From: Chandan Rajendra <chandan@linux.vnet.ibm.com>
To: Amir Goldstein <amir73il@gmail.com>
Cc: overlayfs <linux-unionfs@vger.kernel.org>,
Miklos Szeredi <miklos@szeredi.hu>
Subject: Re: [PATCH V2] ovl: Allocate anonymous devs for lowerdirs
Date: Fri, 30 Jun 2017 16:28:29 +0530 [thread overview]
Message-ID: <2041782.HUJ34Am32x@localhost.localdomain> (raw)
In-Reply-To: <2369642.ypcN4zZqWi@localhost.localdomain>
On Tuesday, June 27, 2017 1:01:17 PM IST Chandan Rajendra wrote:
> On Friday, June 23, 2017 7:04:57 PM IST Amir Goldstein wrote:
> > On Fri, Jun 23, 2017 at 2:01 PM, Chandan Rajendra
> > <chandan@linux.vnet.ibm.com> wrote:
> > > For stat(2) on lowerdir non-dir entries in non-samefs case, this commit
> > > provides unique values for st_dev. The unique values are obtained by
> > > allocating anonymous bdevs for each of the lowerdirs in the overlayfs
> > > instance.
> > >
> > > Signed-off-by: Chandan Rajendra <chandan@linux.vnet.ibm.com>
> > > ---
> >
> > Reviewed-by: Amir Goldstein <amir73il@gmail.com>
> > Tested-by: Amir Goldstein <amir73il@gmail.com>
> >
> > Miklos,
> >
> > I re-created the branch ovl-constino [1] on top of ovl-hardlinks and without
> > the consistent dino patches.
> >
> > Applied Chandan's patch and resolved conflicts with my patches.
> > Then, applied my patch to relax constant st_ino for non-samefs on stat(2).
> >
> > Mutilated unionmount-testsuite layers check [2] to get over the unexpected
> > pseudo dev and now tests pass for non samefs including constant ino
> > verification and persistent ino verification for non-dir.
> >
> > Chandan,
> >
> > If you can fix the mutilated unionmount-testsuite check_layer(), that would be
> > nice.
> >
>
Hi Amir,
In dentry.created(), we have
self.__upper = on_upper or not inode or inode.filetype() == "d"
Do you happen to know why we mark the dentry as being present on the upperdir
filesystem when "inode" evaluates to false?
Consider the case of /lowerdir/a/foo101 file being created in
set_up.py. ctx.record_file() ends up creating a new 'dentry' object with no
inode associated with it. Hence self.__upper is set to the value True. Isn't
this incorrect since the file foo101 is actually being created on lowerdir?
I added the following new test scenario to check_layer() code,
elif dentry.on_upper() and not self.config().is_samefs() and dev != self.upper_fs():
raise TestError(name + ": Upperdir file has incorrect dev id")
In the above code snippet, dentry.on_upper() incorrectly evaluates to True and
hence we end up raising an exception.
Maybe dentry.created should have the following statement instead,
self.__upper = on_upper or (inode and inode.filetype() == "d")
--
chandan
next prev parent reply other threads:[~2017-06-30 10:58 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-06-23 11:01 [PATCH V2] ovl: Allocate anonymous devs for lowerdirs Chandan Rajendra
2017-06-23 13:34 ` Amir Goldstein
2017-06-23 13:42 ` Chandan Rajendra
2017-06-27 7:31 ` Chandan Rajendra
2017-06-27 9:04 ` Amir Goldstein
2017-06-30 10:58 ` Chandan Rajendra [this message]
2017-06-30 14:50 ` Amir Goldstein
2017-07-14 9:38 ` Miklos Szeredi
2017-07-15 14:27 ` Chandan Rajendra
2017-07-24 9:17 ` [PATCH V3] " Chandan Rajendra
2017-07-27 6:24 ` Chandan Rajendra
2017-09-21 17:46 ` Amir Goldstein
2017-09-22 2:40 ` Chandan Rajendra
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=2041782.HUJ34Am32x@localhost.localdomain \
--to=chandan@linux.vnet.ibm.com \
--cc=amir73il@gmail.com \
--cc=linux-unionfs@vger.kernel.org \
--cc=miklos@szeredi.hu \
/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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).