public inbox for [email protected]
 help / color / mirror / Atom feed
From: Al Viro <[email protected]>
To: Jens Axboe <[email protected]>
Cc: [email protected], [email protected],
	[email protected], [email protected]
Subject: Re: [PATCH 5/9] replace do_setxattr() with saner helpers.
Date: Sun, 6 Oct 2024 06:28:59 +0100	[thread overview]
Message-ID: <20241006052859.GD4017910@ZenIV> (raw)
In-Reply-To: <[email protected]>

On Wed, Oct 02, 2024 at 04:55:22PM -0600, Jens Axboe wrote:

> The reason I liked the putname() is that it's unconditional - the caller
> can rely on it being put, regardless of the return value. So I'd say the
> same should be true for ctx.kvalue, and if not, the caller should still
> free it. That's the path of least surprise - no leak for the least
> tested error path, and no UAF in the success case.

The problem with ctx.kvalue is that on the syscall side there's a case when
we do not call either file_setxattr() or filename_setxattr() - -EBADF.
And it's a lot more convenient to do setxattr_copy() first, so we end
up with a lovely landmine:
        filename = getname_xattr(pathname, at_flags);
	if (!filename) {
		CLASS(fd, f)(dfd);
		if (fd_empty(f)) {
			kfree(ctx.kvalue); // lest we leak
			return -EBADF;
		}
		return file_setxattr(fd_file(f), &ctx);
	}
	return filename_setxattr(dfd, filename, lookup_flags, &ctx);

That's asking for trouble, obviously.  So I think we ought to consume
filename (in filename_...()) case, leave struct file reference alone
(we have to - it might have been borrowed rather than cloned) and leave
->kvalue unchanged.  Yes, it ends up being more clumsy, but at least
it's consistent between the cases...

As for consuming filename...  On the syscall side it allows things like
SYSCALL_DEFINE3(mkdirat, int, dfd, const char __user *, pathname, umode_t, mode)
{
        return do_mkdirat(dfd, getname(pathname), mode);
}  
which is better than the alternatives - I mean, that's
SYSCALL_DEFINE3(mkdirat, int, dfd, const char __user *, pathname, umode_t, mode)
{
	struct filename *filename = getname(pathname);
	int res = do_mkdirat(dfd, filename, mode);
	putname(filename);
	return ret;
}  
or 
SYSCALL_DEFINE3(mkdirat, int, dfd, const char __user *, pathname, umode_t, mode)
{
	struct filename *filename __free(putname) = getname(pathname);
	return do_mkdirat(dfd, filename, mode);
}
and both stink, if for different reasons ;-/  Having those things consume
(unconditionally) is better, IMO.

Hell knows; let's go with what I described above for now and see where it leads
when more such helpers are regularized.

> That's a bit different than your putname() case, but I think as long as
> it's consistent regardless of return value, then either approach is
> fine. Maybe just add a comment about that? At least for the consistent
> case, if it blows up, it'll blow up instantly rather than be a surprise
> down the line for "case x,y,z doesn't put it" or "case x,y,z always puts
> in, normal one does not".

Obviously.

> > Questions on the io_uring side:
> > 	* you usually reject REQ_F_FIXED_FILE for ...at() at ->prep() time.
> > Fine, but... what's the point of doing that in IORING_OP_FGETXATTR case?
> > Or IORING_OP_GETXATTR, for that matter, since you pass AT_FDCWD anyway...
> > Am I missing something subtle here?
> 
> Right, it could be allowed for fgetxattr on the io_uring side. Anything
> that passes in a struct file would be fair game to enable it on.
> Anything that passes in a path (eg a non-fd value), it obviously
> wouldn't make sense anyway.

OK, done and force-pushed into #work.xattr.

  reply	other threads:[~2024-10-06  5:29 UTC|newest]

Thread overview: 50+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-10-02  1:10 [RFC][PATCHES] xattr stuff and interactions with io_uring Al Viro
2024-10-02  1:22 ` [PATCH 1/9] xattr: switch to CLASS(fd) Al Viro
2024-10-02  1:22   ` [PATCH 2/9] fs: rename struct xattr_ctx to kernel_xattr_ctx Al Viro
2024-10-02  5:56     ` Christian Brauner
2024-10-02  1:22   ` [PATCH 3/9] io_[gs]etxattr_prep(): just use getname() Al Viro
2024-10-02  5:57     ` Christian Brauner
2024-10-02  1:22   ` [PATCH 4/9] new helper: import_xattr_name() Al Viro
2024-10-02  5:57     ` Christian Brauner
2024-10-02  1:22   ` [PATCH 5/9] replace do_setxattr() with saner helpers Al Viro
2024-10-02  1:34     ` Jens Axboe
2024-10-02  2:08       ` Al Viro
2024-10-02 18:00         ` Jens Axboe
2024-10-02 21:19           ` Al Viro
2024-10-02 22:55             ` Jens Axboe
2024-10-06  5:28               ` Al Viro [this message]
2024-10-07 18:09                 ` Jens Axboe
2024-10-07 18:20                   ` Jens Axboe
2024-10-07 21:20                     ` Al Viro
2024-10-07 22:29                       ` Jens Axboe
2024-10-07 23:58                         ` Al Viro
2024-10-08  1:58                           ` Jens Axboe
2024-10-08  4:08                             ` Al Viro
2024-10-02  1:22   ` [PATCH 6/9] replace do_getxattr() " Al Viro
2024-10-02  5:59     ` Christian Brauner
2024-10-02  1:22   ` [PATCH 7/9] new helpers: file_listxattr(), filename_listxattr() Al Viro
2024-10-02  6:00     ` Christian Brauner
2024-10-02  1:22   ` [PATCH 8/9] new helpers: file_removexattr(), filename_removexattr() Al Viro
2024-10-02  6:01     ` Christian Brauner
2024-10-02  1:22   ` [PATCH 9/9] fs/xattr: add *at family syscalls Al Viro
2024-10-02  6:03     ` Christian Brauner
2024-10-02  5:56   ` [PATCH 1/9] xattr: switch to CLASS(fd) Christian Brauner
2024-11-02  7:28 ` [RFC][PATCHES v2] xattr stuff and interactions with io_uring Al Viro
2024-11-02  7:31   ` [PATCH v2 01/13] teach filename_lookup() to treat NULL filename as "" Al Viro
2024-11-02  7:31     ` [PATCH v2 02/13] getname_maybe_null() - the third variant of pathname copy-in Al Viro
2024-11-02  7:31     ` [PATCH v2 03/13] io_uring: IORING_OP_F[GS]ETXATTR is fine with REQ_F_FIXED_FILE Al Viro
2024-11-02  7:31     ` [PATCH v2 04/13] io_[gs]etxattr_prep(): just use getname() Al Viro
2024-11-02 14:44       ` Jens Axboe
2024-11-02  7:31     ` [PATCH v2 05/13] xattr: switch to CLASS(fd) Al Viro
2024-11-02  7:31     ` [PATCH v2 06/13] fs: rename struct xattr_ctx to kernel_xattr_ctx Al Viro
2024-11-02  7:31     ` [PATCH v2 07/13] new helper: import_xattr_name() Al Viro
2024-11-02  7:31     ` [PATCH v2 08/13] replace do_setxattr() with saner helpers Al Viro
2024-11-02  7:31     ` [PATCH v2 09/13] replace do_getxattr() " Al Viro
2024-11-02  7:31     ` [PATCH v2 10/13] new helpers: file_listxattr(), filename_listxattr() Al Viro
2024-11-02  7:31     ` [PATCH v2 11/13] new helpers: file_removexattr(), filename_removexattr() Al Viro
2024-11-02  7:31     ` [PATCH v2 12/13] fs/xattr: add *at family syscalls Al Viro
2024-11-02  7:31     ` [PATCH v2 13/13] xattr: remove redundant check on variable err Al Viro
2024-11-02 14:43   ` [RFC][PATCHES v2] xattr stuff and interactions with io_uring Jens Axboe
2024-11-03  6:51     ` Al Viro
2024-11-03 13:57       ` Arnd Bergmann
2024-11-03 18:32         ` Al Viro

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=20241006052859.GD4017910@ZenIV \
    [email protected] \
    [email protected] \
    [email protected] \
    [email protected] \
    [email protected] \
    [email protected] \
    /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