Re: [-mm3 patch]Warning fix: check the return value of kobject_add etc.

["Cong WANG" <xiyou.wangcong@gmail.com>]

2007/4/1, Andrew Morton <akpm@linux-foundation.org>:
> On Sat, 31 Mar 2007 10:30:31 +0800 "Cong WANG" <xiyou.wangcong@gmail.com> wrote:
>
> > Since kobject_add, sysfs_create_link and sysfs_create_file are marked
> > as '__must_check', so we must always check their return values, or gcc
> > will give us warnings.
> >
> > Signed-off-by: Cong WANG <xiyou.wangcong@gmail.com>
> >
> > ---
> > --- fs/partitions/check.c.orig        2007-03-30 21:35:45.000000000 +0800
> > +++ fs/partitions/check.c     2007-03-30 21:49:53.000000000 +0800
> > @@ -385,10 +385,16 @@ void add_partition(struct gendisk *disk,
> >       p->kobj.parent = &disk->kobj;
> >       p->kobj.ktype = &ktype_part;
> >       kobject_init(&p->kobj);
> > -     kobject_add(&p->kobj);
> > +     if (kobject_add(&p->kobj)) {
> > +             kfree(p);
> > +             return;
> > +     }
> >       if (!disk->part_uevent_suppress)
> >               kobject_uevent(&p->kobj, KOBJ_ADD);
> > -     sysfs_create_link(&p->kobj, &block_subsys.kset.kobj, "subsystem");
> > +     if (sysfs_create_link(&p->kobj, &block_subsys.kset.kobj, "subsystem")) {
> > +             kfree(p);
> > +             return;
> > +     }
> >       if (flags & ADDPART_FLAG_WHOLEDISK) {
> >               static struct attribute addpartattr = {
> >                       .name = "whole_disk",
> > @@ -396,7 +402,10 @@ void add_partition(struct gendisk *disk,
> >                       .owner = THIS_MODULE,
> >               };
> >
> > -             sysfs_create_file(&p->kobj, &addpartattr);
> > +             if (sysfs_create_file(&p->kobj, &addpartattr)) {
> > +                     kfree(p);
> > +                     return;
> > +             }
> >       }
> >       partition_sysfs_add_subdir(p);
> >       disk->part[part-1] = p;
>
> Well yes, but the point here it to fix kernel bugs, not to just make the
> warnings go away.
>
> The bugs here are that the effects of the kobject_add() and the
> sysfs_create_link() are not undone on the error path.

Thanks for your point.

>
> Also, please always prepare patches in `patch -p1' form, as per
> http://www.zip.com.au/~akpm/linux/patches/stuff/tpp.txt, thanks.
>
>

Sorry. I am confused with this. Does that mean I should make patches
_upon_ the root kernel source directory or first make a copy of the
original source code and then diff against the two dirs? But I was
told that "patches should be based _in_ the root kernel source
directory" and when only one file was modified just to diff it with
the original single file. (See Documentation/SubmittingPatches.)

Can you help out? And should I remake this patch? Thanks again!