| Message ID | 20211119041104.27662-1-songmuchun@bytedance.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | [v2] fs: proc: store PDE()->data into inode->i_private | expand |
On Fri, Nov 19, 2021 at 12:11:04PM +0800, Muchun Song wrote: > + > +/* > + * Obtain the private data passed by user through proc_create_data() or > + * related. > + */ > +static inline void *pde_data(const struct inode *inode) > +{ > + return inode->i_private; > +} > + > +#define PDE_DATA(i) pde_data(i) What is the point of pde_data? If we really think changing to lower case is worth it (I don't think so, using upper case for getting at private data is a common idiom in file systems), we can just do that scripted in one go.
On Thu, 18 Nov 2021 23:23:39 -0800 Christoph Hellwig <hch@infradead.org> wrote: > On Fri, Nov 19, 2021 at 12:11:04PM +0800, Muchun Song wrote: > > + > > +/* > > + * Obtain the private data passed by user through proc_create_data() or > > + * related. > > + */ > > +static inline void *pde_data(const struct inode *inode) > > +{ > > + return inode->i_private; > > +} > > + > > +#define PDE_DATA(i) pde_data(i) > > What is the point of pde_data? It's a regular old C function, hence should be in lower case. I assume the upper case thing is a holdover from when it was implemented as a macro. > If we really think changing to lower > case is worth it (I don't think so, using upper case for getting at > private data is a common idiom in file systems), It is? How odd. I find the upper-case thing to be actively misleading. It's mildly surprising to discover that it's actually a plain old C function. > we can just do that > scripted in one go. Yes, I'd like to see a followup patch which converts the current PDE_DATA() callsites.
On Sat, Nov 20, 2021 at 6:56 AM Andrew Morton <akpm@linux-foundation.org> wrote: > > On Thu, 18 Nov 2021 23:23:39 -0800 Christoph Hellwig <hch@infradead.org> wrote: > > > On Fri, Nov 19, 2021 at 12:11:04PM +0800, Muchun Song wrote: > > > + > > > +/* > > > + * Obtain the private data passed by user through proc_create_data() or > > > + * related. > > > + */ > > > +static inline void *pde_data(const struct inode *inode) > > > +{ > > > + return inode->i_private; > > > +} > > > + > > > +#define PDE_DATA(i) pde_data(i) > > > > What is the point of pde_data? > > It's a regular old C function, hence should be in lower case. > > I assume the upper case thing is a holdover from when it was > implemented as a macro. > > > If we really think changing to lower > > case is worth it (I don't think so, using upper case for getting at > > private data is a common idiom in file systems), > > It is? How odd. > > I find the upper-case thing to be actively misleading. It's mildly > surprising to discover that it's actually a plain old C function. > > > we can just do that > > scripted in one go. > > Yes, I'd like to see a followup patch which converts the current > PDE_DATA() callsites. > You mean replace all PDE_DATA with pde_data in another patch? Thanks.
On Mon, 22 Nov 2021 12:13:33 +0800 Muchun Song <songmuchun@bytedance.com> wrote: > On Sat, Nov 20, 2021 at 6:56 AM Andrew Morton <akpm@linux-foundation.org> wrote: > > > > On Thu, 18 Nov 2021 23:23:39 -0800 Christoph Hellwig <hch@infradead.org> wrote: > > > > > On Fri, Nov 19, 2021 at 12:11:04PM +0800, Muchun Song wrote: > > > > + > > > > +/* > > > > + * Obtain the private data passed by user through proc_create_data() or > > > > + * related. > > > > + */ > > > > +static inline void *pde_data(const struct inode *inode) > > > > +{ > > > > + return inode->i_private; > > > > +} > > > > + > > > > +#define PDE_DATA(i) pde_data(i) > > > > > > What is the point of pde_data? > > > > It's a regular old C function, hence should be in lower case. > > > > I assume the upper case thing is a holdover from when it was > > implemented as a macro. > > > > > If we really think changing to lower > > > case is worth it (I don't think so, using upper case for getting at > > > private data is a common idiom in file systems), > > > > It is? How odd. > > > > I find the upper-case thing to be actively misleading. It's mildly > > surprising to discover that it's actually a plain old C function. > > > > > we can just do that > > > scripted in one go. > > > > Yes, I'd like to see a followup patch which converts the current > > PDE_DATA() callsites. > > > > You mean replace all PDE_DATA with pde_data in another patch? That is indeed what I meant.
diff --git a/fs/proc/generic.c b/fs/proc/generic.c index 5b78739e60e4..f2132407e133 100644 --- a/fs/proc/generic.c +++ b/fs/proc/generic.c @@ -791,12 +791,6 @@ void proc_remove(struct proc_dir_entry *de) } EXPORT_SYMBOL(proc_remove); -void *PDE_DATA(const struct inode *inode) -{ - return __PDE_DATA(inode); -} -EXPORT_SYMBOL(PDE_DATA); - /* * Pull a user buffer into memory and pass it to the file's write handler if * one is supplied. The ->write() method is permitted to modify the diff --git a/fs/proc/inode.c b/fs/proc/inode.c index 599eb724ff2d..f84355c5a36d 100644 --- a/fs/proc/inode.c +++ b/fs/proc/inode.c @@ -650,6 +650,7 @@ struct inode *proc_get_inode(struct super_block *sb, struct proc_dir_entry *de) return NULL; } + inode->i_private = de->data; inode->i_ino = de->low_ino; inode->i_mtime = inode->i_atime = inode->i_ctime = current_time(inode); PROC_I(inode)->pde = de; diff --git a/fs/proc/internal.h b/fs/proc/internal.h index 03415f3fb3a8..06a80f78433d 100644 --- a/fs/proc/internal.h +++ b/fs/proc/internal.h @@ -115,11 +115,6 @@ static inline struct proc_dir_entry *PDE(const struct inode *inode) return PROC_I(inode)->pde; } -static inline void *__PDE_DATA(const struct inode *inode) -{ - return PDE(inode)->data; -} - static inline struct pid *proc_pid(const struct inode *inode) { return PROC_I(inode)->pid; diff --git a/include/linux/proc_fs.h b/include/linux/proc_fs.h index 069c7fd95396..b32fb0ef3308 100644 --- a/include/linux/proc_fs.h +++ b/include/linux/proc_fs.h @@ -110,7 +110,18 @@ extern struct proc_dir_entry *proc_create_data(const char *, umode_t, struct proc_dir_entry *proc_create(const char *name, umode_t mode, struct proc_dir_entry *parent, const struct proc_ops *proc_ops); extern void proc_set_size(struct proc_dir_entry *, loff_t); extern void proc_set_user(struct proc_dir_entry *, kuid_t, kgid_t); -extern void *PDE_DATA(const struct inode *); + +/* + * Obtain the private data passed by user through proc_create_data() or + * related. + */ +static inline void *pde_data(const struct inode *inode) +{ + return inode->i_private; +} + +#define PDE_DATA(i) pde_data(i) + extern void *proc_get_parent_data(const struct inode *); extern void proc_remove(struct proc_dir_entry *); extern void remove_proc_entry(const char *, struct proc_dir_entry *);
PDE_DATA(inode) is introduced to get user private data and hide the layout of struct proc_dir_entry. The inode->i_private is used to do the same thing as well. Save a copy of user private data to inode-> i_private when proc inode is allocated. This means the user also can get their private data by inode->i_private. Introduce pde_data() to wrap inode->i_private so that we can remove PDE_DATA() from fs/proc/generic.c and make PTE_DATE() as a wrapper of pde_data(). It will be easier if we decide to remove PDE_DATE() in the future. Signed-off-by: Muchun Song <songmuchun@bytedance.com> --- Changes in v2: - Drop all drivers related changes (do not remove PDE_DATA() completely). - Introduce pde_data() suggested by Andrew. v1: https://lkml.org/lkml/2021/11/1/575 fs/proc/generic.c | 6 ------ fs/proc/inode.c | 1 + fs/proc/internal.h | 5 ----- include/linux/proc_fs.h | 13 ++++++++++++- 4 files changed, 13 insertions(+), 12 deletions(-)