Jarkko,
On Fri, Oct 16, 2015 at 09:40:22PM +0300, Jarkko Sakkinen wrote:
> Added a new function __compat_only_sysfs_link_group_to_kobj() that adds
> a symlink from attribute or group to a kobject. This needed for
> maintaining backwards compatibility with PPI attributes in the TPM
> driver.
>
> Signed-off-by: Jarkko Sakkinen <[email protected]>
> ---
> fs/sysfs/group.c | 44 ++++++++++++++++++++++++++++++++++++++++++++
> include/linux/sysfs.h | 11 +++++++++++
> 2 files changed, 55 insertions(+)
>
> diff --git a/fs/sysfs/group.c b/fs/sysfs/group.c
> index 39a0199..e123659 100644
> --- a/fs/sysfs/group.c
> +++ b/fs/sysfs/group.c
> @@ -352,3 +352,47 @@ void sysfs_remove_link_from_group(struct kobject *kobj, const char *group_name,
> }
> }
> EXPORT_SYMBOL_GPL(sysfs_remove_link_from_group);
> +
> +/**
> + * __compat_only_sysfs_link_entry_to_kobj - add a symlink to a kobject pointing
> + * to a group or an attribute
> + * @kobj: The kobject containing the group.
> + * @target_kobj: The target kobject.
> + * @target_name: The name of the target group or attribute.
> + */
> +int __compat_only_sysfs_link_entry_to_kobj(struct kobject *kobj,
> + struct kobject *target_kobj,
> + const char *target_name)
> +{
> + struct kernfs_node *target;
> + struct kernfs_node *entry;
> + struct kernfs_node *link;
> +
> + /*
> + * We don't own @target_kobj and it may be removed at any time.
> + * Synchronize using sysfs_symlink_target_lock. See sysfs_remove_dir()
> + * for details.
> + */
> + spin_lock(&sysfs_symlink_target_lock);
> + target = target_kobj->sd;
> + if (target)
> + kernfs_get(target);
> + spin_unlock(&sysfs_symlink_target_lock);
> + if (!target)
> + return -ENOENT;
> +
> + entry = kernfs_find_and_get(target_kobj->sd, target_name);
> + if (!entry) {
> + kernfs_put(target);
> + return -ENOENT;
> + }
> +
> + link = kernfs_create_link(kobj->sd, target_name, entry);
> + if (IS_ERR(link) && PTR_ERR(link) == -EEXIST)
> + sysfs_warn_dup(kobj->sd, target_name);
> +
> + kernfs_put(entry);
> + kernfs_put(target);
> + return IS_ERR(link) ? PTR_ERR(link) : 0;
> +}
> +EXPORT_SYMBOL_GPL(__compat_only_sysfs_link_entry_to_kobj);
> diff --git a/include/linux/sysfs.h b/include/linux/sysfs.h
> index 9f65758..ea090ea 100644
> --- a/include/linux/sysfs.h
> +++ b/include/linux/sysfs.h
> @@ -268,6 +268,9 @@ int sysfs_add_link_to_group(struct kobject *kobj, const char *group_name,
> struct kobject *target, const char *link_name);
> void sysfs_remove_link_from_group(struct kobject *kobj, const char *group_name,
> const char *link_name);
> +int __compat_only_sysfs_link_entry_to_kobj(struct kobject *kobj,
> + struct kobject *target_kobj,
> + const char *target_name);
>
> void sysfs_notify(struct kobject *kobj, const char *dir, const char *attr);
>
> @@ -451,6 +454,14 @@ static inline void sysfs_remove_link_from_group(struct kobject *kobj,
> {
> }
>
> +static inline int __compat_only_sysfs_link_entry_to_kobj(
> + struct kobject *kobj,
> + struct kobject *target_kobj,
> + const char *target_name)
> +{
> + return 0;
> +}
> +
> static inline void sysfs_notify(struct kobject *kobj, const char *dir,
> const char *attr)
> {
> --
> 2.5.0
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
One problem with this patch is that it adds a bunch of code but none of
it is used. So if there is a bug, git bisect will point to a later
patch instead of this one.
--
- Jeremiah Mahler
On Fri, Nov 06, 2015 at 04:23:56PM -0800, Jeremiah Mahler wrote:
> On Fri, Oct 16, 2015 at 09:40:22PM +0300, Jarkko Sakkinen wrote:
> > Added a new function __compat_only_sysfs_link_group_to_kobj() that adds
> > a symlink from attribute or group to a kobject. This needed for
> > maintaining backwards compatibility with PPI attributes in the TPM
> > driver.
> One problem with this patch is that it adds a bunch of code but none of
> it is used. So if there is a bug, git bisect will point to a later
> patch instead of this one.
It is fairly standard practice to split cross-subsystem patches like
this, it aids review. git bisect is not a perfect tool.
Jason