| Message ID | 20180918154434.10299-1-amir73il@gmail.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | fanotify: store fanotify_init() flags in group's fanotify_data | expand |
On Tue 18-09-18 18:44:34, Amir Goldstein wrote: > This averts the need to re-generate flags in fanotify_show_fdinfo() > and sets the scene for addition of more upcoming flags without growing > new members to the fanotify_data struct. > > Signed-off-by: Amir Goldstein <amir73il@gmail.com> Thanks. I've taken this patch to my tree for the next merge window. Honza > --- > > Jan, > > Please consider this small "cleanup" patch for next. > It's part of my fanotify_dentry series, so wasn't going to send it out > yet, but seems like other developers can find it useful [1] in the mean > while. > > Thanks, > Amir. > > [1] https://lkml.org/lkml/2018/9/18/942 > > fs/notify/fanotify/fanotify_user.c | 9 +++++---- > fs/notify/fdinfo.c | 24 +----------------------- > include/linux/fsnotify_backend.h | 4 ++-- > 3 files changed, 8 insertions(+), 29 deletions(-) > > diff --git a/fs/notify/fanotify/fanotify_user.c b/fs/notify/fanotify/fanotify_user.c > index 1347c588f778..0562738a4c51 100644 > --- a/fs/notify/fanotify/fanotify_user.c > +++ b/fs/notify/fanotify/fanotify_user.c > @@ -191,7 +191,8 @@ static int process_access_response(struct fsnotify_group *group, > if (fd < 0) > return -EINVAL; > > - if ((response & FAN_AUDIT) && !group->fanotify_data.audit) > + if ((response & FAN_AUDIT) && > + !(group->fanotify_data.flags & FAN_ENABLE_AUDIT)) > return -EINVAL; > > event = dequeue_event(group, fd); > @@ -701,8 +702,8 @@ SYSCALL_DEFINE2(fanotify_init, unsigned int, flags, unsigned int, event_f_flags) > struct user_struct *user; > struct fanotify_event_info *oevent; > > - pr_debug("%s: flags=%d event_f_flags=%d\n", > - __func__, flags, event_f_flags); > + pr_debug("%s: flags=%x event_f_flags=%x\n", > + __func__, flags, event_f_flags); > > if (!capable(CAP_SYS_ADMIN)) > return -EPERM; > @@ -798,13 +799,13 @@ SYSCALL_DEFINE2(fanotify_init, unsigned int, flags, unsigned int, event_f_flags) > fd = -EPERM; > if (!capable(CAP_AUDIT_WRITE)) > goto out_destroy_group; > - group->fanotify_data.audit = true; > } > > fd = anon_inode_getfd("[fanotify]", &fanotify_fops, group, f_flags); > if (fd < 0) > goto out_destroy_group; > > + group->fanotify_data.flags = flags; > return fd; > > out_destroy_group: > diff --git a/fs/notify/fdinfo.c b/fs/notify/fdinfo.c > index 25385e336ac7..348a184bcdda 100644 > --- a/fs/notify/fdinfo.c > +++ b/fs/notify/fdinfo.c > @@ -142,31 +142,9 @@ static void fanotify_fdinfo(struct seq_file *m, struct fsnotify_mark *mark) > void fanotify_show_fdinfo(struct seq_file *m, struct file *f) > { > struct fsnotify_group *group = f->private_data; > - unsigned int flags = 0; > - > - switch (group->priority) { > - case FS_PRIO_0: > - flags |= FAN_CLASS_NOTIF; > - break; > - case FS_PRIO_1: > - flags |= FAN_CLASS_CONTENT; > - break; > - case FS_PRIO_2: > - flags |= FAN_CLASS_PRE_CONTENT; > - break; > - } > - > - if (group->max_events == UINT_MAX) > - flags |= FAN_UNLIMITED_QUEUE; > - > - if (group->fanotify_data.max_marks == UINT_MAX) > - flags |= FAN_UNLIMITED_MARKS; > - > - if (group->fanotify_data.audit) > - flags |= FAN_ENABLE_AUDIT; > > seq_printf(m, "fanotify flags:%x event-flags:%x\n", > - flags, group->fanotify_data.f_flags); > + group->fanotify_data.flags, group->fanotify_data.f_flags); > > show_fdinfo(m, f, fanotify_fdinfo); > } > diff --git a/include/linux/fsnotify_backend.h b/include/linux/fsnotify_backend.h > index 81b88fc9df31..8e91341cbd8a 100644 > --- a/include/linux/fsnotify_backend.h > +++ b/include/linux/fsnotify_backend.h > @@ -189,10 +189,10 @@ struct fsnotify_group { > /* allows a group to block waiting for a userspace response */ > struct list_head access_list; > wait_queue_head_t access_waitq; > - int f_flags; > + int flags; /* flags from fanotify_init() */ > + int f_flags; /* event_f_flags from fanotify_init() */ > unsigned int max_marks; > struct user_struct *user; > - bool audit; > } fanotify_data; > #endif /* CONFIG_FANOTIFY */ > }; > -- > 2.17.1 >
On Wed, Sep 19, 2018 at 4:47 PM Jan Kara <jack@suse.cz> wrote: > > On Tue 18-09-18 18:44:34, Amir Goldstein wrote: > > This averts the need to re-generate flags in fanotify_show_fdinfo() > > and sets the scene for addition of more upcoming flags without growing > > new members to the fanotify_data struct. > > > > Signed-off-by: Amir Goldstein <amir73il@gmail.com> > > Thanks. I've taken this patch to my tree for the next merge window. > Actually, could you please grab the later version of this patch, which is a bit nicer, from: https://github.com/amir73il/linux/commits/fanotify_unpriv I was going to post v2, but you beat me to it. Below is v1..v2 diff. Thanks! Amir. ---- diff --git a/fs/notify/fanotify/fanotify_user.c b/fs/notify/fanotify/fanotify_user.c index 0562738a4c51..afa89eafe79f 100644 --- a/fs/notify/fanotify/fanotify_user.c +++ b/fs/notify/fanotify/fanotify_user.c @@ -191,8 +191,7 @@ static int process_access_response(struct fsnotify_group *group, if (fd < 0) return -EINVAL; - if ((response & FAN_AUDIT) && - !(group->fanotify_data.flags & FAN_ENABLE_AUDIT)) + if ((response & FAN_AUDIT) && !FAN_GROUP_FLAG(group, FAN_ENABLE_AUDIT)) return -EINVAL; event = dequeue_event(group, fd); diff --git a/include/linux/fanotify.h b/include/linux/fanotify.h index 096c96f4f16a..9c5ea3bdfaa0 100644 --- a/include/linux/fanotify.h +++ b/include/linux/fanotify.h @@ -6,4 +6,8 @@ /* not valid from userspace, only kernel internal */ #define FAN_MARK_ONDIR 0x00000100 + +#define FAN_GROUP_FLAG(group, flag) \ + ((group)->fanotify_data.flags & (flag)) + #endif /* _LINUX_FANOTIFY_H */
diff --git a/fs/notify/fanotify/fanotify_user.c b/fs/notify/fanotify/fanotify_user.c index 1347c588f778..0562738a4c51 100644 --- a/fs/notify/fanotify/fanotify_user.c +++ b/fs/notify/fanotify/fanotify_user.c @@ -191,7 +191,8 @@ static int process_access_response(struct fsnotify_group *group, if (fd < 0) return -EINVAL; - if ((response & FAN_AUDIT) && !group->fanotify_data.audit) + if ((response & FAN_AUDIT) && + !(group->fanotify_data.flags & FAN_ENABLE_AUDIT)) return -EINVAL; event = dequeue_event(group, fd); @@ -701,8 +702,8 @@ SYSCALL_DEFINE2(fanotify_init, unsigned int, flags, unsigned int, event_f_flags) struct user_struct *user; struct fanotify_event_info *oevent; - pr_debug("%s: flags=%d event_f_flags=%d\n", - __func__, flags, event_f_flags); + pr_debug("%s: flags=%x event_f_flags=%x\n", + __func__, flags, event_f_flags); if (!capable(CAP_SYS_ADMIN)) return -EPERM; @@ -798,13 +799,13 @@ SYSCALL_DEFINE2(fanotify_init, unsigned int, flags, unsigned int, event_f_flags) fd = -EPERM; if (!capable(CAP_AUDIT_WRITE)) goto out_destroy_group; - group->fanotify_data.audit = true; } fd = anon_inode_getfd("[fanotify]", &fanotify_fops, group, f_flags); if (fd < 0) goto out_destroy_group; + group->fanotify_data.flags = flags; return fd; out_destroy_group: diff --git a/fs/notify/fdinfo.c b/fs/notify/fdinfo.c index 25385e336ac7..348a184bcdda 100644 --- a/fs/notify/fdinfo.c +++ b/fs/notify/fdinfo.c @@ -142,31 +142,9 @@ static void fanotify_fdinfo(struct seq_file *m, struct fsnotify_mark *mark) void fanotify_show_fdinfo(struct seq_file *m, struct file *f) { struct fsnotify_group *group = f->private_data; - unsigned int flags = 0; - - switch (group->priority) { - case FS_PRIO_0: - flags |= FAN_CLASS_NOTIF; - break; - case FS_PRIO_1: - flags |= FAN_CLASS_CONTENT; - break; - case FS_PRIO_2: - flags |= FAN_CLASS_PRE_CONTENT; - break; - } - - if (group->max_events == UINT_MAX) - flags |= FAN_UNLIMITED_QUEUE; - - if (group->fanotify_data.max_marks == UINT_MAX) - flags |= FAN_UNLIMITED_MARKS; - - if (group->fanotify_data.audit) - flags |= FAN_ENABLE_AUDIT; seq_printf(m, "fanotify flags:%x event-flags:%x\n", - flags, group->fanotify_data.f_flags); + group->fanotify_data.flags, group->fanotify_data.f_flags); show_fdinfo(m, f, fanotify_fdinfo); } diff --git a/include/linux/fsnotify_backend.h b/include/linux/fsnotify_backend.h index 81b88fc9df31..8e91341cbd8a 100644 --- a/include/linux/fsnotify_backend.h +++ b/include/linux/fsnotify_backend.h @@ -189,10 +189,10 @@ struct fsnotify_group { /* allows a group to block waiting for a userspace response */ struct list_head access_list; wait_queue_head_t access_waitq; - int f_flags; + int flags; /* flags from fanotify_init() */ + int f_flags; /* event_f_flags from fanotify_init() */ unsigned int max_marks; struct user_struct *user; - bool audit; } fanotify_data; #endif /* CONFIG_FANOTIFY */ };
This averts the need to re-generate flags in fanotify_show_fdinfo() and sets the scene for addition of more upcoming flags without growing new members to the fanotify_data struct. Signed-off-by: Amir Goldstein <amir73il@gmail.com> --- Jan, Please consider this small "cleanup" patch for next. It's part of my fanotify_dentry series, so wasn't going to send it out yet, but seems like other developers can find it useful [1] in the mean while. Thanks, Amir. [1] https://lkml.org/lkml/2018/9/18/942 fs/notify/fanotify/fanotify_user.c | 9 +++++---- fs/notify/fdinfo.c | 24 +----------------------- include/linux/fsnotify_backend.h | 4 ++-- 3 files changed, 8 insertions(+), 29 deletions(-)