Re: [Qemu-devel] [PATCH v4 02/17] spapr_drc: initial implementation of sPAPRDRConnector device

[Bharata B Rao <bharata.rao@gmail.com>]

On Tue, Dec 23, 2014 at 6:00 PM, Michael Roth <mdroth@linux.vnet.ibm.com> wrote:
> This device emulates a firmware abstraction used by pSeries guests to
> manage hotplug/dynamic-reconfiguration of host-bridges, PCI devices,
> memory, and CPUs. It is conceptually similar to an SHPC device,
> complete with LED indicators to identify individual slots to physical
> physical users and indicate when it is safe to remove a device. In
> some cases it is also used to manage virtualized resources, such a
> memory, CPUs, and physical-host bridges, which in the case of pSeries
> guests are virtualized resources where the physical components are
> managed by the host.
>
> Guests communicate with these DR Connectors using RTAS calls,
> generally by addressing the unique DRC index associated with a
> particular connector for a particular resource. For introspection
> purposes we expose this state initially as QOM properties, and
> in subsequent patches will introduce the RTAS calls that make use of
> it. This constitutes to the 'guest' interface.
>
> On the QEMU side we provide an attach/detach interface to associate
> or cleanup a DeviceState with a particular sPAPRDRConnector in
> response to hotplug/unplug, respectively. This constitutes the
> 'physical' interface to the DR Connector.
>
> Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com>
> ---
>  hw/ppc/Makefile.objs       |   2 +-
>  hw/ppc/spapr_drc.c         | 503 +++++++++++++++++++++++++++++++++++++++++++++
>  include/hw/ppc/spapr_drc.h | 201 ++++++++++++++++++
>  3 files changed, 705 insertions(+), 1 deletion(-)
>  create mode 100644 hw/ppc/spapr_drc.c
>  create mode 100644 include/hw/ppc/spapr_drc.h
>
> diff --git a/hw/ppc/Makefile.objs b/hw/ppc/Makefile.objs
> index 19d9920..ea010fd 100644
> --- a/hw/ppc/Makefile.objs
> +++ b/hw/ppc/Makefile.objs
> @@ -3,7 +3,7 @@ obj-y += ppc.o ppc_booke.o
>  # IBM pSeries (sPAPR)
>  obj-$(CONFIG_PSERIES) += spapr.o spapr_vio.o spapr_events.o
>  obj-$(CONFIG_PSERIES) += spapr_hcall.o spapr_iommu.o spapr_rtas.o
> -obj-$(CONFIG_PSERIES) += spapr_pci.o
> +obj-$(CONFIG_PSERIES) += spapr_pci.o spapr_drc.o
>  ifeq ($(CONFIG_PCI)$(CONFIG_PSERIES)$(CONFIG_LINUX), yyy)
>  obj-y += spapr_pci_vfio.o
>  endif
> diff --git a/hw/ppc/spapr_drc.c b/hw/ppc/spapr_drc.c
> new file mode 100644
> index 0000000..f81c6d1
> --- /dev/null
> +++ b/hw/ppc/spapr_drc.c
> @@ -0,0 +1,503 @@
> +/*
> + * QEMU SPAPR Dynamic Reconfiguration Connector Implementation
> + *
> + * Copyright IBM Corp. 2014
> + *
> + * Authors:
> + *  Michael Roth      <mdroth@linux.vnet.ibm.com>
> + *
> + * This work is licensed under the terms of the GNU GPL, version 2 or later.
> + * See the COPYING file in the top-level directory.
> + */
> +
> +#include "hw/ppc/spapr_drc.h"
> +#include "qom/object.h"
> +#include "hw/qdev.h"
> +#include "qapi/visitor.h"
> +#include "qemu/error-report.h"
> +
> +/* #define DEBUG_SPAPR_DRC */
> +
> +#ifdef DEBUG_SPAPR_DRC
> +#define DPRINTF(fmt, ...) \
> +    do { fprintf(stderr, fmt, ## __VA_ARGS__); } while (0)
> +#define DPRINTFN(fmt, ...) \
> +    do { DPRINTF(fmt, ## __VA_ARGS__); fprintf(stderr, "\n"); } while (0)
> +#else
> +#define DPRINTF(fmt, ...) \
> +    do { } while (0)
> +#define DPRINTFN(fmt, ...) \
> +    do { } while (0)
> +#endif
> +
> +#define DRC_CONTAINER_PATH "/dr-connector"
> +#define DRC_INDEX_TYPE_SHIFT 28
> +#define DRC_INDEX_ID_MASK ~(~0 << DRC_INDEX_TYPE_SHIFT)
> +
> +static int set_isolation_state(sPAPRDRConnector *drc,
> +                               sPAPRDRIsolationState state)
> +{
> +    sPAPRDRConnectorClass *drck = SPAPR_DR_CONNECTOR_GET_CLASS(drc);
> +
> +    DPRINTFN("set_isolation_state: %x", state);
> +    drc->isolation_state = state;
> +    if (drc->awaiting_release &&
> +        drc->isolation_state == SPAPR_DR_ISOLATION_STATE_ISOLATED) {
> +        drck->detach(drc, DEVICE(drc->dev), drc->detach_cb,
> +                     drc->detach_cb_opaque);
> +    }
> +    return 0;
> +}
> +
> +static int set_indicator_state(sPAPRDRConnector *drc,
> +                               sPAPRDRIndicatorState state)
> +{
> +    DPRINTFN("set_indicator_state: %x", state);
> +    drc->indicator_state = state;
> +    return 0;
> +}
> +
> +static int set_allocation_state(sPAPRDRConnector *drc,
> +                                sPAPRDRAllocationState state)
> +{
> +    DPRINTFN("set_allocation_state: %x", state);
> +    drc->indicator_state = state;
> +    return 0;
> +}
> +
> +static uint32_t get_id(sPAPRDRConnector *drc)
> +{
> +    /* this value is unique for DRCs of a particular type, but may
> +     * overlap with the id's of other DRCs. the value is used both
> +     * to calculate a unique (across all DRC types) index, as well
> +     * as generating the ibm,drc-names OFDT property that describes
> +     * DRCs
> +     */
> +    return drc->id;
> +}
> +
> +static sPAPRDRConnectorTypeShift get_type_shift(sPAPRDRConnectorType type)
> +{
> +    uint32_t shift = 0;
> +
> +    g_assert(type != SPAPR_DR_CONNECTOR_TYPE_ANY);
> +    while (type != (1 << shift)) {
> +        shift++;
> +    }
> +    return shift;
> +}
> +
> +static uint32_t get_index(sPAPRDRConnector *drc)
> +{
> +    /* no set format for a drc index: it only needs to be globally
> +     * unique. this is how we encode the DRC type on bare-metal
> +     * however, so might as well do that here
> +     */
> +    return (get_type_shift(drc->type) << DRC_INDEX_TYPE_SHIFT) |
> +            (drc->id & DRC_INDEX_ID_MASK);
> +}
> +
> +static uint32_t get_type(sPAPRDRConnector *drc)
> +{
> +    return drc->type;
> +}
> +
> +/*
> + * dr-entity-sense sensor value
> + * returned via get-sensor-state RTAS calls
> + * as expected by state diagram in PAPR+ 2.7, 13.4
> + * based on the current allocation/indicator/power states
> + * for the DR connector.
> + */
> +static sPAPRDREntitySense entity_sense(sPAPRDRConnector *drc)
> +{
> +    if (drc->dev) {
> +        /* this assumes all PCI devices are assigned to
> +         * a 'live insertion' power domain, where QEMU
> +         * manages power state automatically as opposed
> +         * to the guest. present, non-PCI resources are
> +         * unaffected by power state.
> +         */
> +        return SPAPR_DR_ENTITY_SENSE_PRESENT;
> +    }

Hotplugged CPU comes with drc->dev set (set during
sPAPRDRConnectorClass->attach()) and hence ends up returning PRESENT
but kernel expects UNUSABLE during get-sensor-state rtas call.

> +
> +    if (drc->type == SPAPR_DR_CONNECTOR_TYPE_PCI) {
> +        /* PCI devices, and only PCI devices, use PRESENT
> +         * in cases where we'd otherwise use UNUSABLE
> +         */
> +        return SPAPR_DR_ENTITY_SENSE_EMPTY;
> +    }
> +    return SPAPR_DR_ENTITY_SENSE_UNUSABLE;
> +}