#include <linux/workqueue.h>
#include <linux/mutex.h>
#include <linux/pagemap.h>
-#include <linux/smp_lock.h>
#include <asm/ioctls.h>
#include <linux/sunrpc/types.h>
#include <linux/sunrpc/cache.h>
#include <linux/sunrpc/stats.h>
#include <linux/sunrpc/rpc_pipe_fs.h>
+#include "netns.h"
#define RPCDBG_FACILITY RPCDBG_CACHE
-static int cache_defer_req(struct cache_req *req, struct cache_head *item);
+static void cache_defer_req(struct cache_req *req, struct cache_head *item);
static void cache_revisit_request(struct cache_head *item);
static void cache_init(struct cache_head *h)
}
if (rv == -EAGAIN) {
- if (cache_defer_req(rqstp, h) < 0) {
+ cache_defer_req(rqstp, h);
+ if (!test_bit(CACHE_PENDING, &h->flags)) {
/* Request is not deferred */
rv = cache_is_valid(detail, h);
if (rv == -EAGAIN)
static DEFINE_SPINLOCK(cache_defer_lock);
static LIST_HEAD(cache_defer_list);
-static struct list_head cache_defer_hash[DFR_HASHSIZE];
+static struct hlist_head cache_defer_hash[DFR_HASHSIZE];
static int cache_defer_cnt;
static void __unhash_deferred_req(struct cache_deferred_req *dreq)
{
- list_del_init(&dreq->recent);
- list_del_init(&dreq->hash);
- cache_defer_cnt--;
+ hlist_del_init(&dreq->hash);
+ if (!list_empty(&dreq->recent)) {
+ list_del_init(&dreq->recent);
+ cache_defer_cnt--;
+ }
}
static void __hash_deferred_req(struct cache_deferred_req *dreq, struct cache_head *item)
{
int hash = DFR_HASH(item);
- list_add(&dreq->recent, &cache_defer_list);
- if (cache_defer_hash[hash].next == NULL)
- INIT_LIST_HEAD(&cache_defer_hash[hash]);
- list_add(&dreq->hash, &cache_defer_hash[hash]);
+ INIT_LIST_HEAD(&dreq->recent);
+ hlist_add_head(&dreq->hash, &cache_defer_hash[hash]);
}
-static int setup_deferral(struct cache_deferred_req *dreq, struct cache_head *item)
+static void setup_deferral(struct cache_deferred_req *dreq,
+ struct cache_head *item,
+ int count_me)
{
- struct cache_deferred_req *discard;
dreq->item = item;
__hash_deferred_req(dreq, item);
- /* it is in, now maybe clean up */
- discard = NULL;
- if (++cache_defer_cnt > DFR_MAX) {
- discard = list_entry(cache_defer_list.prev,
- struct cache_deferred_req, recent);
- __unhash_deferred_req(discard);
+ if (count_me) {
+ cache_defer_cnt++;
+ list_add(&dreq->recent, &cache_defer_list);
}
- spin_unlock(&cache_defer_lock);
- if (discard)
- /* there was one too many */
- discard->revisit(discard, 1);
+ spin_unlock(&cache_defer_lock);
- if (!test_bit(CACHE_PENDING, &item->flags)) {
- /* must have just been validated... */
- cache_revisit_request(item);
- return -EAGAIN;
- }
- return 0;
}
struct thread_deferred_req {
complete(&dr->completion);
}
-static int cache_wait_req(struct cache_req *req, struct cache_head *item)
+static void cache_wait_req(struct cache_req *req, struct cache_head *item)
{
struct thread_deferred_req sleeper;
struct cache_deferred_req *dreq = &sleeper.handle;
- int ret;
sleeper.completion = COMPLETION_INITIALIZER_ONSTACK(sleeper.completion);
dreq->revisit = cache_restart_thread;
- ret = setup_deferral(dreq, item);
- if (ret)
- return ret;
+ setup_deferral(dreq, item, 0);
- if (wait_for_completion_interruptible_timeout(
+ if (!test_bit(CACHE_PENDING, &item->flags) ||
+ wait_for_completion_interruptible_timeout(
&sleeper.completion, req->thread_wait) <= 0) {
/* The completion wasn't completed, so we need
* to clean up
*/
spin_lock(&cache_defer_lock);
- if (!list_empty(&sleeper.handle.hash)) {
+ if (!hlist_unhashed(&sleeper.handle.hash)) {
__unhash_deferred_req(&sleeper.handle);
spin_unlock(&cache_defer_lock);
} else {
wait_for_completion(&sleeper.completion);
}
}
- if (test_bit(CACHE_PENDING, &item->flags)) {
- /* item is still pending, try request
- * deferral
- */
- return -ETIMEDOUT;
- }
- /* only return success if we actually deferred the
- * request. In this case we waited until it was
- * answered so no deferral has happened - rather
- * an answer already exists.
+}
+
+static void cache_limit_defers(void)
+{
+ /* Make sure we haven't exceed the limit of allowed deferred
+ * requests.
*/
- return -EEXIST;
+ struct cache_deferred_req *discard = NULL;
+
+ if (cache_defer_cnt <= DFR_MAX)
+ return;
+
+ spin_lock(&cache_defer_lock);
+
+ /* Consider removing either the first or the last */
+ if (cache_defer_cnt > DFR_MAX) {
+ if (net_random() & 1)
+ discard = list_entry(cache_defer_list.next,
+ struct cache_deferred_req, recent);
+ else
+ discard = list_entry(cache_defer_list.prev,
+ struct cache_deferred_req, recent);
+ __unhash_deferred_req(discard);
+ }
+ spin_unlock(&cache_defer_lock);
+ if (discard)
+ discard->revisit(discard, 1);
}
-static int cache_defer_req(struct cache_req *req, struct cache_head *item)
+static void cache_defer_req(struct cache_req *req, struct cache_head *item)
{
struct cache_deferred_req *dreq;
- int ret;
- if (cache_defer_cnt >= DFR_MAX) {
- /* too much in the cache, randomly drop this one,
- * or continue and drop the oldest
- */
- if (net_random()&1)
- return -ENOMEM;
- }
if (req->thread_wait) {
- ret = cache_wait_req(req, item);
- if (ret != -ETIMEDOUT)
- return ret;
+ cache_wait_req(req, item);
+ if (!test_bit(CACHE_PENDING, &item->flags))
+ return;
}
dreq = req->defer(req);
if (dreq == NULL)
- return -ENOMEM;
- return setup_deferral(dreq, item);
+ return;
+ setup_deferral(dreq, item, 1);
+ if (!test_bit(CACHE_PENDING, &item->flags))
+ /* Bit could have been cleared before we managed to
+ * set up the deferral, so need to revisit just in case
+ */
+ cache_revisit_request(item);
+
+ cache_limit_defers();
}
static void cache_revisit_request(struct cache_head *item)
{
struct cache_deferred_req *dreq;
struct list_head pending;
-
- struct list_head *lp;
+ struct hlist_node *lp, *tmp;
int hash = DFR_HASH(item);
INIT_LIST_HEAD(&pending);
spin_lock(&cache_defer_lock);
- lp = cache_defer_hash[hash].next;
- if (lp) {
- while (lp != &cache_defer_hash[hash]) {
- dreq = list_entry(lp, struct cache_deferred_req, hash);
- lp = lp->next;
- if (dreq->item == item) {
- __unhash_deferred_req(dreq);
- list_add(&dreq->recent, &pending);
- }
+ hlist_for_each_entry_safe(dreq, lp, tmp, &cache_defer_hash[hash], hash)
+ if (dreq->item == item) {
+ __unhash_deferred_req(dreq);
+ list_add(&dreq->recent, &pending);
}
- }
+
spin_unlock(&cache_defer_lock);
while (!list_empty(&pending)) {
spin_lock(&cache_defer_lock);
list_for_each_entry_safe(dreq, tmp, &cache_defer_list, recent) {
- if (dreq->owner == owner)
+ if (dreq->owner == owner) {
__unhash_deferred_req(dreq);
+ list_add(&dreq->recent, &pending);
+ }
}
spin_unlock(&cache_defer_lock);
}
}
+static bool cache_listeners_exist(struct cache_detail *detail)
+{
+ if (atomic_read(&detail->readers))
+ return true;
+ if (detail->last_close == 0)
+ /* This cache was never opened */
+ return false;
+ if (detail->last_close < seconds_since_boot() - 30)
+ /*
+ * We allow for the possibility that someone might
+ * restart a userspace daemon without restarting the
+ * server; but after 30 seconds, we give up.
+ */
+ return false;
+ return true;
+}
+
/*
* register an upcall request to user-space and queue it up for read() by the
* upcall daemon.
char *bp;
int len;
- if (atomic_read(&detail->readers) == 0 &&
- detail->last_close < seconds_since_boot() - 30) {
- warn_no_listener(detail);
- return -EINVAL;
+ if (!cache_listeners_exist(detail)) {
+ warn_no_listener(detail);
+ return -EINVAL;
}
buf = kmalloc(PAGE_SIZE, GFP_KERNEL);
if (bp[0] == '\\' && bp[1] == 'x') {
/* HEX STRING */
bp += 2;
- while (isxdigit(bp[0]) && isxdigit(bp[1]) && len < bufsize) {
- int byte = isdigit(*bp) ? *bp-'0' : toupper(*bp)-'A'+10;
- bp++;
- byte <<= 4;
- byte |= isdigit(*bp) ? *bp-'0' : toupper(*bp)-'A'+10;
- *dest++ = byte;
- bp++;
+ while (len < bufsize) {
+ int h, l;
+
+ h = hex_to_bin(bp[0]);
+ if (h < 0)
+ break;
+
+ l = hex_to_bin(bp[1]);
+ if (l < 0)
+ break;
+
+ *dest++ = (h << 4) | l;
+ bp += 2;
len++;
}
} else {
static long cache_ioctl_procfs(struct file *filp,
unsigned int cmd, unsigned long arg)
{
- long ret;
struct inode *inode = filp->f_path.dentry->d_inode;
struct cache_detail *cd = PDE(inode)->data;
- lock_kernel();
- ret = cache_ioctl(inode, filp, cmd, arg, cd);
- unlock_kernel();
-
- return ret;
+ return cache_ioctl(inode, filp, cmd, arg, cd);
}
static int cache_open_procfs(struct inode *inode, struct file *filp)
.read = read_flush_procfs,
.write = write_flush_procfs,
.release = release_flush_procfs,
+ .llseek = no_llseek,
};
-static void remove_cache_proc_entries(struct cache_detail *cd)
+static void remove_cache_proc_entries(struct cache_detail *cd, struct net *net)
{
+ struct sunrpc_net *sn;
+
if (cd->u.procfs.proc_ent == NULL)
return;
if (cd->u.procfs.flush_ent)
if (cd->u.procfs.content_ent)
remove_proc_entry("content", cd->u.procfs.proc_ent);
cd->u.procfs.proc_ent = NULL;
- remove_proc_entry(cd->name, proc_net_rpc);
+ sn = net_generic(net, sunrpc_net_id);
+ remove_proc_entry(cd->name, sn->proc_net_rpc);
}
#ifdef CONFIG_PROC_FS
-static int create_cache_proc_entries(struct cache_detail *cd)
+static int create_cache_proc_entries(struct cache_detail *cd, struct net *net)
{
struct proc_dir_entry *p;
+ struct sunrpc_net *sn;
- cd->u.procfs.proc_ent = proc_mkdir(cd->name, proc_net_rpc);
+ sn = net_generic(net, sunrpc_net_id);
+ cd->u.procfs.proc_ent = proc_mkdir(cd->name, sn->proc_net_rpc);
if (cd->u.procfs.proc_ent == NULL)
goto out_nomem;
cd->u.procfs.channel_ent = NULL;
}
return 0;
out_nomem:
- remove_cache_proc_entries(cd);
+ remove_cache_proc_entries(cd, net);
return -ENOMEM;
}
#else /* CONFIG_PROC_FS */
-static int create_cache_proc_entries(struct cache_detail *cd)
+static int create_cache_proc_entries(struct cache_detail *cd, struct net *net)
{
return 0;
}
INIT_DELAYED_WORK_DEFERRABLE(&cache_cleaner, do_cache_clean);
}
-int cache_register(struct cache_detail *cd)
+int cache_register_net(struct cache_detail *cd, struct net *net)
{
int ret;
sunrpc_init_cache_detail(cd);
- ret = create_cache_proc_entries(cd);
+ ret = create_cache_proc_entries(cd, net);
if (ret)
sunrpc_destroy_cache_detail(cd);
return ret;
}
+
+int cache_register(struct cache_detail *cd)
+{
+ return cache_register_net(cd, &init_net);
+}
EXPORT_SYMBOL_GPL(cache_register);
-void cache_unregister(struct cache_detail *cd)
+void cache_unregister_net(struct cache_detail *cd, struct net *net)
{
- remove_cache_proc_entries(cd);
+ remove_cache_proc_entries(cd, net);
sunrpc_destroy_cache_detail(cd);
}
+
+void cache_unregister(struct cache_detail *cd)
+{
+ cache_unregister_net(cd, &init_net);
+}
EXPORT_SYMBOL_GPL(cache_unregister);
static ssize_t cache_read_pipefs(struct file *filp, char __user *buf,
{
struct inode *inode = filp->f_dentry->d_inode;
struct cache_detail *cd = RPC_I(inode)->private;
- long ret;
-
- lock_kernel();
- ret = cache_ioctl(inode, filp, cmd, arg, cd);
- unlock_kernel();
- return ret;
+ return cache_ioctl(inode, filp, cmd, arg, cd);
}
static int cache_open_pipefs(struct inode *inode, struct file *filp)
.read = read_flush_pipefs,
.write = write_flush_pipefs,
.release = release_flush_pipefs,
+ .llseek = no_llseek,
};
int sunrpc_cache_register_pipefs(struct dentry *parent,
projects / deliverable / linux.git / blobdiff