ext2/ext3:useless code for return value
in [Kernel]
|
Prev: ext4 is faster with LVM than without, and other filesystem benchmarks
Next: c/r (ipc): allow allocation of a desired ipc identifier
From: Jan Kara on 7 May 2010 12:30 > When rsv is the right hand side of goal, we should return NULL, > because now rsv's prev is NULL, or we return rsv. > > Signed-off-by: Peter Pan(潘卫平) <wppan(a)redflag-linux.com> > > --- > fs/ext2/balloc.c | 6 +----- > fs/ext3/balloc.c | 6 +----- > 2 files changed, 2 insertions(+), 10 deletions(-) > > diff --git a/fs/ext2/balloc.c b/fs/ext2/balloc.c > index 3cf038c..023990f 100644 > --- a/fs/ext2/balloc.c > +++ b/fs/ext2/balloc.c > @@ -323,11 +323,7 @@ search_reserve_window(struct rb_root *root, > ext2_fsblk_t goal) > * side of the interval containing the goal. If it's the RHS, > * we need to back up one. > */ > - if (rsv->rsv_start > goal) { > - n = rb_prev(&rsv->rsv_node); > - rsv = rb_entry(n, struct ext2_reserve_window_node, rsv_node); > - } > - return rsv; > + return (rsv->rsv_start < goal) ? rsv : NULL; Hmm, I'm not sure I understand your reasoning. Suppose we have an RB-tree with two intervals 0-10, 20-30. Interval 0-10 is in the root. Now we search for goal 15. In the root we go to right because 10<15, in the next node we go to left because 15 < 20. Then the loop terminates. Now your code would return NULL but previous code would return rb_prev of interval 20-30 which is 0-10. And that is what we want as far as I understand what we expect from the function... Honza -- Jan Kara <jack(a)suse.cz> SuSE CR Labs -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo(a)vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
From: 潘卫平(Peter Pan) on 9 May 2010 22:40
On 05/08/2010 12:23 AM, Jan Kara wrote: >> When rsv is the right hand side of goal, we should return NULL, >> because now rsv's prev is NULL, or we return rsv. >> >> Signed-off-by: Peter Pan(潘卫平)<wppan(a)redflag-linux.com> >> >> --- >> fs/ext2/balloc.c | 6 +----- >> fs/ext3/balloc.c | 6 +----- >> 2 files changed, 2 insertions(+), 10 deletions(-) >> >> diff --git a/fs/ext2/balloc.c b/fs/ext2/balloc.c >> index 3cf038c..023990f 100644 >> --- a/fs/ext2/balloc.c >> +++ b/fs/ext2/balloc.c >> @@ -323,11 +323,7 @@ search_reserve_window(struct rb_root *root, >> ext2_fsblk_t goal) >> * side of the interval containing the goal. If it's the RHS, >> * we need to back up one. >> */ >> - if (rsv->rsv_start> goal) { >> - n = rb_prev(&rsv->rsv_node); >> - rsv = rb_entry(n, struct ext2_reserve_window_node, rsv_node); >> - } >> - return rsv; >> + return (rsv->rsv_start< goal) ? rsv : NULL; > Hmm, I'm not sure I understand your reasoning. Suppose we have an RB-tree > with two intervals 0-10, 20-30. Interval 0-10 is in the root. Now we search > for goal 15. In the root we go to right because 10<15, in the next node we go > to left because 15< 20. Then the loop terminates. Now your code would return > NULL but previous code would return rb_prev of interval 20-30 which is 0-10. > And that is what we want as far as I understand what we expect from the > function... > > Honza You got the point! Many thanks. Regards -- Peter Pan Red Flag Software Co.,Ltd -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo(a)vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/ |