Re: [PATCH] spapr: Improve handling of memory unplug with old guests

[David Gibson <david@gibson.dropbear.id.au>]

[-- Attachment #1: Type: text/plain, Size: 7479 bytes --]

On Fri, Jan 08, 2021 at 06:31:27PM +0100, Greg Kurz wrote:
> Since commit 1e8b5b1aa16b ("spapr: Allow memory unplug to always succeed")
> trying to unplug memory from a guest that doesn't support it (eg. rhel6)
> no longer generates an error like it used to. Instead, it leaves the
> memory around : only a subsequent reboot or manual use of drmgr within
> the guest can complete the hot-unplug sequence. A flag was added to
> SpaprMachineClass so that this new behavior only applies to the default
> machine type.
> 
> We can do better. CAS processes all pending hot-unplug requests. This
> means that we don't really care about what the guest supports if
> the hot-unplug request happens before CAS.
> 
> All guests that we care for, even old ones, set enough bits in OV5
> that lead to a non-empty bitmap in spapr->ov5_cas. Use that as a
> heuristic to decide if CAS has already occured or not.
> 
> Always accept unplug requests that happen before CAS since CAS will
> process them. Restore the previous behavior of rejecting them after
> CAS when we know that the guest doesn't support memory hot-unplug.
> 
> This behavior is suitable for all machine types : this allows to
> drop the pre_6_0_memory_unplug flag.
> 
> Fixes: 1e8b5b1aa16b ("spapr: Allow memory unplug to always succeed")
> Signed-off-by: Greg Kurz <groug@kaod.org>

Applied, sorry it too me so long.

> ---
>  hw/ppc/spapr.c              |   24 +++++++++++++-----------
>  hw/ppc/spapr_events.c       |    3 +--
>  hw/ppc/spapr_ovec.c         |    7 +++++++
>  include/hw/ppc/spapr.h      |    2 +-
>  include/hw/ppc/spapr_ovec.h |    1 +
>  5 files changed, 23 insertions(+), 14 deletions(-)
> 
> diff --git a/hw/ppc/spapr.c b/hw/ppc/spapr.c
> index 2c403b574e37..6c47466fc2f1 100644
> --- a/hw/ppc/spapr.c
> +++ b/hw/ppc/spapr.c
> @@ -4048,6 +4048,18 @@ static void spapr_machine_device_unplug(HotplugHandler *hotplug_dev,
>      }
>  }
>  
> +bool spapr_memory_hot_unplug_supported(SpaprMachineState *spapr)
> +{
> +    return spapr_ovec_test(spapr->ov5_cas, OV5_HP_EVT) ||
> +        /*
> +         * CAS will process all pending unplug requests.
> +         *
> +         * HACK: a guest could theoretically have cleared all bits in OV5,
> +         * but none of the guests we care for do.
> +         */

Hrm.  This is pretty ugly - I thought we had a better canonical way of
determining if CAS had already happened this boot, but it appears
not.  I don't want to delay this patch, since it is an important fix,
but it would be nice if you could do a later cleanup to have a nicer
way of detecting CAS-hasn't-happened.

> +        spapr_ovec_empty(spapr->ov5_cas);
> +}
> +
>  static void spapr_machine_device_unplug_request(HotplugHandler *hotplug_dev,
>                                                  DeviceState *dev, Error **errp)
>  {
> @@ -4056,16 +4068,9 @@ static void spapr_machine_device_unplug_request(HotplugHandler *hotplug_dev,
>      SpaprMachineClass *smc = SPAPR_MACHINE_CLASS(mc);
>  
>      if (object_dynamic_cast(OBJECT(dev), TYPE_PC_DIMM)) {
> -        if (!smc->pre_6_0_memory_unplug ||
> -            spapr_ovec_test(sms->ov5_cas, OV5_HP_EVT)) {
> +        if (spapr_memory_hot_unplug_supported(sms)) {
>              spapr_memory_unplug_request(hotplug_dev, dev, errp);
>          } else {
> -            /* NOTE: this means there is a window after guest reset, prior to
> -             * CAS negotiation, where unplug requests will fail due to the
> -             * capability not being detected yet. This is a bit different than
> -             * the case with PCI unplug, where the events will be queued and
> -             * eventually handled by the guest after boot
> -             */
>              error_setg(errp, "Memory hot unplug not supported for this guest");
>          }
>      } else if (object_dynamic_cast(OBJECT(dev), TYPE_SPAPR_CPU_CORE)) {
> @@ -4543,11 +4548,8 @@ DEFINE_SPAPR_MACHINE(6_0, "6.0", true);
>   */
>  static void spapr_machine_5_2_class_options(MachineClass *mc)
>  {
> -    SpaprMachineClass *smc = SPAPR_MACHINE_CLASS(mc);
> -
>      spapr_machine_6_0_class_options(mc);
>      compat_props_add(mc->compat_props, hw_compat_5_2, hw_compat_5_2_len);
> -    smc->pre_6_0_memory_unplug = true;
>  }
>  
>  DEFINE_SPAPR_MACHINE(5_2, "5.2", false);
> diff --git a/hw/ppc/spapr_events.c b/hw/ppc/spapr_events.c
> index 6aedd988b3d0..d51daedfa6e0 100644
> --- a/hw/ppc/spapr_events.c
> +++ b/hw/ppc/spapr_events.c
> @@ -658,8 +658,7 @@ static void spapr_hotplug_req_event(uint8_t hp_id, uint8_t hp_action,
>          /* we should not be using count_indexed value unless the guest
>           * supports dedicated hotplug event source
>           */
> -        g_assert(!SPAPR_MACHINE_GET_CLASS(spapr)->pre_6_0_memory_unplug ||
> -                 spapr_ovec_test(spapr->ov5_cas, OV5_HP_EVT));
> +        g_assert(spapr_memory_hot_unplug_supported(spapr));
>          hp->drc_id.count_indexed.count =
>              cpu_to_be32(drc_id->count_indexed.count);
>          hp->drc_id.count_indexed.index =
> diff --git a/hw/ppc/spapr_ovec.c b/hw/ppc/spapr_ovec.c
> index dd003f1763fd..b2567caa5cf4 100644
> --- a/hw/ppc/spapr_ovec.c
> +++ b/hw/ppc/spapr_ovec.c
> @@ -125,6 +125,13 @@ bool spapr_ovec_test(SpaprOptionVector *ov, long bitnr)
>      return test_bit(bitnr, ov->bitmap) ? true : false;
>  }
>  
> +bool spapr_ovec_empty(SpaprOptionVector *ov)
> +{
> +    g_assert(ov);
> +
> +    return bitmap_empty(ov->bitmap, OV_MAXBITS);
> +}
> +
>  static void guest_byte_to_bitmap(uint8_t entry, unsigned long *bitmap,
>                                   long bitmap_offset)
>  {
> diff --git a/include/hw/ppc/spapr.h b/include/hw/ppc/spapr.h
> index 1cc19575f548..3ad2ff713279 100644
> --- a/include/hw/ppc/spapr.h
> +++ b/include/hw/ppc/spapr.h
> @@ -142,7 +142,6 @@ struct SpaprMachineClass {
>      hwaddr rma_limit;          /* clamp the RMA to this size */
>      bool pre_5_1_assoc_refpoints;
>      bool pre_5_2_numa_associativity;
> -    bool pre_6_0_memory_unplug;
>  
>      bool (*phb_placement)(SpaprMachineState *spapr, uint32_t index,
>                            uint64_t *buid, hwaddr *pio, 
> @@ -950,4 +949,5 @@ bool spapr_check_pagesize(SpaprMachineState *spapr, hwaddr pagesize,
>  
>  void spapr_set_all_lpcrs(target_ulong value, target_ulong mask);
>  hwaddr spapr_get_rtas_addr(void);
> +bool spapr_memory_hot_unplug_supported(SpaprMachineState *spapr);
>  #endif /* HW_SPAPR_H */
> diff --git a/include/hw/ppc/spapr_ovec.h b/include/hw/ppc/spapr_ovec.h
> index d4dee9e06a01..48b716a060c2 100644
> --- a/include/hw/ppc/spapr_ovec.h
> +++ b/include/hw/ppc/spapr_ovec.h
> @@ -71,6 +71,7 @@ void spapr_ovec_cleanup(SpaprOptionVector *ov);
>  void spapr_ovec_set(SpaprOptionVector *ov, long bitnr);
>  void spapr_ovec_clear(SpaprOptionVector *ov, long bitnr);
>  bool spapr_ovec_test(SpaprOptionVector *ov, long bitnr);
> +bool spapr_ovec_empty(SpaprOptionVector *ov);
>  SpaprOptionVector *spapr_ovec_parse_vector(target_ulong table_addr, int vector);
>  int spapr_dt_ovec(void *fdt, int fdt_offset,
>                    SpaprOptionVector *ov, const char *name);
> 
> 

-- 
David Gibson			| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au	| minimalist, thank you.  NOT _the_ _other_
				| _way_ _around_!
http://www.ozlabs.org/~dgibson

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]