1 files changed, 112 insertions, 106 deletions

diff --git a/fs/gfs2/file.c b/fs/gfs2/file.c
index 3e718cfc19a7..60c6fb91fb58 100644
--- a/fs/gfs2/file.c
+++ b/fs/gfs2/file.c
@@ -704,10 +704,11 @@ static int gfs2_release(struct inode *inode, struct file *file)
 	kfree(file->private_data);
 	file->private_data = NULL;
 
-	if (gfs2_rs_active(&ip->i_res))
-		gfs2_rs_delete(ip, &inode->i_writecount);
-	if (file->f_mode & FMODE_WRITE)
+	if (file->f_mode & FMODE_WRITE) {
+		if (gfs2_rs_active(&ip->i_res))
+			gfs2_rs_delete(ip);
 		gfs2_qa_put(ip);
+	}
 	return 0;
 }
 
@@ -769,33 +770,31 @@ static int gfs2_fsync(struct file *file, loff_t start, loff_t end,
 	return ret ? ret : ret1;
 }
 
-static inline bool should_fault_in_pages(ssize_t ret, struct iov_iter *i,
+static inline bool should_fault_in_pages(struct iov_iter *i,
+					 struct kiocb *iocb,
 					 size_t *prev_count,
 					 size_t *window_size)
 {
 	size_t count = iov_iter_count(i);
-	char __user *p;
-	int pages = 1;
+	size_t size, offs;
 
-	if (likely(!count))
+	if (!count)
 		return false;
-	if (ret <= 0 && ret != -EFAULT)
-		return false;
-	if (!iter_is_iovec(i))
+	if (!user_backed_iter(i))
 		return false;
 
+	size = PAGE_SIZE;
+	offs = offset_in_page(iocb->ki_pos);
 	if (*prev_count != count || !*window_size) {
-		int pages, nr_dirtied;
+		size_t nr_dirtied;
 
-		pages = min_t(int, BIO_MAX_VECS, DIV_ROUND_UP(count, PAGE_SIZE));
 		nr_dirtied = max(current->nr_dirtied_pause -
-				 current->nr_dirtied, 1);
-		pages = min(pages, nr_dirtied);
+				 current->nr_dirtied, 8);
+		size = min_t(size_t, SZ_1M, nr_dirtied << PAGE_SHIFT);
 	}
 
 	*prev_count = count;
-	p = i->iov[0].iov_base + i->iov_offset;
-	*window_size = (size_t)PAGE_SIZE * pages - offset_in_page(p);
+	*window_size = size - offs;
 	return true;
 }
 
@@ -805,7 +804,7 @@ static ssize_t gfs2_file_direct_read(struct kiocb *iocb, struct iov_iter *to,
 	struct file *file = iocb->ki_filp;
 	struct gfs2_inode *ip = GFS2_I(file->f_mapping->host);
 	size_t prev_count = 0, window_size = 0;
-	size_t written = 0;
+	size_t read = 0;
 	ssize_t ret;
 
 	/*
@@ -833,35 +832,33 @@ retry:
 	ret = gfs2_glock_nq(gh);
 	if (ret)
 		goto out_uninit;
-retry_under_glock:
 	pagefault_disable();
 	to->nofault = true;
 	ret = iomap_dio_rw(iocb, to, &gfs2_iomap_ops, NULL,
-			   IOMAP_DIO_PARTIAL, written);
+			   IOMAP_DIO_PARTIAL, NULL, read);
 	to->nofault = false;
 	pagefault_enable();
+	if (ret <= 0 && ret != -EFAULT)
+		goto out_unlock;
+	/* No increment (+=) because iomap_dio_rw returns a cumulative value. */
 	if (ret > 0)
-		written = ret;
-
-	if (should_fault_in_pages(ret, to, &prev_count, &window_size)) {
-		size_t leftover;
+		read = ret;
 
-		gfs2_holder_allow_demote(gh);
-		leftover = fault_in_iov_iter_writeable(to, window_size);
-		gfs2_holder_disallow_demote(gh);
-		if (leftover != window_size) {
-			if (!gfs2_holder_queued(gh))
-				goto retry;
-			goto retry_under_glock;
-		}
+	if (should_fault_in_pages(to, iocb, &prev_count, &window_size)) {
+		gfs2_glock_dq(gh);
+		window_size -= fault_in_iov_iter_writeable(to, window_size);
+		if (window_size)
+			goto retry;
 	}
+out_unlock:
 	if (gfs2_holder_queued(gh))
 		gfs2_glock_dq(gh);
 out_uninit:
 	gfs2_holder_uninit(gh);
+	/* User space doesn't expect partial success. */
 	if (ret < 0)
 		return ret;
-	return written;
+	return read;
 }
 
 static ssize_t gfs2_file_direct_write(struct kiocb *iocb, struct iov_iter *from,
@@ -871,7 +868,7 @@ static ssize_t gfs2_file_direct_write(struct kiocb *iocb, struct iov_iter *from,
 	struct inode *inode = file->f_mapping->host;
 	struct gfs2_inode *ip = GFS2_I(inode);
 	size_t prev_count = 0, window_size = 0;
-	size_t read = 0;
+	size_t written = 0;
 	ssize_t ret;
 
 	/*
@@ -897,41 +894,39 @@ retry:
 	ret = gfs2_glock_nq(gh);
 	if (ret)
 		goto out_uninit;
-retry_under_glock:
 	/* Silently fall back to buffered I/O when writing beyond EOF */
 	if (iocb->ki_pos + iov_iter_count(from) > i_size_read(&ip->i_inode))
-		goto out;
+		goto out_unlock;
 
 	from->nofault = true;
 	ret = iomap_dio_rw(iocb, from, &gfs2_iomap_ops, NULL,
-			   IOMAP_DIO_PARTIAL, read);
+			   IOMAP_DIO_PARTIAL, NULL, written);
 	from->nofault = false;
-
-	if (ret == -ENOTBLK)
-		ret = 0;
+	if (ret <= 0) {
+		if (ret == -ENOTBLK)
+			ret = 0;
+		if (ret != -EFAULT)
+			goto out_unlock;
+	}
+	/* No increment (+=) because iomap_dio_rw returns a cumulative value. */
 	if (ret > 0)
-		read = ret;
-
-	if (should_fault_in_pages(ret, from, &prev_count, &window_size)) {
-		size_t leftover;
+		written = ret;
 
-		gfs2_holder_allow_demote(gh);
-		leftover = fault_in_iov_iter_readable(from, window_size);
-		gfs2_holder_disallow_demote(gh);
-		if (leftover != window_size) {
-			if (!gfs2_holder_queued(gh))
-				goto retry;
-			goto retry_under_glock;
-		}
+	if (should_fault_in_pages(from, iocb, &prev_count, &window_size)) {
+		gfs2_glock_dq(gh);
+		window_size -= fault_in_iov_iter_readable(from, window_size);
+		if (window_size)
+			goto retry;
 	}
-out:
+out_unlock:
 	if (gfs2_holder_queued(gh))
 		gfs2_glock_dq(gh);
 out_uninit:
 	gfs2_holder_uninit(gh);
+	/* User space doesn't expect partial success. */
 	if (ret < 0)
 		return ret;
-	return read;
+	return written;
 }
 
 static ssize_t gfs2_file_read_iter(struct kiocb *iocb, struct iov_iter *to)
@@ -939,7 +934,7 @@ static ssize_t gfs2_file_read_iter(struct kiocb *iocb, struct iov_iter *to)
 	struct gfs2_inode *ip;
 	struct gfs2_holder gh;
 	size_t prev_count = 0, window_size = 0;
-	size_t written = 0;
+	size_t read = 0;
 	ssize_t ret;
 
 	/*
@@ -949,20 +944,19 @@ static ssize_t gfs2_file_read_iter(struct kiocb *iocb, struct iov_iter *to)
 	 * and retry.
 	 */
 
-	if (iocb->ki_flags & IOCB_DIRECT) {
-		ret = gfs2_file_direct_read(iocb, to, &gh);
-		if (likely(ret != -ENOTBLK))
-			return ret;
-		iocb->ki_flags &= ~IOCB_DIRECT;
-	}
+	if (iocb->ki_flags & IOCB_DIRECT)
+		return gfs2_file_direct_read(iocb, to, &gh);
+
+	pagefault_disable();
 	iocb->ki_flags |= IOCB_NOIO;
 	ret = generic_file_read_iter(iocb, to);
 	iocb->ki_flags &= ~IOCB_NOIO;
+	pagefault_enable();
 	if (ret >= 0) {
 		if (!iov_iter_count(to))
 			return ret;
-		written = ret;
-	} else {
+		read = ret;
+	} else if (ret != -EFAULT) {
 		if (ret != -EAGAIN)
 			return ret;
 		if (iocb->ki_flags & IOCB_NOWAIT)
@@ -974,33 +968,26 @@ retry:
 	ret = gfs2_glock_nq(&gh);
 	if (ret)
 		goto out_uninit;
-retry_under_glock:
 	pagefault_disable();
 	ret = generic_file_read_iter(iocb, to);
 	pagefault_enable();
+	if (ret <= 0 && ret != -EFAULT)
+		goto out_unlock;
 	if (ret > 0)
-		written += ret;
+		read += ret;
 
-	if (should_fault_in_pages(ret, to, &prev_count, &window_size)) {
-		size_t leftover;
-
-		gfs2_holder_allow_demote(&gh);
-		leftover = fault_in_iov_iter_writeable(to, window_size);
-		gfs2_holder_disallow_demote(&gh);
-		if (leftover != window_size) {
-			if (!gfs2_holder_queued(&gh)) {
-				if (written)
-					goto out_uninit;
-				goto retry;
-			}
-			goto retry_under_glock;
-		}
+	if (should_fault_in_pages(to, iocb, &prev_count, &window_size)) {
+		gfs2_glock_dq(&gh);
+		window_size -= fault_in_iov_iter_writeable(to, window_size);
+		if (window_size)
+			goto retry;
 	}
+out_unlock:
 	if (gfs2_holder_queued(&gh))
 		gfs2_glock_dq(&gh);
 out_uninit:
 	gfs2_holder_uninit(&gh);
-	return written ? written : ret;
+	return read ? read : ret;
 }
 
 static ssize_t gfs2_file_buffered_write(struct kiocb *iocb,
@@ -1014,7 +1001,7 @@ static ssize_t gfs2_file_buffered_write(struct kiocb *iocb,
 	struct gfs2_holder *statfs_gh = NULL;
 	size_t prev_count = 0, window_size = 0;
 	size_t orig_count = iov_iter_count(from);
-	size_t read = 0;
+	size_t written = 0;
 	ssize_t ret;
 
 	/*
@@ -1032,10 +1019,18 @@ static ssize_t gfs2_file_buffered_write(struct kiocb *iocb,
 
 	gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, gh);
 retry:
+	if (should_fault_in_pages(from, iocb, &prev_count, &window_size)) {
+		window_size -= fault_in_iov_iter_readable(from, window_size);
+		if (!window_size) {
+			ret = -EFAULT;
+			goto out_uninit;
+		}
+		from->count = min(from->count, window_size);
+	}
 	ret = gfs2_glock_nq(gh);
 	if (ret)
 		goto out_uninit;
-retry_under_glock:
+
 	if (inode == sdp->sd_rindex) {
 		struct gfs2_inode *m_ip = GFS2_I(sdp->sd_statfs_inode);
 
@@ -1052,37 +1047,28 @@ retry_under_glock:
 	current->backing_dev_info = NULL;
 	if (ret > 0) {
 		iocb->ki_pos += ret;
-		read += ret;
+		written += ret;
 	}
 
 	if (inode == sdp->sd_rindex)
 		gfs2_glock_dq_uninit(statfs_gh);
 
-	from->count = orig_count - read;
-	if (should_fault_in_pages(ret, from, &prev_count, &window_size)) {
-		size_t leftover;
-
-		gfs2_holder_allow_demote(gh);
-		leftover = fault_in_iov_iter_readable(from, window_size);
-		gfs2_holder_disallow_demote(gh);
-		if (leftover != window_size) {
-			from->count = min(from->count, window_size - leftover);
-			if (!gfs2_holder_queued(gh)) {
-				if (read)
-					goto out_uninit;
-				goto retry;
-			}
-			goto retry_under_glock;
-		}
+	if (ret <= 0 && ret != -EFAULT)
+		goto out_unlock;
+
+	from->count = orig_count - written;
+	if (should_fault_in_pages(from, iocb, &prev_count, &window_size)) {
+		gfs2_glock_dq(gh);
+		goto retry;
 	}
 out_unlock:
 	if (gfs2_holder_queued(gh))
 		gfs2_glock_dq(gh);
 out_uninit:
 	gfs2_holder_uninit(gh);
-	if (statfs_gh)
-		kfree(statfs_gh);
-	return read ? read : ret;
+	kfree(statfs_gh);
+	from->count = orig_count - written;
+	return written ? written : ret;
 }
 
 /**
@@ -1457,6 +1443,22 @@ static int gfs2_lock(struct file *file, int cmd, struct file_lock *fl)
 		return dlm_posix_lock(ls->ls_dlm, ip->i_no_addr, file, cmd, fl);
 }
 
+static void __flock_holder_uninit(struct file *file, struct gfs2_holder *fl_gh)
+{
+	struct gfs2_glock *gl = fl_gh->gh_gl;
+
+	/*
+	 * Make sure gfs2_glock_put() won't sleep under the file->f_lock
+	 * spinlock.
+	 */
+
+	gfs2_glock_hold(gl);
+	spin_lock(&file->f_lock);
+	gfs2_holder_uninit(fl_gh);
+	spin_unlock(&file->f_lock);
+	gfs2_glock_put(gl);
+}
+
 static int do_flock(struct file *file, int cmd, struct file_lock *fl)
 {
 	struct gfs2_file *fp = file->private_data;
@@ -1469,7 +1471,9 @@ static int do_flock(struct file *file, int cmd, struct file_lock *fl)
 	int sleeptime;
 
 	state = (fl->fl_type == F_WRLCK) ? LM_ST_EXCLUSIVE : LM_ST_SHARED;
-	flags = (IS_SETLKW(cmd) ? 0 : LM_FLAG_TRY_1CB) | GL_EXACT;
+	flags = GL_EXACT | GL_NOPID;
+	if (!IS_SETLKW(cmd))
+		flags |= LM_FLAG_TRY_1CB;
 
 	mutex_lock(&fp->f_fl_mutex);
 
@@ -1488,19 +1492,21 @@ static int do_flock(struct file *file, int cmd, struct file_lock *fl)
 				       &gfs2_flock_glops, CREATE, &gl);
 		if (error)
 			goto out;
+		spin_lock(&file->f_lock);
 		gfs2_holder_init(gl, state, flags, fl_gh);
+		spin_unlock(&file->f_lock);
 		gfs2_glock_put(gl);
 	}
 	for (sleeptime = 1; sleeptime <= 4; sleeptime <<= 1) {
 		error = gfs2_glock_nq(fl_gh);
 		if (error != GLR_TRYFAILED)
 			break;
-		fl_gh->gh_flags = LM_FLAG_TRY | GL_EXACT;
-		fl_gh->gh_error = 0;
+		fl_gh->gh_flags &= ~LM_FLAG_TRY_1CB;
+		fl_gh->gh_flags |= LM_FLAG_TRY;
 		msleep(sleeptime);
 	}
 	if (error) {
-		gfs2_holder_uninit(fl_gh);
+		__flock_holder_uninit(file, fl_gh);
 		if (error == GLR_TRYFAILED)
 			error = -EAGAIN;
 	} else {
@@ -1522,7 +1528,7 @@ static void do_unflock(struct file *file, struct file_lock *fl)
 	locks_lock_file_wait(file, fl);
 	if (gfs2_holder_initialized(fl_gh)) {
 		gfs2_glock_dq(fl_gh);
-		gfs2_holder_uninit(fl_gh);
+		__flock_holder_uninit(file, fl_gh);
 	}
 	mutex_unlock(&fp->f_fl_mutex);
 }