Re: [PATCH] Ext4: Fix delalloc enospace handling counter update race

[Mingming Cao <cmm@us.ibm.com>]

在 2008-07-15二的 08:42 +0530，Aneesh Kumar K.V写道：
> On Mon, Jul 14, 2008 at 10:54:36AM -0700, Mingming Cao wrote:
> > 
> > 在 2008-07-14一的 22:23 +0530，Aneesh Kumar K.V写道：
> > > On Mon, Jul 14, 2008 at 09:28:17AM -0700, Mingming Cao wrote:
> > > > 
> > > > Ext4: Fix delalloc enospace handling counter update race
> > > > 
> > > > From: Mingming Cao <cmm@us.ibm.com>
> > 
> > > > Also in the case of truncate, we should not clear the per-allocation allocated
> > > > metablocks counter as that may be in-use by parallel allocation. The patch
> > > > only clear the per-allocation allocated metablocks when allocation is successfully returned.
> > > > 
> > > > 
> > > > Signed-off-by: Mingming Cao <cmm@us.ibm.com>
> > > 
> > > Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
> > 
> > 
> > > > @@ -1519,7 +1529,8 @@ void ext4_da_release_space(struct inode 
> > > > 
> > > >  	BUG_ON(mdb > EXT4_I(inode)->i_reserved_meta_blocks);
> > > >  	EXT4_I(inode)->i_reserved_meta_blocks = mdb;
> > > > -	EXT4_I(inode)->i_allocated_meta_blocks = 0;
> > > > +	if (used)
> > > > +		EXT4_I(inode)->i_allocated_meta_blocks = 0;
> > > >  	spin_unlock(&EXT4_I(inode)->i_block_reservation_lock);
> > > >  }
> > > > 
> > 
> > After think again, this part of race fix may not needed, when the first
> > part of patch is applied. Since truncate is also hold the i_data_sem, so
> > it could be assured that when truncate is releasing the page's  reserved
> > blocks, there is no parallel block allocation.
> 
> But truncate calls invalidate_page without holding i_data_sem.
> 
> vmtruncate -> truncate_inode_pages -> invalidate_page.
> 

Just for the benefit of the readers on the list, we have discussed this
on IRC, the race between truncate and parallel allocation is probably
not cause incorrect free blocks accounting,  but the code is confusing
in the current way. So it's worth a seperate cleanup patch later.

Mingming

--
To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html