linux-kernel - Re: [PATCH v5 2/2] tmpfs: Support 64-bit inums per-sb

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <alpine.LSU.2.11.2002251422370.8029@eggly.anvils>
Date:   Tue, 25 Feb 2020 15:14:05 -0800 (PST)
From:   Hugh Dickins <hughd@...gle.com>
To:     Chris Down <chris@...isdown.name>
cc:     Hugh Dickins <hughd@...gle.com>,
        Dave Chinner <david@...morbit.com>, Chris Mason <clm@...com>,
        Amir Goldstein <amir73il@...il.com>,
        Linux MM <linux-mm@...ck.org>,
        Andrew Morton <akpm@...ux-foundation.org>,
        Al Viro <viro@...iv.linux.org.uk>,
        Matthew Wilcox <willy@...radead.org>,
        Jeff Layton <jlayton@...nel.org>,
        Johannes Weiner <hannes@...xchg.org>,
        Tejun Heo <tj@...nel.org>,
        Mikael Magnusson <mikachu@...il.com>,
        linux-fsdevel <linux-fsdevel@...r.kernel.org>,
        linux-kernel <linux-kernel@...r.kernel.org>,
        Kernel Team <Kernel-team@...com>
Subject: Re: [PATCH v5 2/2] tmpfs: Support 64-bit inums per-sb

On Mon, 20 Jan 2020, Chris Down wrote:
> Hi Hugh,
> 
> Sorry this response took so long, I had some non-work issues that took a lot
> of time last week.

No, clearly it's I who must apologize to you for very slow response.

> 
> Hugh Dickins writes:
> > 
> > So the "inode64" option will be accepted but redundant on mounting,
> > but exists for use as a remount option after mounting or remounting
> > with "inode32": allowing the admin to switch temporarily to mask off
> > the high ino bits with "inode32" when needing to run a limited binary.
> > 
> > Documentation and commit message to alert Andrew and Linus and distros
> > that we are risking some breakage with this, but supplying the antidote
> > (not breakage of any distros themselves, no doubt they're all good;
> > but breakage of what some users might run on them).
> 
> Sounds good.
> 
> > > 
> > > Other than that, the first patch could be similar to how it is now,
> > > incorporating Hugh's improvements to the first patch to put everything
> > > under
> > > the same stat_lock in shmem_reserve_inode.
> > 
> > So, I persuaded Amir to the other aspects my version, but did not
> > persuade you?  Well, I can live with that (or if not, can send mods
> > on top of yours): but please read again why I was uncomfortable with
> > yours, to check that you still prefer it (I agree that your patch is
> > simpler, and none of my discomfort decisive).
> 
> Hmm, which bit were you thinking of? The lack of batching, shmem_encode_fh(),
> or the fact that nr_inodes can now be 0 on non-internal mounts?

I was uncomfortable with tmpfs depleting get_next_ino()'s pool in some
configurations, and wanted the (get_next_ino-like) per-cpu but per-sb
batching for nr_inodes=0, to minimize its stat_lock contention.

I did not have a patch to shmem_encode_fh(), had gone through thinking
that 64-bit inos made an additional fix there necessary; but Amir then
educated us that it is safe as is, though could be cleaned up later.

nr_inodes can be 0 on non-internal mounts, for many years.

> 
> For batching, I'm neutral. I'm happy to use the approach from your patch and
> integrate it (and credit you, of course).

Credit not important, but you may well want to blame me for that
complication :)

> 
> For shmem_encode_fh, I'm not totally sure I understand the concern, if that's
> what you mean.

My concern had been that shmem_encode_fh() builds up an fh from i_ino
and more, looks well prepared for a 64-bit ino, but appeared to be
announcing a 32-bit ino in its return value: Amir reassures us that
that return value does not matter.

> 
> For nr_inodes, I agree that intentional or unintentional, we should at least
> handle this case for now and can adjust later if the behaviour changes.

nr_inodes=0 is an intentional configuration (but 0 denoting infinity:
not very clean, and I've sometimes regretted that choice).

If there's any behavior change, that's a separate matter from these
64-bit ino patches; maybe I mentioned it in passing and confused us -
that we seem to have recently allowed a remounting limited<->unlimited
that was not permitted before, and might or might not need a fix patch.

Sorry again for delaying you, Chris: not at all what I'd wanted to do.
Hugh