static void __free_session(struct nfsd4_session *ses)static void free_session (struct nfsd4_session *ses) {nfsd4_del_conns (ses); -nfsd4_put_drc_mem (&ses->se_fchannel);__free_session (ses); }
@@ -3786,17 +3736,6 @@ static __be32 check_forechannel_attrs(struct nfsd4_channel_attrs *ca, struct nfs ca->maxresp_cached =min_t (u32, ca->maxresp_cached,NFSD_SLOT_CACHE_SIZE +NFSD_MIN_HDR_SEQ_SZ ); ca->maxreqs =min_t (u32, ca->maxreqs,NFSD_MAX_SLOTS_PER_SESSION ); -/* -* Note decreasing slot size below client's request may make it -* difficult for client to function correctly, whereas -* decreasing the number of slots will (just?) affect -* performance. When short on memory we therefore prefer to -* decrease number of slots instead of their size. Clients that -* request larger slots than they need will get poor results: -* Note that we always allow at least one slot, because our -* accounting is soft and provides no guarantees either way. -*/ - ca->maxreqs =nfsd4_get_drc_mem (ca, nn);return nfs_ok; }
@@ -3874,11 +3813,11 @@ nfsd4_create_session(struct svc_rqst *rqstp,return status; status =check_backchannel_attrs (&cr_ses->back_channel);if (status) - goto out_release_drc_mem; + goto out_err; status = nfserr_jukebox; new =alloc_session (&cr_ses->fore_channel, &cr_ses->back_channel);if (!new) - goto out_release_drc_mem; + goto out_err; conn =alloc_conn_from_crses (rqstp, cr_ses);if (!conn) goto out_free_session;
@@ -3987,8 +3926,7 @@ nfsd4_create_session(struct svc_rqst *rqstp,free_conn (conn); out_free_session:__free_session (new); -out_release_drc_mem: -nfsd4_put_drc_mem (&cr_ses->fore_channel); +out_err:return status; }
fs/nfsd/nfsd.h +0 -3
@@ -88,9 +88,6 @@ struct nfsd_genl_rqstp {extern struct svc_program nfsd_programs[];extern const struct svc_version nfsd_version2, nfsd_version3, nfsd_version4;extern struct mutex nfsd_mutex; -extern spinlock_t nfsd_drc_lock; -extern unsigned long nfsd_drc_max_mem; -extern unsigned long nfsd_drc_mem_used;extern atomic_t nfsd_th_cnt;/* number of available threads */ extern const struct seq_operations nfs_exports_op;
fs/nfsd/nfssvc.c +0 -32
@@ -70,16 +70,6 @@ static __be32 nfsd_init_request(struct svc_rqst *,*/ DEFINE_MUTEX (nfsd_mutex); -/* -* nfsd_drc_lock protects nfsd_drc_max_pages and nfsd_drc_pages_used. -* nfsd_drc_max_pages limits the total amount of memory available for -* version 4.1 DRC caches. -* nfsd_drc_pages_used tracks the current version 4.1 DRC memory usage. -*/ -DEFINE_SPINLOCK (nfsd_drc_lock); -unsigned long nfsd_drc_max_mem; -unsigned long nfsd_drc_mem_used; -#if IS_ENABLED (CONFIG_NFS_LOCALIO )static const struct svc_version *localio_versions[] = { [1] = &localio_version1,
@@ -575,27 +565,6 @@ void nfsd_reset_versions(struct nfsd_net *nn) } } -/* -* Each session guarantees a negotiated per slot memory cache for replies -* which in turn consumes memory beyond the v2/v3/v4.0 server. A dedicated -* NFSv4.1 server might want to use more memory for a DRC than a machine -* with mutiple services. -* -* Impose a hard limit on the number of pages for the DRC which varies -* according to the machines free pages. This is of course only a default. -* -* For now this is a #defined shift which could be under admin control -* in the future. -*/ -static void set_max_drc (void ) -{ -#define NFSD_DRC_SIZE_SHIFT 7 - nfsd_drc_max_mem = (nr_free_buffer_pages () - >>NFSD_DRC_SIZE_SHIFT ) *PAGE_SIZE ; - nfsd_drc_mem_used =0 ; -dprintk ("%s nfsd_drc_max_mem %lu \n" , __func__, nfsd_drc_max_mem); -} -static int nfsd_get_default_max_blksize (void ) {struct sysinfo i;
@@ -678,7 +647,6 @@ int nfsd_create_serv(struct net *net) nn->nfsd_serv = serv;spin_unlock (&nfsd_notifier_lock); -set_max_drc ();/* check if the notifier is already set */ if (atomic_inc_return (&nfsd_notifier_refcount) ==1 ) {register_inetaddr_notifier (&nfsd_inetaddr_notifier);
Signature · 1 line
2.47.0
PATCH v5 3/6 nfsd: add session slot count to /proc/fs/nfsd/clients/*/info
In reply to NeilBrownEach client now reports the number of slots allocated in each session.
--- fs/nfsd/nfs4state.c | 8 ++++++++ 1 file changed, 8 insertions(+)
fs/nfsd/nfs4state.c +8 -0
@@ -2643,6 +2643,7 @@ static const char *cb_state2str(int state)static int client_info_show (struct seq_file *m,void *v) {struct inode *inode =file_inode (m->file); +struct nfsd4_session *ses;struct nfs4_client *clp;u64 clid;
@@ -2679,6 +2680,13 @@ static int client_info_show(struct seq_file *m, void *v)seq_printf (m,"callback address: \"%pISpc\"\n" , &clp->cl_cb_conn.cb_addr);seq_printf (m,"admin-revoked states: %d\n" ,atomic_read (&clp->cl_admin_revoked)); +spin_lock (&clp->cl_lock); +seq_printf (m,"session slots:" ); +list_for_each_entry (ses, &clp->cl_sessions, se_perclnt) +seq_printf (m," %u" , ses->se_fchannel.maxreqs); +spin_unlock (&clp->cl_lock); +seq_puts (m,"\n" ); +drop_client (clp);return 0 ;
Signature · 1 line
2.47.0
PATCH v5 4/6 nfsd: allocate new session-based DRC slots on demand.
In reply to NeilBrownIf a client ever uses the highest available slot for a given session, attempt to allocate more slots so there is room for the client to use them if wanted. GFP_NOWAIT is used so if there is not plenty of free memory, failure is expected - which is what we want. It also allows the allocation while holding a spinlock.
Each time we increase the number of slots by 20% (rounded up). This allows fairly quick growth while avoiding excessive over-shoot.
We would expect to stablise with around 10% more slots available than the client actually uses.
--- fs/nfsd/nfs4state.c | 37 ++++++++++++++++++++++++++++++++----- 1 file changed, 32 insertions(+), 5 deletions(-)
fs/nfsd/nfs4state.c +32 -5
@@ -4235,11 +4235,6 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, slot =xa_load (&session->se_slots, seq->slotid);dprintk ("%s: slotid %d\n" , __func__, seq->slotid); -/* We do not negotiate the number of slots yet, so set the -* maxslots to the session maxreqs which is used to encode -* sr_highest_slotid and the sr_target_slot id to maxslots */ - seq->maxslots = session->se_fchannel.maxreqs; -trace_nfsd_slot_seqid_sequence (clp, seq, slot); status =check_slot_seqid (seq->seqid, slot->sl_seqid, slot->sl_flags &NFSD4_SLOT_INUSE );
@@ -4289,6 +4284,38 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, cstate->session = session; cstate->clp = clp; +/* +* If the client ever uses the highest available slot, +* gently try to allocate another 20%. This allows +* fairly quick growth without grossly over-shooting what +* the client might use. +*/ +if (seq->slotid == session->se_fchannel.maxreqs -1 && + session->se_fchannel.maxreqs <NFSD_MAX_SLOTS_PER_SESSION ) { +int s = session->se_fchannel.maxreqs; +int cnt =DIV_ROUND_UP (s,5 ); + +do { +/* +* GFP_NOWAIT both allows allocation under a +* spinlock, and only succeeds if there is +* plenty of memory. +*/ + slot =kzalloc (slot_bytes (&session->se_fchannel), +GFP_NOWAIT ); +if (slot && + !xa_is_err (xa_store (&session->se_slots, s, slot, +GFP_NOWAIT ))) { + s +=1 ; + session->se_fchannel.maxreqs = s; + }else { +kfree (slot); + slot =NULL ; + } + }while (slot && --cnt >0 ); + } + seq->maxslots = session->se_fchannel.maxreqs; + out:switch (clp->cl_cb_state) {case NFSD4_CB_DOWN :
Signature · 1 line
2.47.0
PATCH v5 5/6 nfsd: add support for freeing unused session-DRC slots
In reply to NeilBrownReducing the number of slots in the session slot table requires confirmation from the client. This patch adds reduce_session_slots() which starts the process of getting confirmation, but never calls it. That will come in a later patch.
Before we can free a slot we need to confirm that the client won't try to use it again. This involves returning a lower cr_maxrequests in a SEQUENCE reply and then seeing a ca_maxrequests on the same slot which is not larger than we limit we are trying to impose. So for each slot we need to remember that we have sent a reduced cr_maxrequests.
To achieve this we introduce a concept of request "generations". Each time we decide to reduce cr_maxrequests we increment the generation number, and record this when we return the lower cr_maxrequests to the client. When a slot with the current generation reports a low ca_maxrequests, we commit to that level and free extra slots.
We use an 16 bit generation number (64 seems wasteful) and if it cycles we iterate all slots and reset the generation number to avoid false matches.
When we free a slot we store the seqid in the slot pointer so that it can be restored when we reactivate the slot. The RFC can be read as suggesting that the slot number could restart from one after a slot is retired and reactivated, but also suggests that retiring slots is not required. So when we reactive a slot we accept with the next seqid in sequence, or 1.
When decoding sa_highest_slotid into maxslots we need to add 1 - this matches how it is encoded for the reply.
se_dead is moved in struct nfsd4_session to remove a hole.
--- fs/nfsd/nfs4state.c | 94 ++++++++++++++++++++++++++++++++++++++++----- fs/nfsd/nfs4xdr.c | 5 ++- fs/nfsd/state.h | 6 ++- fs/nfsd/xdr4.h | 2 - 4 files changed, 92 insertions(+), 15 deletions(-)
fs/nfsd/nfs4state.c +84 -10
@@ -1910,17 +1910,69 @@ gen_sessionid(struct nfsd4_session *ses)#define NFSD_MIN_HDR_SEQ_SZ (24 + 12 + 44)static void -free_session_slots (struct nfsd4_session *ses) +free_session_slots (struct nfsd4_session *ses,int from) {int i; -for (i =0 ; i < ses->se_fchannel.maxreqs; i++) { +if (from >= ses->se_fchannel.maxreqs) +return ; + +for (i = from; i < ses->se_fchannel.maxreqs; i++) {struct nfsd4_slot *slot =xa_load (&ses->se_slots, i); -xa_erase (&ses->se_slots, i); +/* +* Save the seqid in case we reactivate this slot. +* This will never require a memory allocation so GFP +* flag is irrelevant +*/ +xa_store (&ses->se_slots, i,xa_mk_value (slot->sl_seqid),0 );free_svc_cred (&slot->sl_cred);kfree (slot); } + ses->se_fchannel.maxreqs = from; +if (ses->se_target_maxslots > from) + ses->se_target_maxslots = from; +} + +/** +* reduce_session_slots - reduce the target max-slots of a session if possible +* @ses: The session to affect +* @dec: how much to decrease the target by +* +* This interface can be used by a shrinker to reduce the target max-slots +* for a session so that some slots can eventually be freed. +* It uses spin_trylock() as it may be called in a context where another +* spinlock is held that has a dependency on client_lock. As shrinkers are +* best-effort, skiping a session is client_lock is already held has no +* great coast +* +* Return value: +* The number of slots that the target was reduced by. +*/ +static int __maybe_unused +reduce_session_slots (struct nfsd4_session *ses,int dec) +{ +struct nfsd_net *nn =net_generic (ses->se_client->net, + nfsd_net_id); +int ret =0 ; + +if (ses->se_target_maxslots <=1 ) +return ret; +if (!spin_trylock (&nn->client_lock)) +return ret; + ret =min (dec, ses->se_target_maxslots-1 ); + ses->se_target_maxslots -= ret; + ses->se_slot_gen +=1 ; +if (ses->se_slot_gen ==0 ) { +int i; + ses->se_slot_gen =1 ; +for (i =0 ; i < ses->se_fchannel.maxreqs; i++) { +struct nfsd4_slot *slot =xa_load (&ses->se_slots, i); + slot->sl_generation =0 ; + } + } +spin_unlock (&nn->client_lock); +return ret; }/*
@@ -1968,6 +2020,7 @@ static struct nfsd4_session *alloc_session(struct nfsd4_channel_attrs *fattrs, } fattrs->maxreqs = i;memcpy (&new->se_fchannel, fattrs,sizeof (struct nfsd4_channel_attrs )); + new->se_target_maxslots = i; new->se_cb_slot_avail = ~0U ; new->se_cb_highest_slot =min (battrs->maxreqs -1 ,NFSD_BC_SLOT_TABLE_SIZE -1 );
@@ -2081,7 +2134,7 @@ static void nfsd4_del_conns(struct nfsd4_session *s)static void __free_session (struct nfsd4_session *ses) { -free_session_slots (ses); +free_session_slots (ses,0 );xa_destroy (&ses->se_slots);kfree (ses); }
@@ -2684,6 +2737,9 @@ static int client_info_show(struct seq_file *m, void *v)seq_printf (m,"session slots:" );list_for_each_entry (ses, &clp->cl_sessions, se_perclnt)seq_printf (m," %u" , ses->se_fchannel.maxreqs); +seq_printf (m,"\nsession target slots:" ); +list_for_each_entry (ses, &clp->cl_sessions, se_perclnt) +seq_printf (m," %u" , ses->se_target_maxslots);spin_unlock (&clp->cl_lock);seq_puts (m,"\n" );
@@ -3674,10 +3730,10 @@ nfsd4_exchange_id_release(union nfsd4_op_u *u)kfree (exid->server_impl_name); } -static __be32 check_slot_seqid (u32 seqid,u32 slot_seqid,bool slot_inuse) +static __be32 check_slot_seqid (u32 seqid,u32 slot_seqid,u8 flags) {/* The slot is in use, and no response has been sent. */ -if (slot_inuse) { +if (flags &NFSD4_SLOT_INUSE ) {if (seqid == slot_seqid)return nfserr_jukebox;else
@@ -3686,6 +3742,8 @@ static __be32 check_slot_seqid(u32 seqid, u32 slot_seqid, bool slot_inuse)/* Note unsigned 32-bit arithmetic handles wraparound: */ if (likely (seqid == slot_seqid +1 ))return nfs_ok; +if ((flags &NFSD4_SLOT_REUSED ) && seqid ==1 ) +return nfs_ok;if (seqid == slot_seqid)return nfserr_replay_cache;return nfserr_seq_misordered;
@@ -4236,8 +4294,7 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,dprintk ("%s: slotid %d\n" , __func__, seq->slotid);trace_nfsd_slot_seqid_sequence (clp, seq, slot); - status =check_slot_seqid (seq->seqid, slot->sl_seqid, - slot->sl_flags &NFSD4_SLOT_INUSE ); + status =check_slot_seqid (seq->seqid, slot->sl_seqid, slot->sl_flags);if (status == nfserr_replay_cache) { status = nfserr_seq_misordered;if (!(slot->sl_flags &NFSD4_SLOT_INITIALIZED ))
@@ -4262,6 +4319,12 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,if (status) goto out_put_session; +if (session->se_target_maxslots < session->se_fchannel.maxreqs && + slot->sl_generation == session->se_slot_gen && + seq->maxslots <= session->se_target_maxslots) +/* Client acknowledged our reduce maxreqs */ +free_session_slots (session, session->se_target_maxslots); + buflen = (seq->cachethis) ? session->se_fchannel.maxresp_cached : session->se_fchannel.maxresp_sz;
@@ -4272,9 +4335,11 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,svc_reserve (rqstp, buflen); status = nfs_ok; -/* Success! bump slot seqid */ +/* Success! accept new slot seqid */ slot->sl_seqid = seq->seqid; + slot->sl_flags &= ~NFSD4_SLOT_REUSED ; slot->sl_flags |=NFSD4_SLOT_INUSE ; + slot->sl_generation = session->se_slot_gen;if (seq->cachethis) slot->sl_flags |=NFSD4_SLOT_CACHETHIS ;else
@@ -4291,9 +4356,11 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,* the client might use. */ if (seq->slotid == session->se_fchannel.maxreqs -1 && + session->se_target_maxslots >= session->se_fchannel.maxreqs && session->se_fchannel.maxreqs <NFSD_MAX_SLOTS_PER_SESSION ) {int s = session->se_fchannel.maxreqs;int cnt =DIV_ROUND_UP (s,5 ); +void *prev_slot; do { /*
@@ -4303,18 +4370,25 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,*/ slot =kzalloc (slot_bytes (&session->se_fchannel),GFP_NOWAIT ); + prev_slot =xa_load (&session->se_slots, s); +if (xa_is_value (prev_slot) && slot) { + slot->sl_seqid =xa_to_value (prev_slot); + slot->sl_flags |=NFSD4_SLOT_REUSED ; + }if (slot && !xa_is_err (xa_store (&session->se_slots, s, slot,GFP_NOWAIT ))) { s +=1 ; session->se_fchannel.maxreqs = s; + session->se_target_maxslots = s; }else {kfree (slot); slot =NULL ; } }while (slot && --cnt >0 ); } - seq->maxslots = session->se_fchannel.maxreqs; + seq->maxslots =max (session->se_target_maxslots, seq->maxslots); + seq->target_maxslots = session->se_target_maxslots; out:switch (clp->cl_cb_state) {
fs/nfsd/nfs4xdr.c +3 -2
@@ -1884,7 +1884,8 @@ nfsd4_decode_sequence(struct nfsd4_compoundargs *argp,return nfserr_bad_xdr; seq->seqid =be32_to_cpup (p++); seq->slotid =be32_to_cpup (p++); - seq->maxslots =be32_to_cpup (p++); +/* sa_highest_slotid counts from 0 but maxslots counts from 1 ... */ + seq->maxslots =be32_to_cpup (p++) +1 ; seq->cachethis =be32_to_cpup (p); seq->status_flags =0 ;
@@ -4968,7 +4969,7 @@ nfsd4_encode_sequence(struct nfsd4_compoundres *resp, __be32 nfserr,if (nfserr != nfs_ok)return nfserr;/* sr_target_highest_slotid */ - nfserr =nfsd4_encode_slotid4 (xdr, seq->maxslots -1 ); + nfserr =nfsd4_encode_slotid4 (xdr, seq->target_maxslots -1 );if (nfserr != nfs_ok)return nfserr;/* sr_status_flags */
fs/nfsd/state.h +5 -1
@@ -245,10 +245,12 @@ struct nfsd4_slot {struct svc_cred sl_cred;u32 sl_datalen;u16 sl_opcnt; +u16 sl_generation;#define NFSD4_SLOT_INUSE (1 << 0)#define NFSD4_SLOT_CACHETHIS (1 << 1)#define NFSD4_SLOT_INITIALIZED (1 << 2)#define NFSD4_SLOT_CACHED (1 << 3) +#define NFSD4_SLOT_REUSED (1 << 4)u8 sl_flags;char sl_data[]; };
@@ -321,7 +323,6 @@ struct nfsd4_session {u32 se_cb_slot_avail;/* bitmap of available slots */ u32 se_cb_highest_slot;/* highest slot client wants */ u32 se_cb_prog; -bool se_dead;struct list_head se_hash;/* hash by sessionid */ struct list_head se_perclnt;struct nfs4_client *se_client;
@@ -331,6 +332,9 @@ struct nfsd4_session {struct list_head se_conns;u32 se_cb_seq_nr[NFSD_BC_SLOT_TABLE_SIZE ];struct xarray se_slots;/* forward channel slots */ +u16 se_slot_gen; +bool se_dead; +u32 se_target_maxslots; };/* formatted contents of nfs4_sessionid */
fs/nfsd/xdr4.h +0 -2
@@ -576,9 +576,7 @@ struct nfsd4_sequence {u32 slotid;/* request/response */ u32 maxslots;/* request/response */ u32 cachethis;/* request */ -#if 0 u32 target_maxslots;/* response */ -#endif /* not yet */ u32 status_flags;/* response */ };
Signature · 1 line
2.47.0
Re: [PATCH 5/6] nfsd: add support for freeing unused session-DRC slots
Quoted 305 lines · NeilBrown
Reducing the number of slots in the session slot table requires confirmation from the client. This patch adds
reduce_session_slots()which starts the process of getting confirmation, but never calls it. That will come in a later patch.Before we can free a slot we need to confirm that the client won't try to use it again. This involves returning a lower
cr_maxrequestsin a SEQUENCE reply and then seeing aca_maxrequestson the same slot which is not larger than we limit we are trying to impose. So for each slot we need to remember that we have sent a reducedcr_maxrequests.To achieve this we introduce a concept of request "generations". Each time we decide to reduce
cr_maxrequestswe increment the generation number, and record this when we return the lowercr_maxrequeststo the client. When a slot with the current generation reports a lowca_maxrequests, we commit to that level and free extra slots.We use an 16 bit generation number (64 seems wasteful) and if it cycles we iterate all slots and reset the generation number to avoid false matches.
When we free a slot we store the seqid in the slot pointer so that it can be restored when we reactivate the slot. The RFC can be read as suggesting that the slot number could restart from one after a slot is retired and reactivated, but also suggests that retiring slots is not required. So when we reactive a slot we accept with the next seqid in sequence, or 1.
When decoding
sa_highest_slotidinto maxslots we need to add 1 - this matches how it is encoded for the reply.
se_deadis moved in structnfsd4_sessionto remove a hole.Reviewed-by Jeff Layton <[email protected]>Signed-off-by NeilBrown <[email protected]>--- fs/nfsd/nfs4state.c | 94 ++++++++++++++++++++++++++++++++++++++++----- fs/nfsd/nfs4xdr.c | 5 ++- fs/nfsd/state.h | 6 ++- fs/nfsd/xdr4.h | 2 - 4 files changed, 92 insertions(+), 15 deletions(-)fs/nfsd/nfs4state.c +84
-10@@ -1910,17 +1910,69 @@ gen_sessionid(struct nfsd4_session *ses)#define NFSD_MIN_HDR_SEQ_SZ (24 + 12 + 44)static void -free_session_slots (struct nfsd4_session *ses) +free_session_slots (struct nfsd4_session *ses,int from) {int i; -for (i =0 ; i < ses->se_fchannel.maxreqs; i++) { +if (from >= ses->se_fchannel.maxreqs) +return ; + +for (i = from; i < ses->se_fchannel.maxreqs; i++) {struct nfsd4_slot *slot =xa_load (&ses->se_slots, i); -xa_erase (&ses->se_slots, i); +/* +* Save the seqid in case we reactivate this slot. +* This will never require a memory allocation so GFP +* flag is irrelevant +*/ +xa_store (&ses->se_slots, i,xa_mk_value (slot->sl_seqid),0 );free_svc_cred (&slot->sl_cred);kfree (slot); } + ses->se_fchannel.maxreqs = from; +if (ses->se_target_maxslots > from) + ses->se_target_maxslots = from; +} + +/** +* reduce_session_slots - reduce the target max-slots of a session if possible +* @ses: The session to affect +* @dec: how much to decrease the target by +* +* This interface can be used by a shrinker to reduce the target max-slots +* for a session so that some slots can eventually be freed. +* It uses spin_trylock() as it may be called in a context where another +* spinlock is held that has a dependency on client_lock. As shrinkers are +* best-effort, skiping a session is client_lock is already held has no +* great coast +* +* Return value: +* The number of slots that the target was reduced by. +*/ +static int __maybe_unused +reduce_session_slots (struct nfsd4_session *ses,int dec) +{ +struct nfsd_net *nn =net_generic (ses->se_client->net, + nfsd_net_id); +int ret =0 ; + +if (ses->se_target_maxslots <=1 ) +return ret; +if (!spin_trylock (&nn->client_lock)) +return ret; + ret =min (dec, ses->se_target_maxslots-1 ); + ses->se_target_maxslots -= ret; + ses->se_slot_gen +=1 ; +if (ses->se_slot_gen ==0 ) { +int i; + ses->se_slot_gen =1 ; +for (i =0 ; i < ses->se_fchannel.maxreqs; i++) { +struct nfsd4_slot *slot =xa_load (&ses->se_slots, i); + slot->sl_generation =0 ; + } + } +spin_unlock (&nn->client_lock); +return ret; }/* @@ -1968,6 +2020,7 @@ static struct nfsd4_session *alloc_session(struct nfsd4_channel_attrs *fattrs, } fattrs->maxreqs = i;memcpy (&new->se_fchannel, fattrs,sizeof (struct nfsd4_channel_attrs )); + new->se_target_maxslots = i; new->se_cb_slot_avail = ~0U ; new->se_cb_highest_slot =min (battrs->maxreqs -1 ,NFSD_BC_SLOT_TABLE_SIZE -1 );@@ -2081,7 +2134,7 @@ static void nfsd4_del_conns(struct nfsd4_session *s)static void __free_session (struct nfsd4_session *ses) { -free_session_slots (ses); +free_session_slots (ses,0 );xa_destroy (&ses->se_slots);kfree (ses); }@@ -2684,6 +2737,9 @@ static int client_info_show(struct seq_file *m, void *v)seq_printf (m,"session slots:" );list_for_each_entry (ses, &clp->cl_sessions, se_perclnt)seq_printf (m," %u" , ses->se_fchannel.maxreqs); +seq_printf (m,"\nsession target slots:" ); +list_for_each_entry (ses, &clp->cl_sessions, se_perclnt) +seq_printf (m," %u" , ses->se_target_maxslots);spin_unlock (&clp->cl_lock);seq_puts (m,"\n" );@@ -3674,10 +3730,10 @@ nfsd4_exchange_id_release(union nfsd4_op_u *u)kfree (exid->server_impl_name); } -static __be32 check_slot_seqid (u32 seqid,u32 slot_seqid,bool slot_inuse) +static __be32 check_slot_seqid (u32 seqid,u32 slot_seqid,u8 flags) {/* The slot is in use, and no response has been sent. */ -if (slot_inuse) { +if (flags &NFSD4_SLOT_INUSE ) {if (seqid == slot_seqid)return nfserr_jukebox;else @@ -3686,6 +3742,8 @@ static __be32 check_slot_seqid(u32 seqid, u32 slot_seqid, bool slot_inuse)/* Note unsigned 32-bit arithmetic handles wraparound: */ if (likely (seqid == slot_seqid +1 ))return nfs_ok; +if ((flags &NFSD4_SLOT_REUSED ) && seqid ==1 ) +return nfs_ok;if (seqid == slot_seqid)return nfserr_replay_cache;return nfserr_seq_misordered;@@ -4236,8 +4294,7 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,dprintk ("%s: slotid %d\n" , __func__, seq->slotid);trace_nfsd_slot_seqid_sequence (clp, seq, slot); - status =check_slot_seqid (seq->seqid, slot->sl_seqid, - slot->sl_flags &NFSD4_SLOT_INUSE ); + status =check_slot_seqid (seq->seqid, slot->sl_seqid, slot->sl_flags);if (status == nfserr_replay_cache) { status = nfserr_seq_misordered;if (!(slot->sl_flags &NFSD4_SLOT_INITIALIZED ))@@ -4262,6 +4319,12 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,if (status) goto out_put_session; +if (session->se_target_maxslots < session->se_fchannel.maxreqs && + slot->sl_generation == session->se_slot_gen && + seq->maxslots <= session->se_target_maxslots) +/* Client acknowledged our reduce maxreqs */ +free_session_slots (session, session->se_target_maxslots); + buflen = (seq->cachethis) ? session->se_fchannel.maxresp_cached : session->se_fchannel.maxresp_sz;@@ -4272,9 +4335,11 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,svc_reserve (rqstp, buflen); status = nfs_ok; -/* Success! bump slot seqid */ +/* Success! accept new slot seqid */ slot->sl_seqid = seq->seqid; + slot->sl_flags &= ~NFSD4_SLOT_REUSED ; slot->sl_flags |=NFSD4_SLOT_INUSE ; + slot->sl_generation = session->se_slot_gen;if (seq->cachethis) slot->sl_flags |=NFSD4_SLOT_CACHETHIS ;else @@ -4291,9 +4356,11 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,* the client might use. */ if (seq->slotid == session->se_fchannel.maxreqs -1 && + session->se_target_maxslots >= session->se_fchannel.maxreqs && session->se_fchannel.maxreqs <NFSD_MAX_SLOTS_PER_SESSION ) {int s = session->se_fchannel.maxreqs;int cnt =DIV_ROUND_UP (s,5 ); +void *prev_slot; do { /*@@ -4303,18 +4370,25 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,*/ slot =kzalloc (slot_bytes (&session->se_fchannel),GFP_NOWAIT ); + prev_slot =xa_load (&session->se_slots, s); +if (xa_is_value (prev_slot) && slot) { + slot->sl_seqid =xa_to_value (prev_slot); + slot->sl_flags |=NFSD4_SLOT_REUSED ; + }if (slot && !xa_is_err (xa_store (&session->se_slots, s, slot,GFP_NOWAIT ))) { s +=1 ; session->se_fchannel.maxreqs = s; + session->se_target_maxslots = s; }else {kfree (slot); slot =NULL ; } }while (slot && --cnt >0 ); } - seq->maxslots = session->se_fchannel.maxreqs; + seq->maxslots =max (session->se_target_maxslots, seq->maxslots); + seq->target_maxslots = session->se_target_maxslots; out:switch (clp->cl_cb_state) {fs/nfsd/nfs4xdr.c +3
-2@@ -1884,7 +1884,8 @@ nfsd4_decode_sequence(struct nfsd4_compoundargs *argp,return nfserr_bad_xdr; seq->seqid =be32_to_cpup (p++); seq->slotid =be32_to_cpup (p++); - seq->maxslots =be32_to_cpup (p++); +/* sa_highest_slotid counts from 0 but maxslots counts from 1 ... */ + seq->maxslots =be32_to_cpup (p++) +1 ; seq->cachethis =be32_to_cpup (p); seq->status_flags =0 ;@@ -4968,7 +4969,7 @@ nfsd4_encode_sequence(struct nfsd4_compoundres *resp, __be32 nfserr,if (nfserr != nfs_ok)return nfserr;/* sr_target_highest_slotid */ - nfserr =nfsd4_encode_slotid4 (xdr, seq->maxslots -1 ); + nfserr =nfsd4_encode_slotid4 (xdr, seq->target_maxslots -1 );if (nfserr != nfs_ok)return nfserr;/* sr_status_flags */ fs/nfsd/state.h +5
-1@@ -245,10 +245,12 @@ struct nfsd4_slot {struct svc_cred sl_cred;u32 sl_datalen;u16 sl_opcnt; +u16 sl_generation;#define NFSD4_SLOT_INUSE (1 << 0)#define NFSD4_SLOT_CACHETHIS (1 << 1)#define NFSD4_SLOT_INITIALIZED (1 << 2)#define NFSD4_SLOT_CACHED (1 << 3) +#define NFSD4_SLOT_REUSED (1 << 4)u8 sl_flags;char sl_data[]; };@@ -321,7 +323,6 @@ struct nfsd4_session {u32 se_cb_slot_avail;/* bitmap of available slots */ u32 se_cb_highest_slot;/* highest slot client wants */ u32 se_cb_prog; -bool se_dead;struct list_head se_hash;/* hash by sessionid */ struct list_head se_perclnt;struct nfs4_client *se_client;@@ -331,6 +332,9 @@ struct nfsd4_session {struct list_head se_conns;u32 se_cb_seq_nr[NFSD_BC_SLOT_TABLE_SIZE ];struct xarray se_slots;/* forward channel slots */ +u16 se_slot_gen; +bool se_dead; +u32 se_target_maxslots; };/* formatted contents of nfs4_sessionid */ fs/nfsd/xdr4.h +0
-0
fs/nfsd/xdr4.h struct nfsd4_sequence { · line 582
Quoted 9 lines
fs/nfsd/xdr4.h +0
-2@@ -576,9 +576,7 @@ struct nfsd4_sequence {u32 slotid;/* request/response */ u32 maxslots;/* request/response */ u32 cachethis;/* request */ -#if 0 u32 target_maxslots;/* response */ -#endif /* not yet */ u32 status_flags;/* response */ };
Hi Neil -
I've found some misbehavior which I've bisected to this commit.
If disconnect injection is set up to break the connection every 25,000 RPCs or so, xfstests running on an NFSv4.1 mount will eventually stall after this commit is applied.
Network capture shows that the server eventually starts returning SEQ_MISORDERED because the client has forgotten an retired slot after a disconnect, and tries to use sequence number 1 for that slot with a new operation.
I've narrowed the issue down to nfs41_is_outlier_target_slotid() on the client. This function uses a bit of calculus to decide when to bump the slot table's generation number. In the failing case, it never bumps the generation, and that results in the client freeing slots it shouldn't. The server's reported { highest, target_highest } slot numbers don't appear to change correctly after the client has reconnected.
If I revert this hunk from 5/6:
@@ -4968,7 +4969,7 @@ nfsd4_encode_sequence(struct nfsd4_compoundres *resp, __be32 nfserr,
if (nfserr != nfs_ok)
return nfserr;
/* sr_target_highest_slotid */
- nfserr = nfsd4_encode_slotid4(xdr, seq->maxslots - 1);
+ nfserr = nfsd4_encode_slotid4(xdr, seq->target_maxslots - 1);
if (nfserr != nfs_ok)
return nfserr;
/* sr_status_flags */
the reproducer above runs to completion in the expected amount of time.
The high order bit here is whether I should drop these patches for v6.14, or whether you believe you can come up with a narrow solution during the early part of v6.14-rc that I can include in an -rc update for NFSD. I can't really tell if a significant amount of surgery will be necessary.
What do you think?
Signature · 1 line
Chuck Lever
Re: [PATCH 5/6] nfsd: add support for freeing unused session-DRC slots
Quoted 352 lines · Chuck Lever
Quoted 305 lines · NeilBrown
Reducing the number of slots in the session slot table requires confirmation from the client. This patch adds
reduce_session_slots()which starts the process of getting confirmation, but never calls it. That will come in a later patch.Before we can free a slot we need to confirm that the client won't try to use it again. This involves returning a lower
cr_maxrequestsin a SEQUENCE reply and then seeing aca_maxrequestson the same slot which is not larger than we limit we are trying to impose. So for each slot we need to remember that we have sent a reducedcr_maxrequests.To achieve this we introduce a concept of request "generations". Each time we decide to reduce
cr_maxrequestswe increment the generation number, and record this when we return the lowercr_maxrequeststo the client. When a slot with the current generation reports a lowca_maxrequests, we commit to that level and free extra slots.We use an 16 bit generation number (64 seems wasteful) and if it cycles we iterate all slots and reset the generation number to avoid false matches.
When we free a slot we store the seqid in the slot pointer so that it can be restored when we reactivate the slot. The RFC can be read as suggesting that the slot number could restart from one after a slot is retired and reactivated, but also suggests that retiring slots is not required. So when we reactive a slot we accept with the next seqid in sequence, or 1.
When decoding
sa_highest_slotidinto maxslots we need to add 1 - this matches how it is encoded for the reply.
se_deadis moved in structnfsd4_sessionto remove a hole.Reviewed-by Jeff Layton <[email protected]>Signed-off-by NeilBrown <[email protected]>--- fs/nfsd/nfs4state.c | 94 ++++++++++++++++++++++++++++++++++++++++----- fs/nfsd/nfs4xdr.c | 5 ++- fs/nfsd/state.h | 6 ++- fs/nfsd/xdr4.h | 2 - 4 files changed, 92 insertions(+), 15 deletions(-)fs/nfsd/nfs4state.c +84
-10@@ -1910,17 +1910,69 @@ gen_sessionid(struct nfsd4_session *ses)#define NFSD_MIN_HDR_SEQ_SZ (24 + 12 + 44)static void -free_session_slots (struct nfsd4_session *ses) +free_session_slots (struct nfsd4_session *ses,int from) {int i; -for (i =0 ; i < ses->se_fchannel.maxreqs; i++) { +if (from >= ses->se_fchannel.maxreqs) +return ; + +for (i = from; i < ses->se_fchannel.maxreqs; i++) {struct nfsd4_slot *slot =xa_load (&ses->se_slots, i); -xa_erase (&ses->se_slots, i); +/* +* Save the seqid in case we reactivate this slot. +* This will never require a memory allocation so GFP +* flag is irrelevant +*/ +xa_store (&ses->se_slots, i,xa_mk_value (slot->sl_seqid),0 );free_svc_cred (&slot->sl_cred);kfree (slot); } + ses->se_fchannel.maxreqs = from; +if (ses->se_target_maxslots > from) + ses->se_target_maxslots = from; +} + +/** +* reduce_session_slots - reduce the target max-slots of a session if possible +* @ses: The session to affect +* @dec: how much to decrease the target by +* +* This interface can be used by a shrinker to reduce the target max-slots +* for a session so that some slots can eventually be freed. +* It uses spin_trylock() as it may be called in a context where another +* spinlock is held that has a dependency on client_lock. As shrinkers are +* best-effort, skiping a session is client_lock is already held has no +* great coast +* +* Return value: +* The number of slots that the target was reduced by. +*/ +static int __maybe_unused +reduce_session_slots (struct nfsd4_session *ses,int dec) +{ +struct nfsd_net *nn =net_generic (ses->se_client->net, + nfsd_net_id); +int ret =0 ; + +if (ses->se_target_maxslots <=1 ) +return ret; +if (!spin_trylock (&nn->client_lock)) +return ret; + ret =min (dec, ses->se_target_maxslots-1 ); + ses->se_target_maxslots -= ret; + ses->se_slot_gen +=1 ; +if (ses->se_slot_gen ==0 ) { +int i; + ses->se_slot_gen =1 ; +for (i =0 ; i < ses->se_fchannel.maxreqs; i++) { +struct nfsd4_slot *slot =xa_load (&ses->se_slots, i); + slot->sl_generation =0 ; + } + } +spin_unlock (&nn->client_lock); +return ret; }/* @@ -1968,6 +2020,7 @@ static struct nfsd4_session *alloc_session(struct nfsd4_channel_attrs *fattrs, } fattrs->maxreqs = i;memcpy (&new->se_fchannel, fattrs,sizeof (struct nfsd4_channel_attrs )); + new->se_target_maxslots = i; new->se_cb_slot_avail = ~0U ; new->se_cb_highest_slot =min (battrs->maxreqs -1 ,NFSD_BC_SLOT_TABLE_SIZE -1 );@@ -2081,7 +2134,7 @@ static void nfsd4_del_conns(struct nfsd4_session *s)static void __free_session (struct nfsd4_session *ses) { -free_session_slots (ses); +free_session_slots (ses,0 );xa_destroy (&ses->se_slots);kfree (ses); }@@ -2684,6 +2737,9 @@ static int client_info_show(struct seq_file *m, void *v)seq_printf (m,"session slots:" );list_for_each_entry (ses, &clp->cl_sessions, se_perclnt)seq_printf (m," %u" , ses->se_fchannel.maxreqs); +seq_printf (m,"\nsession target slots:" ); +list_for_each_entry (ses, &clp->cl_sessions, se_perclnt) +seq_printf (m," %u" , ses->se_target_maxslots);spin_unlock (&clp->cl_lock);seq_puts (m,"\n" );@@ -3674,10 +3730,10 @@ nfsd4_exchange_id_release(union nfsd4_op_u *u)kfree (exid->server_impl_name); } -static __be32 check_slot_seqid (u32 seqid,u32 slot_seqid,bool slot_inuse) +static __be32 check_slot_seqid (u32 seqid,u32 slot_seqid,u8 flags) {/* The slot is in use, and no response has been sent. */ -if (slot_inuse) { +if (flags &NFSD4_SLOT_INUSE ) {if (seqid == slot_seqid)return nfserr_jukebox;else @@ -3686,6 +3742,8 @@ static __be32 check_slot_seqid(u32 seqid, u32 slot_seqid, bool slot_inuse)/* Note unsigned 32-bit arithmetic handles wraparound: */ if (likely (seqid == slot_seqid +1 ))return nfs_ok; +if ((flags &NFSD4_SLOT_REUSED ) && seqid ==1 ) +return nfs_ok;if (seqid == slot_seqid)return nfserr_replay_cache;return nfserr_seq_misordered;@@ -4236,8 +4294,7 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,dprintk ("%s: slotid %d\n" , __func__, seq->slotid);trace_nfsd_slot_seqid_sequence (clp, seq, slot); - status =check_slot_seqid (seq->seqid, slot->sl_seqid, - slot->sl_flags &NFSD4_SLOT_INUSE ); + status =check_slot_seqid (seq->seqid, slot->sl_seqid, slot->sl_flags);if (status == nfserr_replay_cache) { status = nfserr_seq_misordered;if (!(slot->sl_flags &NFSD4_SLOT_INITIALIZED ))@@ -4262,6 +4319,12 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,if (status) goto out_put_session; +if (session->se_target_maxslots < session->se_fchannel.maxreqs && + slot->sl_generation == session->se_slot_gen && + seq->maxslots <= session->se_target_maxslots) +/* Client acknowledged our reduce maxreqs */ +free_session_slots (session, session->se_target_maxslots); + buflen = (seq->cachethis) ? session->se_fchannel.maxresp_cached : session->se_fchannel.maxresp_sz;@@ -4272,9 +4335,11 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,svc_reserve (rqstp, buflen); status = nfs_ok; -/* Success! bump slot seqid */ +/* Success! accept new slot seqid */ slot->sl_seqid = seq->seqid; + slot->sl_flags &= ~NFSD4_SLOT_REUSED ; slot->sl_flags |=NFSD4_SLOT_INUSE ; + slot->sl_generation = session->se_slot_gen;if (seq->cachethis) slot->sl_flags |=NFSD4_SLOT_CACHETHIS ;else @@ -4291,9 +4356,11 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,* the client might use. */ if (seq->slotid == session->se_fchannel.maxreqs -1 && + session->se_target_maxslots >= session->se_fchannel.maxreqs && session->se_fchannel.maxreqs <NFSD_MAX_SLOTS_PER_SESSION ) {int s = session->se_fchannel.maxreqs;int cnt =DIV_ROUND_UP (s,5 ); +void *prev_slot; do { /*@@ -4303,18 +4370,25 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,*/ slot =kzalloc (slot_bytes (&session->se_fchannel),GFP_NOWAIT ); + prev_slot =xa_load (&session->se_slots, s); +if (xa_is_value (prev_slot) && slot) { + slot->sl_seqid =xa_to_value (prev_slot); + slot->sl_flags |=NFSD4_SLOT_REUSED ; + }if (slot && !xa_is_err (xa_store (&session->se_slots, s, slot,GFP_NOWAIT ))) { s +=1 ; session->se_fchannel.maxreqs = s; + session->se_target_maxslots = s; }else {kfree (slot); slot =NULL ; } }while (slot && --cnt >0 ); } - seq->maxslots = session->se_fchannel.maxreqs; + seq->maxslots =max (session->se_target_maxslots, seq->maxslots); + seq->target_maxslots = session->se_target_maxslots; out:switch (clp->cl_cb_state) {fs/nfsd/nfs4xdr.c +3
-2@@ -1884,7 +1884,8 @@ nfsd4_decode_sequence(struct nfsd4_compoundargs *argp,return nfserr_bad_xdr; seq->seqid =be32_to_cpup (p++); seq->slotid =be32_to_cpup (p++); - seq->maxslots =be32_to_cpup (p++); +/* sa_highest_slotid counts from 0 but maxslots counts from 1 ... */ + seq->maxslots =be32_to_cpup (p++) +1 ; seq->cachethis =be32_to_cpup (p); seq->status_flags =0 ;@@ -4968,7 +4969,7 @@ nfsd4_encode_sequence(struct nfsd4_compoundres *resp, __be32 nfserr,if (nfserr != nfs_ok)return nfserr;/* sr_target_highest_slotid */ - nfserr =nfsd4_encode_slotid4 (xdr, seq->maxslots -1 ); + nfserr =nfsd4_encode_slotid4 (xdr, seq->target_maxslots -1 );if (nfserr != nfs_ok)return nfserr;/* sr_status_flags */ fs/nfsd/state.h +5
-1@@ -245,10 +245,12 @@ struct nfsd4_slot {struct svc_cred sl_cred;u32 sl_datalen;u16 sl_opcnt; +u16 sl_generation;#define NFSD4_SLOT_INUSE (1 << 0)#define NFSD4_SLOT_CACHETHIS (1 << 1)#define NFSD4_SLOT_INITIALIZED (1 << 2)#define NFSD4_SLOT_CACHED (1 << 3) +#define NFSD4_SLOT_REUSED (1 << 4)u8 sl_flags;char sl_data[]; };@@ -321,7 +323,6 @@ struct nfsd4_session {u32 se_cb_slot_avail;/* bitmap of available slots */ u32 se_cb_highest_slot;/* highest slot client wants */ u32 se_cb_prog; -bool se_dead;struct list_head se_hash;/* hash by sessionid */ struct list_head se_perclnt;struct nfs4_client *se_client;@@ -331,6 +332,9 @@ struct nfsd4_session {struct list_head se_conns;u32 se_cb_seq_nr[NFSD_BC_SLOT_TABLE_SIZE ];struct xarray se_slots;/* forward channel slots */ +u16 se_slot_gen; +bool se_dead; +u32 se_target_maxslots; };/* formatted contents of nfs4_sessionid */ fs/nfsd/xdr4.h +0
-0fs/nfsd/xdr4.h struct nfsd4_sequence { · line 582
Quoted 9 lines
fs/nfsd/xdr4.h +0
-2@@ -576,9 +576,7 @@ struct nfsd4_sequence {u32 slotid;/* request/response */ u32 maxslots;/* request/response */ u32 cachethis;/* request */ -#if 0 u32 target_maxslots;/* response */ -#endif /* not yet */ u32 status_flags;/* response */ };Hi Neil -
I've found some misbehavior which I've bisected to this commit.
If disconnect injection is set up to break the connection every 25,000 RPCs or so, xfstests running on an NFSv4.1 mount will eventually stall after this commit is applied.
Network capture shows that the server eventually starts returning
SEQ_MISORDEREDbecause the client has forgotten an retired slot after a disconnect, and tries to use sequence number 1 for that slot with a new operation.I've narrowed the issue down to
nfs41_is_outlier_target_slotid()on the client. This function uses a bit of calculus to decide when to bump the slot table's generation number. In the failing case, it never bumps the generation, and that results in the client freeing slots it shouldn't. The server's reported { highest,target_highest} slot numbers don't appear to change correctly after the client has reconnected.If I revert this hunk from 5/6:
fs/nfsd/nfs4xdr.c +1
-1@@ -4968,7 +4969,7 @@ nfsd4_encode_sequence(struct nfsd4_compoundres *resp, __be32 nfserr, if (nfserr != nfs_ok)return nfserr;/* sr_target_highest_slotid */ - nfserr =nfsd4_encode_slotid4 (xdr, seq->maxslots -1 ); + nfserr =nfsd4_encode_slotid4 (xdr, seq->target_maxslots -1 );if (nfserr != nfs_ok)return nfserr;/*
sr_status_flags*/the reproducer above runs to completion in the expected amount of time.
Quoted 7 lines
The high order bit here is whether I should drop these patches for v6.14, or whether you believe you can come up with a narrow solution during the early part of v6.14-rc that I can include in an -rc update for NFSD. I can't really tell if a significant amount of surgery will be necessary.
What do you think?
I think I can fix it.
It sounds like it might be a client bug, or it might be a difference in interpretation of the spec, or it might be an ambiguity in the spec.
But I think I can fix it by setting NFSD_SLOT_REUSED in any slots beyond ->target_maxslots when I reduce target_maxslots. That makes it so that we accept either 1 or the next-in-sequence seqid. Doing that does open up a possible risk of a resend being accepted as a new request. As it is a new connection, resends are a real possibility. I'll have to ponder that a bit more and might need to be careful about retiring slots with a low seqid which have been used recently. Maybe I need to store the time when seqid 1 was used, and not return a slot until some minimum time after that timestamp.
I'll try to get you a patch before the end of next week.
Thanks,
NeilBrown
Re: [PATCH 5/6] nfsd: add support for freeing unused session-DRC slots
Quoted 359 lines · NeilBrown
Quoted 358 lines · Chuck Lever
Quoted 305 lines · NeilBrown
Reducing the number of slots in the session slot table requires confirmation from the client. This patch adds
reduce_session_slots()which starts the process of getting confirmation, but never calls it. That will come in a later patch.Before we can free a slot we need to confirm that the client won't try to use it again. This involves returning a lower
cr_maxrequestsin a SEQUENCE reply and then seeing aca_maxrequestson the same slot which is not larger than we limit we are trying to impose. So for each slot we need to remember that we have sent a reducedcr_maxrequests.To achieve this we introduce a concept of request "generations". Each time we decide to reduce
cr_maxrequestswe increment the generation number, and record this when we return the lowercr_maxrequeststo the client. When a slot with the current generation reports a lowca_maxrequests, we commit to that level and free extra slots.We use an 16 bit generation number (64 seems wasteful) and if it cycles we iterate all slots and reset the generation number to avoid false matches.
When we free a slot we store the seqid in the slot pointer so that it can be restored when we reactivate the slot. The RFC can be read as suggesting that the slot number could restart from one after a slot is retired and reactivated, but also suggests that retiring slots is not required. So when we reactive a slot we accept with the next seqid in sequence, or 1.
When decoding
sa_highest_slotidinto maxslots we need to add 1 - this matches how it is encoded for the reply.
se_deadis moved in structnfsd4_sessionto remove a hole.Reviewed-by Jeff Layton <[email protected]>Signed-off-by NeilBrown <[email protected]>--- fs/nfsd/nfs4state.c | 94 ++++++++++++++++++++++++++++++++++++++++----- fs/nfsd/nfs4xdr.c | 5 ++- fs/nfsd/state.h | 6 ++- fs/nfsd/xdr4.h | 2 - 4 files changed, 92 insertions(+), 15 deletions(-)fs/nfsd/nfs4state.c +84
-10@@ -1910,17 +1910,69 @@ gen_sessionid(struct nfsd4_session *ses)#define NFSD_MIN_HDR_SEQ_SZ (24 + 12 + 44)static void -free_session_slots (struct nfsd4_session *ses) +free_session_slots (struct nfsd4_session *ses,int from) {int i; -for (i =0 ; i < ses->se_fchannel.maxreqs; i++) { +if (from >= ses->se_fchannel.maxreqs) +return ; + +for (i = from; i < ses->se_fchannel.maxreqs; i++) {struct nfsd4_slot *slot =xa_load (&ses->se_slots, i); -xa_erase (&ses->se_slots, i); +/* +* Save the seqid in case we reactivate this slot. +* This will never require a memory allocation so GFP +* flag is irrelevant +*/ +xa_store (&ses->se_slots, i,xa_mk_value (slot->sl_seqid),0 );free_svc_cred (&slot->sl_cred);kfree (slot); } + ses->se_fchannel.maxreqs = from; +if (ses->se_target_maxslots > from) + ses->se_target_maxslots = from; +} + +/** +* reduce_session_slots - reduce the target max-slots of a session if possible +* @ses: The session to affect +* @dec: how much to decrease the target by +* +* This interface can be used by a shrinker to reduce the target max-slots +* for a session so that some slots can eventually be freed. +* It uses spin_trylock() as it may be called in a context where another +* spinlock is held that has a dependency on client_lock. As shrinkers are +* best-effort, skiping a session is client_lock is already held has no +* great coast +* +* Return value: +* The number of slots that the target was reduced by. +*/ +static int __maybe_unused +reduce_session_slots (struct nfsd4_session *ses,int dec) +{ +struct nfsd_net *nn =net_generic (ses->se_client->net, + nfsd_net_id); +int ret =0 ; + +if (ses->se_target_maxslots <=1 ) +return ret; +if (!spin_trylock (&nn->client_lock)) +return ret; + ret =min (dec, ses->se_target_maxslots-1 ); + ses->se_target_maxslots -= ret; + ses->se_slot_gen +=1 ; +if (ses->se_slot_gen ==0 ) { +int i; + ses->se_slot_gen =1 ; +for (i =0 ; i < ses->se_fchannel.maxreqs; i++) { +struct nfsd4_slot *slot =xa_load (&ses->se_slots, i); + slot->sl_generation =0 ; + } + } +spin_unlock (&nn->client_lock); +return ret; }/* @@ -1968,6 +2020,7 @@ static struct nfsd4_session *alloc_session(struct nfsd4_channel_attrs *fattrs, } fattrs->maxreqs = i;memcpy (&new->se_fchannel, fattrs,sizeof (struct nfsd4_channel_attrs )); + new->se_target_maxslots = i; new->se_cb_slot_avail = ~0U ; new->se_cb_highest_slot =min (battrs->maxreqs -1 ,NFSD_BC_SLOT_TABLE_SIZE -1 );@@ -2081,7 +2134,7 @@ static void nfsd4_del_conns(struct nfsd4_session *s)static void __free_session (struct nfsd4_session *ses) { -free_session_slots (ses); +free_session_slots (ses,0 );xa_destroy (&ses->se_slots);kfree (ses); }@@ -2684,6 +2737,9 @@ static int client_info_show(struct seq_file *m, void *v)seq_printf (m,"session slots:" );list_for_each_entry (ses, &clp->cl_sessions, se_perclnt)seq_printf (m," %u" , ses->se_fchannel.maxreqs); +seq_printf (m,"\nsession target slots:" ); +list_for_each_entry (ses, &clp->cl_sessions, se_perclnt) +seq_printf (m," %u" , ses->se_target_maxslots);spin_unlock (&clp->cl_lock);seq_puts (m,"\n" );@@ -3674,10 +3730,10 @@ nfsd4_exchange_id_release(union nfsd4_op_u *u)kfree (exid->server_impl_name); } -static __be32 check_slot_seqid (u32 seqid,u32 slot_seqid,bool slot_inuse) +static __be32 check_slot_seqid (u32 seqid,u32 slot_seqid,u8 flags) {/* The slot is in use, and no response has been sent. */ -if (slot_inuse) { +if (flags &NFSD4_SLOT_INUSE ) {if (seqid == slot_seqid)return nfserr_jukebox;else @@ -3686,6 +3742,8 @@ static __be32 check_slot_seqid(u32 seqid, u32 slot_seqid, bool slot_inuse)/* Note unsigned 32-bit arithmetic handles wraparound: */ if (likely (seqid == slot_seqid +1 ))return nfs_ok; +if ((flags &NFSD4_SLOT_REUSED ) && seqid ==1 ) +return nfs_ok;if (seqid == slot_seqid)return nfserr_replay_cache;return nfserr_seq_misordered;@@ -4236,8 +4294,7 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,dprintk ("%s: slotid %d\n" , __func__, seq->slotid);trace_nfsd_slot_seqid_sequence (clp, seq, slot); - status =check_slot_seqid (seq->seqid, slot->sl_seqid, - slot->sl_flags &NFSD4_SLOT_INUSE ); + status =check_slot_seqid (seq->seqid, slot->sl_seqid, slot->sl_flags);if (status == nfserr_replay_cache) { status = nfserr_seq_misordered;if (!(slot->sl_flags &NFSD4_SLOT_INITIALIZED ))@@ -4262,6 +4319,12 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,if (status) goto out_put_session; +if (session->se_target_maxslots < session->se_fchannel.maxreqs && + slot->sl_generation == session->se_slot_gen && + seq->maxslots <= session->se_target_maxslots) +/* Client acknowledged our reduce maxreqs */ +free_session_slots (session, session->se_target_maxslots); + buflen = (seq->cachethis) ? session->se_fchannel.maxresp_cached : session->se_fchannel.maxresp_sz;@@ -4272,9 +4335,11 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,svc_reserve (rqstp, buflen); status = nfs_ok; -/* Success! bump slot seqid */ +/* Success! accept new slot seqid */ slot->sl_seqid = seq->seqid; + slot->sl_flags &= ~NFSD4_SLOT_REUSED ; slot->sl_flags |=NFSD4_SLOT_INUSE ; + slot->sl_generation = session->se_slot_gen;if (seq->cachethis) slot->sl_flags |=NFSD4_SLOT_CACHETHIS ;else @@ -4291,9 +4356,11 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,* the client might use. */ if (seq->slotid == session->se_fchannel.maxreqs -1 && + session->se_target_maxslots >= session->se_fchannel.maxreqs && session->se_fchannel.maxreqs <NFSD_MAX_SLOTS_PER_SESSION ) {int s = session->se_fchannel.maxreqs;int cnt =DIV_ROUND_UP (s,5 ); +void *prev_slot; do { /*@@ -4303,18 +4370,25 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,*/ slot =kzalloc (slot_bytes (&session->se_fchannel),GFP_NOWAIT ); + prev_slot =xa_load (&session->se_slots, s); +if (xa_is_value (prev_slot) && slot) { + slot->sl_seqid =xa_to_value (prev_slot); + slot->sl_flags |=NFSD4_SLOT_REUSED ; + }if (slot && !xa_is_err (xa_store (&session->se_slots, s, slot,GFP_NOWAIT ))) { s +=1 ; session->se_fchannel.maxreqs = s; + session->se_target_maxslots = s; }else {kfree (slot); slot =NULL ; } }while (slot && --cnt >0 ); } - seq->maxslots = session->se_fchannel.maxreqs; + seq->maxslots =max (session->se_target_maxslots, seq->maxslots); + seq->target_maxslots = session->se_target_maxslots; out:switch (clp->cl_cb_state) {fs/nfsd/nfs4xdr.c +3
-2@@ -1884,7 +1884,8 @@ nfsd4_decode_sequence(struct nfsd4_compoundargs *argp,return nfserr_bad_xdr; seq->seqid =be32_to_cpup (p++); seq->slotid =be32_to_cpup (p++); - seq->maxslots =be32_to_cpup (p++); +/* sa_highest_slotid counts from 0 but maxslots counts from 1 ... */ + seq->maxslots =be32_to_cpup (p++) +1 ; seq->cachethis =be32_to_cpup (p); seq->status_flags =0 ;@@ -4968,7 +4969,7 @@ nfsd4_encode_sequence(struct nfsd4_compoundres *resp, __be32 nfserr,if (nfserr != nfs_ok)return nfserr;/* sr_target_highest_slotid */ - nfserr =nfsd4_encode_slotid4 (xdr, seq->maxslots -1 ); + nfserr =nfsd4_encode_slotid4 (xdr, seq->target_maxslots -1 );if (nfserr != nfs_ok)return nfserr;/* sr_status_flags */ fs/nfsd/state.h +5
-1@@ -245,10 +245,12 @@ struct nfsd4_slot {struct svc_cred sl_cred;u32 sl_datalen;u16 sl_opcnt; +u16 sl_generation;#define NFSD4_SLOT_INUSE (1 << 0)#define NFSD4_SLOT_CACHETHIS (1 << 1)#define NFSD4_SLOT_INITIALIZED (1 << 2)#define NFSD4_SLOT_CACHED (1 << 3) +#define NFSD4_SLOT_REUSED (1 << 4)u8 sl_flags;char sl_data[]; };@@ -321,7 +323,6 @@ struct nfsd4_session {u32 se_cb_slot_avail;/* bitmap of available slots */ u32 se_cb_highest_slot;/* highest slot client wants */ u32 se_cb_prog; -bool se_dead;struct list_head se_hash;/* hash by sessionid */ struct list_head se_perclnt;struct nfs4_client *se_client;@@ -331,6 +332,9 @@ struct nfsd4_session {struct list_head se_conns;u32 se_cb_seq_nr[NFSD_BC_SLOT_TABLE_SIZE ];struct xarray se_slots;/* forward channel slots */ +u16 se_slot_gen; +bool se_dead; +u32 se_target_maxslots; };/* formatted contents of nfs4_sessionid */ fs/nfsd/xdr4.h +0
-0fs/nfsd/xdr4.h struct nfsd4_sequence { · line 582
Quoted 9 lines
fs/nfsd/xdr4.h +0
-2@@ -576,9 +576,7 @@ struct nfsd4_sequence {u32 slotid;/* request/response */ u32 maxslots;/* request/response */ u32 cachethis;/* request */ -#if 0 u32 target_maxslots;/* response */ -#endif /* not yet */ u32 status_flags;/* response */ };Hi Neil -
I've found some misbehavior which I've bisected to this commit.
If disconnect injection is set up to break the connection every 25,000 RPCs or so, xfstests running on an NFSv4.1 mount will eventually stall after this commit is applied.
Network capture shows that the server eventually starts returning
SEQ_MISORDEREDbecause the client has forgotten an retired slot after a disconnect, and tries to use sequence number 1 for that slot with a new operation.I've narrowed the issue down to
nfs41_is_outlier_target_slotid()on the client. This function uses a bit of calculus to decide when to bump the slot table's generation number. In the failing case, it never bumps the generation, and that results in the client freeing slots it shouldn't. The server's reported { highest,target_highest} slot numbers don't appear to change correctly after the client has reconnected.If I revert this hunk from 5/6:
fs/nfsd/nfs4xdr.c +1
-1@@ -4968,7 +4969,7 @@ nfsd4_encode_sequence(struct nfsd4_compoundres *resp, __be32 nfserr, if (nfserr != nfs_ok)return nfserr;/* sr_target_highest_slotid */ - nfserr =nfsd4_encode_slotid4 (xdr, seq->maxslots -1 ); + nfserr =nfsd4_encode_slotid4 (xdr, seq->target_maxslots -1 );if (nfserr != nfs_ok)return nfserr;/* sr_status_flags */the reproducer above runs to completion in the expected amount of time.
The high order bit here is whether I should drop these patches for v6.14, or whether you believe you can come up with a narrow solution during the early part of v6.14-rc that I can include in an -rc update for NFSD. I can't really tell if a significant amount of surgery will be necessary.
Quoted 6 lines
Quoted 1 line
What do you think?
I think I can fix it.
It sounds like it might be a client bug, or it might be a difference in interpretation of the spec, or it might be an ambiguity in the spec.
Perhaps, but consider that the Linux NFS client interoperates successfully with v6.13 NFSD, Solaris's NFSv4.1 implementation, NetApp's implementation, and Hammerspace, and has done for many years.
I agree that the spec language isn't particularly transparent, but somehow all of these implementations have interpreted it the same way.
Quoted 9 lines
But I think I can fix it by setting
NFSD_SLOT_REUSEDin any slots beyond ->target_maxslots when I reducetarget_maxslots. That makes it so that we accept either 1 or the next-in-sequence seqid. Doing that does open up a possible risk of a resend being accepted as a new request. As it is a new connection, resends are a real possibility. I'll have to ponder that a bit more and might need to be careful about retiring slots with a low seqid which have been used recently. Maybe I need to store the time when seqid 1 was used, and not return a slot until some minimum time after that timestamp.
I gathered additional information using a trace_printk() in the client's decode_sequence() function.
In the working case, after reconnect, the server SEQUENCE response has
sr_highest_slotid = 18
sr_target_highest_slot = 18
IIUC 18 is the slot number of the highest slot currently in use.
In the stalling case, after reconnect, the server SEQUENCE response has
sr_highest_slotid = 23
sr_target_highest_slot = 63
The client heuristics decide to free slots 23 and higher, even though one or more of those slots have been active and have sequence numbers larger than 1.
I can provide more detailed reproducer instructions if you wish.
Quoted 1 line
I'll try to get you a patch before the end of next week.
Thanks!
Signature · 1 line
Chuck Lever
Re: [PATCH 5/6] nfsd: add support for freeing unused session-DRC slots
In reply to Chuck LeverQuoted 318 lines · Chuck Lever
Quoted 305 lines · NeilBrown
Reducing the number of slots in the session slot table requires confirmation from the client. This patch adds
reduce_session_slots()which starts the process of getting confirmation, but never calls it. That will come in a later patch.Before we can free a slot we need to confirm that the client won't try to use it again. This involves returning a lower
cr_maxrequestsin a SEQUENCE reply and then seeing aca_maxrequestson the same slot which is not larger than we limit we are trying to impose. So for each slot we need to remember that we have sent a reducedcr_maxrequests.To achieve this we introduce a concept of request "generations". Each time we decide to reduce
cr_maxrequestswe increment the generation number, and record this when we return the lowercr_maxrequeststo the client. When a slot with the current generation reports a lowca_maxrequests, we commit to that level and free extra slots.We use an 16 bit generation number (64 seems wasteful) and if it cycles we iterate all slots and reset the generation number to avoid false matches.
When we free a slot we store the seqid in the slot pointer so that it can be restored when we reactivate the slot. The RFC can be read as suggesting that the slot number could restart from one after a slot is retired and reactivated, but also suggests that retiring slots is not required. So when we reactive a slot we accept with the next seqid in sequence, or 1.
When decoding
sa_highest_slotidinto maxslots we need to add 1 - this matches how it is encoded for the reply.
se_deadis moved in structnfsd4_sessionto remove a hole.Reviewed-by Jeff Layton <[email protected]>Signed-off-by NeilBrown <[email protected]>--- fs/nfsd/nfs4state.c | 94 ++++++++++++++++++++++++++++++++++++++++----- fs/nfsd/nfs4xdr.c | 5 ++- fs/nfsd/state.h | 6 ++- fs/nfsd/xdr4.h | 2 - 4 files changed, 92 insertions(+), 15 deletions(-)fs/nfsd/nfs4state.c +84
-10@@ -1910,17 +1910,69 @@ gen_sessionid(struct nfsd4_session *ses)#define NFSD_MIN_HDR_SEQ_SZ (24 + 12 + 44)static void -free_session_slots (struct nfsd4_session *ses) +free_session_slots (struct nfsd4_session *ses,int from) {int i; -for (i =0 ; i < ses->se_fchannel.maxreqs; i++) { +if (from >= ses->se_fchannel.maxreqs) +return ; + +for (i = from; i < ses->se_fchannel.maxreqs; i++) {struct nfsd4_slot *slot =xa_load (&ses->se_slots, i); -xa_erase (&ses->se_slots, i); +/* +* Save the seqid in case we reactivate this slot. +* This will never require a memory allocation so GFP +* flag is irrelevant +*/ +xa_store (&ses->se_slots, i,xa_mk_value (slot->sl_seqid),0 );free_svc_cred (&slot->sl_cred);kfree (slot); } + ses->se_fchannel.maxreqs = from; +if (ses->se_target_maxslots > from) + ses->se_target_maxslots = from; +} + +/** +* reduce_session_slots - reduce the target max-slots of a session if possible +* @ses: The session to affect +* @dec: how much to decrease the target by +* +* This interface can be used by a shrinker to reduce the target max-slots +* for a session so that some slots can eventually be freed. +* It uses spin_trylock() as it may be called in a context where another +* spinlock is held that has a dependency on client_lock. As shrinkers are +* best-effort, skiping a session is client_lock is already held has no +* great coast +* +* Return value: +* The number of slots that the target was reduced by. +*/ +static int __maybe_unused +reduce_session_slots (struct nfsd4_session *ses,int dec) +{ +struct nfsd_net *nn =net_generic (ses->se_client->net, + nfsd_net_id); +int ret =0 ; + +if (ses->se_target_maxslots <=1 ) +return ret; +if (!spin_trylock (&nn->client_lock)) +return ret; + ret =min (dec, ses->se_target_maxslots-1 ); + ses->se_target_maxslots -= ret; + ses->se_slot_gen +=1 ; +if (ses->se_slot_gen ==0 ) { +int i; + ses->se_slot_gen =1 ; +for (i =0 ; i < ses->se_fchannel.maxreqs; i++) { +struct nfsd4_slot *slot =xa_load (&ses->se_slots, i); + slot->sl_generation =0 ; + } + } +spin_unlock (&nn->client_lock); +return ret; }/* @@ -1968,6 +2020,7 @@ static struct nfsd4_session *alloc_session(struct nfsd4_channel_attrs *fattrs, } fattrs->maxreqs = i;memcpy (&new->se_fchannel, fattrs,sizeof (struct nfsd4_channel_attrs )); + new->se_target_maxslots = i; new->se_cb_slot_avail = ~0U ; new->se_cb_highest_slot =min (battrs->maxreqs -1 ,NFSD_BC_SLOT_TABLE_SIZE -1 );@@ -2081,7 +2134,7 @@ static void nfsd4_del_conns(struct nfsd4_session *s)static void __free_session (struct nfsd4_session *ses) { -free_session_slots (ses); +free_session_slots (ses,0 );xa_destroy (&ses->se_slots);kfree (ses); }@@ -2684,6 +2737,9 @@ static int client_info_show(struct seq_file *m, void *v)seq_printf (m,"session slots:" );list_for_each_entry (ses, &clp->cl_sessions, se_perclnt)seq_printf (m," %u" , ses->se_fchannel.maxreqs); +seq_printf (m,"\nsession target slots:" ); +list_for_each_entry (ses, &clp->cl_sessions, se_perclnt) +seq_printf (m," %u" , ses->se_target_maxslots);spin_unlock (&clp->cl_lock);seq_puts (m,"\n" );@@ -3674,10 +3730,10 @@ nfsd4_exchange_id_release(union nfsd4_op_u *u)kfree (exid->server_impl_name); } -static __be32 check_slot_seqid (u32 seqid,u32 slot_seqid,bool slot_inuse) +static __be32 check_slot_seqid (u32 seqid,u32 slot_seqid,u8 flags) {/* The slot is in use, and no response has been sent. */ -if (slot_inuse) { +if (flags &NFSD4_SLOT_INUSE ) {if (seqid == slot_seqid)return nfserr_jukebox;else @@ -3686,6 +3742,8 @@ static __be32 check_slot_seqid(u32 seqid, u32 slot_seqid, bool slot_inuse)/* Note unsigned 32-bit arithmetic handles wraparound: */ if (likely (seqid == slot_seqid +1 ))return nfs_ok; +if ((flags &NFSD4_SLOT_REUSED ) && seqid ==1 ) +return nfs_ok;if (seqid == slot_seqid)return nfserr_replay_cache;return nfserr_seq_misordered;@@ -4236,8 +4294,7 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,dprintk ("%s: slotid %d\n" , __func__, seq->slotid);trace_nfsd_slot_seqid_sequence (clp, seq, slot); - status =check_slot_seqid (seq->seqid, slot->sl_seqid, - slot->sl_flags &NFSD4_SLOT_INUSE ); + status =check_slot_seqid (seq->seqid, slot->sl_seqid, slot->sl_flags);if (status == nfserr_replay_cache) { status = nfserr_seq_misordered;if (!(slot->sl_flags &NFSD4_SLOT_INITIALIZED ))@@ -4262,6 +4319,12 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,if (status) goto out_put_session; +if (session->se_target_maxslots < session->se_fchannel.maxreqs && + slot->sl_generation == session->se_slot_gen && + seq->maxslots <= session->se_target_maxslots) +/* Client acknowledged our reduce maxreqs */ +free_session_slots (session, session->se_target_maxslots); + buflen = (seq->cachethis) ? session->se_fchannel.maxresp_cached : session->se_fchannel.maxresp_sz;@@ -4272,9 +4335,11 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,svc_reserve (rqstp, buflen); status = nfs_ok; -/* Success! bump slot seqid */ +/* Success! accept new slot seqid */ slot->sl_seqid = seq->seqid; + slot->sl_flags &= ~NFSD4_SLOT_REUSED ; slot->sl_flags |=NFSD4_SLOT_INUSE ; + slot->sl_generation = session->se_slot_gen;if (seq->cachethis) slot->sl_flags |=NFSD4_SLOT_CACHETHIS ;else @@ -4291,9 +4356,11 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,* the client might use. */ if (seq->slotid == session->se_fchannel.maxreqs -1 && + session->se_target_maxslots >= session->se_fchannel.maxreqs && session->se_fchannel.maxreqs <NFSD_MAX_SLOTS_PER_SESSION ) {int s = session->se_fchannel.maxreqs;int cnt =DIV_ROUND_UP (s,5 ); +void *prev_slot; do { /*@@ -4303,18 +4370,25 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,*/ slot =kzalloc (slot_bytes (&session->se_fchannel),GFP_NOWAIT ); + prev_slot =xa_load (&session->se_slots, s); +if (xa_is_value (prev_slot) && slot) { + slot->sl_seqid =xa_to_value (prev_slot); + slot->sl_flags |=NFSD4_SLOT_REUSED ; + }if (slot && !xa_is_err (xa_store (&session->se_slots, s, slot,GFP_NOWAIT ))) { s +=1 ; session->se_fchannel.maxreqs = s; + session->se_target_maxslots = s; }else {kfree (slot); slot =NULL ; } }while (slot && --cnt >0 ); } - seq->maxslots = session->se_fchannel.maxreqs; + seq->maxslots =max (session->se_target_maxslots, seq->maxslots); + seq->target_maxslots = session->se_target_maxslots; out:switch (clp->cl_cb_state) {fs/nfsd/nfs4xdr.c +3
-2@@ -1884,7 +1884,8 @@ nfsd4_decode_sequence(struct nfsd4_compoundargs *argp,return nfserr_bad_xdr; seq->seqid =be32_to_cpup (p++); seq->slotid =be32_to_cpup (p++); - seq->maxslots =be32_to_cpup (p++); +/* sa_highest_slotid counts from 0 but maxslots counts from 1 ... */ + seq->maxslots =be32_to_cpup (p++) +1 ; seq->cachethis =be32_to_cpup (p); seq->status_flags =0 ;@@ -4968,7 +4969,7 @@ nfsd4_encode_sequence(struct nfsd4_compoundres *resp, __be32 nfserr,if (nfserr != nfs_ok)return nfserr;/* sr_target_highest_slotid */ - nfserr =nfsd4_encode_slotid4 (xdr, seq->maxslots -1 ); + nfserr =nfsd4_encode_slotid4 (xdr, seq->target_maxslots -1 );if (nfserr != nfs_ok)return nfserr;/* sr_status_flags */ fs/nfsd/state.h +5
-1@@ -245,10 +245,12 @@ struct nfsd4_slot {struct svc_cred sl_cred;u32 sl_datalen;u16 sl_opcnt; +u16 sl_generation;#define NFSD4_SLOT_INUSE (1 << 0)#define NFSD4_SLOT_CACHETHIS (1 << 1)#define NFSD4_SLOT_INITIALIZED (1 << 2)#define NFSD4_SLOT_CACHED (1 << 3) +#define NFSD4_SLOT_REUSED (1 << 4)u8 sl_flags;char sl_data[]; };@@ -321,7 +323,6 @@ struct nfsd4_session {u32 se_cb_slot_avail;/* bitmap of available slots */ u32 se_cb_highest_slot;/* highest slot client wants */ u32 se_cb_prog; -bool se_dead;struct list_head se_hash;/* hash by sessionid */ struct list_head se_perclnt;struct nfs4_client *se_client;@@ -331,6 +332,9 @@ struct nfsd4_session {struct list_head se_conns;u32 se_cb_seq_nr[NFSD_BC_SLOT_TABLE_SIZE ];struct xarray se_slots;/* forward channel slots */ +u16 se_slot_gen; +bool se_dead; +u32 se_target_maxslots; };/* formatted contents of nfs4_sessionid */ fs/nfsd/xdr4.h +0
-0fs/nfsd/xdr4.h struct nfsd4_sequence { · line 582
Quoted 9 lines
fs/nfsd/xdr4.h +0
-2@@ -576,9 +576,7 @@ struct nfsd4_sequence {u32 slotid;/* request/response */ u32 maxslots;/* request/response */ u32 cachethis;/* request */ -#if 0 u32 target_maxslots;/* response */ -#endif /* not yet */ u32 status_flags;/* response */ };Hi Neil -
Quoted 1 line
I've found some misbehavior which I've bisected to this commit.
Hi Chuck,
could you please confirm that it really was this commit that you bisected to? Not the next one?
Because this commit never reduces ->se_target_maxslots, so the patch which you say removed the symptom should be a no-op.
Even if it was the next commit I'm struggling to pin down the problem. Here is my current analysis - partly to ensure I can present it clearly.
The evidence suggests that the client has retired a slot that the server hasn't. This happens when nfs41_set_server_slotid_locked() calls nfsd4_shrink_slot_table(), and nothing will happen if any slots before the new limit are still in use. If the server reduces its idea of the target when the client isn't even using that many, the slots can be freed immediately that the client gets a reply indicating the new highest_slot number from the server.
The server will not free these slots immediately but will wait to get a confirmation from the client that it has accepted the new limit. But, importantly, the server will not increase the limit that it sends to the client until after it has has a chance to free the retired slots. If the server doesn't increase the limit, then the client won't try to use the retired slots...
Do you still have the network trace which chows the error? Would I be able to look at it?
Thanks,
NeilBrown
Quoted 45 lines
If disconnect injection is set up to break the connection every 25,000 RPCs or so, xfstests running on an NFSv4.1 mount will eventually stall after this commit is applied.
Network capture shows that the server eventually starts returning
SEQ_MISORDEREDbecause the client has forgotten an retired slot after a disconnect, and tries to use sequence number 1 for that slot with a new operation.I've narrowed the issue down to
nfs41_is_outlier_target_slotid()on the client. This function uses a bit of calculus to decide when to bump the slot table's generation number. In the failing case, it never bumps the generation, and that results in the client freeing slots it shouldn't. The server's reported { highest,target_highest} slot numbers don't appear to change correctly after the client has reconnected.If I revert this hunk from 5/6:
fs/nfsd/nfs4xdr.c +1
-1@@ -4968,7 +4969,7 @@ nfsd4_encode_sequence(struct nfsd4_compoundres *resp, __be32 nfserr, if (nfserr != nfs_ok)return nfserr;/* sr_target_highest_slotid */ - nfserr =nfsd4_encode_slotid4 (xdr, seq->maxslots -1 ); + nfserr =nfsd4_encode_slotid4 (xdr, seq->target_maxslots -1 );if (nfserr != nfs_ok)return nfserr;/*
sr_status_flags*/the reproducer above runs to completion in the expected amount of time.
The high order bit here is whether I should drop these patches for v6.14, or whether you believe you can come up with a narrow solution during the early part of v6.14-rc that I can include in an -rc update for NFSD. I can't really tell if a significant amount of surgery will be necessary.
What do you think?
Signature · 1 line
Chuck Lever
Re: [PATCH 5/6] nfsd: add support for freeing unused session-DRC slots
Quoted 319 lines · NeilBrown
Quoted 318 lines · Chuck Lever
Quoted 305 lines · NeilBrown
Reducing the number of slots in the session slot table requires confirmation from the client. This patch adds
reduce_session_slots()which starts the process of getting confirmation, but never calls it. That will come in a later patch.Before we can free a slot we need to confirm that the client won't try to use it again. This involves returning a lower
cr_maxrequestsin a SEQUENCE reply and then seeing aca_maxrequestson the same slot which is not larger than we limit we are trying to impose. So for each slot we need to remember that we have sent a reducedcr_maxrequests.To achieve this we introduce a concept of request "generations". Each time we decide to reduce
cr_maxrequestswe increment the generation number, and record this when we return the lowercr_maxrequeststo the client. When a slot with the current generation reports a lowca_maxrequests, we commit to that level and free extra slots.We use an 16 bit generation number (64 seems wasteful) and if it cycles we iterate all slots and reset the generation number to avoid false matches.
When we free a slot we store the seqid in the slot pointer so that it can be restored when we reactivate the slot. The RFC can be read as suggesting that the slot number could restart from one after a slot is retired and reactivated, but also suggests that retiring slots is not required. So when we reactive a slot we accept with the next seqid in sequence, or 1.
When decoding
sa_highest_slotidinto maxslots we need to add 1 - this matches how it is encoded for the reply.
se_deadis moved in structnfsd4_sessionto remove a hole.Reviewed-by Jeff Layton <[email protected]>Signed-off-by NeilBrown <[email protected]>--- fs/nfsd/nfs4state.c | 94 ++++++++++++++++++++++++++++++++++++++++----- fs/nfsd/nfs4xdr.c | 5 ++- fs/nfsd/state.h | 6 ++- fs/nfsd/xdr4.h | 2 - 4 files changed, 92 insertions(+), 15 deletions(-)fs/nfsd/nfs4state.c +84
-10@@ -1910,17 +1910,69 @@ gen_sessionid(struct nfsd4_session *ses)#define NFSD_MIN_HDR_SEQ_SZ (24 + 12 + 44)static void -free_session_slots (struct nfsd4_session *ses) +free_session_slots (struct nfsd4_session *ses,int from) {int i; -for (i =0 ; i < ses->se_fchannel.maxreqs; i++) { +if (from >= ses->se_fchannel.maxreqs) +return ; + +for (i = from; i < ses->se_fchannel.maxreqs; i++) {struct nfsd4_slot *slot =xa_load (&ses->se_slots, i); -xa_erase (&ses->se_slots, i); +/* +* Save the seqid in case we reactivate this slot. +* This will never require a memory allocation so GFP +* flag is irrelevant +*/ +xa_store (&ses->se_slots, i,xa_mk_value (slot->sl_seqid),0 );free_svc_cred (&slot->sl_cred);kfree (slot); } + ses->se_fchannel.maxreqs = from; +if (ses->se_target_maxslots > from) + ses->se_target_maxslots = from; +} + +/** +* reduce_session_slots - reduce the target max-slots of a session if possible +* @ses: The session to affect +* @dec: how much to decrease the target by +* +* This interface can be used by a shrinker to reduce the target max-slots +* for a session so that some slots can eventually be freed. +* It uses spin_trylock() as it may be called in a context where another +* spinlock is held that has a dependency on client_lock. As shrinkers are +* best-effort, skiping a session is client_lock is already held has no +* great coast +* +* Return value: +* The number of slots that the target was reduced by. +*/ +static int __maybe_unused +reduce_session_slots (struct nfsd4_session *ses,int dec) +{ +struct nfsd_net *nn =net_generic (ses->se_client->net, + nfsd_net_id); +int ret =0 ; + +if (ses->se_target_maxslots <=1 ) +return ret; +if (!spin_trylock (&nn->client_lock)) +return ret; + ret =min (dec, ses->se_target_maxslots-1 ); + ses->se_target_maxslots -= ret; + ses->se_slot_gen +=1 ; +if (ses->se_slot_gen ==0 ) { +int i; + ses->se_slot_gen =1 ; +for (i =0 ; i < ses->se_fchannel.maxreqs; i++) { +struct nfsd4_slot *slot =xa_load (&ses->se_slots, i); + slot->sl_generation =0 ; + } + } +spin_unlock (&nn->client_lock); +return ret; }/* @@ -1968,6 +2020,7 @@ static struct nfsd4_session *alloc_session(struct nfsd4_channel_attrs *fattrs, } fattrs->maxreqs = i;memcpy (&new->se_fchannel, fattrs,sizeof (struct nfsd4_channel_attrs )); + new->se_target_maxslots = i; new->se_cb_slot_avail = ~0U ; new->se_cb_highest_slot =min (battrs->maxreqs -1 ,NFSD_BC_SLOT_TABLE_SIZE -1 );@@ -2081,7 +2134,7 @@ static void nfsd4_del_conns(struct nfsd4_session *s)static void __free_session (struct nfsd4_session *ses) { -free_session_slots (ses); +free_session_slots (ses,0 );xa_destroy (&ses->se_slots);kfree (ses); }@@ -2684,6 +2737,9 @@ static int client_info_show(struct seq_file *m, void *v)seq_printf (m,"session slots:" );list_for_each_entry (ses, &clp->cl_sessions, se_perclnt)seq_printf (m," %u" , ses->se_fchannel.maxreqs); +seq_printf (m,"\nsession target slots:" ); +list_for_each_entry (ses, &clp->cl_sessions, se_perclnt) +seq_printf (m," %u" , ses->se_target_maxslots);spin_unlock (&clp->cl_lock);seq_puts (m,"\n" );@@ -3674,10 +3730,10 @@ nfsd4_exchange_id_release(union nfsd4_op_u *u)kfree (exid->server_impl_name); } -static __be32 check_slot_seqid (u32 seqid,u32 slot_seqid,bool slot_inuse) +static __be32 check_slot_seqid (u32 seqid,u32 slot_seqid,u8 flags) {/* The slot is in use, and no response has been sent. */ -if (slot_inuse) { +if (flags &NFSD4_SLOT_INUSE ) {if (seqid == slot_seqid)return nfserr_jukebox;else @@ -3686,6 +3742,8 @@ static __be32 check_slot_seqid(u32 seqid, u32 slot_seqid, bool slot_inuse)/* Note unsigned 32-bit arithmetic handles wraparound: */ if (likely (seqid == slot_seqid +1 ))return nfs_ok; +if ((flags &NFSD4_SLOT_REUSED ) && seqid ==1 ) +return nfs_ok;if (seqid == slot_seqid)return nfserr_replay_cache;return nfserr_seq_misordered;@@ -4236,8 +4294,7 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,dprintk ("%s: slotid %d\n" , __func__, seq->slotid);trace_nfsd_slot_seqid_sequence (clp, seq, slot); - status =check_slot_seqid (seq->seqid, slot->sl_seqid, - slot->sl_flags &NFSD4_SLOT_INUSE ); + status =check_slot_seqid (seq->seqid, slot->sl_seqid, slot->sl_flags);if (status == nfserr_replay_cache) { status = nfserr_seq_misordered;if (!(slot->sl_flags &NFSD4_SLOT_INITIALIZED ))@@ -4262,6 +4319,12 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,if (status) goto out_put_session; +if (session->se_target_maxslots < session->se_fchannel.maxreqs && + slot->sl_generation == session->se_slot_gen && + seq->maxslots <= session->se_target_maxslots) +/* Client acknowledged our reduce maxreqs */ +free_session_slots (session, session->se_target_maxslots); + buflen = (seq->cachethis) ? session->se_fchannel.maxresp_cached : session->se_fchannel.maxresp_sz;@@ -4272,9 +4335,11 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,svc_reserve (rqstp, buflen); status = nfs_ok; -/* Success! bump slot seqid */ +/* Success! accept new slot seqid */ slot->sl_seqid = seq->seqid; + slot->sl_flags &= ~NFSD4_SLOT_REUSED ; slot->sl_flags |=NFSD4_SLOT_INUSE ; + slot->sl_generation = session->se_slot_gen;if (seq->cachethis) slot->sl_flags |=NFSD4_SLOT_CACHETHIS ;else @@ -4291,9 +4356,11 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,* the client might use. */ if (seq->slotid == session->se_fchannel.maxreqs -1 && + session->se_target_maxslots >= session->se_fchannel.maxreqs && session->se_fchannel.maxreqs <NFSD_MAX_SLOTS_PER_SESSION ) {int s = session->se_fchannel.maxreqs;int cnt =DIV_ROUND_UP (s,5 ); +void *prev_slot; do { /*@@ -4303,18 +4370,25 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,*/ slot =kzalloc (slot_bytes (&session->se_fchannel),GFP_NOWAIT ); + prev_slot =xa_load (&session->se_slots, s); +if (xa_is_value (prev_slot) && slot) { + slot->sl_seqid =xa_to_value (prev_slot); + slot->sl_flags |=NFSD4_SLOT_REUSED ; + }if (slot && !xa_is_err (xa_store (&session->se_slots, s, slot,GFP_NOWAIT ))) { s +=1 ; session->se_fchannel.maxreqs = s; + session->se_target_maxslots = s; }else {kfree (slot); slot =NULL ; } }while (slot && --cnt >0 ); } - seq->maxslots = session->se_fchannel.maxreqs; + seq->maxslots =max (session->se_target_maxslots, seq->maxslots); + seq->target_maxslots = session->se_target_maxslots; out:switch (clp->cl_cb_state) {fs/nfsd/nfs4xdr.c +3
-2@@ -1884,7 +1884,8 @@ nfsd4_decode_sequence(struct nfsd4_compoundargs *argp,return nfserr_bad_xdr; seq->seqid =be32_to_cpup (p++); seq->slotid =be32_to_cpup (p++); - seq->maxslots =be32_to_cpup (p++); +/* sa_highest_slotid counts from 0 but maxslots counts from 1 ... */ + seq->maxslots =be32_to_cpup (p++) +1 ; seq->cachethis =be32_to_cpup (p); seq->status_flags =0 ;@@ -4968,7 +4969,7 @@ nfsd4_encode_sequence(struct nfsd4_compoundres *resp, __be32 nfserr,if (nfserr != nfs_ok)return nfserr;/* sr_target_highest_slotid */ - nfserr =nfsd4_encode_slotid4 (xdr, seq->maxslots -1 ); + nfserr =nfsd4_encode_slotid4 (xdr, seq->target_maxslots -1 );if (nfserr != nfs_ok)return nfserr;/* sr_status_flags */ fs/nfsd/state.h +5
-1@@ -245,10 +245,12 @@ struct nfsd4_slot {struct svc_cred sl_cred;u32 sl_datalen;u16 sl_opcnt; +u16 sl_generation;#define NFSD4_SLOT_INUSE (1 << 0)#define NFSD4_SLOT_CACHETHIS (1 << 1)#define NFSD4_SLOT_INITIALIZED (1 << 2)#define NFSD4_SLOT_CACHED (1 << 3) +#define NFSD4_SLOT_REUSED (1 << 4)u8 sl_flags;char sl_data[]; };@@ -321,7 +323,6 @@ struct nfsd4_session {u32 se_cb_slot_avail;/* bitmap of available slots */ u32 se_cb_highest_slot;/* highest slot client wants */ u32 se_cb_prog; -bool se_dead;struct list_head se_hash;/* hash by sessionid */ struct list_head se_perclnt;struct nfs4_client *se_client;@@ -331,6 +332,9 @@ struct nfsd4_session {struct list_head se_conns;u32 se_cb_seq_nr[NFSD_BC_SLOT_TABLE_SIZE ];struct xarray se_slots;/* forward channel slots */ +u16 se_slot_gen; +bool se_dead; +u32 se_target_maxslots; };/* formatted contents of nfs4_sessionid */ fs/nfsd/xdr4.h +0
-0fs/nfsd/xdr4.h struct nfsd4_sequence { · line 582
Quoted 9 lines
fs/nfsd/xdr4.h +0
-2@@ -576,9 +576,7 @@ struct nfsd4_sequence {u32 slotid;/* request/response */ u32 maxslots;/* request/response */ u32 cachethis;/* request */ -#if 0 u32 target_maxslots;/* response */ -#endif /* not yet */ u32 status_flags;/* response */ };Hi Neil -
Quoted 5 lines
Quoted 1 line
I've found some misbehavior which I've bisected to this commit.
Hi Chuck,
could you please confirm that it really was this commit that you bisected to? Not the next one?
It's this one. I included the hunk that introduces the misbehavior below. It's when the server starts returning a different value for target_highest_slotid in the SEQUENCE result. The target_highest is 63 -- the number the server has in its slot table. The maxslots value is smaller.
In the working case, these two values never differ.
Quoted 2 lines
Because this commit never reduces ->se_target_maxslots, so the patch which you say removed the symptom should be a no-op.
It's not the reducing of target_highest that's the problem. Rather it's that the target_highest and max in-use slot IDs are different values for a brief period after a reconnect.
That triggers the client to think that the server has reduced its slot table size, so the client shrinks its slot table. The server has not actually shrunken it, however, so it continues to expect the client to use the large slot sequence numbers for those slots.
When the client starts to use one of those slots again, it uses a sequence number of 1, and that fails.
Quoted 19 lines
Even if it was the next commit I'm struggling to pin down the problem. Here is my current analysis - partly to ensure I can present it clearly.
The evidence suggests that the client has retired a slot that the server hasn't. This happens when
nfs41_set_server_slotid_locked()callsnfsd4_shrink_slot_table(), and nothing will happen if any slots before the new limit are still in use. If the server reduces its idea of the target when the client isn't even using that many, the slots can be freed immediately that the client gets a reply indicating the newhighest_slotnumber from the server.The server will not free these slots immediately but will wait to get a confirmation from the client that it has accepted the new limit. But, importantly, the server will not increase the limit that it sends to the client until after it has has a chance to free the retired slots. If the server doesn't increase the limit, then the client won't try to use the retired slots...
Quoted 2 lines
Do you still have the network trace which chows the error? Would I be able to look at it?
Sending via WeTransfer.
But also, it's easy enough to reproduce.
Build your server with CONFIG_FAIL_SUNRPC set. Reboot into the new kernel.
Before each test, run this script on the server:
#!/usr/bin/bash
cd /sys/kernel/debug/fail_sunrpc/
echo Y > ignore-cache-wait
echo Y > ignore-client-disconnect
echo 24847 > interval
echo 97 > times
echo 100 > probability
exit 0
On the client, run fstests with an NFSv4.1 mount. It will usually hang within the first 15 tests.
Quoted 45 lines
Quoted 44 lines
If disconnect injection is set up to break the connection every 25,000 RPCs or so, xfstests running on an NFSv4.1 mount will eventually stall after this commit is applied.
Network capture shows that the server eventually starts returning
SEQ_MISORDEREDbecause the client has forgotten an retired slot after a disconnect, and tries to use sequence number 1 for that slot with a new operation.I've narrowed the issue down to
nfs41_is_outlier_target_slotid()on the client. This function uses a bit of calculus to decide when to bump the slot table's generation number. In the failing case, it never bumps the generation, and that results in the client freeing slots it shouldn't. The server's reported { highest,target_highest} slot numbers don't appear to change correctly after the client has reconnected.If I revert this hunk from 5/6:
fs/nfsd/nfs4xdr.c +1
-1@@ -4968,7 +4969,7 @@ nfsd4_encode_sequence(struct nfsd4_compoundres *resp, __be32 nfserr, if (nfserr != nfs_ok)return nfserr;/* sr_target_highest_slotid */ - nfserr =nfsd4_encode_slotid4 (xdr, seq->maxslots -1 ); + nfserr =nfsd4_encode_slotid4 (xdr, seq->target_maxslots -1 );if (nfserr != nfs_ok)return nfserr;/* sr_status_flags */the reproducer above runs to completion in the expected amount of time.
The high order bit here is whether I should drop these patches for v6.14, or whether you believe you can come up with a narrow solution during the early part of v6.14-rc that I can include in an -rc update for NFSD. I can't really tell if a significant amount of surgery will be necessary.
What do you think?
Signature · 1 line
Chuck Lever
Signature · 1 line
Chuck Lever
Re: [PATCH 5/6] nfsd: add support for freeing unused session-DRC slots
Quoted 370 lines · Chuck Lever
Quoted 319 lines · NeilBrown
Quoted 318 lines · Chuck Lever
Quoted 305 lines · NeilBrown
Reducing the number of slots in the session slot table requires confirmation from the client. This patch adds
reduce_session_slots()which starts the process of getting confirmation, but never calls it. That will come in a later patch.Before we can free a slot we need to confirm that the client won't try to use it again. This involves returning a lower
cr_maxrequestsin a SEQUENCE reply and then seeing aca_maxrequestson the same slot which is not larger than we limit we are trying to impose. So for each slot we need to remember that we have sent a reducedcr_maxrequests.To achieve this we introduce a concept of request "generations". Each time we decide to reduce
cr_maxrequestswe increment the generation number, and record this when we return the lowercr_maxrequeststo the client. When a slot with the current generation reports a lowca_maxrequests, we commit to that level and free extra slots.We use an 16 bit generation number (64 seems wasteful) and if it cycles we iterate all slots and reset the generation number to avoid false matches.
When we free a slot we store the seqid in the slot pointer so that it can be restored when we reactivate the slot. The RFC can be read as suggesting that the slot number could restart from one after a slot is retired and reactivated, but also suggests that retiring slots is not required. So when we reactive a slot we accept with the next seqid in sequence, or 1.
When decoding
sa_highest_slotidinto maxslots we need to add 1 - this matches how it is encoded for the reply.
se_deadis moved in structnfsd4_sessionto remove a hole.Reviewed-by Jeff Layton <[email protected]>Signed-off-by NeilBrown <[email protected]>--- fs/nfsd/nfs4state.c | 94 ++++++++++++++++++++++++++++++++++++++++----- fs/nfsd/nfs4xdr.c | 5 ++- fs/nfsd/state.h | 6 ++- fs/nfsd/xdr4.h | 2 - 4 files changed, 92 insertions(+), 15 deletions(-)fs/nfsd/nfs4state.c +84
-10@@ -1910,17 +1910,69 @@ gen_sessionid(struct nfsd4_session *ses)#define NFSD_MIN_HDR_SEQ_SZ (24 + 12 + 44)static void -free_session_slots (struct nfsd4_session *ses) +free_session_slots (struct nfsd4_session *ses,int from) {int i; -for (i =0 ; i < ses->se_fchannel.maxreqs; i++) { +if (from >= ses->se_fchannel.maxreqs) +return ; + +for (i = from; i < ses->se_fchannel.maxreqs; i++) {struct nfsd4_slot *slot =xa_load (&ses->se_slots, i); -xa_erase (&ses->se_slots, i); +/* +* Save the seqid in case we reactivate this slot. +* This will never require a memory allocation so GFP +* flag is irrelevant +*/ +xa_store (&ses->se_slots, i,xa_mk_value (slot->sl_seqid),0 );free_svc_cred (&slot->sl_cred);kfree (slot); } + ses->se_fchannel.maxreqs = from; +if (ses->se_target_maxslots > from) + ses->se_target_maxslots = from; +} + +/** +* reduce_session_slots - reduce the target max-slots of a session if possible +* @ses: The session to affect +* @dec: how much to decrease the target by +* +* This interface can be used by a shrinker to reduce the target max-slots +* for a session so that some slots can eventually be freed. +* It uses spin_trylock() as it may be called in a context where another +* spinlock is held that has a dependency on client_lock. As shrinkers are +* best-effort, skiping a session is client_lock is already held has no +* great coast +* +* Return value: +* The number of slots that the target was reduced by. +*/ +static int __maybe_unused +reduce_session_slots (struct nfsd4_session *ses,int dec) +{ +struct nfsd_net *nn =net_generic (ses->se_client->net, + nfsd_net_id); +int ret =0 ; + +if (ses->se_target_maxslots <=1 ) +return ret; +if (!spin_trylock (&nn->client_lock)) +return ret; + ret =min (dec, ses->se_target_maxslots-1 ); + ses->se_target_maxslots -= ret; + ses->se_slot_gen +=1 ; +if (ses->se_slot_gen ==0 ) { +int i; + ses->se_slot_gen =1 ; +for (i =0 ; i < ses->se_fchannel.maxreqs; i++) { +struct nfsd4_slot *slot =xa_load (&ses->se_slots, i); + slot->sl_generation =0 ; + } + } +spin_unlock (&nn->client_lock); +return ret; }/* @@ -1968,6 +2020,7 @@ static struct nfsd4_session *alloc_session(struct nfsd4_channel_attrs *fattrs, } fattrs->maxreqs = i;memcpy (&new->se_fchannel, fattrs,sizeof (struct nfsd4_channel_attrs )); + new->se_target_maxslots = i; new->se_cb_slot_avail = ~0U ; new->se_cb_highest_slot =min (battrs->maxreqs -1 ,NFSD_BC_SLOT_TABLE_SIZE -1 );@@ -2081,7 +2134,7 @@ static void nfsd4_del_conns(struct nfsd4_session *s)static void __free_session (struct nfsd4_session *ses) { -free_session_slots (ses); +free_session_slots (ses,0 );xa_destroy (&ses->se_slots);kfree (ses); }@@ -2684,6 +2737,9 @@ static int client_info_show(struct seq_file *m, void *v)seq_printf (m,"session slots:" );list_for_each_entry (ses, &clp->cl_sessions, se_perclnt)seq_printf (m," %u" , ses->se_fchannel.maxreqs); +seq_printf (m,"\nsession target slots:" ); +list_for_each_entry (ses, &clp->cl_sessions, se_perclnt) +seq_printf (m," %u" , ses->se_target_maxslots);spin_unlock (&clp->cl_lock);seq_puts (m,"\n" );@@ -3674,10 +3730,10 @@ nfsd4_exchange_id_release(union nfsd4_op_u *u)kfree (exid->server_impl_name); } -static __be32 check_slot_seqid (u32 seqid,u32 slot_seqid,bool slot_inuse) +static __be32 check_slot_seqid (u32 seqid,u32 slot_seqid,u8 flags) {/* The slot is in use, and no response has been sent. */ -if (slot_inuse) { +if (flags &NFSD4_SLOT_INUSE ) {if (seqid == slot_seqid)return nfserr_jukebox;else @@ -3686,6 +3742,8 @@ static __be32 check_slot_seqid(u32 seqid, u32 slot_seqid, bool slot_inuse)/* Note unsigned 32-bit arithmetic handles wraparound: */ if (likely (seqid == slot_seqid +1 ))return nfs_ok; +if ((flags &NFSD4_SLOT_REUSED ) && seqid ==1 ) +return nfs_ok;if (seqid == slot_seqid)return nfserr_replay_cache;return nfserr_seq_misordered;@@ -4236,8 +4294,7 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,dprintk ("%s: slotid %d\n" , __func__, seq->slotid);trace_nfsd_slot_seqid_sequence (clp, seq, slot); - status =check_slot_seqid (seq->seqid, slot->sl_seqid, - slot->sl_flags &NFSD4_SLOT_INUSE ); + status =check_slot_seqid (seq->seqid, slot->sl_seqid, slot->sl_flags);if (status == nfserr_replay_cache) { status = nfserr_seq_misordered;if (!(slot->sl_flags &NFSD4_SLOT_INITIALIZED ))@@ -4262,6 +4319,12 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,if (status) goto out_put_session; +if (session->se_target_maxslots < session->se_fchannel.maxreqs && + slot->sl_generation == session->se_slot_gen && + seq->maxslots <= session->se_target_maxslots) +/* Client acknowledged our reduce maxreqs */ +free_session_slots (session, session->se_target_maxslots); + buflen = (seq->cachethis) ? session->se_fchannel.maxresp_cached : session->se_fchannel.maxresp_sz;@@ -4272,9 +4335,11 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,svc_reserve (rqstp, buflen); status = nfs_ok; -/* Success! bump slot seqid */ +/* Success! accept new slot seqid */ slot->sl_seqid = seq->seqid; + slot->sl_flags &= ~NFSD4_SLOT_REUSED ; slot->sl_flags |=NFSD4_SLOT_INUSE ; + slot->sl_generation = session->se_slot_gen;if (seq->cachethis) slot->sl_flags |=NFSD4_SLOT_CACHETHIS ;else @@ -4291,9 +4356,11 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,* the client might use. */ if (seq->slotid == session->se_fchannel.maxreqs -1 && + session->se_target_maxslots >= session->se_fchannel.maxreqs && session->se_fchannel.maxreqs <NFSD_MAX_SLOTS_PER_SESSION ) {int s = session->se_fchannel.maxreqs;int cnt =DIV_ROUND_UP (s,5 ); +void *prev_slot; do { /*@@ -4303,18 +4370,25 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,*/ slot =kzalloc (slot_bytes (&session->se_fchannel),GFP_NOWAIT ); + prev_slot =xa_load (&session->se_slots, s); +if (xa_is_value (prev_slot) && slot) { + slot->sl_seqid =xa_to_value (prev_slot); + slot->sl_flags |=NFSD4_SLOT_REUSED ; + }if (slot && !xa_is_err (xa_store (&session->se_slots, s, slot,GFP_NOWAIT ))) { s +=1 ; session->se_fchannel.maxreqs = s; + session->se_target_maxslots = s; }else {kfree (slot); slot =NULL ; } }while (slot && --cnt >0 ); } - seq->maxslots = session->se_fchannel.maxreqs; + seq->maxslots =max (session->se_target_maxslots, seq->maxslots); + seq->target_maxslots = session->se_target_maxslots; out:switch (clp->cl_cb_state) {fs/nfsd/nfs4xdr.c +3
-2@@ -1884,7 +1884,8 @@ nfsd4_decode_sequence(struct nfsd4_compoundargs *argp,return nfserr_bad_xdr; seq->seqid =be32_to_cpup (p++); seq->slotid =be32_to_cpup (p++); - seq->maxslots =be32_to_cpup (p++); +/* sa_highest_slotid counts from 0 but maxslots counts from 1 ... */ + seq->maxslots =be32_to_cpup (p++) +1 ; seq->cachethis =be32_to_cpup (p); seq->status_flags =0 ;@@ -4968,7 +4969,7 @@ nfsd4_encode_sequence(struct nfsd4_compoundres *resp, __be32 nfserr,if (nfserr != nfs_ok)return nfserr;/* sr_target_highest_slotid */ - nfserr =nfsd4_encode_slotid4 (xdr, seq->maxslots -1 ); + nfserr =nfsd4_encode_slotid4 (xdr, seq->target_maxslots -1 );if (nfserr != nfs_ok)return nfserr;/* sr_status_flags */ fs/nfsd/state.h +5
-1@@ -245,10 +245,12 @@ struct nfsd4_slot {struct svc_cred sl_cred;u32 sl_datalen;u16 sl_opcnt; +u16 sl_generation;#define NFSD4_SLOT_INUSE (1 << 0)#define NFSD4_SLOT_CACHETHIS (1 << 1)#define NFSD4_SLOT_INITIALIZED (1 << 2)#define NFSD4_SLOT_CACHED (1 << 3) +#define NFSD4_SLOT_REUSED (1 << 4)u8 sl_flags;char sl_data[]; };@@ -321,7 +323,6 @@ struct nfsd4_session {u32 se_cb_slot_avail;/* bitmap of available slots */ u32 se_cb_highest_slot;/* highest slot client wants */ u32 se_cb_prog; -bool se_dead;struct list_head se_hash;/* hash by sessionid */ struct list_head se_perclnt;struct nfs4_client *se_client;@@ -331,6 +332,9 @@ struct nfsd4_session {struct list_head se_conns;u32 se_cb_seq_nr[NFSD_BC_SLOT_TABLE_SIZE ];struct xarray se_slots;/* forward channel slots */ +u16 se_slot_gen; +bool se_dead; +u32 se_target_maxslots; };/* formatted contents of nfs4_sessionid */ fs/nfsd/xdr4.h +0
-0fs/nfsd/xdr4.h struct nfsd4_sequence { · line 582
Quoted 9 lines
fs/nfsd/xdr4.h +0
-2@@ -576,9 +576,7 @@ struct nfsd4_sequence {u32 slotid;/* request/response */ u32 maxslots;/* request/response */ u32 cachethis;/* request */ -#if 0 u32 target_maxslots;/* response */ -#endif /* not yet */ u32 status_flags;/* response */ };Hi Neil -
Quoted 5 lines
Quoted 1 line
I've found some misbehavior which I've bisected to this commit.
Hi Chuck,
could you please confirm that it really was this commit that you bisected to? Not the next one?It's this one. I included the hunk that introduces the misbehavior below. It's when the server starts returning a different value for
target_highest_slotidin the SEQUENCE result. Thetarget_highestis 63 -- the number the server has in its slot table. The maxslots value is smaller.In the working case, these two values never differ.
Quoted 2 lines
Because this commit never reduces ->se_target_maxslots, so the patch which you say removed the symptom should be a no-op.
It's not the reducing of
target_highestthat's the problem. Rather it's that thetarget_highestand max in-use slot IDs are different values for a brief period after a reconnect.That triggers the client to think that the server has reduced its slot table size, so the client shrinks its slot table. The server has not actually shrunken it, however, so it continues to expect the client to use the large slot sequence numbers for those slots.
When the client starts to use one of those slots again, it uses a sequence number of 1, and that fails.
Quoted 19 lines
Even if it was the next commit I'm struggling to pin down the problem. Here is my current analysis - partly to ensure I can present it clearly.
The evidence suggests that the client has retired a slot that the server hasn't. This happens when
nfs41_set_server_slotid_locked()callsnfsd4_shrink_slot_table(), and nothing will happen if any slots before the new limit are still in use. If the server reduces its idea of the target when the client isn't even using that many, the slots can be freed immediately that the client gets a reply indicating the newhighest_slotnumber from the server.The server will not free these slots immediately but will wait to get a confirmation from the client that it has accepted the new limit. But, importantly, the server will not increase the limit that it sends to the client until after it has has a chance to free the retired slots. If the server doesn't increase the limit, then the client won't try to use the retired slots...
Quoted 4 lines
Quoted 2 lines
Do you still have the network trace which chows the error? Would I be able to look at it?
Sending via WeTransfer.
Thanks. I see the problem clearly now. It isn't so much that 'high slot' and 'target' are different, it is that they are both wrong. 'highslot' (which is the max the server will accept) is typically 0 or 4, and 'target' is 63.
The 'highest slot' has been copied from what the client said was the max currently in use. I don't know where the 'target' came from, maybe from the previous reply sent.
This bug was introduced in
nfsd: allocate new session-based DRC slots on demand.
I should have move the seq->maxslots assignment after the "out:", not before.
I'll send a patch.
Quoted 18 lines
But also, it's easy enough to reproduce.
Build your server with
CONFIG_FAIL_SUNRPCset. Reboot into the new kernel.Before each test, run this script on the server:
#!/usr/bin/bash
cd
/sys/kernel/debug/fail_sunrpc/echo Y > ignore-cache-wait
echo Y > ignore-client-disconnect
echo 24847 > interval
echo 97 > times
echo 100 > probability
Quoted 5 lines
exit 0
On the client, run fstests with an NFSv4.1 mount. It will usually hang within the first 15 tests.
I might give that a try - thanks.
NeilBrown
PATCH v5 6/6 nfsd: add shrinker to reduce number of slots allocated per session
In reply to NeilBrownAdd a shrinker which frees unused slots and may ask the clients to use fewer slots on each session.
We keep a global count of the number of freeable slots, which is the sum of one less than the current "target" slots in all sessions in all clients in all net-namespaces. This number is reported by the shrinker.
When the shrinker is asked to free some, we call xxx on each session in a round-robin asking each to reduce the slot count by 1. This will reduce the "target" so the number reported by the shrinker will reduce immediately. The memory will only be freed later when the client confirmed that it is no longer needed.
We use a global list of sessions and move the "head" to after the last session that we asked to reduce, so the next callback from the shrinker will move on to the next session. This pressure should be applied "evenly" across all sessions over time.
--- fs/nfsd/nfs4state.c | 77 ++++++++++++++++++++++++++++++++++++++++++--- fs/nfsd/state.h | 1 + 2 files changed, 73 insertions(+), 5 deletions(-)
fs/nfsd/nfs4state.c +72 -5
@@ -1909,6 +1909,16 @@ gen_sessionid(struct nfsd4_session *ses)*/ #define NFSD_MIN_HDR_SEQ_SZ (24 + 12 + 44) +static struct shrinker *nfsd_slot_shrinker; +static DEFINE_SPINLOCK (nfsd_session_list_lock); +static LIST_HEAD (nfsd_session_list); +/* The sum of "target_slots-1" on every session. The shrinker can push this +* down, though it can take a little while for the memory to actually +* be freed. The "-1" is because we can never free slot 0 while the +* session is active. +*/ +static atomic_t nfsd_total_target_slots =ATOMIC_INIT (0 ); +static void free_session_slots (struct nfsd4_session *ses,int from) {
@@ -1930,8 +1940,11 @@ free_session_slots(struct nfsd4_session *ses, int from)kfree (slot); } ses->se_fchannel.maxreqs = from; -if (ses->se_target_maxslots > from) - ses->se_target_maxslots = from; +if (ses->se_target_maxslots > from) { +int new_target = from ?:1 ; +atomic_sub (ses->se_target_maxslots - new_target, &nfsd_total_target_slots); + ses->se_target_maxslots = new_target; + } }/**
@@ -1949,7 +1962,7 @@ free_session_slots(struct nfsd4_session *ses, int from)* Return value: * The number of slots that the target was reduced by. */ -static int __maybe_unused +static int reduce_session_slots (struct nfsd4_session *ses,int dec) { struct nfsd_net *nn = net_generic(ses->se_client->net,
@@ -1962,6 +1975,7 @@ reduce_session_slots(struct nfsd4_session *ses, int dec)return ret; ret =min (dec, ses->se_target_maxslots-1 ); ses->se_target_maxslots -= ret; +atomic_sub (ret, &nfsd_total_target_slots); ses->se_slot_gen +=1 ;if (ses->se_slot_gen ==0 ) {int i;
@@ -2021,6 +2035,7 @@ static struct nfsd4_session *alloc_session(struct nfsd4_channel_attrs *fattrs, fattrs->maxreqs = i;memcpy (&new->se_fchannel, fattrs,sizeof (struct nfsd4_channel_attrs )); new->se_target_maxslots = i; +atomic_add (i -1 , &nfsd_total_target_slots); new->se_cb_slot_avail = ~0U ; new->se_cb_highest_slot =min (battrs->maxreqs -1 ,NFSD_BC_SLOT_TABLE_SIZE -1 );
@@ -2145,6 +2160,36 @@ static void free_session(struct nfsd4_session *ses)__free_session (ses); } +static unsigned long +nfsd_slot_count (struct shrinker *s,struct shrink_control *sc) +{ +unsigned long cnt =atomic_read (&nfsd_total_target_slots); + +return cnt ? cnt :SHRINK_EMPTY ; +} + +static unsigned long +nfsd_slot_scan (struct shrinker *s,struct shrink_control *sc) +{ +struct nfsd4_session *ses; +unsigned long scanned =0 ; +unsigned long freed =0 ; + +spin_lock (&nfsd_session_list_lock); +list_for_each_entry (ses, &nfsd_session_list, se_all_sessions) { + freed +=reduce_session_slots (ses,1 ); + scanned +=1 ; +if (scanned >= sc->nr_to_scan) { +/* Move starting point for next scan */ +list_move (&nfsd_session_list, &ses->se_all_sessions); +break ; + } + } +spin_unlock (&nfsd_session_list_lock); + sc->nr_scanned = scanned; +return freed; +} +static void init_session (struct svc_rqst *rqstp,struct nfsd4_session *new,struct nfs4_client *clp,struct nfsd4_create_session *cses) {int idx;
@@ -2169,6 +2214,10 @@ static void init_session(struct svc_rqst *rqstp, struct nfsd4_session *new, strulist_add (&new->se_perclnt, &clp->cl_sessions);spin_unlock (&clp->cl_lock); +spin_lock (&nfsd_session_list_lock); +list_add_tail (&new->se_all_sessions, &nfsd_session_list); +spin_unlock (&nfsd_session_list_lock); + {struct sockaddr *sa =svc_addr (rqstp);/*
@@ -2238,6 +2287,9 @@ unhash_session(struct nfsd4_session *ses)spin_lock (&ses->se_client->cl_lock);list_del (&ses->se_perclnt);spin_unlock (&ses->se_client->cl_lock); +spin_lock (&nfsd_session_list_lock); +list_del (&ses->se_all_sessions); +spin_unlock (&nfsd_session_list_lock); }/* SETCLIENTID and SETCLIENTID_CONFIRM Helper functions */
@@ -2373,8 +2425,12 @@ unhash_client_locked(struct nfs4_client *clp) }list_del_init (&clp->cl_lru);spin_lock (&clp->cl_lock); -list_for_each_entry (ses, &clp->cl_sessions, se_perclnt) +spin_lock (&nfsd_session_list_lock); +list_for_each_entry (ses, &clp->cl_sessions, se_perclnt) {list_del_init (&ses->se_hash); +list_del_init (&ses->se_all_sessions); + } +spin_unlock (&nfsd_session_list_lock);spin_unlock (&clp->cl_lock); }
@@ -4380,6 +4436,8 @@ nfsd4_sequence(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, GFP_NOWAIT))) { s += 1; session->se_fchannel.maxreqs = s; + atomic_add(s - session->se_target_maxslots, + &nfsd_total_target_slots); session->se_target_maxslots = s; } else {kfree (slot);
@@ -8776,7 +8834,6 @@ nfs4_state_start_net(struct net *net) }/* initialization to perform when the nfsd service is started: */ -int nfs4_state_start (void ) {
@@ -8786,6 +8843,15 @@ nfs4_state_start(void)if (ret)return ret; + nfsd_slot_shrinker =shrinker_alloc (0 ,"nfsd-DRC-slot" ); +if (!nfsd_slot_shrinker) { +rhltable_destroy (&nfs4_file_rhltable); +return -ENOMEM ; + } + nfsd_slot_shrinker->count_objects = nfsd_slot_count; + nfsd_slot_shrinker->scan_objects = nfsd_slot_scan; +shrinker_register (nfsd_slot_shrinker); +set_max_delegations ();return 0 ; }
@@ -8827,6 +8893,7 @@ voidnfs4_state_shutdown (void) {rhltable_destroy (&nfs4_file_rhltable); +shrinker_free (nfsd_slot_shrinker); } static void
fs/nfsd/state.h +1 -0
@@ -325,6 +325,7 @@ struct nfsd4_session {u32 se_cb_prog;struct list_head se_hash;/* hash by sessionid */ struct list_head se_perclnt; +struct list_head se_all_sessions;/* global list of sessions */ struct nfs4_client *se_client;struct nfs4_sessionid se_sessionid;struct nfsd4_channel_attrs se_fchannel;
Signature · 1 line
2.47.0