From mboxrd@z Thu Jan 1 00:00:00 1970 From: Song Liu Subject: Re: [PATCH V2 1/2] block: Fix a NULL pointer dereference in generic_make_request() Date: Mon, 20 May 2019 22:59:16 -0700 Message-ID: References: <20190520220911.25192-1-gpiccoli@canonical.com> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Return-path: In-Reply-To: <20190520220911.25192-1-gpiccoli@canonical.com> Sender: stable-owner@vger.kernel.org To: "Guilherme G. Piccoli" Cc: linux-block@vger.kernel.org, linux-raid , dm-devel@redhat.com, axboe@kernel.dk, Gavin Guo , Jay Vosburgh , "Guilherme G. Piccoli" , stable@vger.kernel.org, Bart Van Assche , Ming Lei , Eric Ren List-Id: linux-raid.ids On Mon, May 20, 2019 at 3:10 PM Guilherme G. Piccoli wrote: > > Commit 37f9579f4c31 ("blk-mq: Avoid that submitting a bio concurrently > with device removal triggers a crash") introduced a NULL pointer > dereference in generic_make_request(). The patch sets q to NULL and > enter_succeeded to false; right after, there's an 'if (enter_succeeded)' > which is not taken, and then the 'else' will dereference q in > blk_queue_dying(q). > > This patch just moves the 'q = NULL' to a point in which it won't trigger > the oops, although the semantics of this NULLification remains untouched. > > A simple test case/reproducer is as follows: > a) Build kernel v5.2-rc1 with CONFIG_BLK_CGROUP=n. > > b) Create a raid0 md array with 2 NVMe devices as members, and mount it > with an ext4 filesystem. > > c) Run the following oneliner (supposing the raid0 is mounted in /mnt): > (dd of=/mnt/tmp if=/dev/zero bs=1M count=999 &); sleep 0.3; > echo 1 > /sys/block/nvme0n1/device/device/remove > (whereas nvme0n1 is the 2nd array member) > > This will trigger the following oops: > > BUG: unable to handle kernel NULL pointer dereference at 0000000000000078 > PGD 0 P4D 0 > Oops: 0000 [#1] SMP PTI > RIP: 0010:generic_make_request+0x32b/0x400 > Call Trace: > submit_bio+0x73/0x140 > ext4_io_submit+0x4d/0x60 > ext4_writepages+0x626/0xe90 > do_writepages+0x4b/0xe0 > [...] > > This patch has no functional changes and preserves the md/raid0 behavior > when a member is removed before kernel v4.17. > > Cc: stable@vger.kernel.org # v4.17 > Reviewed-by: Bart Van Assche > Reviewed-by: Ming Lei > Tested-by: Eric Ren > Fixes: 37f9579f4c31 ("blk-mq: Avoid that submitting a bio concurrently with device removal triggers a crash") > Signed-off-by: Guilherme G. Piccoli Applied both patches! Thanks for the fix! > --- > > Changes V1->V2: > * Implemented Ming's suggestion (drop {} from if) - thanks Ming! > * Rebased to v5.2-rc1 > * Added Reviewed-by/Tested-by tags > > Also, Ming mentioned a new patch series[0] that will refactor legacy IO > path so probably the bug won't happen anymore. Even in this case, > I consider this patch important specially aiming the stable releases, > in which backporting small bugfixes is much simpler than more complex > patch sets. > > [0] https://lore.kernel.org/linux-block/20190515030310.20393-1-ming.lei@redhat.com/T/#t > > block/blk-core.c | 5 ++--- > 1 file changed, 2 insertions(+), 3 deletions(-) > > diff --git a/block/blk-core.c b/block/blk-core.c > index 419d600e6637..e887915c7804 100644 > --- a/block/blk-core.c > +++ b/block/blk-core.c > @@ -1054,10 +1054,8 @@ blk_qc_t generic_make_request(struct bio *bio) > flags = 0; > if (bio->bi_opf & REQ_NOWAIT) > flags = BLK_MQ_REQ_NOWAIT; > - if (blk_queue_enter(q, flags) < 0) { > + if (blk_queue_enter(q, flags) < 0) > enter_succeeded = false; > - q = NULL; > - } > } > > if (enter_succeeded) { > @@ -1088,6 +1086,7 @@ blk_qc_t generic_make_request(struct bio *bio) > bio_wouldblock_error(bio); > else > bio_io_error(bio); > + q = NULL; > } > bio = bio_list_pop(&bio_list_on_stack[0]); > } while (bio); > -- > 2.21.0 >