Messages in this thread | | | Subject | Re: lessen the impact of a deprecated warning in hugetlbfs | From | Joe Perches <> | Date | Tue, 02 Nov 2010 20:04:22 -0700 |
| |
On Tue, 2010-11-02 at 14:06 -0400, Dave Jones wrote: > WARN_ONCE is a bit strong for a deprecation warning, given that it spews a > huge backtrace. > > Signed-off-by: Dave Jones <davej@redhat.com> > > --- linux-2.6.36.noarch/fs/hugetlbfs/inode.c~ 2010-11-02 14:01:19.489289082 -0400 > +++ linux-2.6.36.noarch/fs/hugetlbfs/inode.c 2010-11-02 14:02:19.947986439 -0400 > @@ -915,8 +915,7 @@ struct file *hugetlb_file_setup(const ch > if (creat_flags == HUGETLB_SHMFS_INODE && !can_do_hugetlb_shm()) { > *user = current_user(); > if (user_shm_lock(size, *user)) { > - WARN_ONCE(1, > - "Using mlock ulimits for SHM_HUGETLB deprecated\n"); > + printk_once("Using mlock ulimits for SHM_HUGETLB deprecated\n");
Please add a KERN_NOTICE or similar. The grammar could be improved via "is deprecated"
| |