From: Christoph Hellwig <hch@infradead.org>
To: Hans de Goede <hdegoede@redhat.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>,
David Howells <dhowells@redhat.com>,
Christoph Hellwig <hch@infradead.org>,
linux-fsdevel@vger.kernel.org
Subject: Re: [PATCH v13] fs: Add VirtualBox guest shared folder (vboxsf) support
Date: Fri, 16 Aug 2019 00:56:54 -0700 [thread overview]
Message-ID: <20190816075654.GA15363@infradead.org> (raw)
In-Reply-To: <20190815131253.237921-2-hdegoede@redhat.com>
A couple minor comments. Otherwise we should be fine, things aren't
going to get much better for such a messed up protocol design.
> + return dir_emit(ctx, d_name, strlen(d_name),
> + fake_ino, d_type);
> + } else {
> + return dir_emit(ctx,
> + info->name.string.utf8,
> + info->name.length,
> + fake_ino, d_type);
> + }
> + }
Nitpick: no need for an else after a return.
> +static int vboxsf_file_release(struct inode *inode, struct file *file)
> +{
> + struct vboxsf_inode *sf_i = VBOXSF_I(inode);
> + struct vboxsf_handle *sf_handle = file->private_data;
> +
> + filemap_write_and_wait(inode->i_mapping);
Normal Linux semantics don't include writing back data on close, so
if you are doing this to follow other things like NFS CTO semantics
it should have a comment explaining that.
> +
> + mutex_lock(&sf_i->handle_list_mutex);
> + list_del(&sf_handle->head);
> + mutex_unlock(&sf_i->handle_list_mutex);
> +
> + kref_put(&sf_handle->refcount, vboxsf_handle_release);
> + file->private_data = NULL;
There is no need to zero ->private_data on release, the file gets
freed and never reused.
> + * Ideally we would wrap generic_file_read_iter with a function which also
> + * does this check, to reduce the chance of us missing writes happening on the
> + * host side after open(). But the vboxsf stat call to the host only works on
> + * filenames, so that would require caching the filename in our
> + * file->private_data and there is no guarantee that filename will still
> + * be valid at read_iter time. So this would be in no way bulletproof.
Well, you can usually generate a file name from file->f_path.dentry.
The only odd case is opened by unliked files. NFS has a special hack
for those called sillyrename (you can grep for that). How similar to
normal posix semantics are expected from this fs?
> +
> + mutex_lock(&sf_i->handle_list_mutex);
> + list_for_each_entry(h, &sf_i->handle_list, head) {
> + if (h->access_flags == SHFL_CF_ACCESS_WRITE ||
> + h->access_flags == SHFL_CF_ACCESS_READWRITE) {
> + kref_get(&h->refcount);
Does this need a kref_get_unless_zero to deal with races during list
removal?
next prev parent reply other threads:[~2019-08-16 7:56 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-08-15 13:12 [PATCH v13] fs: Add VirtualBox guest shared folder (vboxsf) support Hans de Goede
2019-08-15 13:12 ` Hans de Goede
2019-08-16 7:56 ` Christoph Hellwig [this message]
2019-08-16 9:01 ` Hans de Goede
2019-08-16 9:39 ` Hans de Goede
2019-08-16 12:49 ` Christoph Hellwig
2019-08-15 13:33 ` Hans de Goede
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=20190816075654.GA15363@infradead.org \
--to=hch@infradead.org \
--cc=dhowells@redhat.com \
--cc=hdegoede@redhat.com \
--cc=linux-fsdevel@vger.kernel.org \
--cc=viro@zeniv.linux.org.uk \
/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).