Re: [PATCH 09/23] lib: percpu_counter_init error handling

[Peter Zijlstra <a.p.zijlstra@chello.nl>]

On Fri, 2007-08-17 at 11:56 -0400, Josef Sipek wrote:
> On Thu, Aug 16, 2007 at 09:45:34AM +0200, Peter Zijlstra wrote:
> )
> > @@ -996,12 +997,16 @@ static int ext2_fill_super(struct super_
> >  	sbi->s_rsv_window_head.rsv_goal_size = 0;
> >  	ext2_rsv_window_add(sb, &sbi->s_rsv_window_head);
> >  
> > -	percpu_counter_init(&sbi->s_freeblocks_counter,
> > +	err = percpu_counter_init(&sbi->s_freeblocks_counter,
> >  				ext2_count_free_blocks(sb));
> > -	percpu_counter_init(&sbi->s_freeinodes_counter,
> > +	err |= percpu_counter_init(&sbi->s_freeinodes_counter,
> >  				ext2_count_free_inodes(sb));
> > -	percpu_counter_init(&sbi->s_dirs_counter,
> > +	err |= percpu_counter_init(&sbi->s_dirs_counter,
> >  				ext2_count_dirs(sb));
> > +	if (err) {
> > +		printk(KERN_ERR "EXT2-fs: insufficient memory\n");
> > +		goto failed_mount3;
> > +	}
> 
> Can percpu_counter_init fail with only one error code? If not, the error
> code potentially used in future at failed_mount3 could be nonsensical
> because of the bitwise or-ing.

I guess I could have written saner code :-/ will try to come up with
something that is both clear and short.

Attachment: signature.asc
Description: This is a digitally signed message part