Merge with /home/shaggy/git/linus-clean/

5 files changed, 73 insertions, 20 deletions

diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c
index 4526da8907..0c9a2ee54c 100644
--- a/fs/fuse/dev.c
+++ b/fs/fuse/dev.c
@@ -66,6 +66,12 @@ static void restore_sigs(sigset_t *oldset)
         sigprocmask(SIG_SETMASK, oldset, NULL);
 }
 
+/*
+ * Reset request, so that it can be reused
+ *
+ * The caller must be _very_ careful to make sure, that it is holding
+ * the only reference to req
+ */
 void fuse_reset_request(struct fuse_req *req)
 {
         int preallocated = req->preallocated;
@@ -120,9 +126,9 @@ struct fuse_req *fuse_get_request(struct fuse_conn *fc)
         return do_get_request(fc);
 }
 
+/* Must be called with fuse_lock held */
 static void fuse_putback_request(struct fuse_conn *fc, struct fuse_req *req)
 {
-        spin_lock(&fuse_lock);
         if (req->preallocated) {
                 atomic_dec(&fc->num_waiting);
                 list_add(&req->list, &fc->unused_list);
@@ -134,11 +140,19 @@ static void fuse_putback_request(struct fuse_conn *fc, struct fuse_req *req)
                 fc->outstanding_debt--;
         else
                 up(&fc->outstanding_sem);
-        spin_unlock(&fuse_lock);
 }
 
 void fuse_put_request(struct fuse_conn *fc, struct fuse_req *req)
 {
+        if (atomic_dec_and_test(&req->count)) {
+                spin_lock(&fuse_lock);
+                fuse_putback_request(fc, req);
+                spin_unlock(&fuse_lock);
+        }
+}
+
+static void fuse_put_request_locked(struct fuse_conn *fc, struct fuse_req *req)
+{
         if (atomic_dec_and_test(&req->count))
                 fuse_putback_request(fc, req);
 }
@@ -163,26 +177,36 @@ void fuse_release_background(struct fuse_req *req)
  * still waiting), the 'end' callback is called if given, else the
  * reference to the request is released
  *
+ * Releasing extra reference for foreground requests must be done
+ * within the same locked region as setting state to finished.  This
+ * is because fuse_reset_request() may be called after request is
+ * finished and it must be the sole possessor.  If request is
+ * interrupted and put in the background, it will return with an error
+ * and hence never be reset and reused.
+ *
  * Called with fuse_lock, unlocks it
  */
 static void request_end(struct fuse_conn *fc, struct fuse_req *req)
 {
-        void (*end) (struct fuse_conn *, struct fuse_req *) = req->end;
-        req->end = NULL;
         list_del(&req->list);
         req->state = FUSE_REQ_FINISHED;
-        spin_unlock(&fuse_lock);
-        if (req->background) {
+        if (!req->background) {
+                wake_up(&req->waitq);
+                fuse_put_request_locked(fc, req);
+                spin_unlock(&fuse_lock);
+        } else {
+                void (*end) (struct fuse_conn *, struct fuse_req *) = req->end;
+                req->end = NULL;
+                spin_unlock(&fuse_lock);
                 down_read(&fc->sbput_sem);
                 if (fc->mounted)
                         fuse_release_background(req);
                 up_read(&fc->sbput_sem);
+                if (end)
+                        end(fc, req);
+                else
+                        fuse_put_request(fc, req);
         }
-        wake_up(&req->waitq);
-        if (end)
-                end(fc, req);
-        else
-                fuse_put_request(fc, req);
 }
 
 /*
diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c
index 21fd59c7bc..c72a8a9793 100644
--- a/fs/fuse/dir.c
+++ b/fs/fuse/dir.c
@@ -111,6 +111,8 @@ static int fuse_dentry_revalidate(struct dentry *entry, struct nameidata *nd)
 
                 /* Doesn't hurt to "reset" the validity timeout */
                 fuse_invalidate_entry_cache(entry);
+
+                /* For negative dentries, always do a fresh lookup */
                 if (!inode)
                         return 0;
 
@@ -122,6 +124,9 @@ static int fuse_dentry_revalidate(struct dentry *entry, struct nameidata *nd)
                 fuse_lookup_init(req, entry->d_parent->d_inode, entry, &outarg);
                 request_send(fc, req);
                 err = req->out.h.error;
+                /* Zero nodeid is same as -ENOENT */
+                if (!err && !outarg.nodeid)
+                        err = -ENOENT;
                 if (!err) {
                         struct fuse_inode *fi = get_fuse_inode(inode);
                         if (outarg.nodeid != get_node_id(inode)) {
@@ -190,8 +195,9 @@ static struct dentry *fuse_lookup(struct inode *dir, struct dentry *entry,
         fuse_lookup_init(req, dir, entry, &outarg);
         request_send(fc, req);
         err = req->out.h.error;
-        if (!err && ((outarg.nodeid && invalid_nodeid(outarg.nodeid)) ||
-                     !valid_mode(outarg.attr.mode)))
+        /* Zero nodeid is same as -ENOENT, but with valid timeout */
+        if (!err && outarg.nodeid &&
+            (invalid_nodeid(outarg.nodeid) || !valid_mode(outarg.attr.mode)))
                 err = -EIO;
         if (!err && outarg.nodeid) {
                 inode = fuse_iget(dir->i_sb, outarg.nodeid, outarg.generation,
diff --git a/fs/fuse/file.c b/fs/fuse/file.c
index a7ef5e716f..6f05379b0a 100644
--- a/fs/fuse/file.c
+++ b/fs/fuse/file.c
@@ -116,9 +116,14 @@ int fuse_open_common(struct inode *inode, struct file *file, int isdir)
 /* Special case for failed iget in CREATE */
 static void fuse_release_end(struct fuse_conn *fc, struct fuse_req *req)
 {
-        u64 nodeid = req->in.h.nodeid;
-        fuse_reset_request(req);
-        fuse_send_forget(fc, req, nodeid, 1);
+        /* If called from end_io_requests(), req has more than one
+           reference and fuse_reset_request() cannot work */
+        if (fc->connected) {
+                u64 nodeid = req->in.h.nodeid;
+                fuse_reset_request(req);
+                fuse_send_forget(fc, req, nodeid, 1);
+        } else
+                fuse_put_request(fc, req);
 }
 
 void fuse_send_release(struct fuse_conn *fc, struct fuse_file *ff,
@@ -335,9 +340,14 @@ static void fuse_send_readpages(struct fuse_req *req, struct file *file,
         loff_t pos = page_offset(req->pages[0]);
         size_t count = req->num_pages << PAGE_CACHE_SHIFT;
         req->out.page_zeroing = 1;
-        req->end = fuse_readpages_end;
         fuse_read_fill(req, file, inode, pos, count, FUSE_READ);
-        request_send_background(fc, req);
+        if (fc->async_read) {
+                req->end = fuse_readpages_end;
+                request_send_background(fc, req);
+        } else {
+                request_send(fc, req);
+                fuse_readpages_end(fc, req);
+        }
 }
 
 struct fuse_readpages_data {
diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h
index 46cf933aa3..4a83adfec9 100644
--- a/fs/fuse/fuse_i.h
+++ b/fs/fuse/fuse_i.h
@@ -272,6 +272,9 @@ struct fuse_conn {
             reply, before any other request, and never cleared */
         unsigned conn_error : 1;
 
+        /** Do readpages asynchronously?  Only set in INIT */
+        unsigned async_read : 1;
+
         /*
          * The following bitfields are only for optimization purposes
          * and hence races in setting them will not cause malfunction
diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c
index c755a0440a..879e6fba94 100644
--- a/fs/fuse/inode.c
+++ b/fs/fuse/inode.c
@@ -473,6 +473,16 @@ static void process_init_reply(struct fuse_conn *fc, struct fuse_req *req)
         if (req->out.h.error || arg->major != FUSE_KERNEL_VERSION)
                 fc->conn_error = 1;
         else {
+                unsigned long ra_pages;
+
+                if (arg->minor >= 6) {
+                        ra_pages = arg->max_readahead / PAGE_CACHE_SIZE;
+                        if (arg->flags & FUSE_ASYNC_READ)
+                                fc->async_read = 1;
+                } else
+                        ra_pages = fc->max_read / PAGE_CACHE_SIZE;
+
+                fc->bdi.ra_pages = min(fc->bdi.ra_pages, ra_pages);
                 fc->minor = arg->minor;
                 fc->max_write = arg->minor < 5 ? 4096 : arg->max_write;
         }
@@ -496,6 +506,8 @@ static void fuse_send_init(struct fuse_conn *fc)
 
         arg->major = FUSE_KERNEL_VERSION;
         arg->minor = FUSE_KERNEL_MINOR_VERSION;
+        arg->max_readahead = fc->bdi.ra_pages * PAGE_CACHE_SIZE;
+        arg->flags |= FUSE_ASYNC_READ;
         req->in.h.opcode = FUSE_INIT;
         req->in.numargs = 1;
         req->in.args[0].size = sizeof(*arg);
@@ -552,8 +564,6 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent)
         fc->user_id = d.user_id;
         fc->group_id = d.group_id;
         fc->max_read = d.max_read;
-        if (fc->max_read / PAGE_CACHE_SIZE < fc->bdi.ra_pages)
-                fc->bdi.ra_pages = fc->max_read / PAGE_CACHE_SIZE;
 
         /* Used by get_root_inode() */
         sb->s_fs_info = fc;