Re: [PATCH for-6.12 0/4] block, bfq: fix corner cases related to bfqq merging

[Ming Lei <ming.lei@redhat.com>]

On Wed, Sep 04, 2024 at 09:32:26AM +0800, Yu Kuai wrote:
> Hi,
> 
> 在 2024/09/03 23:51, Jens Axboe 写道:
> > On 9/2/24 7:03 AM, Yu Kuai wrote:
> > > From: Yu Kuai <yukuai3@huawei.com>
> > > 
> > > Our syzkaller report a UAF problem(details in patch 1), however it can't
> > > be reporduced. And this set are some corner cases fix that might be
> > > related, and they are found by code review.
> > > 
> > > Yu Kuai (4):
> > >    block, bfq: fix possible UAF for bfqq->bic with merge chain
> > >    block, bfq: choose the last bfqq from merge chain in
> > >      bfq_setup_cooperator()
> > >    block, bfq: don't break merge chain in bfq_split_bfqq()
> > >    block, bfq: use bfq_reassign_last_bfqq() in bfq_bfqq_move()
> > > 
> > >   block/bfq-cgroup.c  |  7 +------
> > >   block/bfq-iosched.c | 17 +++++++++++------
> > >   block/bfq-iosched.h |  2 ++
> > >   3 files changed, 14 insertions(+), 12 deletions(-)
> > 
> > BFQ is effectively unmaintained, and has been for quite a while at
> > this point. I'll apply these, thanks for looking into it, but I think we
> > should move BFQ to an unmaintained state at this point.
> 
> Sorry to hear that, we would be willing to take on the responsibility of
> maintaining this code, please let me know if there are any specific
> guidelines or processes we should follow. We do have customers are using
> bfq in downstream kernels, and we are still running lots of test for
> bfq.

Hi Yukuai,
 
BTW, BFQ is default IO scheduler for SCSI/MMC in Fedora. That is great if you'd
like to volunteer to take over maintaining it.


thanks,
Ming