static kset (and kobject)...

static kset (and kobject)...

[Arthur Jones]

Hi Zack,  FYI:  I just learned that static ksets and kobjects
are bad (see Documentation/kobject.txt for details).  The
following commit:

commit 52fd3d6fea441835fe3a35b7280e5e128bdeca9b
Author: Zach Brown <zach.brown@oracle.com>
Date:   Thu Dec 15 14:31:23 2005 -0800

    [PATCH] OCFS2: The Second Oracle Cluster Filesystem
    
    Very simple printk wrapper which adds the ability to enable various
    sets of debug messages at run-time.
    
    Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
    Signed-off-by: Kurt Hackel <kurt.hackel@oracle.com>

Introduces:

+static struct kset mlog_kset = {
+       .kobj   = {.name = "logmask", .ktype = &mlog_ktype},
+};
+
+int mlog_sys_init(struct subsystem *o2cb_subsys)
+{
+       int i = 0;
+
+       while (mlog_attrs[i].attr.mode) {
+               mlog_attr_ptrs[i] = &mlog_attrs[i].attr;
+               i++;
+       }
+       mlog_attr_ptrs[i] = NULL;
+
+       mlog_kset.subsys = o2cb_subsys;
+       return kset_register(&mlog_kset);
+}
+
+void mlog_sys_shutdown(void)
+{
+       kset_unregister(&mlog_kset);
+}

which I've learned the hard way probably won't
work as expected.  I'm surprised that you don't
see some sort of issue w/ this on module reload...

Arthur

Re: static kset (and kobject)...

[Zach Brown]

Arthur Jones wrote:
> Hi Zack,  FYI:  I just learned that static ksets and kobjects
> are bad (see Documentation/kobject.txt for details).

Hi!  Mark is the OCFS2 maintainer and I'll let him respond to your
observation.  (Note his @suse address that I've used in this reply.)

- z