* [fake patch] un-inline sysfs_get_kobject and release_sysfs_dirent
@ 2004-12-09 5:26 Adam J. Richter
2004-12-30 13:35 ` Maneesh Soni
0 siblings, 1 reply; 2+ messages in thread
From: Adam J. Richter @ 2004-12-09 5:26 UTC (permalink / raw)
To: maneesh; +Cc: akpm, chrisw, greg, linux-kernel, viro
Hi Maneesh,
The following fake patch against a heavily hacked
linux-2.6.10-rc2-bk16, which moves sysfs_get_kobject and release_sysfs_dirent
from fs/sysfs/sysfs.h as inline functions to .c files as non-inline
functions. These routines are big enough, that they should not be
inline functions. Although my previous patches have contributed
to this, I believe that the routines were still big enough that making
them non-inline would be justified even without any of my previous
patches.
There were no changes to the contents of the functions other
than making them non-inline, although a previous patch that I posted
added SYSFS_ATTR_GROUP, which resulted in a change to sysfs_get_kobject
which affects this patch, as does my change to avoid allocating
the "s_children" field for non-directories.
I am posting this patch now to simplify my patch for
unpinning sysfs directories, which I expect to post in about
an hour. I am posting it against 2.6.10-rc2-bk16 for consistency
with my stream of patches from a few days ago, which this patch
is based on. After posting the patch for unpinning sysfs directories,
I expect to update to the latest snapshot, and I will then
be able to regenerate patches against that if that would be
helpful.
__ ______________
Adam J. Richter \ /
adam@yggdrasil.com | g g d r a s i l
Signed-off-by: Adam J. Richter <adam@yggdrasil.com>
diff -u linux/fs/sysfs.old/dir.c linux/fs/sysfs/dir.c
--- linux/fs/sysfs.old/dir.c 2004-12-03 12:04:45.000000000 +0800
+++ linux/fs/sysfs/dir.c 2004-12-09 13:14:03.000000000 +0800
@@ -57,6 +57,20 @@
return sd;
}
+void release_sysfs_dirent(struct sysfs_dirent * sd)
+{
+ if (sd->s_type & SYSFS_KOBJ_LINK) {
+ struct sysfs_symlink * sl = sd->s_element;
+ kfree(sl->link_name);
+ kobject_put(sl->target_kobj);
+ kfree(sl);
+ }
+ if (sysfs_type_dir(sd->s_type))
+ kmem_cache_free(sysfs_dir_cachep, to_sysfs_dir(sd));
+ else
+ kmem_cache_free(sysfs_dirent_cachep, sd);
+}
+
int sysfs_make_dirent(struct sysfs_dir * parent_sd, struct dentry * dentry,
void * element, umode_t mode, int type)
{
diff -u linux/fs/sysfs.old/inode.c linux/fs/sysfs/inode.c
--- linux/fs/sysfs.old/inode.c 2004-12-03 12:04:45.000000000 +0800
+++ linux/fs/sysfs/inode.c 2004-12-09 13:14:21.000000000 +0800
@@ -164,4 +164,24 @@
up(&dir->d_inode->i_sem);
}
+struct kobject *sysfs_get_kobject(struct dentry *dentry)
+{
+ struct kobject * kobj = NULL;
+ spin_lock(&dcache_lock);
+ if (!d_unhashed(dentry)) {
+ struct sysfs_dirent * sd = dentry->d_fsdata;
+ if (sd->s_type == SYSFS_ATTR_GROUP) {
+ sd = dentry->d_parent->d_fsdata;
+ BUG_ON(sd->s_type != SYSFS_DIR);
+ }
+ if (sd->s_type & SYSFS_KOBJ_LINK) {
+ struct sysfs_symlink * sl = sd->s_element;
+ kobj = kobject_get(sl->target_kobj);
+ } else
+ kobj = kobject_get(sd->s_element);
+ }
+ spin_unlock(&dcache_lock);
+
+ return kobj;
+}
diff -u linux/fs/sysfs.old/sysfs.h linux/fs/sysfs/sysfs.h
--- linux/fs/sysfs.old/sysfs.h 2004-12-03 12:04:45.000000000 +0800
+++ linux/fs/sysfs/sysfs.h 2004-12-09 13:09:40.000000000 +0800
@@ -98,41 +98,8 @@
return ((struct bin_attribute *) sd->s_element);
}
-static inline struct kobject *sysfs_get_kobject(struct dentry *dentry)
-{
- struct kobject * kobj = NULL;
-
- spin_lock(&dcache_lock);
- if (!d_unhashed(dentry)) {
- struct sysfs_dirent * sd = dentry->d_fsdata;
- if (sd->s_type == SYSFS_ATTR_GROUP) {
- sd = dentry->d_parent->d_fsdata;
- BUG_ON(sd->s_type != SYSFS_DIR);
- }
- if (sd->s_type & SYSFS_KOBJ_LINK) {
- struct sysfs_symlink * sl = sd->s_element;
- kobj = kobject_get(sl->target_kobj);
- } else
- kobj = kobject_get(sd->s_element);
- }
- spin_unlock(&dcache_lock);
-
- return kobj;
-}
-
-static inline void release_sysfs_dirent(struct sysfs_dirent * sd)
-{
- if (sd->s_type & SYSFS_KOBJ_LINK) {
- struct sysfs_symlink * sl = sd->s_element;
- kfree(sl->link_name);
- kobject_put(sl->target_kobj);
- kfree(sl);
- }
- if (sysfs_type_dir(sd->s_type))
- kmem_cache_free(sysfs_dir_cachep, to_sysfs_dir(sd));
- else
- kmem_cache_free(sysfs_dirent_cachep, sd);
-}
+extern struct kobject *sysfs_get_kobject(struct dentry *dentry);
+extern void release_sysfs_dirent(struct sysfs_dirent * sd);
static inline void sysfs_put(struct sysfs_dirent * sd)
{
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: [fake patch] un-inline sysfs_get_kobject and release_sysfs_dirent
2004-12-09 5:26 [fake patch] un-inline sysfs_get_kobject and release_sysfs_dirent Adam J. Richter
@ 2004-12-30 13:35 ` Maneesh Soni
0 siblings, 0 replies; 2+ messages in thread
From: Maneesh Soni @ 2004-12-30 13:35 UTC (permalink / raw)
To: Adam J. Richter; +Cc: akpm, chrisw, greg, linux-kernel, viro
On Wed, Dec 08, 2004 at 09:26:32PM -0800, Adam J. Richter wrote:
> Hi Maneesh,
>
> The following fake patch against a heavily hacked
> linux-2.6.10-rc2-bk16, which moves sysfs_get_kobject and release_sysfs_dirent
> from fs/sysfs/sysfs.h as inline functions to .c files as non-inline
> functions. These routines are big enough, that they should not be
> inline functions. Although my previous patches have contributed
> to this, I believe that the routines were still big enough that making
> them non-inline would be justified even without any of my previous
> patches.
>
> There were no changes to the contents of the functions other
> than making them non-inline, although a previous patch that I posted
> added SYSFS_ATTR_GROUP, which resulted in a change to sysfs_get_kobject
> which affects this patch, as does my change to avoid allocating
> the "s_children" field for non-directories.
>
> I am posting this patch now to simplify my patch for
> unpinning sysfs directories, which I expect to post in about
> an hour. I am posting it against 2.6.10-rc2-bk16 for consistency
> with my stream of patches from a few days ago, which this patch
> is based on. After posting the patch for unpinning sysfs directories,
> I expect to update to the latest snapshot, and I will then
> be able to regenerate patches against that if that would be
> helpful.
>
No problem in un-inlining.
Thanks
Maneesh
>
> Signed-off-by: Adam J. Richter <adam@yggdrasil.com>
>
> diff -u linux/fs/sysfs.old/dir.c linux/fs/sysfs/dir.c
> --- linux/fs/sysfs.old/dir.c 2004-12-03 12:04:45.000000000 +0800
> +++ linux/fs/sysfs/dir.c 2004-12-09 13:14:03.000000000 +0800
> @@ -57,6 +57,20 @@
> return sd;
> }
>
> +void release_sysfs_dirent(struct sysfs_dirent * sd)
> +{
> + if (sd->s_type & SYSFS_KOBJ_LINK) {
> + struct sysfs_symlink * sl = sd->s_element;
> + kfree(sl->link_name);
> + kobject_put(sl->target_kobj);
> + kfree(sl);
> + }
> + if (sysfs_type_dir(sd->s_type))
> + kmem_cache_free(sysfs_dir_cachep, to_sysfs_dir(sd));
> + else
> + kmem_cache_free(sysfs_dirent_cachep, sd);
> +}
> +
> int sysfs_make_dirent(struct sysfs_dir * parent_sd, struct dentry * dentry,
> void * element, umode_t mode, int type)
> {
> diff -u linux/fs/sysfs.old/inode.c linux/fs/sysfs/inode.c
> --- linux/fs/sysfs.old/inode.c 2004-12-03 12:04:45.000000000 +0800
> +++ linux/fs/sysfs/inode.c 2004-12-09 13:14:21.000000000 +0800
> @@ -164,4 +164,24 @@
> up(&dir->d_inode->i_sem);
> }
>
> +struct kobject *sysfs_get_kobject(struct dentry *dentry)
> +{
> + struct kobject * kobj = NULL;
>
> + spin_lock(&dcache_lock);
> + if (!d_unhashed(dentry)) {
> + struct sysfs_dirent * sd = dentry->d_fsdata;
> + if (sd->s_type == SYSFS_ATTR_GROUP) {
> + sd = dentry->d_parent->d_fsdata;
> + BUG_ON(sd->s_type != SYSFS_DIR);
> + }
> + if (sd->s_type & SYSFS_KOBJ_LINK) {
> + struct sysfs_symlink * sl = sd->s_element;
> + kobj = kobject_get(sl->target_kobj);
> + } else
> + kobj = kobject_get(sd->s_element);
> + }
> + spin_unlock(&dcache_lock);
> +
> + return kobj;
> +}
> diff -u linux/fs/sysfs.old/sysfs.h linux/fs/sysfs/sysfs.h
> --- linux/fs/sysfs.old/sysfs.h 2004-12-03 12:04:45.000000000 +0800
> +++ linux/fs/sysfs/sysfs.h 2004-12-09 13:09:40.000000000 +0800
> @@ -98,41 +98,8 @@
> return ((struct bin_attribute *) sd->s_element);
> }
>
> -static inline struct kobject *sysfs_get_kobject(struct dentry *dentry)
> -{
> - struct kobject * kobj = NULL;
> -
> - spin_lock(&dcache_lock);
> - if (!d_unhashed(dentry)) {
> - struct sysfs_dirent * sd = dentry->d_fsdata;
> - if (sd->s_type == SYSFS_ATTR_GROUP) {
> - sd = dentry->d_parent->d_fsdata;
> - BUG_ON(sd->s_type != SYSFS_DIR);
> - }
> - if (sd->s_type & SYSFS_KOBJ_LINK) {
> - struct sysfs_symlink * sl = sd->s_element;
> - kobj = kobject_get(sl->target_kobj);
> - } else
> - kobj = kobject_get(sd->s_element);
> - }
> - spin_unlock(&dcache_lock);
> -
> - return kobj;
> -}
> -
> -static inline void release_sysfs_dirent(struct sysfs_dirent * sd)
> -{
> - if (sd->s_type & SYSFS_KOBJ_LINK) {
> - struct sysfs_symlink * sl = sd->s_element;
> - kfree(sl->link_name);
> - kobject_put(sl->target_kobj);
> - kfree(sl);
> - }
> - if (sysfs_type_dir(sd->s_type))
> - kmem_cache_free(sysfs_dir_cachep, to_sysfs_dir(sd));
> - else
> - kmem_cache_free(sysfs_dirent_cachep, sd);
> -}
> +extern struct kobject *sysfs_get_kobject(struct dentry *dentry);
> +extern void release_sysfs_dirent(struct sysfs_dirent * sd);
>
> static inline void sysfs_put(struct sysfs_dirent * sd)
> {
>
--
Maneesh Soni
Linux Technology Center,
IBM India Software Labs,
Bangalore, India
email: maneesh@in.ibm.com
Phone: 91-80-25044990
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2004-12-30 13:39 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2004-12-09 5:26 [fake patch] un-inline sysfs_get_kobject and release_sysfs_dirent Adam J. Richter
2004-12-30 13:35 ` Maneesh Soni
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox