Remove fcntl f_op
Subsystems (4)
PATCH 1/1 Remove fcntl f_op
The newly introduced ->fcntl file_operation is badly thought out, not to mention undocumented. This patch replaces it with two better defined operations -- check_flags and dir_notify. Any other fcntl()s that filesystems are interested in can have their own properly typed f_op method when they need it.
Documentation/filesystems/Locking +4 -0
@@ -349,6 +349,8 @@ prototypes: loff_t *, int); unsigned long (*get_unmapped_area)(struct file *, unsigned long, unsigned long, unsigned long, unsigned long); + int (*check_flags)(int); + int (*dir_notify)(struct file *, unsigned long); }; locking rules:
@@ -375,6 +377,8 @@ writev: no sendfile: no sendpage: no get_unmapped_area: no +check_flags: no +dir_notify: no ->llseek() locking has moved from llseek to the individual llseek implementations. If your fs is not using generic_file_llseek, you
fs/cifs/cifsfs.c +2 -6
@@ -540,18 +540,14 @@ struct file_operations cifs_file_ops = { .flush = cifs_flush, .mmap = cifs_file_mmap, .sendfile = generic_file_sendfile, -#ifdef CONFIG_CIFS_FCNTL - .fcntl = cifs_fcntl, -#endif + .dir_notify = cifs_dir_notify, };struct file_operations cifs_dir_ops = { .readdir = cifs_readdir, .release = cifs_closedir, .read = generic_read_dir, -#ifdef CONFIG_CIFS_FCNTL - .fcntl = cifs_fcntl, -#endif + .dir_notify = cifs_dir_notify, }; static void
fs/cifs/cifsfs.h +1 -1
@@ -78,7 +78,7 @@ extern int cifs_file_mmap(struct file *extern struct file_operations cifs_dir_ops;extern int cifs_dir_open (struct inode *inode,struct file *file);extern int cifs_readdir (struct file *file,void *direntry,filldir_t filldir); -extern long cifs_fcntl (int ,unsigned int ,unsigned long ,struct file *); +extern int cifs_dir_notify (struct file *,unsigned long arg);/* Functions related to dir entries */ extern struct dentry_operations cifs_dentry_ops;
fs/cifs/fcntl.c +1 -51
61 lines not on this page · raw message
fs/dnotify.c +3 -0
@@ -104,6 +104,9 @@ int fcntl_dirnotify(int fd, struct file dn->dn_next = inode->i_dnotify; inode->i_dnotify = dn;spin_unlock (&inode->i_lock); + +if (filp->f_op && filp->f_op->dir_notify) +return filp->f_op->dir_notify (filp, arg);return 0 ; out_free:
fs/fcntl.c +7 -11
@@ -239,6 +239,11 @@ static int setfl(int fd, struct file * freturn -EINVAL ; } +if (filp->f_op && filp->f_op->check_flags) + error = filp->f_op->check_flags (arg); +if (error) +return error; +lock_kernel ();if ((arg ^ filp->f_flags) &FASYNC ) {if (filp->f_op && filp->f_op->fasync) {
@@ -287,8 +292,8 @@ void f_delown(struct file *filp)EXPORT_SYMBOL (f_delown); -long generic_file_fcntl (int fd,unsigned int cmd, -unsigned long arg,struct file *filp) +static long do_fcntl (int fd,unsigned int cmd,unsigned long arg, +struct file *filp) {long err = -EINVAL ;
@@ -356,15 +359,6 @@ long generic_file_fcntl(int fd, unsigned }return err; } -EXPORT_SYMBOL (generic_file_fcntl); - -static long do_fcntl (int fd,unsigned int cmd, -unsigned long arg,struct file *filp) -{ -if (filp->f_op && filp->f_op->fcntl) -return filp->f_op->fcntl (fd, cmd, arg, filp); -return generic_file_fcntl (fd, cmd, arg, filp); -}asmlinkage long sys_fcntl (int fd,unsigned int cmd,unsigned long arg) {
fs/nfs/file.c +11 -25
@@ -33,8 +33,6 @@#define NFSDBG_FACILITY NFSDBG_FILE -static long nfs_file_fcntl (int fd,unsigned int cmd, -unsigned long arg,struct file *filp);static int nfs_file_open (struct inode *,struct file *);static int nfs_file_release (struct inode *,struct file *);static int nfs_file_mmap (struct file *,struct vm_area_struct *);
@@ -43,6 +41,7 @@ static ssize_t nfs_file_read(struct kiocstatic ssize_t nfs_file_write (struct kiocb *,const char __user *,size_t ,loff_t );static int nfs_file_flush (struct file *);static int nfs_fsync (struct file *,struct dentry *dentry,int datasync); +static int nfs_check_flags (int flags);struct file_operations nfs_file_operations = { .llseek = remote_llseek,
@@ -57,7 +56,7 @@ struct file_operations nfs_file_operatio .fsync = nfs_fsync, .lock = nfs_lock, .sendfile = nfs_file_sendfile, - .fcntl = nfs_file_fcntl, + .check_flags = nfs_check_flags, };struct inode_operations nfs_file_inode_operations = {
@@ -71,26 +70,12 @@ struct inode_operations nfs_file_inode_o# define IS_SWAPFILE (inode) (0)#endif -#define nfs_invalid_flags (O_APPEND | O_DIRECT) - -/* -* Check for special cases that NFS doesn't support, and -* pass the rest to the generic fcntl function. -*/ -static long -nfs_file_fcntl (int fd,unsigned int cmd, -unsigned long arg,struct file *filp) -{ -switch (cmd) { -case F_SETFL : -if ((filp->f_flags & nfs_invalid_flags) == nfs_invalid_flags) -return -EINVAL ; -break ; -default : -break ; - } +static int nfs_check_flags (int flags) +{ +if (flags & (O_APPEND |O_DIRECT )) +return -EINVAL ; -return generic_file_fcntl (fd, cmd, arg, filp); +return 0 ; }/*
@@ -101,10 +86,11 @@ nfs_file_open(struct inode *inode, struc {struct nfs_server *server =NFS_SERVER (inode);int (*open)(struct inode *,struct file *); -int res =0 ; +int res; -if ((filp->f_flags & nfs_invalid_flags) == nfs_invalid_flags) -return -EINVAL ; + res =nfs_check_flags (filp->f_flags); +if (!res) +return res;lock_kernel ();/* Do NFSv4 open() call */
include/linux/fs.h +2 -5
@@ -659,9 +679,6 @@ extern struct list_head file_lock_list;#include <linux/fcntl.h> -extern long generic_file_fcntl (int fd,unsigned int cmd, -unsigned long arg,struct file *filp); -extern int fcntl_getlk (struct file *,struct flock __user *);extern int fcntl_setlk (struct file *,unsigned int ,struct flock __user *);
@@ -889,8 +907,8 @@ struct file_operations {ssize_t (*sendfile) (struct file *,loff_t *,size_t ,read_actor_t ,void *);ssize_t (*sendpage) (struct file *,struct page *,int ,size_t ,loff_t *,int );unsigned long (*get_unmapped_area)(struct file *,unsigned long ,unsigned long ,unsigned long ,unsigned long ); -long (*fcntl)(int fd,unsigned int cmd, -unsigned long arg,struct file *filp); +int (*check_flags)(int ); +int (*dir_notify)(struct file *filp,unsigned long arg); };struct inode_operations {
Signature · 6 lines
"Next the statesmen will invent cheap lies, putting the blame upon the nation that is attacked, and every man will be glad of those conscience-soothing falsities, and will diligently study them, and refuse to examine any refutations of them; and thus he will by and by convince himself that the war is just, and will thank God for the better sleep he enjoys after this process of grotesque self-deception." -- Mark Twain
Quoted 6 lines · Matthew Wilcox
The newly introduced ->fcntl
file_operationis badly thought out, not to mention undocumented. This patch replaces it with two better defined operations --check_flagsanddir_notify. Any other fcntl()s that filesystems are interested in can have their own properly typedf_opmethod when they need it.
ACK
Quoted 2 lines · Matthew Wilcox
The newly introduced ->fcntl
file_operationis badly thought out,
Not really - it's very simple.
Quoted 1 line
not to mention undocumented.
That's because it's simple.
Quoted 2 lines
This patch replaces it with two better defined operations --
check_flagsanddir_notify.
Except these _do_ need to be documented, and you didn't do that.
Seems like a fairly pointless patch to me but whatever - it's a trivial matter. Please send an updated patch which documents the new API.
Quoted 4 lines · Andrew Morton
Quoted 1 line · Matthew Wilcox
The newly introduced ->fcntl
file_operationis badly thought out,Not really - it's very simple.
Mmm. It's typeless, passes the fd number down to the filesystem, and pretty much requires a switch() in every filesystem that wants to implement it. It's also a bad API in that it allows filesystems to introduce private fcntls or do something completely different in response to a standard fcntl.
Quoted 7 lines
Quoted 2 lines
This patch replaces it with two better defined operations --
check_flagsanddir_notify.Except these _do_ need to be documented, and you didn't do that.
Seems like a fairly pointless patch to me but whatever - it's a trivial matter. Please send an updated patch which documents the new API.
I documented them where the other file_operations are documented, in Documentation/filesystems/Locking.
If you wanted Documentation/filesystems/vfs.txt to mention them, well, that needs to be updated to something from this decade first -- I can't believe anyone's actually reading that. Just from file_operations, it's missing aio_read, aio_write, aio_fsync, readv, writev, sendfile, sendpage and get_unmapped_area. It also claims check_media_change and revalidate are file_operations, when they aren't.
If there's somewhere else to document them, I'd be only too happy to.
Signature · 6 lines
"Next the statesmen will invent cheap lies, putting the blame upon the nation that is attacked, and every man will be glad of those conscience-soothing falsities, and will diligently study them, and refuse to examine any refutations of them; and thus he will by and by convince himself that the war is just, and will thank God for the better sleep he enjoys after this process of grotesque self-deception." -- Mark Twain
Quoted 3 lines · Matthew Wilcox
I documented them where the other
file_operationsare documented, inDocumentation/filesystems/Locking.
Documentation/filesystems/Locking does document address_space_ops. But not, it seems, file_operations.
Quoted 8 lines
If you wanted
Documentation/filesystems/vfs.txtto mention them, well, that needs to be updated to something from this decade first -- I can't believe anyone's actually reading that. Just fromfile_operations, it's missingaio_read,aio_write,aio_fsync, readv, writev, sendfile, sendpage andget_unmapped_area. It also claimscheck_media_changeand revalidate arefile_operations, when they aren't.If there's somewhere else to document them, I'd be only too happy to.
Documentation in Documentation/ goes out of date - we've seen this again and again and again. Documentation in code comments gets a lot more eyeball time, and people spot errors and omissions.
For this reason, I'd suggest that we prefer to document the code inside the code. External interfaces and things which are distributed all over the tree (proc.txt, kernel-parameters.txt, etc) are more appropriately handled in Documentation/.