1 files changed, 95 insertions, 92 deletions

diff --git a/fs/dlm/user.c b/fs/dlm/user.c
index e73a4bb572a..142e21655ee 100644
--- a/fs/dlm/user.c
+++ b/fs/dlm/user.c
@@ -1,5 +1,5 @@
 /*
- * Copyright (C) 2006-2009 Red Hat, Inc.  All rights reserved.
+ * Copyright (C) 2006-2010 Red Hat, Inc.  All rights reserved.
  *
  * This copyrighted material is made available to anyone wishing to use,
  * modify, copy, or redistribute it subject to the terms and conditions
@@ -17,12 +17,14 @@
 #include <linux/spinlock.h>
 #include <linux/dlm.h>
 #include <linux/dlm_device.h>
+#include <linux/slab.h>
 
 #include "dlm_internal.h"
 #include "lockspace.h"
 #include "lock.h"
 #include "lvb_table.h"
 #include "user.h"
+#include "ast.h"
 
 static const char name_prefix[] = "dlm";
 static const struct file_operations device_fops;
@@ -151,19 +153,16 @@ static void compat_output(struct dlm_lock_result *res,
    not related to the lifetime of the lkb struct which is managed
    entirely by refcount. */
 
-static int lkb_is_endoflife(struct dlm_lkb *lkb, int sb_status, int type)
+static int lkb_is_endoflife(int mode, int status)
 {
-	switch (sb_status) {
+	switch (status) {
 	case -DLM_EUNLOCK:
 		return 1;
 	case -DLM_ECANCEL:
 	case -ETIMEDOUT:
 	case -EDEADLK:
-		if (lkb->lkb_grmode == DLM_LOCK_IV)
-			return 1;
-		break;
 	case -EAGAIN:
-		if (type == AST_COMP && lkb->lkb_grmode == DLM_LOCK_IV)
+		if (mode == DLM_LOCK_IV)
 			return 1;
 		break;
 	}
@@ -173,12 +172,13 @@ static int lkb_is_endoflife(struct dlm_lkb *lkb, int sb_status, int type)
 /* we could possibly check if the cancel of an orphan has resulted in the lkb
    being removed and then remove that lkb from the orphans list and free it */
 
-void dlm_user_add_ast(struct dlm_lkb *lkb, int type, int bastmode)
+void dlm_user_add_ast(struct dlm_lkb *lkb, uint32_t flags, int mode,
+		      int status, uint32_t sbflags, uint64_t seq)
 {
 	struct dlm_ls *ls;
 	struct dlm_user_args *ua;
 	struct dlm_user_proc *proc;
-	int eol = 0, ast_type;
+	int rv;
 
 	if (lkb->lkb_flags & (DLM_IFL_ORPHAN | DLM_IFL_DEAD))
 		return;
@@ -199,47 +199,29 @@ void dlm_user_add_ast(struct dlm_lkb *lkb, int type, int bastmode)
 	ua = lkb->lkb_ua;
 	proc = ua->proc;
 
-	if (type == AST_BAST && ua->bastaddr == NULL)
+	if ((flags & DLM_CB_BAST) && ua->bastaddr == NULL)
 		goto out;
 
+	if ((flags & DLM_CB_CAST) && lkb_is_endoflife(mode, status))
+		lkb->lkb_flags |= DLM_IFL_ENDOFLIFE;
+
 	spin_lock(&proc->asts_spin);
 
-	ast_type = lkb->lkb_ast_type;
-	lkb->lkb_ast_type |= type;
-	if (bastmode)
-		lkb->lkb_bastmode = bastmode;
+	rv = dlm_add_lkb_callback(lkb, flags, mode, status, sbflags, seq);
+	if (rv < 0) {
+		spin_unlock(&proc->asts_spin);
+		goto out;
+	}
 
-	if (!ast_type) {
+	if (list_empty(&lkb->lkb_cb_list)) {
 		kref_get(&lkb->lkb_ref);
-		list_add_tail(&lkb->lkb_astqueue, &proc->asts);
+		list_add_tail(&lkb->lkb_cb_list, &proc->asts);
 		wake_up_interruptible(&proc->wait);
 	}
-	if (type == AST_COMP && (ast_type & AST_COMP))
-		log_debug(ls, "ast overlap %x status %x %x",
-			  lkb->lkb_id, ua->lksb.sb_status, lkb->lkb_flags);
-
-	eol = lkb_is_endoflife(lkb, ua->lksb.sb_status, type);
-	if (eol) {
-		lkb->lkb_ast_type &= ~AST_BAST;
-		lkb->lkb_flags |= DLM_IFL_ENDOFLIFE;
-	}
-
-	/* We want to copy the lvb to userspace when the completion
-	   ast is read if the status is 0, the lock has an lvb and
-	   lvb_ops says we should.  We could probably have set_lvb_lock()
-	   set update_user_lvb instead and not need old_mode */
-
-	if ((lkb->lkb_ast_type & AST_COMP) &&
-	    (lkb->lkb_lksb->sb_status == 0) &&
-	    lkb->lkb_lksb->sb_lvbptr &&
-	    dlm_lvb_operations[ua->old_mode + 1][lkb->lkb_grmode + 1])
-		ua->update_user_lvb = 1;
-	else
-		ua->update_user_lvb = 0;
-
 	spin_unlock(&proc->asts_spin);
 
-	if (eol) {
+	if (lkb->lkb_flags & DLM_IFL_ENDOFLIFE) {
+		/* N.B. spin_lock locks_spin, not asts_spin */
 		spin_lock(&proc->locks_spin);
 		if (!list_empty(&lkb->lkb_ownqueue)) {
 			list_del_init(&lkb->lkb_ownqueue);
@@ -410,8 +392,9 @@ static int device_create_lockspace(struct dlm_lspace_params *params)
 	if (!capable(CAP_SYS_ADMIN))
 		return -EPERM;
 
-	error = dlm_new_lockspace(params->name, strlen(params->name),
-				  &lockspace, params->flags, DLM_USER_LVB_LEN);
+	error = dlm_new_lockspace(params->name, NULL, params->flags,
+				  DLM_USER_LVB_LEN, NULL, NULL, NULL,
+				  &lockspace);
 	if (error)
 		return error;
 
@@ -510,7 +493,6 @@ static ssize_t device_write(struct file *file, const char __user *buf,
 {
 	struct dlm_user_proc *proc = file->private_data;
 	struct dlm_write_request *kbuf;
-	sigset_t tmpsig, allsigs;
 	int error;
 
 #ifdef CONFIG_COMPAT
@@ -520,6 +502,13 @@ static ssize_t device_write(struct file *file, const char __user *buf,
 #endif
 		return -EINVAL;
 
+	/*
+	 * can't compare against COMPAT/dlm_write_request32 because
+	 * we don't yet know if is64bit is zero
+	 */
+	if (count > sizeof(struct dlm_write_request) + DLM_RESNAME_MAXLEN)
+		return -EINVAL;
+
 	kbuf = kzalloc(count + 1, GFP_NOFS);
 	if (!kbuf)
 		return -ENOMEM;
@@ -567,9 +556,6 @@ static ssize_t device_write(struct file *file, const char __user *buf,
 		goto out_free;
 	}
 
-	sigfillset(&allsigs);
-	sigprocmask(SIG_BLOCK, &allsigs, &tmpsig);
-
 	error = -EINVAL;
 
 	switch (kbuf->cmd)
@@ -577,7 +563,7 @@ static ssize_t device_write(struct file *file, const char __user *buf,
 	case DLM_USER_LOCK:
 		if (!proc) {
 			log_print("no locking on control device");
-			goto out_sig;
+			goto out_free;
 		}
 		error = device_user_lock(proc, &kbuf->i.lock);
 		break;
@@ -585,7 +571,7 @@ static ssize_t device_write(struct file *file, const char __user *buf,
 	case DLM_USER_UNLOCK:
 		if (!proc) {
 			log_print("no locking on control device");
-			goto out_sig;
+			goto out_free;
 		}
 		error = device_user_unlock(proc, &kbuf->i.lock);
 		break;
@@ -593,7 +579,7 @@ static ssize_t device_write(struct file *file, const char __user *buf,
 	case DLM_USER_DEADLOCK:
 		if (!proc) {
 			log_print("no locking on control device");
-			goto out_sig;
+			goto out_free;
 		}
 		error = device_user_deadlock(proc, &kbuf->i.lock);
 		break;
@@ -601,7 +587,7 @@ static ssize_t device_write(struct file *file, const char __user *buf,
 	case DLM_USER_CREATE_LOCKSPACE:
 		if (proc) {
 			log_print("create/remove only on control device");
-			goto out_sig;
+			goto out_free;
 		}
 		error = device_create_lockspace(&kbuf->i.lspace);
 		break;
@@ -609,7 +595,7 @@ static ssize_t device_write(struct file *file, const char __user *buf,
 	case DLM_USER_REMOVE_LOCKSPACE:
 		if (proc) {
 			log_print("create/remove only on control device");
-			goto out_sig;
+			goto out_free;
 		}
 		error = device_remove_lockspace(&kbuf->i.lspace);
 		break;
@@ -617,7 +603,7 @@ static ssize_t device_write(struct file *file, const char __user *buf,
 	case DLM_USER_PURGE:
 		if (!proc) {
 			log_print("no locking on control device");
-			goto out_sig;
+			goto out_free;
 		}
 		error = device_user_purge(proc, &kbuf->i.purge);
 		break;
@@ -627,9 +613,6 @@ static ssize_t device_write(struct file *file, const char __user *buf,
 			  kbuf->cmd);
 	}
 
- out_sig:
-	sigprocmask(SIG_SETMASK, &tmpsig, NULL);
-	recalc_sigpending();
  out_free:
 	kfree(kbuf);
 	return error;
@@ -670,15 +653,11 @@ static int device_close(struct inode *inode, struct file *file)
 {
 	struct dlm_user_proc *proc = file->private_data;
 	struct dlm_ls *ls;
-	sigset_t tmpsig, allsigs;
 
 	ls = dlm_find_lockspace_local(proc->lockspace);
 	if (!ls)
 		return -ENOENT;
 
-	sigfillset(&allsigs);
-	sigprocmask(SIG_BLOCK, &allsigs, &tmpsig);
-
 	set_bit(DLM_PROC_FLAGS_CLOSING, &proc->flags);
 
 	dlm_clear_proc_locks(ls, proc);
@@ -696,14 +675,12 @@ static int device_close(struct inode *inode, struct file *file)
 	/* FIXME: AUTOFREE: if this ls is no longer used do
 	   device_remove_lockspace() */
 
-	sigprocmask(SIG_SETMASK, &tmpsig, NULL);
-	recalc_sigpending();
-
 	return 0;
 }
 
-static int copy_result_to_user(struct dlm_user_args *ua, int compat, int type,
-			       int bmode, char __user *buf, size_t count)
+static int copy_result_to_user(struct dlm_user_args *ua, int compat,
+			       uint32_t flags, int mode, int copy_lvb,
+			       char __user *buf, size_t count)
 {
 #ifdef CONFIG_COMPAT
 	struct dlm_lock_result32 result32;
@@ -727,10 +704,10 @@ static int copy_result_to_user(struct dlm_user_args *ua, int compat, int type,
 	   notes that a new blocking AST address and parameter are set even if
 	   the conversion fails, so maybe we should just do that. */
 
-	if (type == AST_BAST) {
+	if (flags & DLM_CB_BAST) {
 		result.user_astaddr = ua->bastaddr;
 		result.user_astparam = ua->bastparam;
-		result.bast_mode = bmode;
+		result.bast_mode = mode;
 	} else {
 		result.user_astaddr = ua->castaddr;
 		result.user_astparam = ua->castparam;
@@ -747,8 +724,7 @@ static int copy_result_to_user(struct dlm_user_args *ua, int compat, int type,
 	/* copy lvb to userspace if there is one, it's been updated, and
 	   the user buffer has space for it */
 
-	if (ua->update_user_lvb && ua->lksb.sb_lvbptr &&
-	    count >= len + DLM_USER_LVB_LEN) {
+	if (copy_lvb && ua->lksb.sb_lvbptr && count >= len + DLM_USER_LVB_LEN) {
 		if (copy_to_user(buf+len, ua->lksb.sb_lvbptr,
 				 DLM_USER_LVB_LEN)) {
 			error = -EFAULT;
@@ -798,11 +774,12 @@ static ssize_t device_read(struct file *file, char __user *buf, size_t count,
 	struct dlm_user_proc *proc = file->private_data;
 	struct dlm_lkb *lkb;
 	DECLARE_WAITQUEUE(wait, current);
-	int error, type=0, bmode=0, removed = 0;
+	struct dlm_callback cb;
+	int rv, resid, copy_lvb = 0;
 
 	if (count == sizeof(struct dlm_device_version)) {
-		error = copy_version_to_user(buf, count);
-		return error;
+		rv = copy_version_to_user(buf, count);
+		return rv;
 	}
 
 	if (!proc) {
@@ -817,6 +794,8 @@ static ssize_t device_read(struct file *file, char __user *buf, size_t count,
 #endif
 		return -EINVAL;
 
+ try_another:
+
 	/* do we really need this? can a read happen after a close? */
 	if (test_bit(DLM_PROC_FLAGS_CLOSING, &proc->flags))
 		return -EINVAL;
@@ -847,36 +826,57 @@ static ssize_t device_read(struct file *file, char __user *buf, size_t count,
 		}
 	}
 
-	/* there may be both completion and blocking asts to return for
-	   the lkb, don't remove lkb from asts list unless no asts remain */
+	/* if we empty lkb_callbacks, we don't want to unlock the spinlock
+	   without removing lkb_cb_list; so empty lkb_cb_list is always
+	   consistent with empty lkb_callbacks */
+
+	lkb = list_entry(proc->asts.next, struct dlm_lkb, lkb_cb_list);
 
-	lkb = list_entry(proc->asts.next, struct dlm_lkb, lkb_astqueue);
+	rv = dlm_rem_lkb_callback(lkb->lkb_resource->res_ls, lkb, &cb, &resid);
+	if (rv < 0) {
+		/* this shouldn't happen; lkb should have been removed from
+		   list when resid was zero */
+		log_print("dlm_rem_lkb_callback empty %x", lkb->lkb_id);
+		list_del_init(&lkb->lkb_cb_list);
+		spin_unlock(&proc->asts_spin);
+		/* removes ref for proc->asts, may cause lkb to be freed */
+		dlm_put_lkb(lkb);
+		goto try_another;
+	}
+	if (!resid)
+		list_del_init(&lkb->lkb_cb_list);
+	spin_unlock(&proc->asts_spin);
 
-	if (lkb->lkb_ast_type & AST_COMP) {
-		lkb->lkb_ast_type &= ~AST_COMP;
-		type = AST_COMP;
-	} else if (lkb->lkb_ast_type & AST_BAST) {
-		lkb->lkb_ast_type &= ~AST_BAST;
-		type = AST_BAST;
-		bmode = lkb->lkb_bastmode;
+	if (cb.flags & DLM_CB_SKIP) {
+		/* removes ref for proc->asts, may cause lkb to be freed */
+		if (!resid)
+			dlm_put_lkb(lkb);
+		goto try_another;
 	}
 
-	if (!lkb->lkb_ast_type) {
-		list_del(&lkb->lkb_astqueue);
-		removed = 1;
+	if (cb.flags & DLM_CB_CAST) {
+		int old_mode, new_mode;
+
+		old_mode = lkb->lkb_last_cast.mode;
+		new_mode = cb.mode;
+
+		if (!cb.sb_status && lkb->lkb_lksb->sb_lvbptr &&
+		    dlm_lvb_operations[old_mode + 1][new_mode + 1])
+			copy_lvb = 1;
+
+		lkb->lkb_lksb->sb_status = cb.sb_status;
+		lkb->lkb_lksb->sb_flags = cb.sb_flags;
 	}
-	spin_unlock(&proc->asts_spin);
 
-	error = copy_result_to_user(lkb->lkb_ua,
-			 	test_bit(DLM_PROC_FLAGS_COMPAT, &proc->flags),
-				type, bmode, buf, count);
+	rv = copy_result_to_user(lkb->lkb_ua,
+				 test_bit(DLM_PROC_FLAGS_COMPAT, &proc->flags),
+				 cb.flags, cb.mode, copy_lvb, buf, count);
 
-	/* removes reference for the proc->asts lists added by
-	   dlm_user_add_ast() and may result in the lkb being freed */
-	if (removed)
+	/* removes ref for proc->asts, may cause lkb to be freed */
+	if (!resid)
 		dlm_put_lkb(lkb);
 
-	return error;
+	return rv;
 }
 
 static unsigned int device_poll(struct file *file, poll_table *wait)
@@ -946,6 +946,7 @@ static const struct file_operations device_fops = {
 	.write   = device_write,
 	.poll    = device_poll,
 	.owner   = THIS_MODULE,
+	.llseek  = noop_llseek,
 };
 
 static const struct file_operations ctl_device_fops = {
@@ -954,6 +955,7 @@ static const struct file_operations ctl_device_fops = {
 	.read    = device_read,
 	.write   = device_write,
 	.owner   = THIS_MODULE,
+	.llseek  = noop_llseek,
 };
 
 static struct miscdevice ctl_device = {
@@ -966,6 +968,7 @@ static const struct file_operations monitor_device_fops = {
 	.open    = monitor_device_open,
 	.release = monitor_device_close,
 	.owner   = THIS_MODULE,
+	.llseek  = noop_llseek,
 };
 
 static struct miscdevice monitor_device = {