blob: 96352ab7bd8100af89f2fadfa50825e31221dd7b [file] [log] [blame]
Andrew Scullb4b6d4a2019-01-02 15:54:55 +00001// SPDX-License-Identifier: GPL-2.0
2/*
3 * Request reply cache. This is currently a global cache, but this may
4 * change in the future and be a per-client cache.
5 *
6 * This code is heavily inspired by the 44BSD implementation, although
7 * it does things a bit differently.
8 *
9 * Copyright (C) 1995, 1996 Olaf Kirch <okir@monad.swb.de>
10 */
11
David Brazdil0f672f62019-12-10 10:32:29 +000012#include <linux/sunrpc/svc_xprt.h>
Andrew Scullb4b6d4a2019-01-02 15:54:55 +000013#include <linux/slab.h>
14#include <linux/vmalloc.h>
15#include <linux/sunrpc/addr.h>
16#include <linux/highmem.h>
17#include <linux/log2.h>
18#include <linux/hash.h>
19#include <net/checksum.h>
20
21#include "nfsd.h"
22#include "cache.h"
23
24#define NFSDDBG_FACILITY NFSDDBG_REPCACHE
25
26/*
27 * We use this value to determine the number of hash buckets from the max
28 * cache size, the idea being that when the cache is at its maximum number
29 * of entries, then this should be the average number of entries per bucket.
30 */
31#define TARGET_BUCKET_SIZE 64
32
33struct nfsd_drc_bucket {
David Brazdil0f672f62019-12-10 10:32:29 +000034 struct rb_root rb_head;
Andrew Scullb4b6d4a2019-01-02 15:54:55 +000035 struct list_head lru_head;
36 spinlock_t cache_lock;
37};
38
Andrew Scullb4b6d4a2019-01-02 15:54:55 +000039static int nfsd_cache_append(struct svc_rqst *rqstp, struct kvec *vec);
40static unsigned long nfsd_reply_cache_count(struct shrinker *shrink,
41 struct shrink_control *sc);
42static unsigned long nfsd_reply_cache_scan(struct shrinker *shrink,
43 struct shrink_control *sc);
44
Andrew Scullb4b6d4a2019-01-02 15:54:55 +000045/*
46 * Put a cap on the size of the DRC based on the amount of available
47 * low memory in the machine.
48 *
49 * 64MB: 8192
50 * 128MB: 11585
51 * 256MB: 16384
52 * 512MB: 23170
53 * 1GB: 32768
54 * 2GB: 46340
55 * 4GB: 65536
56 * 8GB: 92681
57 * 16GB: 131072
58 *
59 * ...with a hard cap of 256k entries. In the worst case, each entry will be
60 * ~1k, so the above numbers should give a rough max of the amount of memory
61 * used in k.
David Brazdil0f672f62019-12-10 10:32:29 +000062 *
63 * XXX: these limits are per-container, so memory used will increase
64 * linearly with number of containers. Maybe that's OK.
Andrew Scullb4b6d4a2019-01-02 15:54:55 +000065 */
66static unsigned int
67nfsd_cache_size_limit(void)
68{
69 unsigned int limit;
David Brazdil0f672f62019-12-10 10:32:29 +000070 unsigned long low_pages = totalram_pages() - totalhigh_pages();
Andrew Scullb4b6d4a2019-01-02 15:54:55 +000071
72 limit = (16 * int_sqrt(low_pages)) << (PAGE_SHIFT-10);
73 return min_t(unsigned int, limit, 256*1024);
74}
75
76/*
77 * Compute the number of hash buckets we need. Divide the max cachesize by
78 * the "target" max bucket size, and round up to next power of two.
79 */
80static unsigned int
81nfsd_hashsize(unsigned int limit)
82{
83 return roundup_pow_of_two(limit / TARGET_BUCKET_SIZE);
84}
85
86static u32
David Brazdil0f672f62019-12-10 10:32:29 +000087nfsd_cache_hash(__be32 xid, struct nfsd_net *nn)
Andrew Scullb4b6d4a2019-01-02 15:54:55 +000088{
David Brazdil0f672f62019-12-10 10:32:29 +000089 return hash_32(be32_to_cpu(xid), nn->maskbits);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +000090}
91
92static struct svc_cacherep *
David Brazdil0f672f62019-12-10 10:32:29 +000093nfsd_reply_cache_alloc(struct svc_rqst *rqstp, __wsum csum,
94 struct nfsd_net *nn)
Andrew Scullb4b6d4a2019-01-02 15:54:55 +000095{
96 struct svc_cacherep *rp;
97
David Brazdil0f672f62019-12-10 10:32:29 +000098 rp = kmem_cache_alloc(nn->drc_slab, GFP_KERNEL);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +000099 if (rp) {
100 rp->c_state = RC_UNUSED;
101 rp->c_type = RC_NOCACHE;
David Brazdil0f672f62019-12-10 10:32:29 +0000102 RB_CLEAR_NODE(&rp->c_node);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000103 INIT_LIST_HEAD(&rp->c_lru);
David Brazdil0f672f62019-12-10 10:32:29 +0000104
105 memset(&rp->c_key, 0, sizeof(rp->c_key));
106 rp->c_key.k_xid = rqstp->rq_xid;
107 rp->c_key.k_proc = rqstp->rq_proc;
108 rpc_copy_addr((struct sockaddr *)&rp->c_key.k_addr, svc_addr(rqstp));
109 rpc_set_port((struct sockaddr *)&rp->c_key.k_addr, rpc_get_port(svc_addr(rqstp)));
110 rp->c_key.k_prot = rqstp->rq_prot;
111 rp->c_key.k_vers = rqstp->rq_vers;
112 rp->c_key.k_len = rqstp->rq_arg.len;
113 rp->c_key.k_csum = csum;
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000114 }
115 return rp;
116}
117
118static void
David Brazdil0f672f62019-12-10 10:32:29 +0000119nfsd_reply_cache_free_locked(struct nfsd_drc_bucket *b, struct svc_cacherep *rp,
120 struct nfsd_net *nn)
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000121{
122 if (rp->c_type == RC_REPLBUFF && rp->c_replvec.iov_base) {
David Brazdil0f672f62019-12-10 10:32:29 +0000123 nn->drc_mem_usage -= rp->c_replvec.iov_len;
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000124 kfree(rp->c_replvec.iov_base);
125 }
David Brazdil0f672f62019-12-10 10:32:29 +0000126 if (rp->c_state != RC_UNUSED) {
127 rb_erase(&rp->c_node, &b->rb_head);
128 list_del(&rp->c_lru);
129 atomic_dec(&nn->num_drc_entries);
130 nn->drc_mem_usage -= sizeof(*rp);
131 }
132 kmem_cache_free(nn->drc_slab, rp);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000133}
134
135static void
David Brazdil0f672f62019-12-10 10:32:29 +0000136nfsd_reply_cache_free(struct nfsd_drc_bucket *b, struct svc_cacherep *rp,
137 struct nfsd_net *nn)
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000138{
139 spin_lock(&b->cache_lock);
David Brazdil0f672f62019-12-10 10:32:29 +0000140 nfsd_reply_cache_free_locked(b, rp, nn);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000141 spin_unlock(&b->cache_lock);
142}
143
David Brazdil0f672f62019-12-10 10:32:29 +0000144int nfsd_reply_cache_init(struct nfsd_net *nn)
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000145{
146 unsigned int hashsize;
147 unsigned int i;
148 int status = 0;
149
David Brazdil0f672f62019-12-10 10:32:29 +0000150 nn->max_drc_entries = nfsd_cache_size_limit();
151 atomic_set(&nn->num_drc_entries, 0);
152 hashsize = nfsd_hashsize(nn->max_drc_entries);
153 nn->maskbits = ilog2(hashsize);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000154
David Brazdil0f672f62019-12-10 10:32:29 +0000155 nn->nfsd_reply_cache_shrinker.scan_objects = nfsd_reply_cache_scan;
156 nn->nfsd_reply_cache_shrinker.count_objects = nfsd_reply_cache_count;
157 nn->nfsd_reply_cache_shrinker.seeks = 1;
158 status = register_shrinker(&nn->nfsd_reply_cache_shrinker);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000159 if (status)
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000160 goto out_nomem;
161
David Brazdil0f672f62019-12-10 10:32:29 +0000162 nn->drc_slab = kmem_cache_create("nfsd_drc",
163 sizeof(struct svc_cacherep), 0, 0, NULL);
164 if (!nn->drc_slab)
165 goto out_shrinker;
166
167 nn->drc_hashtbl = kcalloc(hashsize,
168 sizeof(*nn->drc_hashtbl), GFP_KERNEL);
169 if (!nn->drc_hashtbl) {
170 nn->drc_hashtbl = vzalloc(array_size(hashsize,
171 sizeof(*nn->drc_hashtbl)));
172 if (!nn->drc_hashtbl)
173 goto out_slab;
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000174 }
175
176 for (i = 0; i < hashsize; i++) {
David Brazdil0f672f62019-12-10 10:32:29 +0000177 INIT_LIST_HEAD(&nn->drc_hashtbl[i].lru_head);
178 spin_lock_init(&nn->drc_hashtbl[i].cache_lock);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000179 }
David Brazdil0f672f62019-12-10 10:32:29 +0000180 nn->drc_hashsize = hashsize;
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000181
182 return 0;
David Brazdil0f672f62019-12-10 10:32:29 +0000183out_slab:
184 kmem_cache_destroy(nn->drc_slab);
185out_shrinker:
186 unregister_shrinker(&nn->nfsd_reply_cache_shrinker);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000187out_nomem:
188 printk(KERN_ERR "nfsd: failed to allocate reply cache\n");
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000189 return -ENOMEM;
190}
191
David Brazdil0f672f62019-12-10 10:32:29 +0000192void nfsd_reply_cache_shutdown(struct nfsd_net *nn)
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000193{
194 struct svc_cacherep *rp;
195 unsigned int i;
196
David Brazdil0f672f62019-12-10 10:32:29 +0000197 unregister_shrinker(&nn->nfsd_reply_cache_shrinker);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000198
David Brazdil0f672f62019-12-10 10:32:29 +0000199 for (i = 0; i < nn->drc_hashsize; i++) {
200 struct list_head *head = &nn->drc_hashtbl[i].lru_head;
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000201 while (!list_empty(head)) {
202 rp = list_first_entry(head, struct svc_cacherep, c_lru);
David Brazdil0f672f62019-12-10 10:32:29 +0000203 nfsd_reply_cache_free_locked(&nn->drc_hashtbl[i],
204 rp, nn);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000205 }
206 }
207
David Brazdil0f672f62019-12-10 10:32:29 +0000208 kvfree(nn->drc_hashtbl);
209 nn->drc_hashtbl = NULL;
210 nn->drc_hashsize = 0;
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000211
David Brazdil0f672f62019-12-10 10:32:29 +0000212 kmem_cache_destroy(nn->drc_slab);
213 nn->drc_slab = NULL;
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000214}
215
216/*
217 * Move cache entry to end of LRU list, and queue the cleaner to run if it's
218 * not already scheduled.
219 */
220static void
221lru_put_end(struct nfsd_drc_bucket *b, struct svc_cacherep *rp)
222{
223 rp->c_timestamp = jiffies;
224 list_move_tail(&rp->c_lru, &b->lru_head);
225}
226
227static long
David Brazdil0f672f62019-12-10 10:32:29 +0000228prune_bucket(struct nfsd_drc_bucket *b, struct nfsd_net *nn)
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000229{
230 struct svc_cacherep *rp, *tmp;
231 long freed = 0;
232
233 list_for_each_entry_safe(rp, tmp, &b->lru_head, c_lru) {
234 /*
235 * Don't free entries attached to calls that are still
236 * in-progress, but do keep scanning the list.
237 */
238 if (rp->c_state == RC_INPROG)
239 continue;
David Brazdil0f672f62019-12-10 10:32:29 +0000240 if (atomic_read(&nn->num_drc_entries) <= nn->max_drc_entries &&
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000241 time_before(jiffies, rp->c_timestamp + RC_EXPIRE))
242 break;
David Brazdil0f672f62019-12-10 10:32:29 +0000243 nfsd_reply_cache_free_locked(b, rp, nn);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000244 freed++;
245 }
246 return freed;
247}
248
249/*
250 * Walk the LRU list and prune off entries that are older than RC_EXPIRE.
251 * Also prune the oldest ones when the total exceeds the max number of entries.
252 */
253static long
David Brazdil0f672f62019-12-10 10:32:29 +0000254prune_cache_entries(struct nfsd_net *nn)
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000255{
256 unsigned int i;
257 long freed = 0;
258
David Brazdil0f672f62019-12-10 10:32:29 +0000259 for (i = 0; i < nn->drc_hashsize; i++) {
260 struct nfsd_drc_bucket *b = &nn->drc_hashtbl[i];
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000261
262 if (list_empty(&b->lru_head))
263 continue;
264 spin_lock(&b->cache_lock);
David Brazdil0f672f62019-12-10 10:32:29 +0000265 freed += prune_bucket(b, nn);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000266 spin_unlock(&b->cache_lock);
267 }
268 return freed;
269}
270
271static unsigned long
272nfsd_reply_cache_count(struct shrinker *shrink, struct shrink_control *sc)
273{
David Brazdil0f672f62019-12-10 10:32:29 +0000274 struct nfsd_net *nn = container_of(shrink,
275 struct nfsd_net, nfsd_reply_cache_shrinker);
276
277 return atomic_read(&nn->num_drc_entries);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000278}
279
280static unsigned long
281nfsd_reply_cache_scan(struct shrinker *shrink, struct shrink_control *sc)
282{
David Brazdil0f672f62019-12-10 10:32:29 +0000283 struct nfsd_net *nn = container_of(shrink,
284 struct nfsd_net, nfsd_reply_cache_shrinker);
285
286 return prune_cache_entries(nn);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000287}
288/*
289 * Walk an xdr_buf and get a CRC for at most the first RC_CSUMLEN bytes
290 */
291static __wsum
292nfsd_cache_csum(struct svc_rqst *rqstp)
293{
294 int idx;
295 unsigned int base;
296 __wsum csum;
297 struct xdr_buf *buf = &rqstp->rq_arg;
298 const unsigned char *p = buf->head[0].iov_base;
299 size_t csum_len = min_t(size_t, buf->head[0].iov_len + buf->page_len,
300 RC_CSUMLEN);
301 size_t len = min(buf->head[0].iov_len, csum_len);
302
303 /* rq_arg.head first */
304 csum = csum_partial(p, len, 0);
305 csum_len -= len;
306
307 /* Continue into page array */
308 idx = buf->page_base / PAGE_SIZE;
309 base = buf->page_base & ~PAGE_MASK;
310 while (csum_len) {
311 p = page_address(buf->pages[idx]) + base;
312 len = min_t(size_t, PAGE_SIZE - base, csum_len);
313 csum = csum_partial(p, len, csum);
314 csum_len -= len;
315 base = 0;
316 ++idx;
317 }
318 return csum;
319}
320
David Brazdil0f672f62019-12-10 10:32:29 +0000321static int
322nfsd_cache_key_cmp(const struct svc_cacherep *key,
323 const struct svc_cacherep *rp, struct nfsd_net *nn)
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000324{
David Brazdil0f672f62019-12-10 10:32:29 +0000325 if (key->c_key.k_xid == rp->c_key.k_xid &&
326 key->c_key.k_csum != rp->c_key.k_csum)
327 ++nn->payload_misses;
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000328
David Brazdil0f672f62019-12-10 10:32:29 +0000329 return memcmp(&key->c_key, &rp->c_key, sizeof(key->c_key));
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000330}
331
332/*
333 * Search the request hash for an entry that matches the given rqstp.
334 * Must be called with cache_lock held. Returns the found entry or
David Brazdil0f672f62019-12-10 10:32:29 +0000335 * inserts an empty key on failure.
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000336 */
337static struct svc_cacherep *
David Brazdil0f672f62019-12-10 10:32:29 +0000338nfsd_cache_insert(struct nfsd_drc_bucket *b, struct svc_cacherep *key,
339 struct nfsd_net *nn)
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000340{
David Brazdil0f672f62019-12-10 10:32:29 +0000341 struct svc_cacherep *rp, *ret = key;
342 struct rb_node **p = &b->rb_head.rb_node,
343 *parent = NULL;
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000344 unsigned int entries = 0;
David Brazdil0f672f62019-12-10 10:32:29 +0000345 int cmp;
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000346
David Brazdil0f672f62019-12-10 10:32:29 +0000347 while (*p != NULL) {
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000348 ++entries;
David Brazdil0f672f62019-12-10 10:32:29 +0000349 parent = *p;
350 rp = rb_entry(parent, struct svc_cacherep, c_node);
351
352 cmp = nfsd_cache_key_cmp(key, rp, nn);
353 if (cmp < 0)
354 p = &parent->rb_left;
355 else if (cmp > 0)
356 p = &parent->rb_right;
357 else {
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000358 ret = rp;
David Brazdil0f672f62019-12-10 10:32:29 +0000359 goto out;
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000360 }
361 }
David Brazdil0f672f62019-12-10 10:32:29 +0000362 rb_link_node(&key->c_node, parent, p);
363 rb_insert_color(&key->c_node, &b->rb_head);
364out:
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000365 /* tally hash chain length stats */
David Brazdil0f672f62019-12-10 10:32:29 +0000366 if (entries > nn->longest_chain) {
367 nn->longest_chain = entries;
368 nn->longest_chain_cachesize = atomic_read(&nn->num_drc_entries);
369 } else if (entries == nn->longest_chain) {
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000370 /* prefer to keep the smallest cachesize possible here */
David Brazdil0f672f62019-12-10 10:32:29 +0000371 nn->longest_chain_cachesize = min_t(unsigned int,
372 nn->longest_chain_cachesize,
373 atomic_read(&nn->num_drc_entries));
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000374 }
375
David Brazdil0f672f62019-12-10 10:32:29 +0000376 lru_put_end(b, ret);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000377 return ret;
378}
379
380/*
381 * Try to find an entry matching the current call in the cache. When none
382 * is found, we try to grab the oldest expired entry off the LRU list. If
383 * a suitable one isn't there, then drop the cache_lock and allocate a
384 * new one, then search again in case one got inserted while this thread
385 * didn't hold the lock.
386 */
387int
388nfsd_cache_lookup(struct svc_rqst *rqstp)
389{
David Brazdil0f672f62019-12-10 10:32:29 +0000390 struct nfsd_net *nn = net_generic(SVC_NET(rqstp), nfsd_net_id);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000391 struct svc_cacherep *rp, *found;
392 __be32 xid = rqstp->rq_xid;
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000393 __wsum csum;
David Brazdil0f672f62019-12-10 10:32:29 +0000394 u32 hash = nfsd_cache_hash(xid, nn);
395 struct nfsd_drc_bucket *b = &nn->drc_hashtbl[hash];
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000396 int type = rqstp->rq_cachetype;
397 int rtn = RC_DOIT;
398
399 rqstp->rq_cacherep = NULL;
400 if (type == RC_NOCACHE) {
401 nfsdstats.rcnocache++;
402 return rtn;
403 }
404
405 csum = nfsd_cache_csum(rqstp);
406
407 /*
408 * Since the common case is a cache miss followed by an insert,
409 * preallocate an entry.
410 */
David Brazdil0f672f62019-12-10 10:32:29 +0000411 rp = nfsd_reply_cache_alloc(rqstp, csum, nn);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000412 if (!rp) {
413 dprintk("nfsd: unable to allocate DRC entry!\n");
David Brazdil0f672f62019-12-10 10:32:29 +0000414 return rtn;
415 }
416
417 spin_lock(&b->cache_lock);
418 found = nfsd_cache_insert(b, rp, nn);
419 if (found != rp) {
420 nfsd_reply_cache_free_locked(NULL, rp, nn);
421 rp = found;
422 goto found_entry;
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000423 }
424
425 nfsdstats.rcmisses++;
426 rqstp->rq_cacherep = rp;
427 rp->c_state = RC_INPROG;
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000428
David Brazdil0f672f62019-12-10 10:32:29 +0000429 atomic_inc(&nn->num_drc_entries);
430 nn->drc_mem_usage += sizeof(*rp);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000431
David Brazdil0f672f62019-12-10 10:32:29 +0000432 /* go ahead and prune the cache */
433 prune_bucket(b, nn);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000434 out:
435 spin_unlock(&b->cache_lock);
436 return rtn;
437
438found_entry:
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000439 /* We found a matching entry which is either in progress or done. */
David Brazdil0f672f62019-12-10 10:32:29 +0000440 nfsdstats.rchits++;
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000441 rtn = RC_DROPIT;
David Brazdil0f672f62019-12-10 10:32:29 +0000442
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000443 /* Request being processed */
444 if (rp->c_state == RC_INPROG)
445 goto out;
446
447 /* From the hall of fame of impractical attacks:
448 * Is this a user who tries to snoop on the cache? */
449 rtn = RC_DOIT;
450 if (!test_bit(RQ_SECURE, &rqstp->rq_flags) && rp->c_secure)
451 goto out;
452
453 /* Compose RPC reply header */
454 switch (rp->c_type) {
455 case RC_NOCACHE:
456 break;
457 case RC_REPLSTAT:
458 svc_putu32(&rqstp->rq_res.head[0], rp->c_replstat);
459 rtn = RC_REPLY;
460 break;
461 case RC_REPLBUFF:
462 if (!nfsd_cache_append(rqstp, &rp->c_replvec))
463 goto out; /* should not happen */
464 rtn = RC_REPLY;
465 break;
466 default:
467 printk(KERN_WARNING "nfsd: bad repcache type %d\n", rp->c_type);
David Brazdil0f672f62019-12-10 10:32:29 +0000468 nfsd_reply_cache_free_locked(b, rp, nn);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000469 }
470
471 goto out;
472}
473
474/*
475 * Update a cache entry. This is called from nfsd_dispatch when
476 * the procedure has been executed and the complete reply is in
477 * rqstp->rq_res.
478 *
479 * We're copying around data here rather than swapping buffers because
480 * the toplevel loop requires max-sized buffers, which would be a waste
481 * of memory for a cache with a max reply size of 100 bytes (diropokres).
482 *
483 * If we should start to use different types of cache entries tailored
484 * specifically for attrstat and fh's, we may save even more space.
485 *
486 * Also note that a cachetype of RC_NOCACHE can legally be passed when
487 * nfsd failed to encode a reply that otherwise would have been cached.
488 * In this case, nfsd_cache_update is called with statp == NULL.
489 */
490void
491nfsd_cache_update(struct svc_rqst *rqstp, int cachetype, __be32 *statp)
492{
David Brazdil0f672f62019-12-10 10:32:29 +0000493 struct nfsd_net *nn = net_generic(SVC_NET(rqstp), nfsd_net_id);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000494 struct svc_cacherep *rp = rqstp->rq_cacherep;
495 struct kvec *resv = &rqstp->rq_res.head[0], *cachv;
496 u32 hash;
497 struct nfsd_drc_bucket *b;
498 int len;
499 size_t bufsize = 0;
500
501 if (!rp)
502 return;
503
David Brazdil0f672f62019-12-10 10:32:29 +0000504 hash = nfsd_cache_hash(rp->c_key.k_xid, nn);
505 b = &nn->drc_hashtbl[hash];
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000506
507 len = resv->iov_len - ((char*)statp - (char*)resv->iov_base);
508 len >>= 2;
509
510 /* Don't cache excessive amounts of data and XDR failures */
511 if (!statp || len > (256 >> 2)) {
David Brazdil0f672f62019-12-10 10:32:29 +0000512 nfsd_reply_cache_free(b, rp, nn);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000513 return;
514 }
515
516 switch (cachetype) {
517 case RC_REPLSTAT:
518 if (len != 1)
519 printk("nfsd: RC_REPLSTAT/reply len %d!\n",len);
520 rp->c_replstat = *statp;
521 break;
522 case RC_REPLBUFF:
523 cachv = &rp->c_replvec;
524 bufsize = len << 2;
525 cachv->iov_base = kmalloc(bufsize, GFP_KERNEL);
526 if (!cachv->iov_base) {
David Brazdil0f672f62019-12-10 10:32:29 +0000527 nfsd_reply_cache_free(b, rp, nn);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000528 return;
529 }
530 cachv->iov_len = bufsize;
531 memcpy(cachv->iov_base, statp, bufsize);
532 break;
533 case RC_NOCACHE:
David Brazdil0f672f62019-12-10 10:32:29 +0000534 nfsd_reply_cache_free(b, rp, nn);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000535 return;
536 }
537 spin_lock(&b->cache_lock);
David Brazdil0f672f62019-12-10 10:32:29 +0000538 nn->drc_mem_usage += bufsize;
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000539 lru_put_end(b, rp);
540 rp->c_secure = test_bit(RQ_SECURE, &rqstp->rq_flags);
541 rp->c_type = cachetype;
542 rp->c_state = RC_DONE;
543 spin_unlock(&b->cache_lock);
544 return;
545}
546
547/*
548 * Copy cached reply to current reply buffer. Should always fit.
549 * FIXME as reply is in a page, we should just attach the page, and
550 * keep a refcount....
551 */
552static int
553nfsd_cache_append(struct svc_rqst *rqstp, struct kvec *data)
554{
555 struct kvec *vec = &rqstp->rq_res.head[0];
556
557 if (vec->iov_len + data->iov_len > PAGE_SIZE) {
558 printk(KERN_WARNING "nfsd: cached reply too large (%zd).\n",
559 data->iov_len);
560 return 0;
561 }
562 memcpy((char*)vec->iov_base + vec->iov_len, data->iov_base, data->iov_len);
563 vec->iov_len += data->iov_len;
564 return 1;
565}
566
567/*
568 * Note that fields may be added, removed or reordered in the future. Programs
569 * scraping this file for info should test the labels to ensure they're
570 * getting the correct field.
571 */
572static int nfsd_reply_cache_stats_show(struct seq_file *m, void *v)
573{
David Brazdil0f672f62019-12-10 10:32:29 +0000574 struct nfsd_net *nn = m->private;
575
576 seq_printf(m, "max entries: %u\n", nn->max_drc_entries);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000577 seq_printf(m, "num entries: %u\n",
David Brazdil0f672f62019-12-10 10:32:29 +0000578 atomic_read(&nn->num_drc_entries));
579 seq_printf(m, "hash buckets: %u\n", 1 << nn->maskbits);
580 seq_printf(m, "mem usage: %u\n", nn->drc_mem_usage);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000581 seq_printf(m, "cache hits: %u\n", nfsdstats.rchits);
582 seq_printf(m, "cache misses: %u\n", nfsdstats.rcmisses);
583 seq_printf(m, "not cached: %u\n", nfsdstats.rcnocache);
David Brazdil0f672f62019-12-10 10:32:29 +0000584 seq_printf(m, "payload misses: %u\n", nn->payload_misses);
585 seq_printf(m, "longest chain len: %u\n", nn->longest_chain);
586 seq_printf(m, "cachesize at longest: %u\n", nn->longest_chain_cachesize);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000587 return 0;
588}
589
590int nfsd_reply_cache_stats_open(struct inode *inode, struct file *file)
591{
David Brazdil0f672f62019-12-10 10:32:29 +0000592 struct nfsd_net *nn = net_generic(file_inode(file)->i_sb->s_fs_info,
593 nfsd_net_id);
594
595 return single_open(file, nfsd_reply_cache_stats_show, nn);
Andrew Scullb4b6d4a2019-01-02 15:54:55 +0000596}