Message-ID: <07171afd91dbd05b425d92e54f9832f9.pc@manguebit.com>
From:   Paulo Alcantara <pc@manguebit.com>
To:     David Howells <dhowells@redhat.com>,
        Steve French <smfrench@gmail.com>
Cc:     David Howells <dhowells@redhat.com>,
        Shyam Prasad N <nspmangalore@gmail.com>,
        Rohith Surabattula <rohiths.msft@gmail.com>,
        Tom Talpey <tom@talpey.com>,
        Stefan Metzmacher <metze@samba.org>,
        Jeff Layton <jlayton@kernel.org>, linux-cifs@vger.kernel.org,
        linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org,
        Murphy Zhou <jencce.kernel@gmail.com>,
        Steve French <sfrench@samba.org>
Subject: Re: [PATCH 1/1] cifs: Fix memory leak in direct I/O
In-Reply-To: <20230228223838.3794807-2-dhowells@redhat.com>
References: <20230228223838.3794807-1-dhowells@redhat.com>
 <20230228223838.3794807-2-dhowells@redhat.com>
Date:   Wed, 01 Mar 2023 12:11:09 -0300
MIME-Version: 1.0
Content-Type: text/plain
Precedence: bulk

David Howells <dhowells@redhat.com> writes:

> When __cifs_readv() and __cifs_writev() extract pages from a user-backed
> iterator into a BVEC-type iterator, they set ->bv_need_unpin to note
> whether they need to unpin the pages later.  However, in both cases they
> examine the BVEC-type iterator and not the source iterator - and so
> bv_need_unpin doesn't get set and the pages are leaked.
>
> I think this may be responsible for the generic/208 xfstest failing
> occasionally with:
>
> 	WARNING: CPU: 0 PID: 3064 at mm/gup.c:218 try_grab_page+0x65/0x100
> 	RIP: 0010:try_grab_page+0x65/0x100
> 	follow_page_pte+0x1a7/0x570
> 	__get_user_pages+0x1a2/0x650
> 	__gup_longterm_locked+0xdc/0xb50
> 	internal_get_user_pages_fast+0x17f/0x310
> 	pin_user_pages_fast+0x46/0x60
> 	iov_iter_extract_pages+0xc9/0x510
> 	? __kmalloc_large_node+0xb1/0x120
> 	? __kmalloc_node+0xbe/0x130
> 	netfs_extract_user_iter+0xbf/0x200 [netfs]
> 	__cifs_writev+0x150/0x330 [cifs]
> 	vfs_write+0x2a8/0x3c0
> 	ksys_pwrite64+0x65/0xa0
>
> with the page refcount going negative.  This is less unlikely than it seems
> because the page is being pinned, not simply got, and so the refcount
> increased by 1024 each time, and so only needs to be called around ~2097152
> for the refcount to go negative.
>
> Further, the test program (aio-dio-invalidate-failure) uses a 32MiB static
> buffer and all the PTEs covering it refer to the same page because it's
> never written to.
>
> The warning in try_grab_page():
>
> 	if (WARN_ON_ONCE(folio_ref_count(folio) <= 0))
> 		return -ENOMEM;
>
> then trips and prevents us ever using the page again for DIO at least.
>
> Fixes: d08089f649a0 ("cifs: Change the I/O paths to use an iterator rather than a page list")
> Reported-by: Murphy Zhou <jencce.kernel@gmail.com>
> Link: https://lore.kernel.org/r/CAH2r5mvaTsJ---n=265a4zqRA7pP+o4MJ36WCQUS6oPrOij8cw@mail.gmail.com
> Signed-off-by: David Howells <dhowells@redhat.com>
> cc: Steve French <sfrench@samba.org>
> cc: Shyam Prasad N <nspmangalore@gmail.com>
> cc: Rohith Surabattula <rohiths.msft@gmail.com>
> cc: Paulo Alcantara <pc@cjr.nz>
> cc: Jeff Layton <jlayton@kernel.org>
> cc: linux-cifs@vger.kernel.org
> ---
>  fs/cifs/file.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)

Reviewed-by: Paulo Alcantara (SUSE) <pc@manguebit.com>