lkml.org 
[lkml]   [2013]   [Apr]   [24]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH v3 17/18] ext4: make punch hole code path work with bigalloc
On Tue, 23 Apr 2013, Zheng Liu wrote:

> Date: Tue, 23 Apr 2013 17:19:28 +0800
> From: Zheng Liu <gnehzuil.liu@gmail.com>
> To: Jan Kara <jack@suse.cz>
> Cc: Lukas Czerner <lczerner@redhat.com>, linux-mm@kvack.org,
> linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org,
> linux-ext4@vger.kernel.org
> Subject: Re: [PATCH v3 17/18] ext4: make punch hole code path work with
> bigalloc
>
> On Sat, Apr 20, 2013 at 03:42:41PM +0200, Jan Kara wrote:
> > On Tue 09-04-13 11:14:26, Lukas Czerner wrote:
> > > Currently punch hole is disabled in file systems with bigalloc
> > > feature enabled. However the recent changes in punch hole patch should
> > > make it easier to support punching holes on bigalloc enabled file
> > > systems.
> > >
> > > This commit changes partial_cluster handling in ext4_remove_blocks(),
> > > ext4_ext_rm_leaf() and ext4_ext_remove_space(). Currently
> > > partial_cluster is unsigned long long type and it makes sure that we
> > > will free the partial cluster if all extents has been released from that
> > > cluster. However it has been specifically designed only for truncate.
> > >
> > > With punch hole we can be freeing just some extents in the cluster
> > > leaving the rest untouched. So we have to make sure that we will notice
> > > cluster which still has some extents. To do this I've changed
> > > partial_cluster to be signed long long type. The only scenario where
> > > this could be a problem is when cluster_size == block size, however in
> > > that case there would not be any partial clusters so we're safe. For
> > > bigger clusters the signed type is enough. Now we use the negative value
> > > in partial_cluster to mark such cluster used, hence we know that we must
> > > not free it even if all other extents has been freed from such cluster.
> > >
> > > This scenario can be described in simple diagram:
> > >
> > > |FFF...FF..FF.UUU|
> > > ^----------^
> > > punch hole
> > >
> > > . - free space
> > > | - cluster boundary
> > > F - freed extent
> > > U - used extent
> > >
> > > Also update respective tracepoints to use signed long long type for
> > > partial_cluster.
> > The patch looks OK. You can add:
> > Reviewed-by: Jan Kara <jack@suse.cz>
> >
> > Just a minor nit - sometimes you use 'signed long long', sometimes 'long
> > long int', sometimes just 'long long'. In kernel we tend to always use just
> > 'long long' so it would be good to clean that up.
>
> Another question is that in patch 01/18 invalidatepage signature is
> changed from
> int (*invalidatepage) (struct page *, unsigned long);
> to
> void (*invalidatepage) (struct page *, unsigned int, unsigned int);
>
> The argument type is changed from 'unsigned long' to 'unsigned int'. My
> question is why we need to change it.
>
> Thanks,
> - Zheng
>

Hi Zheng,

this was changed on Hugh Dickins request because it makes it clearer
that those args are indeed intended to be offsets within a page
(0..PAGE_CACHE_SIZE).

Even though PAGE_CACHE_SIZE can be defined as unsigned long, this is
only for convenience. Here is quote from Hugh:

"
They would be defined as unsigned long so that they can be used in
masks like ~(PAGE_SIZE - 1), and behave as expected on addresses,
without needing casts to be added all over.

We do not (currently!) expect PAGE_SIZE or PAGE_CACHE_SIZE to grow
beyond an unsigned int - but indeed they can be larger than what's
held in an unsigned short (look no further than ia64 or ppc64).

For more reassurance, see include/linux/highmem.h, which declares
zero_user_segments() and others: unsigned int (well, unsigned with
the int implicit) for offsets within a page.

Hugh
"

I should probably mention that in the description.

Thanks!
-Lukas


\
 
 \ /
  Last update: 2013-04-24 13:41    [W:0.096 / U:0.496 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site