From: Andrey Ignatov <rdna@fb.com>
To: Christoph Hellwig <hch@lst.de>
Cc: Kees Cook <keescook@chromium.org>,
Iurii Zaikin <yzaikin@google.com>,
Luis Chamberlain <mcgrof@kernel.org>,
Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
"Rafael J. Wysocki" <rafael@kernel.org>,
Alexei Starovoitov <ast@kernel.org>,
Daniel Borkmann <daniel@iogearbox.net>,
<linux-kernel@vger.kernel.org>, <linux-mm@kvack.org>,
<linux-fsdevel@vger.kernel.org>, <netdev@vger.kernel.org>,
<bpf@vger.kernel.org>
Subject: Re: [Potential Spoof] Re: [PATCH 6/6] sysctl: pass kernel pointers to ->proc_handler
Date: Fri, 17 Apr 2020 15:36:20 -0700 [thread overview]
Message-ID: <20200417223620.GB15155@rdna-mbp> (raw)
In-Reply-To: <20200417193910.GA7011@rdna-mbp>
Andrey Ignatov <rdna@fb.com> [Fri, 2020-04-17 12:41 -0700]:
> Christoph Hellwig <hch@lst.de> [Thu, 2020-04-16 23:42 -0700]:
...
> > @@ -564,27 +564,36 @@ static ssize_t proc_sys_call_handler(struct file *filp, void __user *buf,
> > if (!table->proc_handler)
> > goto out;
> >
> > - error = BPF_CGROUP_RUN_PROG_SYSCTL(head, table, write, buf, &count,
> > - ppos, &new_buf);
> > + if (write) {
> > + kbuf = memdup_user_nul(ubuf, count);
> > + if (IS_ERR(kbuf)) {
> > + error = PTR_ERR(kbuf);
> > + goto out;
> > + }
> > + } else {
> > + error = -ENOMEM;
> > + kbuf = kzalloc(count, GFP_KERNEL);
> > + if (!kbuf)
> > + goto out;
> > + }
> > +
> > + error = BPF_CGROUP_RUN_PROG_SYSCTL(head, table, write, &kbuf, &count,
> > + ppos);
> > if (error)
> > - goto out;
> > + goto out_free_buf;
> >
> > /* careful: calling conventions are nasty here */
> > - if (new_buf) {
> > - mm_segment_t old_fs;
> > -
> > - old_fs = get_fs();
> > - set_fs(KERNEL_DS);
> > - error = table->proc_handler(table, write, (void __user *)new_buf,
> > - &count, ppos);
> > - set_fs(old_fs);
> > - kfree(new_buf);
> > - } else {
> > - error = table->proc_handler(table, write, buf, &count, ppos);
> > - }
> > + error = table->proc_handler(table, write, kbuf, &count, ppos);
> > + if (error)
> > + goto out_free_buf;
> > +
> > + error = -EFAULT;
> > + if (copy_to_user(ubuf, kbuf, count))
> > + goto out_free_buf;
This copy_to_user is where the last failing test I mentioned in the
previous email was failing:
> Test case: sysctl_set_new_value sysctl:write ok .. [FAIL]
What the test does is it attaches BPF program that overrides the value
that user is trying to write to sysctl net/ipv4/route/mtu_expires.
User tries to write "606", BPF program overrides it with "600" using
bpf_sysctl_set_new_value() helper.
This leads to kbuf being replaced in BPF_CGROUP_RUN_PROG_SYSCTL call
above with a new buffer allocated inside __cgroup_bpf_run_filter_sysctl.
And when this new buffer is tried to be copied to user here it fails.
In `strace -e ./test_sysctl` it can be seen as:
write(5, "606", 3) = -1 EFAULT (Bad address)
I also verified same with printk.
Changing it to:
if (!write && copy_to_user(ubuf, kbuf, count))
(basically what Matthew Wilcox suggested earlier) fixes the problem.
> >
> > - if (!error)
> > - error = count;
> > + error = count;
> > +out_free_buf:
> > + kfree(kbuf);
> > out:
> > sysctl_head_finish(head);
> >
...
> I applied the whole patchset to bpf-next tree and run selftests. This
> patch breaks 4 of them:
>
> % cd tools/testing/selftests/bpf/
> % ./test_sysctl
> ...
> Test case: sysctl_get_new_value sysctl:write ok .. [FAIL]
> Test case: sysctl_get_new_value sysctl:write ok long .. [FAIL]
> Test case: sysctl_get_new_value sysctl:write E2BIG .. [FAIL]
> Test case: sysctl_set_new_value sysctl:read EINVAL .. [PASS]
> Test case: sysctl_set_new_value sysctl:write ok .. [FAIL]
> ...
> Summary: 36 PASSED, 4 FAILED
>
> I applied both changes I suggested above and it reduces number of broken
> selftests to one:
>
> Test case: sysctl_set_new_value sysctl:write ok .. [FAIL]
>
> I haven't debugged this last one though yet ..
--
Andrey Ignatov
next prev parent reply other threads:[~2020-04-17 22:36 UTC|newest]
Thread overview: 20+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-04-17 6:41 pass kernel pointers to the sysctl ->proc_handler method Christoph Hellwig
2020-04-17 6:41 ` [PATCH 1/6] bpf-cgroup: remove unused exports Christoph Hellwig
2020-04-17 6:41 ` [PATCH 2/6] firmware_loader: " Christoph Hellwig
2020-04-17 7:43 ` Greg Kroah-Hartman
2020-04-17 7:48 ` Christoph Hellwig
2020-04-17 6:41 ` [PATCH 3/6] mm: remove watermark_boost_factor_sysctl_handler Christoph Hellwig
2020-04-17 6:41 ` [PATCH 4/6] sysctl: remove all extern declaration from sysctl.c Christoph Hellwig
2020-04-22 12:33 ` Eric W. Biederman
2020-04-22 17:26 ` Christoph Hellwig
2020-04-17 6:41 ` [PATCH 5/6] sysctl: avoid forward declarations Christoph Hellwig
2020-04-17 6:41 ` [PATCH 6/6] sysctl: pass kernel pointers to ->proc_handler Christoph Hellwig
2020-04-17 7:45 ` Greg Kroah-Hartman
2020-04-17 18:17 ` Matthew Wilcox
2020-04-21 7:42 ` Christoph Hellwig
2020-04-17 19:39 ` Andrey Ignatov
2020-04-17 19:50 ` Matthew Wilcox
2020-04-17 22:38 ` Andrey Ignatov
2020-04-21 7:55 ` Christoph Hellwig
2020-04-17 22:36 ` Andrey Ignatov [this message]
2020-04-17 18:00 ` pass kernel pointers to the sysctl ->proc_handler method Luis Chamberlain
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=20200417223620.GB15155@rdna-mbp \
--to=rdna@fb.com \
--cc=ast@kernel.org \
--cc=bpf@vger.kernel.org \
--cc=daniel@iogearbox.net \
--cc=gregkh@linuxfoundation.org \
--cc=hch@lst.de \
--cc=keescook@chromium.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mcgrof@kernel.org \
--cc=netdev@vger.kernel.org \
--cc=rafael@kernel.org \
--cc=yzaikin@google.com \
/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).