Re: [RFC PATCH 8/9] debugfs: defer debugfs_fsdata allocation to first usage

[Nicolai Stange]

On Tue, Apr 18 2017, Johannes Berg wrote:

> On Sun, 2017-04-16 at 11:51 +0200, Nicolai Stange wrote:
>> 
>> +++ b/fs/debugfs/file.c
>> @@ -53,6 +53,7 @@ const struct file_operations
>> *debugfs_real_fops(const struct file *filp)
>> Â{
>> Â	struct debugfs_fsdata *fsd = F_DENTRY(filp)->d_fsdata;
>> Â
>> +	WARN_ON((unsigned long)fsd &
>> DEBUGFS_FSDATA_IS_REAL_FOPS_BIT);
>> Â	return fsd->real_fops;
>
> I'm not a fan of BUG_ON(), but in this case, if you have a completely
> bogus pointer here, and then you return fsd->real_fops which will be
> even more bogus, and *then* you call a function from within it... that
> seems like a recipe for disaster.

Hmm. Note that debugfs_real_fops() is provided for users outside of the
debugfs core which encapsulate their file_operations in a larger struct
and access this one through a container_of(debugfs_real_fops(dentry), ...).
Now, what the above WARN_ON() really checks for is whether the
debugfs_real_fops() has been called under a debugfs_file_get()/-_put()
pair. The point of requiring this isn't protection ([1]) or anything, but
merely to be able to streamline the implementation of
debugfs_real_fops(): knowing that a debugfs_file_get() is active allows
for the omission of the WARNed_ON case, namely that the fops are
encoded in ->d_fsdata itself.


> So either you could return some valid ops (perhaps
> debugfs_noop_file_operations although those don't have .name or .poll,
> so it doesn't cover everything), or you can just BUG_ON() here
> directly, saving the incomprehensible crash later.

The purpose of that WARN_ON() there was to turn a potentially
incomprehensible crash into a comprehensible one ;)

In order to avoid a new BUG_ON(), what about keeping the WARN_ON() as is
and returning NULL instead of the garbage? That would crash current on
first access and the earlier warning would hopefully give some clue?

Thanks,

Nicolai


[1] "protection" in the sense of "protection against file removal".  With
    the proposed [9/9] ("debugfs: free debugfs_fsdata instances"), the
    patch introducing occasional freeing of ->d_fsdata, an additional
    RCU read side section would be needed if debugfs_real_fops() were
    allowed to be called outside of a debugfs_file_get()/-_put() pair.


>> ÂEXPORT_SYMBOL_GPL(debugfs_real_fops);
>> @@ -74,9 +75,35 @@ EXPORT_SYMBOL_GPL(debugfs_real_fops);
>> Â */
>> Âint debugfs_file_get(struct dentry *dentry)
>> Â{
>> -	struct debugfs_fsdata *fsd = dentry->d_fsdata;
>> +	struct debugfs_fsdata *fsd;
>> +	void *d_fsd;
>> +
>> +	d_fsd = READ_ONCE(dentry->d_fsdata);
>> +	if (!((unsigned long)d_fsd &
>> DEBUGFS_FSDATA_IS_REAL_FOPS_BIT)) {
>> +		fsd = d_fsd;
>> +	} else {
>> +		fsd = kmalloc(sizeof(*fsd), GFP_KERNEL);
>> +		if (!fsd)
>> +			return -ENOMEM;
>> +
>> +		fsd->real_fops = (void *)((unsigned long)d_fsd &
>> +					~DEBUGFS_FSDATA_IS_REAL_FOPS
>> _BIT);
>> +		refcount_set(&fsd->active_users, 1);
>> +		init_completion(&fsd->active_users_drained);
>> +		if (cmpxchg(&dentry->d_fsdata, d_fsd, fsd) != d_fsd)
>> {
>> +			kfree(fsd);
>> +			fsd = READ_ONCE(dentry->d_fsdata);
>> +		}
>> +	}
>> Â
>> -	/* Avoid starvation of removers. */
>> +	/*
>> +	Â* In case of a successful cmpxchg() above, this check is
>> +	Â* strictly necessary and must follow it, see the comment in
>> +	Â* __debugfs_remove_file().
>> +	Â* OTOH, if the cmpxchg() hasn't been executed or wasn't
>> +	Â* successful, this serves the purpose of not starving
>> +	Â* removers.
>> +	Â*/
>> Â	if (d_unlinked(dentry))
>> Â		return -EIO;
>> Â
>> @@ -98,7 +125,7 @@ EXPORT_SYMBOL_GPL(debugfs_file_get);
>> Â */
>> Âvoid debugfs_file_put(struct dentry *dentry)
>> Â{
>> -	struct debugfs_fsdata *fsd = dentry->d_fsdata;
>> +	struct debugfs_fsdata *fsd = READ_ONCE(dentry->d_fsdata);
>> Â
>> Â	if (refcount_dec_and_test(&fsd->active_users))
>> Â		complete(&fsd->active_users_drained);
>> @@ -109,10 +136,11 @@ static int open_proxy_open(struct inode *inode,
>> struct file *filp)
>> Â{
>> Â	struct dentry *dentry = F_DENTRY(filp);
>> Â	const struct file_operations *real_fops = NULL;
>> -	int r = 0;
>> +	int r;
>> Â
>> -	if (debugfs_file_get(dentry))
>> -		return -ENOENT;
>> +	r = debugfs_file_get(dentry);
>> +	if (r)
>> +		return r == -EIO ? -ENOENT : r;
>> Â
>> Â	real_fops = debugfs_real_fops(filp);
>> Â	real_fops = fops_get(real_fops);
>> @@ -233,10 +261,11 @@ static int full_proxy_open(struct inode *inode,
>> struct file *filp)
>> Â	struct dentry *dentry = F_DENTRY(filp);
>> Â	const struct file_operations *real_fops = NULL;
>> Â	struct file_operations *proxy_fops = NULL;
>> -	int r = 0;
>> +	int r;
>> Â
>> -	if (debugfs_file_get(dentry))
>> -		return -ENOENT;
>> +	r = debugfs_file_get(dentry);
>> +	if (r)
>> +		return r == -EIO ? -ENOENT : r;
>> Â
>> Â	real_fops = debugfs_real_fops(filp);
>> Â	real_fops = fops_get(real_fops);
>> diff --git a/fs/debugfs/inode.c b/fs/debugfs/inode.c
>> index 5550f11d60bd..2360c17ec00a 100644
>> --- a/fs/debugfs/inode.c
>> +++ b/fs/debugfs/inode.c
>> @@ -184,7 +184,10 @@ static const struct super_operations
>> debugfs_super_operations = {
>> Â
>> Âstatic void debugfs_release_dentry(struct dentry *dentry)
>> Â{
>> -	kfree(dentry->d_fsdata);
>> +	void *fsd = dentry->d_fsdata;
>> +
>> +	if (!((unsigned long)fsd & DEBUGFS_FSDATA_IS_REAL_FOPS_BIT))
>> +		kfree(dentry->d_fsdata);
>> Â}
>> Â
>> Âstatic struct vfsmount *debugfs_automount(struct path *path)
>> @@ -346,35 +349,25 @@ static struct dentry
>> *__debugfs_create_file(const char *name, umode_t mode,
>> Â{
>> Â	struct dentry *dentry;
>> Â	struct inode *inode;
>> -	struct debugfs_fsdata *fsd;
>> -
>> -	fsd = kmalloc(sizeof(*fsd), GFP_KERNEL);
>> -	if (!fsd)
>> -		return NULL;
>> Â
>> Â	if (!(mode & S_IFMT))
>> Â		mode |= S_IFREG;
>> Â	BUG_ON(!S_ISREG(mode));
>> Â	dentry = start_creating(name, parent);
>> Â
>> -	if (IS_ERR(dentry)) {
>> -		kfree(fsd);
>> +	if (IS_ERR(dentry))
>> Â		return NULL;
>> -	}
>> Â
>> Â	inode = debugfs_get_inode(dentry->d_sb);
>> -	if (unlikely(!inode)) {
>> -		kfree(fsd);
>> +	if (unlikely(!inode))
>> Â		return failed_creating(dentry);
>> -	}
>> Â
>> Â	inode->i_mode = mode;
>> Â	inode->i_private = data;
>> Â
>> Â	inode->i_fop = proxy_fops;
>> -	fsd->real_fops = real_fops;
>> -	refcount_set(&fsd->active_users, 1);
>> -	dentry->d_fsdata = fsd;
>> +	dentry->d_fsdata = (void *)((unsigned long)real_fops |
>> +				DEBUGFS_FSDATA_IS_REAL_FOPS_BIT);
>> Â
>> Â	d_instantiate(dentry, inode);
>> Â	fsnotify_create(d_inode(dentry->d_parent), dentry);
>> @@ -637,8 +630,17 @@ static void __debugfs_remove_file(struct dentry
>> *dentry, struct dentry *parent)
>> Â
>> Â	simple_unlink(d_inode(parent), dentry);
>> Â	d_delete(dentry);
>> -	fsd = dentry->d_fsdata;
>> -	init_completion(&fsd->active_users_drained);
>> +
>> +	/*
>> +	Â* Paired with the closing smp_mb() implied by a successful
>> +	Â* cmpxchg() in debugfs_file_get(): either
>> +	Â* debugfs_file_get() must see a dead dentry or we must see
>> a
>> +	Â* debugfs_fsdata instance at ->d_fsdata here (or both).
>> +	Â*/
>> +	smp_mb();
>> +	fsd = READ_ONCE(dentry->d_fsdata);
>> +	if ((unsigned long)fsd & DEBUGFS_FSDATA_IS_REAL_FOPS_BIT)
>> +		return;
>> Â	if (!refcount_dec_and_test(&fsd->active_users))
>> Â		wait_for_completion(&fsd->active_users_drained);
>> Â}
>> diff --git a/fs/debugfs/internal.h b/fs/debugfs/internal.h
>> index 0eea99432840..cb1e8139c398 100644
>> --- a/fs/debugfs/internal.h
>> +++ b/fs/debugfs/internal.h
>> @@ -25,4 +25,12 @@ struct debugfs_fsdata {
>> Â	struct completion active_users_drained;
>> Â};
>> Â
>> +/*
>> + * A dentry's ->d_fsdata either points to the real fops or to a
>> + * dynamically allocated debugfs_fsdata instance.
>> + * In order to distinguish between these two cases, a real fops
>> + * pointer gets its lowest bit set.
>> + */
>> +#define DEBUGFS_FSDATA_IS_REAL_FOPS_BIT BIT(0)
>> +
>> Â#endif /* _DEBUGFS_INTERNAL_H_ */