From: Omar Sandoval <osandov@osandov.com>
To: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>,
Dave Chinner <david@fromorbit.com>,
linux-fsdevel <linux-fsdevel@vger.kernel.org>,
linux-btrfs <linux-btrfs@vger.kernel.org>,
Linux API <linux-api@vger.kernel.org>,
Kernel Team <kernel-team@fb.com>,
Dave Chinner <dchinner@redhat.com>
Subject: Re: [PATCH RESEND x3 v9 1/9] iov_iter: add copy_struct_from_iter()
Date: Wed, 23 Jun 2021 12:33:40 -0700 [thread overview]
Message-ID: <YNOMlLZWdaNlEJtZ@relinquished.localdomain> (raw)
In-Reply-To: <CAHk-=wjptRD=dzst-=O0D_X2q0kU2ijdTEjrg0=vvtqdjJ_x8g@mail.gmail.com>
On Wed, Jun 23, 2021 at 11:28:15AM -0700, Linus Torvalds wrote:
> On Wed, Jun 23, 2021 at 10:49 AM Omar Sandoval <osandov@osandov.com> wrote:
> >
> > Al, Linus, what do you think? Is there a path forward for this series as
> > is?
>
> So the "read from user space in order to write" is a no-go for me. It
> completely violates what a "read()" system call should do. It also
> entirely violates what an iovec can and should do.
>
> And honestly, if Al hates the "first iov entry" model, I'm not sure I
> want to merge that version - I personally find it fine, but Al is
> effectively the iov-iter maintainer.
>
> I do worry a bit about the "first iov entry" simply because it might
> work for "writev2()" when given virtual user space addresses - but I
> think it's conceptually broken for things like direct-IO which might
> do things by physical address, and what is a contiguous user space
> virtual address is not necessarily a contiguous physical address.
>
> Yes, the filesystem can - and does - hide that path by basically not
> doing direct-IO on the first entry at all, and just treat is very
> specially in the front end of the IO access, but that only reinforces
> the whole "this is not at all like read/write".
>
> Similar issues might crop up in other situations, ie splice etc, where
> it's not at all obvious that the iov_iter boundaries would be
> maintained as it moves through the system.
>
> So while I personally find the "first iov entry" model fairly
> reasonable, I think Dave is being disingenuous when he says that it
> looks like a normal read/write. It very much does not. The above is
> quite fundamental.
>
> > I'd be happy to have this functionality merged in any form, but I do
> > think that this approach with preadv2/pwritev2 using iov_len is decent
> > relative to the alternatives.
>
> As mentioned, I find it acceptable. I'm completely unimpressed with
> Dave's argument, but ioctl's aren't perfect either, so weak or not,
> that argument being bogus doesn't necessarily mean that the iovec
> entry model is wrong.
>
> That said, thinking about exactly the fact that I don't think a
> translation from iovec to anything else can be truly valid, I find the
> iter_is_iovec() case to be the only obviously valid one.
>
> Which gets me back to: how can any of the non-iovec alternatives ever
> be valid? You did mention having missed ITER_XARRAY, but my question
> is more fundamental than that. How could a non-iter_is_iovec ever be
> valid? There are no possible interfaces that can generate such a thing
> sanely.
I only implemented the bvec and kvec cases for completeness, since
copy_struct_from_iter() would appear to be a generic helper. At least
for RWF_ENCODED, a bvec seems pretty bogus, but it doesn't seem too
far-flung to imagine an in-kernel user of RWF_ENCODED that uses a kvec.
One other option that we haven't considered is ditching the
copy_struct_from_user() semantics and going the simpler route of adding
some reserved space to the end of struct encoded_iov:
struct encoded_iov {
__aligned_u64 len;
__aligned_u64 unencoded_len;
__aligned_u64 unencoded_offset;
__u32 compression;
__u32 encryption;
__u8 reserved[32];
};
Then we can do an unconditional copy_from_user_full(sizeof(struct
encoded_iov)) and check the reserved space in the typical fashion.
(And in the unlikely case that we use up all of that space with
extensions, I suppose we could have an RWF_ENCODED2 with a matching
struct encoded_iov2.)
next prev parent reply other threads:[~2021-06-23 19:33 UTC|newest]
Thread overview: 50+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-06-17 23:51 [PATCH RESEND x3 v9 0/9] fs: interface for directly reading/writing compressed data Omar Sandoval
2021-06-17 23:51 ` [PATCH RESEND x3 v9 1/9] iov_iter: add copy_struct_from_iter() Omar Sandoval
2021-06-18 18:50 ` Linus Torvalds
2021-06-18 19:42 ` Al Viro
2021-06-18 19:49 ` Al Viro
2021-06-18 20:33 ` Omar Sandoval
2021-06-18 20:32 ` Omar Sandoval
2021-06-18 20:58 ` Al Viro
2021-06-18 21:10 ` Linus Torvalds
2021-06-18 21:32 ` Al Viro
2021-06-18 21:40 ` Linus Torvalds
2021-06-18 22:10 ` Omar Sandoval
2021-06-18 22:32 ` Al Viro
2021-06-19 0:43 ` Omar Sandoval
2021-06-21 18:46 ` Omar Sandoval
2021-06-21 19:33 ` Linus Torvalds
2021-06-21 20:46 ` Omar Sandoval
2021-06-21 20:53 ` Omar Sandoval
2021-06-21 20:55 ` Omar Sandoval
2021-06-22 22:06 ` Dave Chinner
2021-06-23 17:49 ` Omar Sandoval
2021-06-23 18:28 ` Linus Torvalds
2021-06-23 19:33 ` Omar Sandoval [this message]
2021-06-23 19:45 ` Al Viro
2021-06-23 20:46 ` Omar Sandoval
2021-06-23 21:39 ` Al Viro
2021-06-23 21:58 ` Omar Sandoval
2021-06-23 22:26 ` Al Viro
2021-06-24 2:00 ` Matthew Wilcox
2021-06-24 6:14 ` Omar Sandoval
2021-06-24 17:52 ` Linus Torvalds
2021-06-24 18:28 ` Omar Sandoval
2021-06-24 21:07 ` Linus Torvalds
2021-06-24 22:41 ` Martin K. Petersen
2021-06-25 3:38 ` Matthew Wilcox
2021-06-25 16:16 ` Linus Torvalds
2021-06-25 21:07 ` Omar Sandoval
2021-07-07 17:59 ` Omar Sandoval
2021-07-19 15:44 ` Josef Bacik
2021-06-24 6:41 ` Christoph Hellwig
2021-06-24 7:50 ` Omar Sandoval
2021-06-18 22:14 ` Al Viro
2021-06-17 23:51 ` [PATCH RESEND x3 v9 2/9] fs: add O_ALLOW_ENCODED open flag Omar Sandoval
2021-06-17 23:51 ` [PATCH RESEND x3 v9 3/9] fs: add RWF_ENCODED for reading/writing compressed data Omar Sandoval
2021-06-17 23:51 ` [PATCH RESEND x3 v9 4/9] btrfs: don't advance offset for compressed bios in btrfs_csum_one_bio() Omar Sandoval
2021-06-17 23:51 ` [PATCH RESEND x3 v9 5/9] btrfs: add ram_bytes and offset to btrfs_ordered_extent Omar Sandoval
2021-06-17 23:51 ` [PATCH RESEND x3 v9 6/9] btrfs: support different disk extent size for delalloc Omar Sandoval
2021-06-17 23:51 ` [PATCH RESEND x3 v9 7/9] btrfs: optionally extend i_size in cow_file_range_inline() Omar Sandoval
2021-06-17 23:51 ` [PATCH RESEND x3 v9 8/9] btrfs: implement RWF_ENCODED reads Omar Sandoval
2021-06-17 23:51 ` [PATCH RESEND x3 v9 9/9] btrfs: implement RWF_ENCODED writes Omar Sandoval
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=YNOMlLZWdaNlEJtZ@relinquished.localdomain \
--to=osandov@osandov.com \
--cc=david@fromorbit.com \
--cc=dchinner@redhat.com \
--cc=kernel-team@fb.com \
--cc=linux-api@vger.kernel.org \
--cc=linux-btrfs@vger.kernel.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=torvalds@linux-foundation.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).