From: Kees Cook <keescook@chromium.org>
To: David Laight <David.Laight@ACULAB.COM>
Cc: "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
Sargun Dhillon <sargun@sargun.me>,
Christian Brauner <christian@brauner.io>,
"David S. Miller" <davem@davemloft.net>,
Christoph Hellwig <hch@lst.de>, Tycho Andersen <tycho@tycho.ws>,
Jakub Kicinski <kuba@kernel.org>,
Alexander Viro <viro@zeniv.linux.org.uk>,
Aleksa Sarai <cyphar@cyphar.com>,
Matt Denton <mpdenton@google.com>, Jann Horn <jannh@google.com>,
Chris Palmer <palmer@google.com>,
Robert Sesek <rsesek@google.com>,
Giuseppe Scrivano <gscrivan@redhat.com>,
Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
Andy Lutomirski <luto@amacapital.net>,
Will Drewry <wad@chromium.org>, Shuah Khan <shuah@kernel.org>,
"netdev@vger.kernel.org" <netdev@vger.kernel.org>,
"containers@lists.linux-foundation.org"
<containers@lists.linux-foundation.org>,
"linux-api@vger.kernel.org" <linux-api@vger.kernel.org>,
"linux-fsdevel@vger.kernel.org" <linux-fsdevel@vger.kernel.org>,
"linux-kselftest@vger.kernel.org"
<linux-kselftest@vger.kernel.org>
Subject: Re: [PATCH v4 03/11] fs: Add fd_install_received() wrapper for __fd_install_received()
Date: Wed, 17 Jun 2020 12:58:00 -0700 [thread overview]
Message-ID: <202006171141.4DA1174979@keescook> (raw)
In-Reply-To: <6de12195ec3244b99e6026b4b46e5be2@AcuMS.aculab.com>
On Wed, Jun 17, 2020 at 03:35:20PM +0000, David Laight wrote:
> From: Kees Cook
> > Sent: 16 June 2020 04:25
> >
> > For both pidfd and seccomp, the __user pointer is not used. Update
> > __fd_install_received() to make writing to ufd optional. (ufd
> > itself cannot checked for NULL because this changes the SCM_RIGHTS
> > interface behavior.) In these cases, the new fd needs to be returned
> > on success. Update the existing callers to handle it. Add new wrapper
> > fd_install_received() for pidfd and seccomp that does not use the ufd
> > argument.
> ...>
> > static inline int fd_install_received_user(struct file *file, int __user *ufd,
> > unsigned int o_flags)
> > {
> > - return __fd_install_received(file, ufd, o_flags);
> > + return __fd_install_received(file, true, ufd, o_flags);
> > +}
>
> Can you get rid of the 'return user' parameter by adding
> if (!ufd) return -EFAULT;
> to the above wrapper, then checking for NULL in the function?
>
> Or does that do the wrong horrid things in the fail path?
Oh, hm. No, that shouldn't break the failure path, since everything gets
unwound in __fd_install_received if the ufd write fails.
Effectively this (I'll chop it up into the correct patches):
diff --git a/fs/file.c b/fs/file.c
index b583e7c60571..3b80324a31cc 100644
--- a/fs/file.c
+++ b/fs/file.c
@@ -939,18 +939,16 @@ int replace_fd(unsigned fd, struct file *file, unsigned flags)
*
* @fd: fd to install into (if negative, a new fd will be allocated)
* @file: struct file that was received from another process
- * @ufd_required: true to use @ufd for writing fd number to userspace
* @ufd: __user pointer to write new fd number to
* @o_flags: the O_* flags to apply to the new fd entry
*
* Installs a received file into the file descriptor table, with appropriate
* checks and count updates. Optionally writes the fd number to userspace, if
- * @ufd_required is true (@ufd cannot just be tested for NULL because NULL may
- * actually get passed into SCM_RIGHTS).
+ * @ufd is non-NULL.
*
* Returns newly install fd or -ve on error.
*/
-int __fd_install_received(int fd, struct file *file, bool ufd_required,
+int __fd_install_received(int fd, struct file *file,
int __user *ufd, unsigned int o_flags)
{
struct socket *sock;
@@ -967,7 +965,7 @@ int __fd_install_received(int fd, struct file *file, bool ufd_required,
return new_fd;
}
- if (ufd_required) {
+ if (ufd) {
error = put_user(new_fd, ufd);
if (error) {
put_unused_fd(new_fd);
diff --git a/include/linux/file.h b/include/linux/file.h
index f1d16e24a12e..2ade0d90bc5e 100644
--- a/include/linux/file.h
+++ b/include/linux/file.h
@@ -91,20 +91,22 @@ extern void put_unused_fd(unsigned int fd);
extern void fd_install(unsigned int fd, struct file *file);
-extern int __fd_install_received(int fd, struct file *file, bool ufd_required,
+extern int __fd_install_received(int fd, struct file *file,
int __user *ufd, unsigned int o_flags);
static inline int fd_install_received_user(struct file *file, int __user *ufd,
unsigned int o_flags)
{
- return __fd_install_received(-1, file, true, ufd, o_flags);
+ if (ufd == NULL)
+ return -EFAULT;
+ return __fd_install_received(-1, file, ufd, o_flags);
}
static inline int fd_install_received(struct file *file, unsigned int o_flags)
{
- return __fd_install_received(-1, file, false, NULL, o_flags);
+ return __fd_install_received(-1, file, NULL, o_flags);
}
static inline int fd_replace_received(int fd, struct file *file, unsigned int o_flags)
{
- return __fd_install_received(fd, file, false, NULL, o_flags);
+ return __fd_install_received(fd, file, NULL, o_flags);
}
extern void flush_delayed_fput(void);
--
Kees Cook
next prev parent reply other threads:[~2020-06-17 19:58 UTC|newest]
Thread overview: 28+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-06-16 3:25 [PATCH v4 00/11] Add seccomp notifier ioctl that enables adding fds Kees Cook
2020-06-16 3:25 ` [PATCH v4 01/11] net/scm: Regularize compat handling of scm_detach_fds() Kees Cook
2020-06-16 3:25 ` [PATCH v4 02/11] fs: Move __scm_install_fd() to __fd_install_received() Kees Cook
2020-06-16 5:29 ` Sargun Dhillon
2020-06-16 5:48 ` Kees Cook
2020-06-17 15:25 ` David Laight
2020-06-17 18:40 ` Kees Cook
2020-06-18 8:56 ` Christian Brauner
2020-06-18 20:05 ` Kees Cook
2020-06-16 3:25 ` [PATCH v4 03/11] fs: Add fd_install_received() wrapper for __fd_install_received() Kees Cook
2020-06-17 15:35 ` David Laight
2020-06-17 19:58 ` Kees Cook [this message]
2020-06-18 8:19 ` David Laight
2020-06-16 3:25 ` [PATCH v4 04/11] pidfd: Replace open-coded partial fd_install_received() Kees Cook
2020-06-16 3:25 ` [PATCH v4 05/11] fs: Expand __fd_install_received() to accept fd Kees Cook
2020-06-16 3:25 ` [PATCH v4 06/11] seccomp: Introduce addfd ioctl to seccomp user notifier Kees Cook
2020-06-16 3:25 ` [PATCH v4 07/11] selftests/seccomp: Test SECCOMP_IOCTL_NOTIF_ADDFD Kees Cook
2020-06-16 3:25 ` [PATCH v4 08/11] selftests/seccomp: Make kcmp() less required Kees Cook
2020-06-16 14:57 ` Tycho Andersen
2020-06-16 16:03 ` Kees Cook
2020-06-16 3:25 ` [PATCH v4 09/11] selftests/seccomp: Rename user_trap_syscall() to user_notif_syscall() Kees Cook
2020-06-16 14:56 ` Tycho Andersen
2020-06-16 3:25 ` [PATCH v4 10/11] seccomp: Switch addfd to Extensible Argument ioctl Kees Cook
2020-06-16 14:55 ` Tycho Andersen
2020-06-16 16:05 ` Kees Cook
2020-06-16 16:18 ` Tycho Andersen
2020-06-16 3:25 ` [PATCH v4 11/11] seccomp: Fix ioctl number for SECCOMP_IOCTL_NOTIF_ID_VALID Kees Cook
2020-06-18 22:16 ` [PATCH v4 00/11] Add seccomp notifier ioctl that enables adding fds Sargun Dhillon
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=202006171141.4DA1174979@keescook \
--to=keescook@chromium.org \
--cc=David.Laight@ACULAB.COM \
--cc=christian@brauner.io \
--cc=containers@lists.linux-foundation.org \
--cc=cyphar@cyphar.com \
--cc=davem@davemloft.net \
--cc=gregkh@linuxfoundation.org \
--cc=gscrivan@redhat.com \
--cc=hch@lst.de \
--cc=jannh@google.com \
--cc=kuba@kernel.org \
--cc=linux-api@vger.kernel.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-kselftest@vger.kernel.org \
--cc=luto@amacapital.net \
--cc=mpdenton@google.com \
--cc=netdev@vger.kernel.org \
--cc=palmer@google.com \
--cc=rsesek@google.com \
--cc=sargun@sargun.me \
--cc=shuah@kernel.org \
--cc=tycho@tycho.ws \
--cc=viro@zeniv.linux.org.uk \
--cc=wad@chromium.org \
/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).