Re: [Qemu-devel] [PATCH 17/33] spapr_drc: initial implementation of sPAPRDRConnector device

[David Gibson]

On Fri, May 08, 2015 at 12:39:20AM +0200, Alexander Graf wrote:
> 
> 
> On 07.05.15 07:33, David Gibson wrote:
> > From: Michael Roth <address@hidden>
> > 
> > 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 <address@hidden>
> > Reviewed-by: David Gibson <address@hidden>
> > Signed-off-by: David Gibson <address@hidden>
> > ---
> >  hw/ppc/Makefile.objs       |   2 +-
> >  hw/ppc/spapr_drc.c         | 588 
> > +++++++++++++++++++++++++++++++++++++++++++++
> >  include/hw/ppc/spapr_drc.h | 199 +++++++++++++++
> >  3 files changed, 788 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 437955d..c8ab06e 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-$(CONFIG_PSERIES) += spapr_pci.o spapr_rtc.o
> > +obj-$(CONFIG_PSERIES) += spapr_pci.o spapr_rtc.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..047c6c7
> > --- /dev/null
> > +++ b/hw/ppc/spapr_drc.c
> > @@ -0,0 +1,588 @@
> > +/*
> > + * QEMU SPAPR Dynamic Reconfiguration Connector Implementation
> > + *
> > + * Copyright IBM Corp. 2014
> > + *
> > + * Authors:
> > + *  Michael Roth      <address@hidden>
> > + *
> > + * 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 sPAPRDRConnectorTypeShift get_type_shift(sPAPRDRConnectorType type)
> > +{
> > +    uint32_t shift = 0;
> > +
> > +    /* make sure this isn't SPAPR_DR_CONNECTOR_TYPE_ANY, or some
> > +     * other wonky value.
> > +     */
> > +    g_assert(is_power_of_2(type));
> > +
> > +    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 int set_isolation_state(sPAPRDRConnector *drc,
> > +                               sPAPRDRIsolationState state)
> > +{
> > +    sPAPRDRConnectorClass *drck = SPAPR_DR_CONNECTOR_GET_CLASS(drc);
> > +
> > +    DPRINTFN("drc: %x, set_isolation_state: %x", get_index(drc), state);
> > +
> > +    drc->isolation_state = state;
> > +
> > +    if (drc->isolation_state == SPAPR_DR_ISOLATION_STATE_ISOLATED) {
> > +        /* if we're awaiting release, but still in an unconfigured state,
> > +         * it's likely the guest is still in the process of configuring
> > +         * the device and is transitioning the devices to an ISOLATED
> > +         * state as a part of that process. so we only complete the
> > +         * removal when this transition happens for a device in a
> > +         * configured state, as suggested by the state diagram from
> > +         * PAPR+ 2.7, 13.4
> > +         */
> > +        if (drc->awaiting_release) {
> > +            if (drc->configured) {
> > +                DPRINTFN("finalizing device removal");
> > +                drck->detach(drc, DEVICE(drc->dev), drc->detach_cb,
> > +                             drc->detach_cb_opaque, NULL);
> > +            } else {
> > +                DPRINTFN("deferring device removal on unconfigured 
> > device\n");
> > +            }
> > +        }
> > +        drc->configured = false;
> > +    }
> > +
> > +    return 0;
> > +}
> > +
> > +static int set_indicator_state(sPAPRDRConnector *drc,
> > +                               sPAPRDRIndicatorState state)
> > +{
> > +    DPRINTFN("drc: %x, set_indicator_state: %x", get_index(drc), state);
> > +    drc->indicator_state = state;
> > +    return 0;
> > +}
> > +
> > +static int set_allocation_state(sPAPRDRConnector *drc,
> > +                                sPAPRDRAllocationState state)
> > +{
> > +    sPAPRDRConnectorClass *drck = SPAPR_DR_CONNECTOR_GET_CLASS(drc);
> > +
> > +    DPRINTFN("drc: %x, set_allocation_state: %x", get_index(drc), state);
> > +
> > +    if (drc->type != SPAPR_DR_CONNECTOR_TYPE_PCI) {
> > +        drc->allocation_state = state;
> > +        if (drc->awaiting_release &&
> > +            drc->allocation_state == SPAPR_DR_ALLOCATION_STATE_UNUSABLE) {
> > +            DPRINTFN("finalizing device removal");
> > +            drck->detach(drc, DEVICE(drc->dev), drc->detach_cb,
> > +                         drc->detach_cb_opaque, NULL);
> > +        }
> > +    }
> > +    return 0;
> > +}
> > +
> > +static uint32_t get_type(sPAPRDRConnector *drc)
> > +{
> > +    return drc->type;
> > +}
> > +
> > +static const char *get_name(sPAPRDRConnector *drc)
> > +{
> > +    return drc->name;
> > +}
> > +
> > +static const void *get_fdt(sPAPRDRConnector *drc, int *fdt_start_offset)
> > +{
> > +    if (fdt_start_offset) {
> > +        *fdt_start_offset = drc->fdt_start_offset;
> > +    }
> > +    return drc->fdt;
> > +}
> > +
> > +static void set_configured(sPAPRDRConnector *drc)
> > +{
> > +    DPRINTFN("drc: %x, set_configured", get_index(drc));
> > +
> > +    if (drc->isolation_state != SPAPR_DR_ISOLATION_STATE_UNISOLATED) {
> > +        /* guest should be not configuring an isolated device */
> > +        DPRINTFN("drc: %x, set_configured: skipping isolated device",
> > +                 get_index(drc));
> > +        return;
> > +    }
> > +    drc->configured = true;
> > +}
> > +
> > +/*
> > + * 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)
> > +{
> > +    sPAPRDREntitySense state;
> > +
> > +    if (drc->dev) {
> > +        if (drc->type != SPAPR_DR_CONNECTOR_TYPE_PCI &&
> > +            drc->allocation_state == SPAPR_DR_ALLOCATION_STATE_UNUSABLE) {
> > +            /* for logical DR, we return a state of UNUSABLE
> > +             * iff the allocation state UNUSABLE.
> > +             * Otherwise, report the state as USABLE/PRESENT,
> > +             * as we would for PCI.
> > +             */
> > +            state = SPAPR_DR_ENTITY_SENSE_UNUSABLE;
> > +        } else {
> > +            /* 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.
> > +             */
> > +            state = SPAPR_DR_ENTITY_SENSE_PRESENT;
> > +        }
> > +    } else {
> > +        if (drc->type == SPAPR_DR_CONNECTOR_TYPE_PCI) {
> > +            /* PCI devices, and only PCI devices, use EMPTY
> > +             * in cases where we'd otherwise use UNUSABLE
> > +             */
> > +            state = SPAPR_DR_ENTITY_SENSE_EMPTY;
> > +        } else {
> > +            state = SPAPR_DR_ENTITY_SENSE_UNUSABLE;
> > +        }
> > +    }
> > +
> > +    DPRINTFN("drc: %x, entity_sense: %x", get_index(drc), state);
> > +    return state;
> > +}
> > +
> > +static void prop_get_index(Object *obj, Visitor *v, void *opaque,
> > +                                  const char *name, Error **errp)
> > +{
> > +    sPAPRDRConnector *drc = SPAPR_DR_CONNECTOR(obj);
> > +    sPAPRDRConnectorClass *drck = SPAPR_DR_CONNECTOR_GET_CLASS(drc);
> > +    uint32_t value = (uint32_t)drck->get_index(drc);
> > +    visit_type_uint32(v, &value, name, errp);
> > +}
> > +
> > +static void prop_get_type(Object *obj, Visitor *v, void *opaque,
> > +                          const char *name, Error **errp)
> > +{
> > +    sPAPRDRConnector *drc = SPAPR_DR_CONNECTOR(obj);
> > +    sPAPRDRConnectorClass *drck = SPAPR_DR_CONNECTOR_GET_CLASS(drc);
> > +    uint32_t value = (uint32_t)drck->get_type(drc);
> > +    visit_type_uint32(v, &value, name, errp);
> > +}
> > +
> > +static char *prop_get_name(Object *obj, Error **errp)
> > +{
> > +    sPAPRDRConnector *drc = SPAPR_DR_CONNECTOR(obj);
> > +    sPAPRDRConnectorClass *drck = SPAPR_DR_CONNECTOR_GET_CLASS(drc);
> > +    return g_strdup(drck->get_name(drc));
> > +}
> > +
> > +static void prop_get_entity_sense(Object *obj, Visitor *v, void *opaque,
> > +                                  const char *name, Error **errp)
> > +{
> > +    sPAPRDRConnector *drc = SPAPR_DR_CONNECTOR(obj);
> > +    sPAPRDRConnectorClass *drck = SPAPR_DR_CONNECTOR_GET_CLASS(drc);
> > +    uint32_t value = (uint32_t)drck->entity_sense(drc);
> > +    visit_type_uint32(v, &value, name, errp);
> > +}
> > +
> > +static void prop_get_fdt(Object *obj, Visitor *v, void *opaque,
> > +                        const char *name, Error **errp)
> > +{
> > +    sPAPRDRConnector *drc = SPAPR_DR_CONNECTOR(obj);
> > +    int fdt_offset_next, fdt_offset, fdt_depth;
> > +    void *fdt;
> > +
> > +    if (!drc->fdt) {
> > +        return;
> > +    }
> > +
> > +    fdt = drc->fdt;
> > +    fdt_offset = drc->fdt_start_offset;
> > +    fdt_depth = 0;
> > +
> > +    do {
> > +        const char *name = NULL;
> > +        const struct fdt_property *prop = NULL;
> > +        int prop_len = 0, name_len = 0;
> > +        uint32_t tag;
> > +
> > +        tag = fdt_next_tag(fdt, fdt_offset, &fdt_offset_next);
> > +        switch (tag) {
> > +        case FDT_BEGIN_NODE:
> > +            fdt_depth++;
> > +            name = fdt_get_name(fdt, fdt_offset, &name_len);
> > +            visit_start_struct(v, NULL, NULL, name, 0, NULL);
> > +            break;
> > +        case FDT_END_NODE:
> > +            /* shouldn't ever see an FDT_END_NODE before FDT_BEGIN_NODE */
> > +            g_assert(fdt_depth > 0);
> > +            visit_end_struct(v, NULL);
> > +            fdt_depth--;
> > +            break;
> > +        case FDT_PROP: {
> > +            int i;
> > +            prop = fdt_get_property_by_offset(fdt, fdt_offset, &prop_len);
> 
> Did this fdt function only get introduced recently? My test system has
> 1.3.0 installed which doesn't contain it apparently.
> 
> Please add a configure check to verify it actually exists if it's really
> necessary.

Erm.. that's odd.  That function went in with commit 73dca9ae, which
is included in dtc 1.3.0..

-- 
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

pgp4obVzuEVIk.pgp
Description: PGP signature