Merge branch 'afs-dh' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs

Pull AFS updates from Al Viro:
 "The AFS series posted by dhowells depended upon lookup_one_len()
  rework; now that prereq is in the mainline, that series had been
  rebased on top of it and got some exposure and testing..."

* 'afs-dh' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
  afs: Do better accretion of small writes on newly created content
  afs: Add stats for data transfer operations
  afs: Trace protocol errors
  afs: Locally edit directory data for mkdir/create/unlink/...
  afs: Adjust the directory XDR structures
  afs: Split the directory content defs into a header
  afs: Fix directory handling
  afs: Split the dynroot stuff out and give it its own ops tables
  afs: Keep track of invalid-before version for dentry coherency
  afs: Rearrange status mapping
  afs: Make it possible to get the data version in readpage
  afs: Init inode before accessing cache
  afs: Introduce a statistics proc file
  afs: Dump bad status record
  afs: Implement @cell substitution handling
  afs: Implement @sys substitution handling
  afs: Prospectively look up extra files when doing a single lookup
  afs: Don't over-increment the cell usage count when pinning it
  afs: Fix checker warnings
  vfs: Remove the const from dir_context::actor

Documentation/filesystems/afs.txt

@@ -11,7 +11,7 @@ Contents:
  - Proc filesystem.
  - The cell database.
  - Security.
- - Examples.
+ - The @sys substitution.
 
 
 ========
@@ -230,3 +230,29 @@ If a file is opened with a particular key and then the file descriptor is
 passed to a process that doesn't have that key (perhaps over an AF_UNIX
 socket), then the operations on the file will be made with key that was used to
 open the file.
+
+
+=====================
+THE @SYS SUBSTITUTION
+=====================
+
+The list of up to 16 @sys substitutions for the current network namespace can
+be configured by writing a list to /proc/fs/afs/sysname:
+
+	[root@andromeda ~]# echo foo amd64_linux_26 >/proc/fs/afs/sysname
+
+or cleared entirely by writing an empty list:
+
+	[root@andromeda ~]# echo >/proc/fs/afs/sysname
+
+The current list for current network namespace can be retrieved by:
+
+	[root@andromeda ~]# cat /proc/fs/afs/sysname
+	foo
+	amd64_linux_26
+
+When @sys is being substituted for, each element of the list is tried in the
+order given.
+
+By default, the list will contain one item that conforms to the pattern
+"<arch>_linux_26", amd64 being the name for x86_64.

fs/afs/Makefile

@@ -12,6 +12,8 @@ kafs-objs := \
 	cell.o \
 	cmservice.o \
 	dir.o \
+	dir_edit.o \
+	dynroot.o \
 	file.o \
 	flock.o \
 	fsclient.o \

fs/afs/addr_list.c

@@ -243,9 +243,9 @@ void afs_merge_fs_addr4(struct afs_addr_list *alist, __be32 xdr, u16 port)
 		    xport == a->sin6_port)
 			return;
 		if (xdr == a->sin6_addr.s6_addr32[3] &&
-		    xport < a->sin6_port)
+		    (u16 __force)xport < (u16 __force)a->sin6_port)
 			break;
-		if (xdr < a->sin6_addr.s6_addr32[3])
+		if ((u32 __force)xdr < (u32 __force)a->sin6_addr.s6_addr32[3])
 			break;
 	}
 
@@ -280,7 +280,7 @@ void afs_merge_fs_addr6(struct afs_addr_list *alist, __be32 *xdr, u16 port)
 		    xport == a->sin6_port)
 			return;
 		if (diff == 0 &&
-		    xport < a->sin6_port)
+		    (u16 __force)xport < (u16 __force)a->sin6_port)
 			break;
 		if (diff < 0)
 			break;

fs/afs/afs.h

@@ -67,10 +67,14 @@ typedef enum {
 } afs_callback_type_t;
 
 struct afs_callback {
-	struct afs_fid		fid;		/* file identifier */
-	unsigned		version;	/* callback version */
-	unsigned		expiry;		/* time at which expires */
-	afs_callback_type_t	type;		/* type of callback */
+	unsigned		version;	/* Callback version */
+	unsigned		expiry;		/* Time at which expires */
+	afs_callback_type_t	type;		/* Type of callback */
+};
+
+struct afs_callback_break {
+	struct afs_fid		fid;		/* File identifier */
+	struct afs_callback	cb;		/* Callback details */
 };
 
 #define AFSCBMAX 50	/* maximum callbacks transferred per bulk op */
@@ -123,21 +127,20 @@ typedef u32 afs_access_t;
  * AFS file status information
  */
 struct afs_file_status {
-	unsigned		if_version;	/* interface version */
-#define AFS_FSTATUS_VERSION	1
+	u64			size;		/* file size */
+	afs_dataversion_t	data_version;	/* current data version */
+	time_t			mtime_client;	/* last time client changed data */
+	time_t			mtime_server;	/* last time server changed data */
+	unsigned		abort_code;	/* Abort if bulk-fetching this failed */
 
 	afs_file_type_t		type;		/* file type */
 	unsigned		nlink;		/* link count */
-	u64			size;		/* file size */
-	afs_dataversion_t	data_version;	/* current data version */
 	u32			author;		/* author ID */
-	kuid_t			owner;		/* owner ID */
-	kgid_t			group;		/* group ID */
+	u32			owner;		/* owner ID */
+	u32			group;		/* group ID */
 	afs_access_t		caller_access;	/* access rights for authenticated caller */
 	afs_access_t		anon_access;	/* access rights for unauthenticated caller */
 	umode_t			mode;		/* UNIX mode */
-	time_t			mtime_client;	/* last time client changed data */
-	time_t			mtime_server;	/* last time server changed data */
 	s32			lock_count;	/* file lock count (0=UNLK -1=WRLCK +ve=#RDLCK */
 };
 

fs/afs/afs_fs.h

@@ -31,10 +31,12 @@ enum AFS_FS_Operations {
 	FSGETVOLUMEINFO		= 148,	/* AFS Get information about a volume */
 	FSGETVOLUMESTATUS	= 149,	/* AFS Get volume status information */
 	FSGETROOTVOLUME		= 151,	/* AFS Get root volume name */
+	FSBULKSTATUS		= 155,	/* AFS Fetch multiple file statuses */
 	FSSETLOCK		= 156,	/* AFS Request a file lock */
 	FSEXTENDLOCK		= 157,	/* AFS Extend a file lock */
 	FSRELEASELOCK		= 158,	/* AFS Release a file lock */
 	FSLOOKUP		= 161,	/* AFS lookup file in directory */
+	FSINLINEBULKSTATUS	= 65536, /* AFS Fetch multiple file statuses with inline errors */
 	FSFETCHDATA64		= 65537, /* AFS Fetch file data */
 	FSSTOREDATA64		= 65538, /* AFS Store file data */
 	FSGIVEUPALLCALLBACKS	= 65539, /* AFS Give up all outstanding callbacks on a server */

fs/afs/callback.c

@@ -96,26 +96,6 @@ int afs_register_server_cb_interest(struct afs_vnode *vnode,
 	return 0;
 }
 
-/*
- * Set a vnode's interest on a server.
- */
-void afs_set_cb_interest(struct afs_vnode *vnode, struct afs_cb_interest *cbi)
-{
-	struct afs_cb_interest *old_cbi = NULL;
-
-	if (vnode->cb_interest == cbi)
-		return;
-
-	write_seqlock(&vnode->cb_lock);
-	if (vnode->cb_interest != cbi) {
-		afs_get_cb_interest(cbi);
-		old_cbi = vnode->cb_interest;
-		vnode->cb_interest = cbi;
-	}
-	write_sequnlock(&vnode->cb_lock);
-	afs_put_cb_interest(afs_v2net(vnode), cbi);
-}
-
 /*
  * Remove an interest on a server.
  */
@@ -150,6 +130,7 @@ void afs_break_callback(struct afs_vnode *vnode)
 
 	write_seqlock(&vnode->cb_lock);
 
+	clear_bit(AFS_VNODE_NEW_CONTENT, &vnode->flags);
 	if (test_and_clear_bit(AFS_VNODE_CB_PROMISED, &vnode->flags)) {
 		vnode->cb_break++;
 		afs_clear_permits(vnode);
@@ -207,7 +188,7 @@ static void afs_break_one_callback(struct afs_server *server,
  * allow the fileserver to break callback promises
  */
 void afs_break_callbacks(struct afs_server *server, size_t count,
-			 struct afs_callback callbacks[])
+			 struct afs_callback_break *callbacks)
 {
 	_enter("%p,%zu,", server, count);
 
@@ -219,9 +200,9 @@ void afs_break_callbacks(struct afs_server *server, size_t count,
 		       callbacks->fid.vid,
 		       callbacks->fid.vnode,
 		       callbacks->fid.unique,
-		       callbacks->version,
-		       callbacks->expiry,
-		       callbacks->type
+		       callbacks->cb.version,
+		       callbacks->cb.expiry,
+		       callbacks->cb.type
 		       );
 		afs_break_one_callback(server, &callbacks->fid);
 	}

fs/afs/cell.c

@@ -18,7 +18,7 @@
 #include <keys/rxrpc-type.h>
 #include "internal.h"
 
-unsigned __read_mostly afs_cell_gc_delay = 10;
+static unsigned __read_mostly afs_cell_gc_delay = 10;
 
 static void afs_manage_cell(struct work_struct *);
 
@@ -75,7 +75,7 @@ struct afs_cell *afs_lookup_cell_rcu(struct afs_net *net,
 			cell = rcu_dereference_raw(net->ws_cell);
 			if (cell) {
 				afs_get_cell(cell);
-				continue;
+				break;
 			}
 			ret = -EDESTADDRREQ;
 			continue;
@@ -130,6 +130,8 @@ static struct afs_cell *afs_alloc_cell(struct afs_net *net,
 		_leave(" = -ENAMETOOLONG");
 		return ERR_PTR(-ENAMETOOLONG);
 	}
+	if (namelen == 5 && memcmp(name, "@cell", 5) == 0)
+		return ERR_PTR(-EINVAL);
 
 	_enter("%*.*s,%s", namelen, namelen, name, vllist);
 
@@ -334,8 +336,8 @@ int afs_cell_init(struct afs_net *net, const char *rootcell)
 		return PTR_ERR(new_root);
 	}
 
-	set_bit(AFS_CELL_FL_NO_GC, &new_root->flags);
-	afs_get_cell(new_root);
+	if (!test_and_set_bit(AFS_CELL_FL_NO_GC, &new_root->flags))
+		afs_get_cell(new_root);
 
 	/* install the new cell */
 	write_seqlock(&net->cells_lock);
@@ -411,7 +413,7 @@ static void afs_cell_destroy(struct rcu_head *rcu)
 
 	ASSERTCMP(atomic_read(&cell->usage), ==, 0);
 
-	afs_put_addrlist(cell->vl_addrs);
+	afs_put_addrlist(rcu_access_pointer(cell->vl_addrs));
 	key_put(cell->anonymous_key);
 	kfree(cell);
 

fs/afs/cmservice.c

@@ -178,8 +178,8 @@ static void SRXAFSCB_CallBack(struct work_struct *work)
  */
 static int afs_deliver_cb_callback(struct afs_call *call)
 {
+	struct afs_callback_break *cb;
 	struct sockaddr_rxrpc srx;
-	struct afs_callback *cb;
 	struct afs_server *server;
 	__be32 *bp;
 	int ret, loop;
@@ -201,7 +201,7 @@ static int afs_deliver_cb_callback(struct afs_call *call)
 		call->count = ntohl(call->tmp);
 		_debug("FID count: %u", call->count);
 		if (call->count > AFSCBMAX)
-			return -EBADMSG;
+			return afs_protocol_error(call, -EBADMSG);
 
 		call->buffer = kmalloc(call->count * 3 * 4, GFP_KERNEL);
 		if (!call->buffer)
@@ -218,7 +218,7 @@ static int afs_deliver_cb_callback(struct afs_call *call)
 
 		_debug("unmarshall FID array");
 		call->request = kcalloc(call->count,
-					sizeof(struct afs_callback),
+					sizeof(struct afs_callback_break),
 					GFP_KERNEL);
 		if (!call->request)
 			return -ENOMEM;
@@ -229,7 +229,7 @@ static int afs_deliver_cb_callback(struct afs_call *call)
 			cb->fid.vid	= ntohl(*bp++);
 			cb->fid.vnode	= ntohl(*bp++);
 			cb->fid.unique	= ntohl(*bp++);
-			cb->type	= AFSCM_CB_UNTYPED;
+			cb->cb.type	= AFSCM_CB_UNTYPED;
 		}
 
 		call->offset = 0;
@@ -245,7 +245,7 @@ static int afs_deliver_cb_callback(struct afs_call *call)
 		call->count2 = ntohl(call->tmp);
 		_debug("CB count: %u", call->count2);
 		if (call->count2 != call->count && call->count2 != 0)
-			return -EBADMSG;
+			return afs_protocol_error(call, -EBADMSG);
 		call->offset = 0;
 		call->unmarshall++;
 
@@ -260,9 +260,9 @@ static int afs_deliver_cb_callback(struct afs_call *call)
 		cb = call->request;
 		bp = call->buffer;
 		for (loop = call->count2; loop > 0; loop--, cb++) {
-			cb->version	= ntohl(*bp++);
-			cb->expiry	= ntohl(*bp++);
-			cb->type	= ntohl(*bp++);
+			cb->cb.version	= ntohl(*bp++);
+			cb->cb.expiry	= ntohl(*bp++);
+			cb->cb.type	= ntohl(*bp++);
 		}
 
 		call->offset = 0;
@@ -500,9 +500,9 @@ static int afs_deliver_cb_probe_uuid(struct afs_call *call)
 
 		b = call->buffer;
 		r = call->request;
-		r->time_low			= ntohl(b[0]);
-		r->time_mid			= ntohl(b[1]);
-		r->time_hi_and_version		= ntohl(b[2]);
+		r->time_low			= b[0];
+		r->time_mid			= htons(ntohl(b[1]));
+		r->time_hi_and_version		= htons(ntohl(b[2]));
 		r->clock_seq_hi_and_reserved 	= ntohl(b[3]);
 		r->clock_seq_low		= ntohl(b[4]);
 

fs/afs/dynroot.c

+/* dir.c: AFS dynamic root handling
+ *
+ * Copyright (C) 2018 Red Hat, Inc. All Rights Reserved.
+ * Written by David Howells (dhowells@redhat.com)
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public Licence
+ * as published by the Free Software Foundation; either version
+ * 2 of the Licence, or (at your option) any later version.
+ */
+
+#include <linux/fs.h>
+#include <linux/namei.h>
+#include <linux/dns_resolver.h>
+#include "internal.h"
+
+const struct file_operations afs_dynroot_file_operations = {
+	.open		= dcache_dir_open,
+	.release	= dcache_dir_close,
+	.iterate_shared	= dcache_readdir,
+	.llseek		= dcache_dir_lseek,
+};
+
+/*
+ * Probe to see if a cell may exist.  This prevents positive dentries from
+ * being created unnecessarily.
+ */
+static int afs_probe_cell_name(struct dentry *dentry)
+{
+	struct afs_cell *cell;
+	const char *name = dentry->d_name.name;
+	size_t len = dentry->d_name.len;
+	int ret;
+
+	/* Names prefixed with a dot are R/W mounts. */
+	if (name[0] == '.') {
+		if (len == 1)
+			return -EINVAL;
+		name++;
+		len--;
+	}
+
+	cell = afs_lookup_cell_rcu(afs_d2net(dentry), name, len);
+	if (!IS_ERR(cell)) {
+		afs_put_cell(afs_d2net(dentry), cell);
+		return 0;
+	}
+
+	ret = dns_query("afsdb", name, len, "ipv4", NULL, NULL);
+	if (ret == -ENODATA)
+		ret = -EDESTADDRREQ;
+	return ret;
+}
+
+/*
+ * Try to auto mount the mountpoint with pseudo directory, if the autocell
+ * operation is setted.
+ */
+struct inode *afs_try_auto_mntpt(struct dentry *dentry, struct inode *dir)
+{
+	struct afs_vnode *vnode = AFS_FS_I(dir);
+	struct inode *inode;
+	int ret = -ENOENT;
+
+	_enter("%p{%pd}, {%x:%u}",
+	       dentry, dentry, vnode->fid.vid, vnode->fid.vnode);
+
+	if (!test_bit(AFS_VNODE_AUTOCELL, &vnode->flags))
+		goto out;
+
+	ret = afs_probe_cell_name(dentry);
+	if (ret < 0)
+		goto out;
+
+	inode = afs_iget_pseudo_dir(dir->i_sb, false);
+	if (IS_ERR(inode)) {
+		ret = PTR_ERR(inode);
+		goto out;
+	}
+
+	_leave("= %p", inode);
+	return inode;
+
+out:
+	_leave("= %d", ret);
+	return ERR_PTR(ret);
+}
+
+/*
+ * Look up @cell in a dynroot directory.  This is a substitution for the
+ * local cell name for the net namespace.
+ */
+static struct dentry *afs_lookup_atcell(struct dentry *dentry)
+{
+	struct afs_cell *cell;
+	struct afs_net *net = afs_d2net(dentry);
+	struct dentry *ret;
+	unsigned int seq = 0;
+	char *name;
+	int len;
+
+	if (!net->ws_cell)
+		return ERR_PTR(-ENOENT);
+
+	ret = ERR_PTR(-ENOMEM);
+	name = kmalloc(AFS_MAXCELLNAME + 1, GFP_KERNEL);
+	if (!name)
+		goto out_p;
+
+	rcu_read_lock();
+	do {
+		read_seqbegin_or_lock(&net->cells_lock, &seq);
+		cell = rcu_dereference_raw(net->ws_cell);
+		if (cell) {
+			len = cell->name_len;
+			memcpy(name, cell->name, len + 1);
+		}
+	} while (need_seqretry(&net->cells_lock, seq));
+	done_seqretry(&net->cells_lock, seq);
+	rcu_read_unlock();
+
+	ret = ERR_PTR(-ENOENT);
+	if (!cell)
+		goto out_n;
+
+	ret = lookup_one_len(name, dentry->d_parent, len);
+
+	/* We don't want to d_add() the @cell dentry here as we don't want to
+	 * the cached dentry to hide changes to the local cell name.
+	 */
+
+out_n:
+	kfree(name);
+out_p:
+	return ret;
+}
+
+/*
+ * Look up an entry in a dynroot directory.
+ */
+static struct dentry *afs_dynroot_lookup(struct inode *dir, struct dentry *dentry,
+					 unsigned int flags)
+{
+	struct afs_vnode *vnode;
+	struct inode *inode;
+	int ret;
+
+	vnode = AFS_FS_I(dir);
+
+	_enter("%pd", dentry);
+
+	ASSERTCMP(d_inode(dentry), ==, NULL);
+
+	if (dentry->d_name.len >= AFSNAMEMAX) {
+		_leave(" = -ENAMETOOLONG");
+		return ERR_PTR(-ENAMETOOLONG);
+	}
+
+	if (dentry->d_name.len == 5 &&
+	    memcmp(dentry->d_name.name, "@cell", 5) == 0)
+		return afs_lookup_atcell(dentry);
+
+	inode = afs_try_auto_mntpt(dentry, dir);
+	if (IS_ERR(inode)) {
+		ret = PTR_ERR(inode);
+		if (ret == -ENOENT) {
+			d_add(dentry, NULL);
+			_leave(" = NULL [negative]");
+			return NULL;
+		}
+		_leave(" = %d [do]", ret);
+		return ERR_PTR(ret);
+	}
+
+	d_add(dentry, inode);
+	_leave(" = 0 { ino=%lu v=%u }",
+	       d_inode(dentry)->i_ino, d_inode(dentry)->i_generation);
+	return NULL;
+}
+
+const struct inode_operations afs_dynroot_inode_operations = {
+	.lookup		= afs_dynroot_lookup,
+};
+
+/*
+ * Dirs in the dynamic root don't need revalidation.
+ */
+static int afs_dynroot_d_revalidate(struct dentry *dentry, unsigned int flags)
+{
+	return 1;
+}
+
+/*
+ * Allow the VFS to enquire as to whether a dentry should be unhashed (mustn't
+ * sleep)
+ * - called from dput() when d_count is going to 0.
+ * - return 1 to request dentry be unhashed, 0 otherwise
+ */
+static int afs_dynroot_d_delete(const struct dentry *dentry)
+{
+	return d_really_is_positive(dentry);
+}
+
+const struct dentry_operations afs_dynroot_dentry_operations = {
+	.d_revalidate	= afs_dynroot_d_revalidate,
+	.d_delete	= afs_dynroot_d_delete,
+	.d_release	= afs_d_release,
+	.d_automount	= afs_d_automount,
+};

fs/afs/file.c

@@ -30,7 +30,6 @@ static int afs_readpages(struct file *filp, struct address_space *mapping,
 
 const struct file_operations afs_file_operations = {
 	.open		= afs_open,
-	.flush		= afs_flush,
 	.release	= afs_release,
 	.llseek		= generic_file_llseek,
 	.read_iter	= generic_file_read_iter,
@@ -146,6 +145,9 @@ int afs_open(struct inode *inode, struct file *file)
 		if (ret < 0)
 			goto error_af;
 	}
+
+	if (file->f_flags & O_TRUNC)
+		set_bit(AFS_VNODE_NEW_CONTENT, &vnode->flags);
 	
 	file->private_data = af;
 	_leave(" = 0");
@@ -170,6 +172,9 @@ int afs_release(struct inode *inode, struct file *file)
 
 	_enter("{%x:%u},", vnode->fid.vid, vnode->fid.vnode);
 
+	if ((file->f_mode & FMODE_WRITE))
+		return vfs_fsync(file, 0);
+
 	file->private_data = NULL;
 	if (af->wb)
 		afs_put_wb_key(af->wb);
@@ -187,10 +192,12 @@ void afs_put_read(struct afs_read *req)
 {
 	int i;
 
-	if (atomic_dec_and_test(&req->usage)) {
+	if (refcount_dec_and_test(&req->usage)) {
 		for (i = 0; i < req->nr_pages; i++)
 			if (req->pages[i])
 				put_page(req->pages[i]);
+		if (req->pages != req->array)
+			kfree(req->pages);
 		kfree(req);
 	}
 }
@@ -240,6 +247,12 @@ int afs_fetch_data(struct afs_vnode *vnode, struct key *key, struct afs_read *de
 		ret = afs_end_vnode_operation(&fc);
 	}
 
+	if (ret == 0) {
+		afs_stat_v(vnode, n_fetches);
+		atomic_long_add(desc->actual_len,
+				&afs_v2net(vnode)->n_fetch_bytes);
+	}
+
 	_leave(" = %d", ret);
 	return ret;
 }
@@ -297,10 +310,11 @@ int afs_page_filler(void *data, struct page *page)
 		 * end of the file, the server will return a short read and the
 		 * unmarshalling code will clear the unfilled space.
 		 */
-		atomic_set(&req->usage, 1);
+		refcount_set(&req->usage, 1);
 		req->pos = (loff_t)page->index << PAGE_SHIFT;
 		req->len = PAGE_SIZE;
 		req->nr_pages = 1;
+		req->pages = req->array;
 		req->pages[0] = page;
 		get_page(page);
 
@@ -309,10 +323,6 @@ int afs_page_filler(void *data, struct page *page)
 		ret = afs_fetch_data(vnode, key, req);
 		afs_put_read(req);
 
-		if (ret >= 0 && S_ISDIR(inode->i_mode) &&
-		    !afs_dir_check_page(inode, page))
-			ret = -EIO;
-
 		if (ret < 0) {
 			if (ret == -ENOENT) {
 				_debug("got NOENT from server"
@@ -447,10 +457,11 @@ static int afs_readpages_one(struct file *file, struct address_space *mapping,
 	if (!req)
 		return -ENOMEM;
 
-	atomic_set(&req->usage, 1);
+	refcount_set(&req->usage, 1);
 	req->page_done = afs_readpages_page_done;
 	req->pos = first->index;
 	req->pos <<= PAGE_SHIFT;
+	req->pages = req->array;
 
 	/* Transfer the pages to the request.  We add them in until one fails
 	 * to add to the LRU and then we stop (as that'll make a hole in the

fs/afs/flock.c

@@ -613,7 +613,7 @@ static int afs_do_getlk(struct file *file, struct file_lock *fl)
 	posix_test_lock(file, fl);
 	if (fl->fl_type == F_UNLCK) {
 		/* no local locks; consult the server */
-		ret = afs_fetch_status(vnode, key);
+		ret = afs_fetch_status(vnode, key, false);
 		if (ret < 0)
 			goto error;
 

fs/afs/inode.c

@@ -30,12 +30,11 @@ static const struct inode_operations afs_symlink_inode_operations = {
 };
 
 /*
- * map the AFS file status to the inode member variables
+ * Initialise an inode from the vnode status.
  */
-static int afs_inode_map_status(struct afs_vnode *vnode, struct key *key)
+static int afs_inode_init_from_status(struct afs_vnode *vnode, struct key *key)
 {
 	struct inode *inode = AFS_VNODE_TO_I(vnode);
-	bool changed;
 
 	_debug("FS: ft=%d lk=%d sz=%llu ver=%Lu mod=%hu",
 	       vnode->status.type,
@@ -46,16 +45,21 @@ static int afs_inode_map_status(struct afs_vnode *vnode, struct key *key)
 
 	read_seqlock_excl(&vnode->cb_lock);
 
+	afs_update_inode_from_status(vnode, &vnode->status, NULL,
+				     AFS_VNODE_NOT_YET_SET);
+
 	switch (vnode->status.type) {
 	case AFS_FTYPE_FILE:
 		inode->i_mode	= S_IFREG | vnode->status.mode;
 		inode->i_op	= &afs_file_inode_operations;
 		inode->i_fop	= &afs_file_operations;
+		inode->i_mapping->a_ops	= &afs_fs_aops;
 		break;
 	case AFS_FTYPE_DIR:
 		inode->i_mode	= S_IFDIR | vnode->status.mode;
 		inode->i_op	= &afs_dir_inode_operations;
 		inode->i_fop	= &afs_dir_file_operations;
+		inode->i_mapping->a_ops	= &afs_dir_aops;
 		break;
 	case AFS_FTYPE_SYMLINK:
 		/* Symlinks with a mode of 0644 are actually mountpoints. */
@@ -67,45 +71,31 @@ static int afs_inode_map_status(struct afs_vnode *vnode, struct key *key)
 			inode->i_mode	= S_IFDIR | 0555;
 			inode->i_op	= &afs_mntpt_inode_operations;
 			inode->i_fop	= &afs_mntpt_file_operations;
+			inode->i_mapping->a_ops	= &afs_fs_aops;
 		} else {
 			inode->i_mode	= S_IFLNK | vnode->status.mode;
 			inode->i_op	= &afs_symlink_inode_operations;
+			inode->i_mapping->a_ops	= &afs_fs_aops;
 		}
 		inode_nohighmem(inode);
 		break;
 	default:
 		printk("kAFS: AFS vnode with undefined type\n");
 		read_sequnlock_excl(&vnode->cb_lock);
-		return -EBADMSG;
+		return afs_protocol_error(NULL, -EBADMSG);
 	}
 
-	changed = (vnode->status.size != inode->i_size);
-
-	set_nlink(inode, vnode->status.nlink);
-	inode->i_uid		= vnode->status.owner;
-	inode->i_gid            = vnode->status.group;
-	inode->i_size		= vnode->status.size;
-	inode->i_ctime.tv_sec	= vnode->status.mtime_client;
-	inode->i_ctime.tv_nsec	= 0;
-	inode->i_atime		= inode->i_mtime = inode->i_ctime;
 	inode->i_blocks		= 0;
-	inode->i_generation	= vnode->fid.unique;
-	inode_set_iversion_raw(inode, vnode->status.data_version);
-	inode->i_mapping->a_ops	= &afs_fs_aops;
+	vnode->invalid_before	= vnode->status.data_version;
 
 	read_sequnlock_excl(&vnode->cb_lock);
-
-#ifdef CONFIG_AFS_FSCACHE
-	if (changed)
-		fscache_attr_changed(vnode->cache);
-#endif
 	return 0;
 }
 
 /*
  * Fetch file status from the volume.
  */
-int afs_fetch_status(struct afs_vnode *vnode, struct key *key)
+int afs_fetch_status(struct afs_vnode *vnode, struct key *key, bool new_inode)
 {
 	struct afs_fs_cursor fc;
 	int ret;
@@ -119,7 +109,7 @@ int afs_fetch_status(struct afs_vnode *vnode, struct key *key)
 	if (afs_begin_vnode_operation(&fc, vnode, key)) {
 		while (afs_select_fileserver(&fc)) {
 			fc.cb_break = vnode->cb_break + vnode->cb_s_break;
-			afs_fs_fetch_file_status(&fc, NULL);
+			afs_fs_fetch_file_status(&fc, NULL, new_inode);
 		}
 
 		afs_check_for_remote_deletion(&fc, fc.vnode);
@@ -255,6 +245,11 @@ static void afs_get_inode_cache(struct afs_vnode *vnode)
 	} __packed key;
 	struct afs_vnode_cache_aux aux;
 
+	if (vnode->status.type == AFS_FTYPE_DIR) {
+		vnode->cache = NULL;
+		return;
+	}
+
 	key.vnode_id		= vnode->fid.vnode;
 	key.unique		= vnode->fid.unique;
 	key.vnode_id_ext[0]	= 0;
@@ -307,7 +302,7 @@ struct inode *afs_iget(struct super_block *sb, struct key *key,
 
 	if (!status) {
 		/* it's a remotely extant inode */
-		ret = afs_fetch_status(vnode, key);
+		ret = afs_fetch_status(vnode, key, true);
 		if (ret < 0)
 			goto bad_inode;
 	} else {
@@ -331,15 +326,12 @@ struct inode *afs_iget(struct super_block *sb, struct key *key,
 		vnode->cb_expires_at += ktime_get_real_seconds();
 	}
 
-	/* set up caching before mapping the status, as map-status reads the
-	 * first page of symlinks to see if they're really mountpoints */
-	inode->i_size = vnode->status.size;
-	afs_get_inode_cache(vnode);
-
-	ret = afs_inode_map_status(vnode, key);
+	ret = afs_inode_init_from_status(vnode, key);
 	if (ret < 0)
 		goto bad_inode;
 
+	afs_get_inode_cache(vnode);
+
 	/* success */
 	clear_bit(AFS_VNODE_UNSET, &vnode->flags);
 	inode->i_flags |= S_NOATIME;
@@ -349,10 +341,6 @@ struct inode *afs_iget(struct super_block *sb, struct key *key,
 
 	/* failure */
 bad_inode:
-#ifdef CONFIG_AFS_FSCACHE
-	fscache_relinquish_cookie(vnode->cache, NULL, ret == -ENOENT);
-	vnode->cache = NULL;
-#endif
 	iget_failed(inode);
 	_leave(" = %d [bad]", ret);
 	return ERR_PTR(ret);
@@ -407,8 +395,11 @@ int afs_validate(struct afs_vnode *vnode, struct key *key)
 	if (test_bit(AFS_VNODE_CB_PROMISED, &vnode->flags)) {
 		if (vnode->cb_s_break != vnode->cb_interest->server->cb_s_break) {
 			vnode->cb_s_break = vnode->cb_interest->server->cb_s_break;
-		} else if (!test_bit(AFS_VNODE_DIR_MODIFIED, &vnode->flags) &&
-			   !test_bit(AFS_VNODE_ZAP_DATA, &vnode->flags) &&
+		} else if (vnode->status.type == AFS_FTYPE_DIR &&
+			   test_bit(AFS_VNODE_DIR_VALID, &vnode->flags) &&
+			   vnode->cb_expires_at - 10 > now) {
+				valid = true;
+		} else if (!test_bit(AFS_VNODE_ZAP_DATA, &vnode->flags) &&
 			   vnode->cb_expires_at - 10 > now) {
 				valid = true;
 		}
@@ -432,7 +423,7 @@ int afs_validate(struct afs_vnode *vnode, struct key *key)
 	 * access */
 	if (!test_bit(AFS_VNODE_CB_PROMISED, &vnode->flags)) {
 		_debug("not promised");
-		ret = afs_fetch_status(vnode, key);
+		ret = afs_fetch_status(vnode, key, false);
 		if (ret < 0) {
 			if (ret == -ENOENT) {
 				set_bit(AFS_VNODE_DELETED, &vnode->flags);
@@ -453,8 +444,6 @@ int afs_validate(struct afs_vnode *vnode, struct key *key)
 	 * different */
 	if (test_and_clear_bit(AFS_VNODE_ZAP_DATA, &vnode->flags))
 		afs_zap_data(vnode);
-
-	clear_bit(AFS_VNODE_DIR_MODIFIED, &vnode->flags);
 	mutex_unlock(&vnode->validate_lock);
 valid:
 	_leave(" = 0");
@@ -544,7 +533,7 @@ void afs_evict_inode(struct inode *inode)
 	}
 #endif
 
-	afs_put_permits(vnode->permit_cache);
+	afs_put_permits(rcu_access_pointer(vnode->permit_cache));
 	_leave("");
 }
 

fs/afs/main.c

@@ -34,11 +34,42 @@ MODULE_PARM_DESC(rootcell, "root AFS cell name and VL server IP addr list");
 struct workqueue_struct *afs_wq;
 struct afs_net __afs_net;
 
+#if defined(CONFIG_ALPHA)
+const char afs_init_sysname[] = "alpha_linux26";
+#elif defined(CONFIG_X86_64)
+const char afs_init_sysname[] = "amd64_linux26";
+#elif defined(CONFIG_ARM)
+const char afs_init_sysname[] = "arm_linux26";
+#elif defined(CONFIG_ARM64)
+const char afs_init_sysname[] = "aarch64_linux26";
+#elif defined(CONFIG_X86_32)
+const char afs_init_sysname[] = "i386_linux26";
+#elif defined(CONFIG_IA64)
+const char afs_init_sysname[] = "ia64_linux26";
+#elif defined(CONFIG_PPC64)
+const char afs_init_sysname[] = "ppc64_linux26";
+#elif defined(CONFIG_PPC32)
+const char afs_init_sysname[] = "ppc_linux26";
+#elif defined(CONFIG_S390)
+#ifdef CONFIG_64BIT
+const char afs_init_sysname[] = "s390x_linux26";
+#else
+const char afs_init_sysname[] = "s390_linux26";
+#endif
+#elif defined(CONFIG_SPARC64)
+const char afs_init_sysname[] = "sparc64_linux26";
+#elif defined(CONFIG_SPARC32)
+const char afs_init_sysname[] = "sparc_linux26";
+#else
+const char afs_init_sysname[] = "unknown_linux26";
+#endif
+
 /*
  * Initialise an AFS network namespace record.
  */
 static int __net_init afs_net_init(struct afs_net *net)
 {
+	struct afs_sysnames *sysnames;
 	int ret;
 
 	net->live = true;
@@ -67,6 +98,16 @@ static int __net_init afs_net_init(struct afs_net *net)
 	INIT_WORK(&net->fs_manager, afs_manage_servers);
 	timer_setup(&net->fs_timer, afs_servers_timer, 0);
 
+	ret = -ENOMEM;
+	sysnames = kzalloc(sizeof(*sysnames), GFP_KERNEL);
+	if (!sysnames)
+		goto error_sysnames;
+	sysnames->subs[0] = (char *)&afs_init_sysname;
+	sysnames->nr = 1;
+	refcount_set(&sysnames->usage, 1);
+	net->sysnames = sysnames;
+	rwlock_init(&net->sysnames_lock);
+
 	/* Register the /proc stuff */
 	ret = afs_proc_init(net);
 	if (ret < 0)
@@ -92,6 +133,8 @@ static int __net_init afs_net_init(struct afs_net *net)
 	net->live = false;
 	afs_proc_cleanup(net);
 error_proc:
+	afs_put_sysnames(net->sysnames);
+error_sysnames:
 	net->live = false;
 	return ret;
 }
@@ -106,6 +149,7 @@ static void __net_exit afs_net_exit(struct afs_net *net)
 	afs_purge_servers(net);
 	afs_close_socket(net);
 	afs_proc_cleanup(net);
+	afs_put_sysnames(net->sysnames);
 }
 
 /*

fs/afs/rotate.c

@@ -21,7 +21,7 @@
 /*
  * Initialise a filesystem server cursor for iterating over FS servers.
  */
-void afs_init_fs_cursor(struct afs_fs_cursor *fc, struct afs_vnode *vnode)
+static void afs_init_fs_cursor(struct afs_fs_cursor *fc, struct afs_vnode *vnode)
 {
 	memset(fc, 0, sizeof(*fc));
 }

fs/afs/rxrpc.c

@@ -926,3 +926,12 @@ int afs_extract_data(struct afs_call *call, void *buf, size_t count,
 	afs_set_call_complete(call, ret, remote_abort);
 	return ret;
 }
+
+/*
+ * Log protocol error production.
+ */
+noinline int afs_protocol_error(struct afs_call *call, int error)
+{
+	trace_afs_protocol_error(call, error, __builtin_return_address(0));
+	return error;
+}

fs/afs/security.c

@@ -178,18 +178,14 @@ void afs_cache_permit(struct afs_vnode *vnode, struct key *key,
 		}
 	}
 
-	if (cb_break != (vnode->cb_break + vnode->cb_interest->server->cb_s_break)) {
-		rcu_read_unlock();
+	if (cb_break != (vnode->cb_break + vnode->cb_interest->server->cb_s_break))
 		goto someone_else_changed_it;
-	}
 
 	/* We need a ref on any permits list we want to copy as we'll have to
 	 * drop the lock to do memory allocation.
 	 */
-	if (permits && !refcount_inc_not_zero(&permits->usage)) {
-		rcu_read_unlock();
+	if (permits && !refcount_inc_not_zero(&permits->usage))
 		goto someone_else_changed_it;
-	}
 
 	rcu_read_unlock();
 
@@ -278,6 +274,7 @@ void afs_cache_permit(struct afs_vnode *vnode, struct key *key,
 	/* Someone else changed the cache under us - don't recheck at this
 	 * time.
 	 */
+	rcu_read_unlock();
 	return;
 }
 
@@ -296,8 +293,6 @@ int afs_check_permit(struct afs_vnode *vnode, struct key *key,
 	_enter("{%x:%u},%x",
 	       vnode->fid.vid, vnode->fid.vnode, key_serial(key));
 
-	permits = vnode->permit_cache;
-
 	/* check the permits to see if we've got one yet */
 	if (key == vnode->volume->cell->anonymous_key) {
 		_debug("anon");
@@ -327,7 +322,7 @@ int afs_check_permit(struct afs_vnode *vnode, struct key *key,
 		 */
 		_debug("no valid permit");
 
-		ret = afs_fetch_status(vnode, key);
+		ret = afs_fetch_status(vnode, key, false);
 		if (ret < 0) {
 			*_access = 0;
 			_leave(" = %d", ret);

fs/afs/server.c

@@ -59,7 +59,8 @@ struct afs_server *afs_find_server(struct afs_net *net,
 				alist = rcu_dereference(server->addresses);
 				for (i = alist->nr_ipv4; i < alist->nr_addrs; i++) {
 					b = &alist->addrs[i].transport.sin6;
-					diff = (u16)a->sin6_port - (u16)b->sin6_port;
+					diff = ((u16 __force)a->sin6_port -
+						(u16 __force)b->sin6_port);
 					if (diff == 0)
 						diff = memcmp(&a->sin6_addr,
 							      &b->sin6_addr,
@@ -79,10 +80,11 @@ struct afs_server *afs_find_server(struct afs_net *net,
 				alist = rcu_dereference(server->addresses);
 				for (i = 0; i < alist->nr_ipv4; i++) {
 					b = &alist->addrs[i].transport.sin6;
-					diff = (u16)a->sin6_port - (u16)b->sin6_port;
+					diff = ((u16 __force)a->sin6_port -
+						(u16 __force)b->sin6_port);
 					if (diff == 0)
-						diff = ((u32)a->sin6_addr.s6_addr32[3] -
-							(u32)b->sin6_addr.s6_addr32[3]);
+						diff = ((u32 __force)a->sin6_addr.s6_addr32[3] -
+							(u32 __force)b->sin6_addr.s6_addr32[3]);
 					if (diff == 0)
 						goto found;
 					if (diff < 0) {
@@ -381,7 +383,7 @@ static void afs_server_rcu(struct rcu_head *rcu)
 {
 	struct afs_server *server = container_of(rcu, struct afs_server, rcu);
 
-	afs_put_addrlist(server->addresses);
+	afs_put_addrlist(rcu_access_pointer(server->addresses));
 	kfree(server);
 }
 
@@ -390,7 +392,7 @@ static void afs_server_rcu(struct rcu_head *rcu)
  */
 static void afs_destroy_server(struct afs_net *net, struct afs_server *server)
 {
-	struct afs_addr_list *alist = server->addresses;
+	struct afs_addr_list *alist = rcu_access_pointer(server->addresses);
 	struct afs_addr_cursor ac = {
 		.alist	= alist,
 		.addr	= &alist->addrs[0],

fs/afs/super.c

@@ -154,7 +154,7 @@ static int afs_show_devname(struct seq_file *m, struct dentry *root)
 		seq_puts(m, "none");
 		return 0;
 	}
-	
+
 	switch (volume->type) {
 	case AFSVL_RWVOL:
 		break;
@@ -269,7 +269,7 @@ static int afs_parse_device_name(struct afs_mount_params *params,
 	int cellnamesz;
 
 	_enter(",%s", name);
-	
+
 	if (!name) {
 		printk(KERN_ERR "kAFS: no volume name specified\n");
 		return -EINVAL;
@@ -418,7 +418,10 @@ static int afs_fill_super(struct super_block *sb,
 	if (!sb->s_root)
 		goto error;
 
-	sb->s_d_op = &afs_fs_dentry_operations;
+	if (params->dyn_root)
+		sb->s_d_op = &afs_dynroot_dentry_operations;
+	else
+		sb->s_d_op = &afs_fs_dentry_operations;
 
 	_leave(" = 0");
 	return 0;
@@ -676,7 +679,7 @@ static int afs_statfs(struct dentry *dentry, struct kstatfs *buf)
 		buf->f_bfree	= 0;
 		return 0;
 	}
-	
+
 	key = afs_request_key(vnode->volume->cell);
 	if (IS_ERR(key))
 		return PTR_ERR(key);

fs/afs/vlclient.c

@@ -303,7 +303,7 @@ struct afs_addr_list *afs_vl_get_addrs_u(struct afs_net *net,
 	r->uuid.clock_seq_hi_and_reserved 	= htonl(u->clock_seq_hi_and_reserved);
 	r->uuid.clock_seq_low			= htonl(u->clock_seq_low);
 	for (i = 0; i < 6; i++)
-		r->uuid.node[i] = ntohl(u->node[i]);
+		r->uuid.node[i] = htonl(u->node[i]);
 
 	trace_afs_make_vl_call(call);
 	return (struct afs_addr_list *)afs_make_call(ac, call, GFP_KERNEL, false);
@@ -450,7 +450,7 @@ static int afs_deliver_yfsvl_get_endpoints(struct afs_call *call)
 		call->count2	= ntohl(*bp); /* Type or next count */
 
 		if (call->count > YFS_MAXENDPOINTS)
-			return -EBADMSG;
+			return afs_protocol_error(call, -EBADMSG);
 
 		alist = afs_alloc_addrlist(call->count, FS_SERVICE, AFS_FS_PORT);
 		if (!alist)
@@ -474,7 +474,7 @@ static int afs_deliver_yfsvl_get_endpoints(struct afs_call *call)
 			size = sizeof(__be32) * (1 + 4 + 1);
 			break;
 		default:
-			return -EBADMSG;
+			return afs_protocol_error(call, -EBADMSG);
 		}
 
 		size += sizeof(__be32);
@@ -487,24 +487,24 @@ static int afs_deliver_yfsvl_get_endpoints(struct afs_call *call)
 		switch (call->count2) {
 		case YFS_ENDPOINT_IPV4:
 			if (ntohl(bp[0]) != sizeof(__be32) * 2)
-				return -EBADMSG;
+				return afs_protocol_error(call, -EBADMSG);
 			afs_merge_fs_addr4(alist, bp[1], ntohl(bp[2]));
 			bp += 3;
 			break;
 		case YFS_ENDPOINT_IPV6:
 			if (ntohl(bp[0]) != sizeof(__be32) * 5)
-				return -EBADMSG;
+				return afs_protocol_error(call, -EBADMSG);
 			afs_merge_fs_addr6(alist, bp + 1, ntohl(bp[5]));
 			bp += 6;
 			break;
 		default:
-			return -EBADMSG;
+			return afs_protocol_error(call, -EBADMSG);
 		}
 
 		/* Got either the type of the next entry or the count of
 		 * volEndpoints if no more fsEndpoints.
 		 */
-		call->count2 = htonl(*bp++);
+		call->count2 = ntohl(*bp++);
 
 		call->offset = 0;
 		call->count--;
@@ -517,7 +517,7 @@ static int afs_deliver_yfsvl_get_endpoints(struct afs_call *call)
 		if (!call->count)
 			goto end;
 		if (call->count > YFS_MAXENDPOINTS)
-			return -EBADMSG;
+			return afs_protocol_error(call, -EBADMSG);
 
 		call->unmarshall = 3;
 
@@ -531,7 +531,7 @@ static int afs_deliver_yfsvl_get_endpoints(struct afs_call *call)
 			return ret;
 
 		bp = call->buffer;
-		call->count2 = htonl(*bp++);
+		call->count2 = ntohl(*bp++);
 		call->offset = 0;
 		call->unmarshall = 4;
 
@@ -545,7 +545,7 @@ static int afs_deliver_yfsvl_get_endpoints(struct afs_call *call)
 			size = sizeof(__be32) * (1 + 4 + 1);
 			break;
 		default:
-			return -EBADMSG;
+			return afs_protocol_error(call, -EBADMSG);
 		}
 
 		if (call->count > 1)
@@ -558,16 +558,16 @@ static int afs_deliver_yfsvl_get_endpoints(struct afs_call *call)
 		switch (call->count2) {
 		case YFS_ENDPOINT_IPV4:
 			if (ntohl(bp[0]) != sizeof(__be32) * 2)
-				return -EBADMSG;
+				return afs_protocol_error(call, -EBADMSG);
 			bp += 3;
 			break;
 		case YFS_ENDPOINT_IPV6:
 			if (ntohl(bp[0]) != sizeof(__be32) * 5)
-				return -EBADMSG;
+				return afs_protocol_error(call, -EBADMSG);
 			bp += 6;
 			break;
 		default:
-			return -EBADMSG;
+			return afs_protocol_error(call, -EBADMSG);
 		}
 
 		/* Got either the type of the next entry or the count of
@@ -576,7 +576,7 @@ static int afs_deliver_yfsvl_get_endpoints(struct afs_call *call)
 		call->offset = 0;
 		call->count--;
 		if (call->count > 0) {
-			call->count2 = htonl(*bp++);
+			call->count2 = ntohl(*bp++);
 			goto again;
 		}
 

fs/afs/write.c

@@ -42,10 +42,11 @@ static int afs_fill_page(struct afs_vnode *vnode, struct key *key,
 	if (!req)
 		return -ENOMEM;
 
-	atomic_set(&req->usage, 1);
+	refcount_set(&req->usage, 1);
 	req->pos = pos;
 	req->len = len;
 	req->nr_pages = 1;
+	req->pages = req->array;
 	req->pages[0] = page;
 	get_page(page);
 
@@ -124,7 +125,12 @@ int afs_write_begin(struct file *file, struct address_space *mapping,
 					     page->index, priv);
 			goto flush_conflicting_write;
 		}
-		if (to < f || from > t)
+		/* If the file is being filled locally, allow inter-write
+		 * spaces to be merged into writes.  If it's not, only write
+		 * back what the user gives us.
+		 */
+		if (!test_bit(AFS_VNODE_NEW_CONTENT, &vnode->flags) &&
+		    (to < f || from > t))
 			goto flush_conflicting_write;
 		if (from < f)
 			f = from;
@@ -355,6 +361,12 @@ static int afs_store_data(struct address_space *mapping,
 	}
 
 	switch (ret) {
+	case 0:
+		afs_stat_v(vnode, n_stores);
+		atomic_long_add((last * PAGE_SIZE + to) -
+				(first * PAGE_SIZE + offset),
+				&afs_v2net(vnode)->n_store_bytes);
+		break;
 	case -EACCES:
 	case -EPERM:
 	case -ENOKEY:
@@ -412,7 +424,8 @@ static int afs_write_back_from_locked_page(struct address_space *mapping,
 		trace_afs_page_dirty(vnode, tracepoint_string("WARN"),
 				     primary_page->index, priv);
 
-	if (start >= final_page || to < PAGE_SIZE)
+	if (start >= final_page ||
+	    (to < PAGE_SIZE && !test_bit(AFS_VNODE_NEW_CONTENT, &vnode->flags)))
 		goto no_more;
 
 	start++;
@@ -433,9 +446,10 @@ static int afs_write_back_from_locked_page(struct address_space *mapping,
 		}
 
 		for (loop = 0; loop < n; loop++) {
-			if (to != PAGE_SIZE)
-				break;
 			page = pages[loop];
+			if (to != PAGE_SIZE &&
+			    !test_bit(AFS_VNODE_NEW_CONTENT, &vnode->flags))
+				break;
 			if (page->index > final_page)
 				break;
 			if (!trylock_page(page))
@@ -448,7 +462,8 @@ static int afs_write_back_from_locked_page(struct address_space *mapping,
 			priv = page_private(page);
 			f = priv & AFS_PRIV_MAX;
 			t = priv >> AFS_PRIV_SHIFT;
-			if (f != 0) {
+			if (f != 0 &&
+			    !test_bit(AFS_VNODE_NEW_CONTENT, &vnode->flags)) {
 				unlock_page(page);
 				break;
 			}
@@ -734,20 +749,6 @@ int afs_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 	return file_write_and_wait_range(file, start, end);
 }
 
-/*
- * Flush out all outstanding writes on a file opened for writing when it is
- * closed.
- */
-int afs_flush(struct file *file, fl_owner_t id)
-{
-	_enter("");
-
-	if ((file->f_mode & FMODE_WRITE) == 0)
-		return 0;
-
-	return vfs_fsync(file, 0);
-}
-
 /*
  * notification that a previously read-only page is about to become writable
  * - if it returns an error, the caller will deliver a bus error signal

fs/afs/xdr_fs.h

+/* AFS fileserver XDR types
+ *
+ * Copyright (C) 2018 Red Hat, Inc. All Rights Reserved.
+ * Written by David Howells (dhowells@redhat.com)
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public Licence
+ * as published by the Free Software Foundation; either version
+ * 2 of the Licence, or (at your option) any later version.
+ */
+
+#ifndef XDR_FS_H
+#define XDR_FS_H
+
+struct afs_xdr_AFSFetchStatus {
+	__be32	if_version;
+#define AFS_FSTATUS_VERSION	1
+	__be32	type;
+	__be32	nlink;
+	__be32	size_lo;
+	__be32	data_version_lo;
+	__be32	author;
+	__be32	owner;
+	__be32	caller_access;
+	__be32	anon_access;
+	__be32	mode;
+	__be32	parent_vnode;
+	__be32	parent_unique;
+	__be32	seg_size;
+	__be32	mtime_client;
+	__be32	mtime_server;
+	__be32	group;
+	__be32	sync_counter;
+	__be32	data_version_hi;
+	__be32	lock_count;
+	__be32	size_hi;
+	__be32	abort_code;
+} __packed;
+
+#define AFS_DIR_HASHTBL_SIZE	128
+#define AFS_DIR_DIRENT_SIZE	32
+#define AFS_DIR_SLOTS_PER_BLOCK	64
+#define AFS_DIR_BLOCK_SIZE	2048
+#define AFS_DIR_BLOCKS_PER_PAGE	(PAGE_SIZE / AFS_DIR_BLOCK_SIZE)
+#define AFS_DIR_MAX_SLOTS	65536
+#define AFS_DIR_BLOCKS_WITH_CTR	128
+#define AFS_DIR_MAX_BLOCKS	1023
+#define AFS_DIR_RESV_BLOCKS	1
+#define AFS_DIR_RESV_BLOCKS0	13
+
+/*
+ * Directory entry structure.
+ */
+union afs_xdr_dirent {
+	struct {
+		u8		valid;
+		u8		unused[1];
+		__be16		hash_next;
+		__be32		vnode;
+		__be32		unique;
+		u8		name[16];
+		u8		overflow[4];	/* if any char of the name (inc
+						 * NUL) reaches here, consume
+						 * the next dirent too */
+	} u;
+	u8			extended_name[32];
+} __packed;
+
+/*
+ * Directory block header (one at the beginning of every 2048-byte block).
+ */
+struct afs_xdr_dir_hdr {
+	__be16		npages;
+	__be16		magic;
+#define AFS_DIR_MAGIC htons(1234)
+	u8		reserved;
+	u8		bitmap[8];
+	u8		pad[19];
+} __packed;
+
+/*
+ * Directory block layout
+ */
+union afs_xdr_dir_block {
+	struct afs_xdr_dir_hdr		hdr;
+
+	struct {
+		struct afs_xdr_dir_hdr	hdr;
+		u8			alloc_ctrs[AFS_DIR_MAX_BLOCKS];
+		__be16			hashtable[AFS_DIR_HASHTBL_SIZE];
+	} meta;
+
+	union afs_xdr_dirent	dirents[AFS_DIR_SLOTS_PER_BLOCK];
+} __packed;
+
+/*
+ * Directory layout on a linux VM page.
+ */
+struct afs_xdr_dir_page {
+	union afs_xdr_dir_block	blocks[AFS_DIR_BLOCKS_PER_PAGE];
+};
+
+#endif /* XDR_FS_H */

include/linux/fs.h

@@ -1667,7 +1667,7 @@ typedef int (*filldir_t)(struct dir_context *, const char *, int, loff_t, u64,
 			 unsigned);
 
 struct dir_context {
-	const filldir_t actor;
+	filldir_t actor;
 	loff_t pos;
 };
 

include/trace/events/afs.h

@@ -49,6 +49,7 @@ enum afs_fs_operation {
 	afs_FS_ExtendLock		= 157,	/* AFS Extend a file lock */
 	afs_FS_ReleaseLock		= 158,	/* AFS Release a file lock */
 	afs_FS_Lookup			= 161,	/* AFS lookup file in directory */
+	afs_FS_InlineBulkStatus		= 65536, /* AFS Fetch multiple file statuses with errors */
 	afs_FS_FetchData64		= 65537, /* AFS Fetch file data */
 	afs_FS_StoreData64		= 65538, /* AFS Store file data */
 	afs_FS_GiveUpAllCallBacks	= 65539, /* AFS Give up all our callbacks on a server */
@@ -62,6 +63,27 @@ enum afs_vl_operation {
 	afs_VL_GetCapabilities	= 65537,	/* AFS Get VL server capabilities */
 };
 
+enum afs_edit_dir_op {
+	afs_edit_dir_create,
+	afs_edit_dir_create_error,
+	afs_edit_dir_create_inval,
+	afs_edit_dir_create_nospc,
+	afs_edit_dir_delete,
+	afs_edit_dir_delete_error,
+	afs_edit_dir_delete_inval,
+	afs_edit_dir_delete_noent,
+};
+
+enum afs_edit_dir_reason {
+	afs_edit_dir_for_create,
+	afs_edit_dir_for_link,
+	afs_edit_dir_for_mkdir,
+	afs_edit_dir_for_rename,
+	afs_edit_dir_for_rmdir,
+	afs_edit_dir_for_symlink,
+	afs_edit_dir_for_unlink,
+};
+
 #endif /* end __AFS_DECLARE_TRACE_ENUMS_ONCE_ONLY */
 
 /*
@@ -93,6 +115,7 @@ enum afs_vl_operation {
 	EM(afs_FS_ExtendLock,			"FS.ExtendLock") \
 	EM(afs_FS_ReleaseLock,			"FS.ReleaseLock") \
 	EM(afs_FS_Lookup,			"FS.Lookup") \
+	EM(afs_FS_InlineBulkStatus,		"FS.InlineBulkStatus") \
 	EM(afs_FS_FetchData64,			"FS.FetchData64") \
 	EM(afs_FS_StoreData64,			"FS.StoreData64") \
 	EM(afs_FS_GiveUpAllCallBacks,		"FS.GiveUpAllCallBacks") \
@@ -104,6 +127,25 @@ enum afs_vl_operation {
 	EM(afs_YFSVL_GetEndpoints,		"YFSVL.GetEndpoints") \
 	E_(afs_VL_GetCapabilities,		"VL.GetCapabilities")
 
+#define afs_edit_dir_ops				  \
+	EM(afs_edit_dir_create,			"create") \
+	EM(afs_edit_dir_create_error,		"c_fail") \
+	EM(afs_edit_dir_create_inval,		"c_invl") \
+	EM(afs_edit_dir_create_nospc,		"c_nspc") \
+	EM(afs_edit_dir_delete,			"delete") \
+	EM(afs_edit_dir_delete_error,		"d_err ") \
+	EM(afs_edit_dir_delete_inval,		"d_invl") \
+	E_(afs_edit_dir_delete_noent,		"d_nent")
+
+#define afs_edit_dir_reasons				  \
+	EM(afs_edit_dir_for_create,		"Create") \
+	EM(afs_edit_dir_for_link,		"Link  ") \
+	EM(afs_edit_dir_for_mkdir,		"MkDir ") \
+	EM(afs_edit_dir_for_rename,		"Rename") \
+	EM(afs_edit_dir_for_rmdir,		"RmDir ") \
+	EM(afs_edit_dir_for_symlink,		"Symlnk") \
+	E_(afs_edit_dir_for_unlink,		"Unlink")
+
 
 /*
  * Export enum symbols via userspace.
@@ -116,6 +158,8 @@ enum afs_vl_operation {
 afs_call_traces;
 afs_fs_operations;
 afs_vl_operations;
+afs_edit_dir_ops;
+afs_edit_dir_reasons;
 
 /*
  * Now redefine the EM() and E_() macros to map the enums to the strings that
@@ -462,6 +506,75 @@ TRACE_EVENT(afs_call_state,
 		      __entry->ret, __entry->abort)
 	    );
 
+TRACE_EVENT(afs_edit_dir,
+	    TP_PROTO(struct afs_vnode *dvnode,
+		     enum afs_edit_dir_reason why,
+		     enum afs_edit_dir_op op,
+		     unsigned int block,
+		     unsigned int slot,
+		     unsigned int f_vnode,
+		     unsigned int f_unique,
+		     const char *name),
+
+	    TP_ARGS(dvnode, why, op, block, slot, f_vnode, f_unique, name),
+
+	    TP_STRUCT__entry(
+		    __field(unsigned int,		vnode		)
+		    __field(unsigned int,		unique		)
+		    __field(enum afs_edit_dir_reason,	why		)
+		    __field(enum afs_edit_dir_op,	op		)
+		    __field(unsigned int,		block		)
+		    __field(unsigned short,		slot		)
+		    __field(unsigned int,		f_vnode		)
+		    __field(unsigned int,		f_unique	)
+		    __array(char,			name, 18	)
+			     ),
+
+	    TP_fast_assign(
+		    int __len = strlen(name);
+		    __len = min(__len, 17);
+		    __entry->vnode	= dvnode->fid.vnode;
+		    __entry->unique	= dvnode->fid.unique;
+		    __entry->why	= why;
+		    __entry->op		= op;
+		    __entry->block	= block;
+		    __entry->slot	= slot;
+		    __entry->f_vnode	= f_vnode;
+		    __entry->f_unique	= f_unique;
+		    memcpy(__entry->name, name, __len);
+		    __entry->name[__len] = 0;
+			   ),
+
+	    TP_printk("d=%x:%x %s %s %u[%u] f=%x:%x %s",
+		      __entry->vnode, __entry->unique,
+		      __print_symbolic(__entry->why, afs_edit_dir_reasons),
+		      __print_symbolic(__entry->op, afs_edit_dir_ops),
+		      __entry->block, __entry->slot,
+		      __entry->f_vnode, __entry->f_unique,
+		      __entry->name)
+	    );
+
+TRACE_EVENT(afs_protocol_error,
+	    TP_PROTO(struct afs_call *call, int error, const void *where),
+
+	    TP_ARGS(call, error, where),
+
+	    TP_STRUCT__entry(
+		    __field(unsigned int,	call		)
+		    __field(int,		error		)
+		    __field(const void *,	where		)
+			     ),
+
+	    TP_fast_assign(
+		    __entry->call = call ? call->debug_id : 0;
+		    __entry->error = error;
+		    __entry->where = where;
+			   ),
+
+	    TP_printk("c=%08x r=%d sp=%pSR",
+		      __entry->call, __entry->error, __entry->where)
+	    );
+
 #endif /* _TRACE_AFS_H */
 
 /* This part must be outside protection */