Re: [PATCH] laptop-mode for 2.6, version 2

[Nikita Danilov <Nikita@Namesys.COM>]

Hugang writes:
 > On Thu, 25 Dec 2003 12:17:43 +0300
 > Nikita Danilov <Nikita@Namesys.COM> wrote:
 > 
 > > >    unsigned long blocks;
 > > >    unsigned long mount_options = REISERFS_SB(s)->s_mount_opt;
 > > >    unsigned long safe_mask = 0;
 > > > +  unsigned int commit_max_age = -1;
 > > 
 > > Assigning -1 to the unsigned int looks strange. Let's use 0, it is
 > > invalid anyway.
 > Yes, must change to -1, fixed.
 > 
 > > I think that it would be better to first
 > > 
 > > SB_JOURNAL_MAX_COMMIT_AGE(p_s_sb) = val
 > > 
 > > in the parse_options() (after checking for validity), and in
 > > journal_init() do something like
 > > 
 > > if (SB_JOURNAL_MAX_COMMIT_AGE(p_s_sb) == 0) {
 > >   SB_JOURNAL_MAX_COMMIT_AGE(p_s_sb) = le32_to_cpu (jh->jh_journal.jp_journal_max_commit_age);
 > > }
 > > 
 > > This will also get rid of
 > > 
 > > +  if(commit_max_age != -1) {
 > > +	  SB_JOURNAL_MAX_COMMIT_AGE(s) = commit_max_age;
 > > +  }
 > > +
 > > 
 > > piece in reiserfs_remount.
 > > 
 > > Otherwise patch looks ok. Have you tested it?
 > > 
 > In the parse_options() can not assigning commit max age to super block, 
 > the journal memory not malloc, so I pass a it to journal_init.
 > 
 > Yes, It works in my laptop for 1 days. Every thinks is fine.

+		if ( val > 0 ) {
+			*commit_max_age = val;
+		}

here warning should be issued, and mount refused, if val == 0.

 > 
 > Thanks.

Nikita.