Re: general protection fault in override_creds

From: Jens Axboe <axboe@kernel.dk>
To: Dmitry Vyukov <dvyukov@google.com>,
	syzbot <syzbot+5320383e16029ba057ff@syzkaller.appspotmail.com>,
	Al Viro <viro@zeniv.linux.org.uk>,
	io-uring@vger.kernel.org,
	linux-fsdevel <linux-fsdevel@vger.kernel.org>
Cc: Anna.Schumaker@netapp.com,
	Casey Schaufler <casey@schaufler-ca.com>,
	David Howells <dhowells@redhat.com>, Jann Horn <jannh@google.com>,
	Kees Cook <keescook@chromium.org>,
	Kate Stewart <kstewart@linuxfoundation.org>,
	LKML <linux-kernel@vger.kernel.org>, NeilBrown <neilb@suse.com>,
	syzkaller-bugs <syzkaller-bugs@googlegroups.com>,
	Thomas Gleixner <tglx@linutronix.de>
Subject: Re: general protection fault in override_creds
Date: Mon, 2 Dec 2019 08:49:04 -0700	[thread overview]
Message-ID: <7be77044-5390-7df1-33d4-e92166fb927a@kernel.dk> (raw)
In-Reply-To: <39ea5e00-3278-a84f-25ae-c7a93a4b8ab5@kernel.dk>

On 12/2/19 7:40 AM, Jens Axboe wrote:
> On 12/1/19 10:40 PM, Dmitry Vyukov wrote:
>> On Mon, Dec 2, 2019 at 7:35 AM syzbot
>> <syzbot+5320383e16029ba057ff@syzkaller.appspotmail.com> wrote:
>>>
>>> Hello,
>>>
>>> syzbot found the following crash on:
>>>
>>> HEAD commit:    b94ae8ad Merge tag 'seccomp-v5.5-rc1' of git://git.kernel...
>>> git tree:       upstream
>>> console output: https://syzkaller.appspot.com/x/log.txt?x=10f9ffcee00000
>>> kernel config:  https://syzkaller.appspot.com/x/.config?x=ff560c3de405258c
>>> dashboard link: https://syzkaller.appspot.com/bug?extid=5320383e16029ba057ff
>>> compiler:       gcc (GCC) 9.0.0 20181231 (experimental)
>>> syz repro:      https://syzkaller.appspot.com/x/repro.syz?x=12dd682ae00000
>>> C reproducer:   https://syzkaller.appspot.com/x/repro.c?x=16290abce00000
>>>
>>> IMPORTANT: if you fix the bug, please add the following tag to the commit:
>>> Reported-by: syzbot+5320383e16029ba057ff@syzkaller.appspotmail.com
>>
>> I think this relates to fs/io_uring.c rather than kernel/cred.c.
>> +io_uring maintainers
> 
> Yeah that's my fault, guessing the below will fix it. I'll test and
> queue it up.

This is cleaner, tested fine.

diff --git a/fs/io-wq.c b/fs/io-wq.c
index 91b85df0861e..74b40506c5d9 100644
--- a/fs/io-wq.c
+++ b/fs/io-wq.c
@@ -111,7 +111,7 @@ struct io_wq {
  
  	struct task_struct *manager;
  	struct user_struct *user;
-	struct cred *creds;
+	const struct cred *creds;
  	struct mm_struct *mm;
  	refcount_t refs;
  	struct completion done;
diff --git a/fs/io-wq.h b/fs/io-wq.h
index 600e0158cba7..dd0af0d7376c 100644
--- a/fs/io-wq.h
+++ b/fs/io-wq.h
@@ -87,7 +87,7 @@ typedef void (put_work_fn)(struct io_wq_work *);
  struct io_wq_data {
  	struct mm_struct *mm;
  	struct user_struct *user;
-	struct cred *creds;
+	const struct cred *creds;
  
  	get_work_fn *get_work;
  	put_work_fn *put_work;
diff --git a/fs/io_uring.c b/fs/io_uring.c
index ec53aa7cdc94..5cab7a047317 100644
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -238,7 +238,7 @@ struct io_ring_ctx {
  
  	struct user_struct	*user;
  
-	struct cred		*creds;
+	const struct cred	*creds;
  
  	/* 0 is for ctx quiesce/reinit/free, 1 is for sqo_thread started */
  	struct completion	*completions;
@@ -4759,7 +4759,7 @@ static int io_uring_create(unsigned entries, struct io_uring_params *p)
  	ctx->compat = in_compat_syscall();
  	ctx->account_mem = account_mem;
  	ctx->user = user;
-	ctx->creds = prepare_creds();
+	ctx->creds = get_current_cred();
  
  	ret = io_allocate_scq_urings(ctx, p);
  	if (ret)

-- 
Jens Axboe

