[PATCH v11 1/1] nvmet: support reservation feature

[Dmitry Bogdanov]

On Tue, Oct 08, 2024 at 03:27:46PM +0800, Guixin Liu wrote:
> 在 2024/10/7 16:08, Dmitry Bogdanov 写道:
> > On Sat, Oct 05, 2024 at 10:26:58PM +0800, Guixin Liu wrote:
> > > 在 2024/9/30 20:53, Dmitry Bogdanov 写道:
> > > > On Sun, Sep 29, 2024 at 11:14:10AM +0800, Guixin Liu wrote:
> > > > > This patch implements the reservation feature, includes:
> > > > > 1. reservation register(register, unregister and replace).
> > > > > 2. reservation acquire(acquire, preempt, preempt and abort).
> > > > > 3. reservation release(release and clear).
> > > > > 4. reservation report.
> > > > > 5. set feature and get feature of reservation notify mask.
> > > > > 6. get log page of reservation event.
> > > > > 
> > > > > And also make reservation configurable, one can set ns to support
> > > > > reservation before enable ns. The default of resv_enable is false.
> > > > > 
> > > > > Signed-off-by: Guixin Liu <kanie at linux.alibaba.com>
> > > > > ---
> > > > > +
> > > > > +static void nvmet_execute_pr_report(struct nvmet_req *req)
> > > > > +{
> > > > > +       u32 cdw11 = le32_to_cpu(req->cmd->common.cdw11);
> > > > > +       u32 cdw10 = le32_to_cpu(req->cmd->common.cdw10);
> > > > > +       u32 num_bytes = 4 * (cdw10 + 1); /* cdw10 is number of dwords */
> > > > > +       u8 eds = cdw11 & 1; /* Extended data structure, bit 00 */
> > > > > +       struct nvme_registered_ctrl_ext *ctrl_eds;
> > > > > +       struct nvme_reservation_status_ext *data;
> > > > > +       struct nvmet_pr *pr = &req->ns->pr;
> > > > > +       struct nvmet_pr_registrant *holder;
> > > > > +       struct nvmet_pr_registrant *reg;
> > > > > +       u16 num_ctrls = 0;
> > > > > +       u16 status;
> > > > > +       u8 rtype;
> > > > > +
> > > > > +       /* nvmet hostid(uuid_t) is 128 bit. */
> > > > > +       if (!eds) {
> > > > > +               req->error_loc = offsetof(struct nvme_common_command, cdw11);
> > > > > +               status = NVME_SC_HOST_ID_INCONSIST | NVME_SC_DNR;
> > > > > +               goto out;
> > > > > +       }
> > > > > +
> > > > > +       if (num_bytes < sizeof(struct nvme_reservation_status_ext)) {
> > > > > +               req->error_loc = offsetof(struct nvme_common_command, cdw10);
> > > > > +               status = NVME_SC_INVALID_FIELD | NVME_SC_DNR;
> > > > > +               goto out;
> > > > > +       }
> > > > > +
> > > > > +       data = kmalloc(num_bytes, GFP_KERNEL);
> > > > > +       if (!data) {
> > > > > +               status = NVME_SC_INTERNAL;
> > > > > +               goto out;
> > > > > +       }
> > > > > +       memset(data, 0, num_bytes);
> > > > > +       data->gen = cpu_to_le32(atomic_read(&pr->generation));
> > > > > +       data->ptpls = 0;
> > > > > +       ctrl_eds = data->regctl_eds;
> > > > > +
> > > > > +       rcu_read_lock();
> > > > > +       holder = rcu_dereference(pr->holder);
> > > > > +       rtype = holder ? holder->rtype : 0;
> > > > > +       data->rtype = rtype;
> > > > > +
> > > > > +       list_for_each_entry_rcu(reg, &pr->registrant_list, entry) {
> > > > > +               if ((void *)ctrl_eds >= (void *)(data + num_bytes))
> > > > > +                       break;
> > > > > +               /*
> > > > > +                * Dynamic controller, set cntlid to 0xffff.
> > > > > +                */
> > > > > +               ctrl_eds->cntlid = 0xffff;
> > > > > +               if (rtype == NVME_PR_WRITE_EXCLUSIVE_ALL_REGS ||
> > > > > +                   rtype == NVME_PR_EXCLUSIVE_ACCESS_ALL_REGS)
> > > > > +                       ctrl_eds->rcsts = 1;
> > > > > +               if (holder && uuid_equal(&reg->hostid, &holder->hostid))
> > > > is just (reg == holder) not enough here?
> > > Yes.
> > > > > +                       ctrl_eds->rcsts = 1;
> > > > > +               uuid_copy((uuid_t *)&ctrl_eds->hostid, &reg->hostid);
> > > > > +               ctrl_eds->rkey = cpu_to_le64(reg->rkey);
> > > > > +               ctrl_eds++;
> > > > > +               num_ctrls++;
> > > > > +       }
> > > > > +       rcu_read_unlock();
> > > > > +
> > > > > +       put_unaligned_le16(num_ctrls, data->regctl);
> > > > Here you should report the number of all registered registrants in ns,
> > > > not the only reported. It is used by the host to understand that it got
> > > > not a full response.
> > > > NUMD field (cdw11) is about transfering the response. But the structure itsef in
> > > > the reposnse has to be complete.
> > > The NVMe spec says that:
> > > 
> > >      "This indicates the number of Registered Controller data structures
> > > and/or Registered Controller extended data structures contained in this
> > > data structure."
> > You read not the latest Spec again. Spec v2.1 has another statement:
> >    Number of Registrants (REGSTRNT): This field indicates the number of registrants of the
> >    namespace. This indicates the number of Registrant data structures or Registrant Extended data
> >    structures contained in this data structure.
> 
> Well, still, "contained in this data structure" means that we should
> report how many registrants we reported.

No, vice-versa :), that means that "this data structure" shall contain "the number
of registrants of the namespace".

> 
> I do a research on nvme-cli and SPDK:
> 
> 1. nvme-cli: In json_nvme_resv_report, use reported regctls to traverse
> reported registrants.

I saw, and that is a security violation - to trust a data from a network.
Try to set a small NUMD and ask ReservationRepost from SPDK nvme target -
you will get a coredump of nvme-cli due to aouut-of-bound access of its
buffer (NUMD sized).

Instead, it shall send ReservationReport twice - the first with small
buffer to get just REGSTRNT, and the second with a full buffer to get
the whole list. That is a common handling of a variable length response
in SCSI, and NVME too.

> 
> 2. SPDK: In nvmf_ns_reservation_report, set regctl to the number of
> reported registrants.

Looks like you see again not the last version of the reference.
The latest SPDK has a correct implementation - It reports a whole
registrants list since Feb 13, 2023.
Please, follow it.

> Best Regards,
> 
> Guixin Liu
> 
> > > It means that host use this filed to obtain the reported registrants number.
> > > 
> > Think from the host point of view - how are you going to understand
> > how much registrants are there in the namespace?

BR,
 Dmitry