Re: [PATCH for-next 2/3] RDMA/rtrs: Fix rxe_dealloc_pd warning

[Leon Romanovsky <leon@kernel.org>]

On Tue, Apr 18, 2023 at 07:04:00AM +0000, Zhijian Li (Fujitsu) wrote:
> 
> 
> On 18/04/2023 02:04, Leon Romanovsky wrote:
> > On Mon, Apr 17, 2023 at 02:18:24AM +0000, Zhijian Li (Fujitsu) wrote:
> >>
> >>
> >> On 14/04/2023 23:58, Zhu Yanjun wrote:
> >>> 在 2023/4/13 21:24, Leon Romanovsky 写道:
> >>>> On Thu, Apr 13, 2023 at 08:12:15AM +0000, Zhijian Li (Fujitsu) wrote:
> >>>>>
> >>>>>
> >>>>> On 13/04/2023 15:35, Guoqing Jiang wrote:
> >>>>>> Hi,
> >>>>>>
> >>>>>> I take a closer look today.
> >>>>>>
> >>>>>> On 4/12/23 09:15, Zhijian Li (Fujitsu) wrote:
> >>>>>>>
> >>>>>>> On 11/04/2023 20:26, Leon Romanovsky wrote:
> >>>>>>>> On Tue, Apr 11, 2023 at 02:43:46AM +0000, Zhijian Li (Fujitsu) wrote:
> >>>>>>>>>
> >>>>>>>>> On 10/04/2023 21:10, Guoqing Jiang wrote:
> >>>>>>>>>>
> >>>>>>>>>> On 4/10/23 20:08, Leon Romanovsky wrote:
> >>>>>>>>>>> On Mon, Apr 10, 2023 at 06:43:03AM +0000, Li Zhijian wrote:
> >>>>>>>>>>>> The warning occurs when destroying PD whose reference count is not zero.
> >>>>>>>>>>>>
> >>>>>>>>>>>> Precodition: clt_path->s.con_num is 2.
> >>>>>>>>>>>> So 2 cm connection will be created as below:
> >>>>>>>>>>>> CPU0                                              CPU1
> >>>>>>>>>>>> init_conns {                              |
> >>>>>>>>>>>>        create_cm() // a. con[0] created        |
> >>>>>>>>>>>>                                                |  a'. rtrs_clt_rdma_cm_handler() {
> >>>>>>>>>>>>                                                |    rtrs_rdma_addr_resolved()
> >>>>>>>>>>>>                                                |      create_con_cq_qp(con); << con[0]
> >>>>>>>>>>>>                                                |  }
> >>>>>>>>>>>>                                                | in this moment, refcnt of PD was increased to 2+
> >>>>>>
> >>>>>> What do you mean "refcnt of PD"? usecnt in struct ib_pd or dev_ref.
> >>>>>
> >>>>> I mean usecnt in struct ib_pd
> >>>>>
> >>>>>
> >>>>>
> >>>>>>
> >>>>>>>>>>>>                                                |
> >>>>>>>>>>>>        create_cm() // b. cid = 1, failed       |
> >>>>>>>>>>>>          destroy_con_cq_qp()                   |
> >>>>>>>>>>>>            rtrs_ib_dev_put()                   |
> >>>>>>>>>>>>              dev_free()                        |
> >>>>>>>>>>>>                ib_dealloc_pd(dev->ib_pd) << PD |
> >>>>>>>>>>>>                 is destroyed, but refcnt is    |
> >>>>>>>>>>>>                 still greater than 0           |
> >>>>>>
> >>>>>> Assuming you mean "pd->usecnt". We only allocate pd in con[0] by rtrs_ib_dev_find_or_add,
> >>>>>> if con[1] failed to create cm, then alloc_path_reqs -> ib_alloc_mr -> atomic_inc(&pd->usecnt)
> >>>>>> can't be triggered. Is there other places could increase the refcnt?
> >>>>>
> >>>>>
> >>>>> Yes, when create a qp, it will also associate to this PD, that also mean refcnt of PD will be increased.
> >>>>>
> >>>>> When con[0](create_con_cq_qp) succeeded, refcnt of PD will be 2. and then when con[1] failed, since
> >>>>> QP didn't create, refcnt of PD is still 2. con[1]'s cleanup will destroy the PD(ib_dealloc_pd) since dev_ref = 1, after that its
> >>>>> refcnt is still 1.
> >>>>
> >>>> Why is refcnt 1 in con[1] destruction phase? It seems to me like a bug.
> >>
> >>
> >>
> >>> +	if (!con->has_dev)
> >>> +		return;
> >>>   	if (clt_path->s.dev_ref && !--clt_path->s.dev_ref) {
> >>>   		rtrs_ib_dev_put(clt_path->s.dev);
> >>>   		clt_path->s.dev = NULL;
> >>
> >> Currently, without this patch:
> >> 1. PD and clt_path->s.dev are shared among connections.
> >> 2. every con[n]'s cleanup phase will call destroy_con_cq_qp()
> >> 3. clt_path->s.dev will be always decreased in destroy_con_cq_qp(), and when
> >>      clt_path->s.dev become zero, it will destroy PD.
> >> 4. when con[1] failed to create, con[1] will not take clt_path->s.dev, but it try to decreased clt_path->s.dev <<< it's wrong to do that.
> > 
> > So please fix it by making sure that failure to create con[1] will
> > release resources which were allocated. If con[1] didn't increase
> > s.dev_ref, it shouldn't decrease it either.
> 
> You are right, the current patch did exactly that.
> It introduced a con owning flag 'has_dev' to indicate whether this con has taken s.dev.
> so that its cleanup phase will only decrease its s.dev properly.

The has_dev is a workaround and not a solution. In proper error unwind
sequence, you won't need extra flag.

Thanks

> 
> Thanks
> Zhijian
> 
> 
> > 
> > Thanks
> > 
> >>
> >>
> >> Thanks
> >> Zhijian
> >>
> >>> Agree. We should find out why refcnt 1 and fix this problem.
> >>
> >>
> >>
> >>
> >>>
> >>> Zhu Yanjun
> >>>>
> >>>> Thanks
> >>>