From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 8E283C77B61 for ; Thu, 27 Apr 2023 19:28:42 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1344054AbjD0T2l (ORCPT ); Thu, 27 Apr 2023 15:28:41 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:53262 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S245269AbjD0T2Z (ORCPT ); Thu, 27 Apr 2023 15:28:25 -0400 Received: from dfw.source.kernel.org (dfw.source.kernel.org [IPv6:2604:1380:4641:c500::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id BE0FB6197; Thu, 27 Apr 2023 12:27:54 -0700 (PDT) Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by dfw.source.kernel.org (Postfix) with ESMTPS id 58C7C6194C; Thu, 27 Apr 2023 19:26:33 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id 0C41EC4339B; Thu, 27 Apr 2023 19:26:31 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1682623592; bh=zNqPMBxxUkpB1RVwJpyCKxHKynhK30FtWh637sHuqXI=; h=Subject:From:To:Cc:Date:In-Reply-To:References:From; b=qe2kWBnOFLparTgg62MkmCmjqc3DAJi6fIXyEkxWsXKbH16/aafmuvw8hD52MemFW txIA5kzQIveYVqWdnXfQx9AZ2GJpZ2ajdb3CRY7V5bs5th70Mdi8MtaztnALJnn4Xu Abf4xWFAIgu1RPWqBYkcZZ4qXV2eOynHsuazQ2/Yu6RYM1rCHZuTki2E3AxVmWMnBG Fd7rBYvKSiiLpX/9SBwO7TD8UnEXgPXQYygii/KV+jchhp5nnfxyKvBGK+cN3x/uQc zXRSK6QiqJiP7cmL4QlOpp4Bnz1KgWsHZLp000o9/QBjy+i4QEsRcH2UTLayIKa7W8 z5Ok4tMnkY+zg== Message-ID: <62c2c23184719c4cf0b590b02d799e3d13c43f36.camel@kernel.org> Subject: Re: [RFC][PATCH 0/4] Prepare for supporting more filesystems with fanotify From: Jeff Layton To: Amir Goldstein Cc: Jan Kara , Christian Brauner , Miklos Szeredi , linux-unionfs@vger.kernel.org, linux-fsdevel@vger.kernel.org, linux-api@vger.kernel.org Date: Thu, 27 Apr 2023 15:26:30 -0400 In-Reply-To: References: <20230425130105.2606684-1-amir73il@gmail.com> Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable User-Agent: Evolution 3.48.1 (3.48.1-1.fc38) MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: linux-api@vger.kernel.org On Thu, 2023-04-27 at 22:11 +0300, Amir Goldstein wrote: > handle_bytes >=20 > On Thu, Apr 27, 2023 at 7:36=E2=80=AFPM Jeff Layton = wrote: > >=20 > > On Thu, 2023-04-27 at 18:52 +0300, Amir Goldstein wrote: > > > On Thu, Apr 27, 2023 at 6:13=E2=80=AFPM Jeff Layton wrote: > > > >=20 > > > > On Tue, 2023-04-25 at 16:01 +0300, Amir Goldstein wrote: > > > > > Jan, > > > > >=20 > > > > > Following up on the FAN_REPORT_ANY_FID proposal [1], here is a sh= ot at an > > > > > alternative proposal to seamlessly support more filesystems. > > > > >=20 > > > > > While fanotify relaxes the requirements for filesystems to suppor= t > > > > > reporting fid to require only the ->encode_fh() operation, there = are > > > > > currently no new filesystems that meet the relaxed requirements. > > > > >=20 > > > > > I will shortly post patches that allow overlayfs to meet the new > > > > > requirements with default overlay configurations. > > > > >=20 > > > > > The overlay and vfs/fanotify patch sets are completely independen= t. > > > > > The are both available on my github branch [2] and there is a sim= ple > > > > > LTP test variant that tests reporting fid from overlayfs [3], whi= ch > > > > > also demonstrates the minor UAPI change of name_to_handle_at(2) f= or > > > > > requesting a non-decodeable file handle by userspace. > > > > >=20 > > > > > Thanks, > > > > > Amir. > > > > >=20 > > > > > [1] https://lore.kernel.org/linux-fsdevel/20230417162721.ouzs33oh= 6mb7vtft@quack3/ > > > > > [2] https://github.com/amir73il/linux/commits/exportfs_encode_fid > > > > > [3] https://github.com/amir73il/ltp/commits/exportfs_encode_fid > > > > >=20 > > > > > Amir Goldstein (4): > > > > > exportfs: change connectable argument to bit flags > > > > > exportfs: add explicit flag to request non-decodeable file hand= les > > > > > exportfs: allow exporting non-decodeable file handles to usersp= ace > > > > > fanotify: support reporting non-decodeable file handles > > > > >=20 > > > > > Documentation/filesystems/nfs/exporting.rst | 4 +-- > > > > > fs/exportfs/expfs.c | 29 +++++++++++++++= +++--- > > > > > fs/fhandle.c | 20 ++++++++------ > > > > > fs/nfsd/nfsfh.c | 5 ++-- > > > > > fs/notify/fanotify/fanotify.c | 4 +-- > > > > > fs/notify/fanotify/fanotify_user.c | 6 ++--- > > > > > fs/notify/fdinfo.c | 2 +- > > > > > include/linux/exportfs.h | 18 ++++++++++--- > > > > > include/uapi/linux/fcntl.h | 5 ++++ > > > > > 9 files changed, 67 insertions(+), 26 deletions(-) > > > > >=20 > > > >=20 > > > > This set looks fairly benign to me, so ACK on the general concept. > > >=20 > > > Thanks! > > >=20 > > > >=20 > > > > I am starting to dislike how the AT_* flags are turning into a bunc= h of > > > > flags that only have meanings on certain syscalls. I don't see a cl= eaner > > > > way to handle it though. > > >=20 > > > Yeh, it's not great. > > >=20 > > > There is also a way to extend the existing API with: > > >=20 > > > Perhstruct file_handle { > > > unsigned int handle_bytes:8; > > > unsigned int handle_flags:24; > > > int handle_type; > > > unsigned char f_handle[]; > > > }; > > >=20 > > > AFAICT, this is guaranteed to be backward compat > > > with old kernels and old applications. > > >=20 > >=20 > > That could work. It would probably look cleaner as a union though. > > Something like this maybe? > >=20 > > union { > > unsigned int legacy_handle_bytes; > > struct { > > u8 handle_bytes; > > u8 __reserved; > > u16 handle_flags; > > }; > > } >=20 > I have no problem with the union, but does this struct > guarantee that the lowest byte of legacy_handle_bytes > is in handle_bytes for all architectures? >=20 That is a very good point.=20 > That's the reason I went with >=20 > struct { > unsigned int handle_bytes:8; > unsigned int handle_flags:24; > } >=20 > Is there a problem with this approach? >=20 I just have a natural aversion to bitfields. What you're proposing would work fine, I think. You won't be able to take a pointer into the bitfield of course, but that's not necessarily a showstopper for an "interface struct" like file_handle. > > > unsigned int handle_bytes:8; > > > unsigned int handle_flags:24; > >=20 > > __reserved must be zeroed (for now). You could consider using it for > > some other purpose later. > >=20 > > It's a little ugly as an API but it would be backward compatible, given > > that we never use the high bits today anyway. > >=20 > > Callers might need to deal with an -EINVAL when they try to pass non- > > zero handle_flags to existing kernels, since you'd trip the > > MAX_HANDLE_SZ check that's there today. > >=20 >=20 > Exactly. >=20 > > > It also may not be a bad idea that the handle_flags could > > > be used to request specific fh properties (FID) and can also > > > describe the properties of the returned fh (i.e. non-decodeable) > > > that could also be respected by open_by_handle_at(). > > >=20 > > > For backward compact, kernel will only set handle_flags in > > > response if new flags were set in the request. > > >=20 > > > Do you consider this extension better than AT_HANDLE_FID > > > or worse? At least it is an API change that is contained within the > > > exportfs subsystem, without polluting the AT_ flags global namespace. > > >=20 > >=20 > > Personally, yes. I think adding a struct file_handle_v2 would be cleane= r > > and allows for expanding the API later through new flags. >=20 > I agree. > I will give it a try. Cool. --=20 Jeff Layton