All of lore.kernel.org
 help / color / mirror / Atom feed
From: Amir Goldstein <amir73il@gmail.com>
To: Vivek Goyal <vgoyal@redhat.com>
Cc: overlayfs <linux-unionfs@vger.kernel.org>,
	Miklos Szeredi <miklos@szeredi.hu>,
	"Eric W. Biederman" <ebiederm@xmission.com>
Subject: Re: [PATCH 5/9] ovl: Set xattr OVL_XATTR_METACOPY on upper file
Date: Wed, 11 Oct 2017 23:44:26 +0300	[thread overview]
Message-ID: <CAOQ4uxi-3y2ENXhv_=nQcLk8rKN20bz7cRuRQXuDuP+jhZBpsg@mail.gmail.com> (raw)
In-Reply-To: <20171011201630.GE8086@redhat.com>

On Wed, Oct 11, 2017 at 11:16 PM, Vivek Goyal <vgoyal@redhat.com> wrote:
> On Tue, Oct 10, 2017 at 08:03:05PM +0300, Amir Goldstein wrote:
>
> [..]
>> > @@ -511,6 +564,10 @@ static int ovl_copy_up_locked(struct ovl_copy_up_ctx *c)
>> >                 goto out_cleanup;
>> >
>> >         ovl_inode_update(d_inode(c->dentry), newdentry);
>> > +       if (c->metadata_only) {
>> > +               smp_wmb();
>>
>> This wmb does not address the only problem imo.
>> The problem is you must not allow inode to appear to have
>> An upper dentry before you made sure metacopy flag is visible.
>> So first you need to set metacopy flag before updating upper.
>> Then you also need to test them is correct order with rmb.
>
> Ok, so conceptually it makse sense to me that OVL_METACOPY flag should
> be set before upper dentry is made visible.
>
> So ovl_set_flag(OVL_METACOPY) should move before ovl_inode_update().
> ovl_inode_update() already has a smp_wmb(). So write side is probably
> fine.
>
> But read side, ovl_upperdentry_dereference() only has data dependency
> barrier and not smp_rmb(). (lockless_dereference(oi->__upperdentry)). I
> am not sure if that's sufficient for ovl_inode->flags read or not. May
> be we need to replace data dependency barrier with smp_rmb() in in
> ovl_upperdentry_dereference()?

I think we do need smp_rmb() between ovl_upperdentry_dereference()
and testing OVL_METACOPY, but not inside ovl_upperdentry_dereference()
somewhere in d_real() path maybe in ovl_open_need_copy_up()

>
> Even if we do this, I think we need barriers around setting/resetting
> of OVL_METACOPY for the reset case. I will explain this in other patch.
>

I think the spinlock in atomic bit ops of set_bit/clear_bit is got you covered
w.r.t. memory barriers. Only need to worry about lockless test_bit() IMO.

Amir.

  reply	other threads:[~2017-10-11 20:44 UTC|newest]

Thread overview: 31+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-10-10 15:32 [RFC PATCH 0/9][V3] overlayfs: Delayed copy up of data Vivek Goyal
2017-10-10 15:32 ` [PATCH 1/9] ovl: ovl_check_setxattr() get rid of redundant -EOPNOTSUPP check Vivek Goyal
2017-10-10 15:32 ` [PATCH 2/9] ovl: During copy up, first copy up metadata and then data Vivek Goyal
2017-10-10 15:32 ` [PATCH 3/9] ovl: Provide a mount option metacopy=on/off for metadata copyup Vivek Goyal
2017-10-11  1:36   ` Amir Goldstein
2017-10-11 13:57     ` Vivek Goyal
2017-10-11 16:29       ` Amir Goldstein
2017-10-11 16:53         ` Vivek Goyal
2017-10-11 17:36           ` Amir Goldstein
2017-10-11 18:34             ` Vivek Goyal
2017-10-11 20:29               ` Amir Goldstein
2017-10-12 13:23                 ` Vivek Goyal
2017-10-12 13:39                   ` Amir Goldstein
2017-10-10 15:32 ` [PATCH 4/9] ovl: Copy up only metadata during copy up where it makes sense Vivek Goyal
2017-10-10 15:32 ` [PATCH 5/9] ovl: Set xattr OVL_XATTR_METACOPY on upper file Vivek Goyal
2017-10-10 17:03   ` Amir Goldstein
2017-10-11 20:16     ` Vivek Goyal
2017-10-11 20:44       ` Amir Goldstein [this message]
2017-10-10 15:32 ` [PATCH 6/9] ovl: Fix ovl_getattr() to get number of blocks from lower Vivek Goyal
2017-10-10 15:32 ` [PATCH 7/9] ovl: Introduce read/write barriers around metacopy flag update Vivek Goyal
2017-10-10 17:12   ` Amir Goldstein
2017-10-11 20:27     ` Vivek Goyal
2017-10-11 21:08       ` Amir Goldstein
2017-10-13 18:27         ` Vivek Goyal
2017-10-14  6:05           ` Amir Goldstein
2017-10-14  7:00             ` Amir Goldstein
2017-10-16 13:24               ` Vivek Goyal
2017-10-16 13:24             ` Vivek Goyal
2017-10-16 13:31               ` Amir Goldstein
2017-10-10 15:32 ` [PATCH 8/9] ovl: Set OVL_METACOPY flag during ovl_lookup() Vivek Goyal
2017-10-10 15:32 ` [PATCH 9/9] ovl: Return lower dentry if only metadata copy up took place Vivek Goyal

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to='CAOQ4uxi-3y2ENXhv_=nQcLk8rKN20bz7cRuRQXuDuP+jhZBpsg@mail.gmail.com' \
    --to=amir73il@gmail.com \
    --cc=ebiederm@xmission.com \
    --cc=linux-unionfs@vger.kernel.org \
    --cc=miklos@szeredi.hu \
    --cc=vgoyal@redhat.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body. 
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.