Re: [PATCH 1/8] quota: Allow to pass mount path to quotactl

[Jan Kara <jack@suse.cz>]

[-- Attachment #1: Type: text/plain, Size: 3472 bytes --]

I'm sorry for the late reply, I was busy with other things and I wasn't
quite sure how I'd like this to be handled :)

On Tue 04-02-20 11:35:23, Sascha Hauer wrote:
> On Wed, Jan 29, 2020 at 01:29:29AM +0000, Al Viro wrote:
> > On Tue, Jan 28, 2020 at 11:06:31AM +0100, Sascha Hauer wrote:
> > > Hi Jan,
> > 
> > > @@ -810,6 +811,36 @@ static struct super_block *quotactl_block(const char __user *special, int cmd)
> > >  #endif
> > >  }
> > >  
> > > +static struct super_block *quotactl_path(const char __user *special, int cmd,
> > > +					 struct path *path)
> > > +{
> > > +	struct super_block *sb;
> > > +	int ret;
> > > +
> > > +	ret = user_path_at(AT_FDCWD, special, LOOKUP_FOLLOW | LOOKUP_AUTOMOUNT,
> > > +			   path);
> > > +	if (ret)
> > > +		return ERR_PTR(ret);
> > > +
> > > +	sb = path->mnt->mnt_sb;
> > > +restart:
> > > +	if (quotactl_cmd_onoff(cmd))
> > > +		down_write(&sb->s_umount);
> > > +	else
> > > +		down_read(&sb->s_umount);
> > > +
> > > +	if (quotactl_cmd_write(cmd) && sb->s_writers.frozen != SB_UNFROZEN) {
> > > +		if (quotactl_cmd_onoff(cmd))
> > > +			up_write(&sb->s_umount);
> > > +		else
> > > +			up_read(&sb->s_umount);
> > > +		wait_event(sb->s_writers.wait_unfrozen,
> > > +			   sb->s_writers.frozen == SB_UNFROZEN);
> > > +		goto restart;
> > > +	}
> > > +
> > > +	return sb;
> > > +}
> > 
> > This partial duplicate of __get_super_thawed() guts does *not* belong here,
> > especially not interleaved with quota-specific checks.
> > 
> > > +	if (q_path) {
> > > +		if (quotactl_cmd_onoff(cmd))
> > > +			up_write(&sb->s_umount);
> > > +		else
> > > +			up_read(&sb->s_umount);
> > > +
> > > +		path_put(&sb_path);
> > > +	} else {
> > > +		if (!quotactl_cmd_onoff(cmds))
> > > +			drop_super(sb);
> > > +		else
> > > +			drop_super_exclusive(sb);
> > > +	}
> > 
> > Er...  Why not have the same code that you've used to lock the damn thing
> > (needs to be moved to fs/super.c) simply get a passive ref to it?  Then
> > you could do the same thing, q_path or no q_path...
> 
> I am getting confused here. To an earlier version of this series you
> responded:
> 
> > And for path-based you don't need to mess with superblock
> > references - just keep the struct path until the end.  That
> > will keep the superblock alive and active just fine.
> 
> I did that and got the objection from Jan:
> 
> > So I've realized that just looking up superblock with user_path_at() is not
> > enough. Quota code also expects that the superblock will be locked
> > (sb->s_umount) and filesystem will not be frozen (in case the quota
> > operation is going to modify the filesystem). This is needed to serialize
> > e.g. remount and quota operations or quota operations among themselves.

Yes, using passive reference is not necessary. On the other hand the
symmetry with how get_super() and friends work has some appeal too so if Al
wants that, well, he's the maintainer ;)

> So after drawing circles we now seem to be back at passive references.
> What I have now in my tree is this in fs/super.c, untested currently:

I was thinking how to make the API most sensible. In the end I've decided
for a variant that is attached - we pass in struct path which enforces
active reference to a superblock and thus we don't have to be afraid of the
superblock going away or similar problems. Also the operation "get me
superblock for a path" kind of makes sense...

Guys, what do you think?

								Honza
-- 
Jan Kara <jack@suse.com>
SUSE Labs, CR

[-- Attachment #2: 0001-fs-Provide-functions-for-getting-locked-and-thawed-s.patch --]
[-- Type: text/x-patch, Size: 3714 bytes --]

From 18c7913475342f10b4723c7e22409acc573e6436 Mon Sep 17 00:00:00 2001
From: Jan Kara <jack@suse.cz>
Date: Tue, 18 Feb 2020 10:05:21 +0100
Subject: [PATCH] fs: Provide functions for getting locked and thawed
 superblock from a path

Provide functions to get locked (with s_umount) and if desired also thawed
superblock for a given struct path. We additionally also get passive
reference count of the superblock so that the superblock can be unlocked
with put_super() / put_super_exlusive() similarly to how get_super_*()
class of functions operates.

Signed-off-by: Jan Kara <jack@suse.cz>
---
 fs/super.c         | 64 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
 include/linux/fs.h |  3 +++
 2 files changed, 67 insertions(+)

diff --git a/fs/super.c b/fs/super.c
index cd352530eca9..2c78eac57fa5 100644
--- a/fs/super.c
+++ b/fs/super.c
@@ -836,6 +836,70 @@ struct super_block *get_super_exclusive_thawed(struct block_device *bdev)
 }
 EXPORT_SYMBOL(get_super_exclusive_thawed);
 
+static struct super_block *__hold_super(struct path *path, bool excl,
+					bool thawed)
+{
+	struct super_block *sb = path->mnt->mnt_sb;
+
+	while (1) {
+		if (excl)
+			down_write(&sb->s_umount);
+		else
+			down_read(&sb->s_umount);
+		if (!thawed || sb->s_writers.frozen == SB_UNFROZEN) {
+			spin_lock(&sb_lock);
+			sb->s_count++;
+			spin_unlock(&sb_lock);
+			return sb;
+		}
+		if (excl)
+			up_write(&sb->s_umount);
+		else
+			up_read(&sb->s_umount);
+		wait_event(sb->s_writers.wait_unfrozen,
+					sb->s_writers.frozen == SB_UNFROZEN);
+	}
+}
+
+/**
+ *	hold_super - get locked superblock for a path
+ *	@path: path to get superblock for
+ *
+ *	This function gets superblock for @path and returns it with s_umount
+ *	held in shared mode and superblock's passive refcount (sb->s_count)
+ *	incremented.
+ */
+struct super_block *hold_super(struct path *path)
+{
+	return __hold_super(path, false, false);
+}
+
+/**
+ *	hold_super_thawed - get locked thawed superblock for a path
+ *	@path: path to get superblock for
+ *
+ *	This function gets superblock for @path, makes sure it is not frozen,
+ *	and returns it with s_umount held in shared mode and superblock's
+ *	passive refcount (sb->s_count) incremented.
+ */
+struct super_block *hold_super_thawed(struct path *path)
+{
+	return __hold_super(path, false, true);
+}
+
+/**
+ *	hold_super_thawed_exclusive - get locked thawed superblock for a path
+ *	@path: path to get superblock for
+ *
+ *	This function gets superblock for @path, makes sure it is not frozen,
+ *	and returns it with s_umount held in exclusive mode and superblock's
+ *	passive refcount (sb->s_count) incremented.
+ */
+struct super_block *hold_super_thawed_exclusive(struct path *path)
+{
+	return __hold_super(path, true, true);
+}
+
 /**
  * get_active_super - get an active reference to the superblock of a device
  * @bdev: device to get the superblock for
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 3cd4fe6b845e..2406494d2f54 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -3305,6 +3305,9 @@ extern struct file_system_type *get_fs_type(const char *name);
 extern struct super_block *get_super(struct block_device *);
 extern struct super_block *get_super_thawed(struct block_device *);
 extern struct super_block *get_super_exclusive_thawed(struct block_device *bdev);
+struct super_block *hold_super(struct path *path);
+struct super_block *hold_super_thawed(struct path *path);
+struct super_block *hold_super_thawed_exclusive(struct path *path);
 extern struct super_block *get_active_super(struct block_device *bdev);
 extern void drop_super(struct super_block *sb);
 extern void drop_super_exclusive(struct super_block *sb);
-- 
2.16.4


[-- Attachment #3: Type: text/plain, Size: 144 bytes --]

______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/