static int physdev_unmap_pirq(struct physdev_unmap_pirq *unmap)
{
struct domain *d;
int ret;
if ( !IS_PRIV(current->domain) )
return -EPERM;
if ( unmap->domid == DOMID_SELF )
d = rcu_lock_domain(current->domain);
else
d = rcu_lock_domain_by_id(unmap->domid);
if ( d == NULL )
return -ESRCH;
spin_lock(&pcidevs_lock);
spin_lock(&d->event_lock);
ret = unmap_domain_pirq(d, unmap->pirq);
spin_unlock(&d->event_lock);
spin_unlock(&pcidevs_lock);
rcu_unlock_domain(d);
return ret;
}
static inline int is_free_domid(domid_t dom)
{
struct domain *d;
if ( dom >= DOMID_FIRST_RESERVED )
return 0;
if ( (d = rcu_lock_domain_by_id(dom)) == NULL )
return 1;
rcu_unlock_domain(d);
return 0;
}
int rcu_lock_remote_target_domain_by_id(domid_t dom, struct domain **d)
{
if ( (*d = rcu_lock_domain_by_id(dom)) == NULL )
return -ESRCH;
if ( (*d == current->domain) || !IS_PRIV_FOR(current->domain, *d) )
{
rcu_unlock_domain(*d);
return -EPERM;
}
return 0;
}
static int late_hwdom_init(struct domain *d)
{
#ifdef CONFIG_LATE_HWDOM
struct domain *dom0;
int rv;
if ( d != hardware_domain || d->domain_id == 0 )
return 0;
rv = xsm_init_hardware_domain(XSM_HOOK, d);
if ( rv )
return rv;
printk("Initialising hardware domain %d\n", hardware_domid);
dom0 = rcu_lock_domain_by_id(0);
ASSERT(dom0 != NULL);
/*
* Hardware resource ranges for domain 0 have been set up from
* various sources intended to restrict the hardware domain's
* access. Apply these ranges to the actual hardware domain.
*
* Because the lists are being swapped, a side effect of this
* operation is that Domain 0's rangesets are cleared. Since
* domain 0 should not be accessing the hardware when it constructs
* a hardware domain, this should not be a problem. Both lists
* may be modified after this hypercall returns if a more complex
* device model is desired.
*/
rangeset_swap(d->irq_caps, dom0->irq_caps);
rangeset_swap(d->iomem_caps, dom0->iomem_caps);
#ifdef CONFIG_X86
rangeset_swap(d->arch.ioport_caps, dom0->arch.ioport_caps);
#endif
rcu_unlock_domain(dom0);
iommu_hwdom_init(d);
return rv;
#else
return 0;
#endif
}
int rcu_lock_target_domain_by_id(domid_t dom, struct domain **d)
{
if ( dom == DOMID_SELF )
{
*d = rcu_lock_current_domain();
return 0;
}
if ( (*d = rcu_lock_domain_by_id(dom)) == NULL )
return -ESRCH;
if ( !IS_PRIV_FOR(current->domain, *d) )
{
rcu_unlock_domain(*d);
return -EPERM;
}
return 0;
}
static long evtchn_alloc_unbound(evtchn_alloc_unbound_t *alloc)
{
struct evtchn *chn;
struct domain *d;
int port;
domid_t dom = alloc->dom;
long rc;
if ( dom == DOMID_SELF )
dom = current->domain->domain_id;
else if ( !IS_PRIV(current->domain) )
return -EPERM;
if ( (d = rcu_lock_domain_by_id(dom)) == NULL )
return -ESRCH;
spin_lock(&d->evtchn_lock);
if ( (port = get_free_port(d)) < 0 )
ERROR_EXIT_DOM(port, d);
chn = evtchn_from_port(d, port);
rc = xsm_evtchn_unbound(d, chn, alloc->remote_dom);
if ( rc )
goto out;
chn->state = ECS_UNBOUND;
if ( (chn->u.unbound.remote_domid = alloc->remote_dom) == DOMID_SELF )
chn->u.unbound.remote_domid = current->domain->domain_id;
alloc->port = port;
out:
spin_unlock(&d->evtchn_lock);
rcu_unlock_domain(d);
return rc;
}
static long evtchn_bind_interdomain(evtchn_bind_interdomain_t *bind)
{
struct evtchn *lchn, *rchn;
struct domain *ld = current->domain, *rd;
int lport, rport = bind->remote_port;
domid_t rdom = bind->remote_dom;
long rc;
if ( rdom == DOMID_SELF )
rdom = current->domain->domain_id;
if ( (rd = rcu_lock_domain_by_id(rdom)) == NULL )
return -ESRCH;
/* Avoid deadlock by first acquiring lock of domain with smaller id. */
if ( ld < rd )
{
spin_lock(&ld->evtchn_lock);
spin_lock(&rd->evtchn_lock);
}
else
{
if ( ld != rd )
spin_lock(&rd->evtchn_lock);
spin_lock(&ld->evtchn_lock);
}
if ( (lport = get_free_port(ld)) < 0 )
ERROR_EXIT(lport);
lchn = evtchn_from_port(ld, lport);
if ( !port_is_valid(rd, rport) )
ERROR_EXIT_DOM(-EINVAL, rd);
rchn = evtchn_from_port(rd, rport);
if ( (rchn->state != ECS_UNBOUND) ||
(rchn->u.unbound.remote_domid != ld->domain_id) )
ERROR_EXIT_DOM(-EINVAL, rd);
rc = xsm_evtchn_interdomain(ld, lchn, rd, rchn);
if ( rc )
goto out;
lchn->u.interdomain.remote_dom = rd;
lchn->u.interdomain.remote_port = (u16)rport;
lchn->state = ECS_INTERDOMAIN;
rchn->u.interdomain.remote_dom = ld;
rchn->u.interdomain.remote_port = (u16)lport;
rchn->state = ECS_INTERDOMAIN;
/*
* We may have lost notifications on the remote unbound port. Fix that up
* here by conservatively always setting a notification on the local port.
*/
evtchn_set_pending(ld->vcpu[lchn->notify_vcpu_id], lport);
bind->local_port = lport;
out:
spin_unlock(&ld->evtchn_lock);
if ( ld != rd )
spin_unlock(&rd->evtchn_lock);
rcu_unlock_domain(rd);
return rc;
}
/*
* Returns 0 if TLB flush / invalidate required by caller.
* va will indicate the address to be invalidated.
*
* addr is _either_ a host virtual address, or the address of the pte to
* update, as indicated by the GNTMAP_contains_pte flag.
*/
static void
__gnttab_map_grant_ref(
struct gnttab_map_grant_ref *op)
{
struct domain *ld, *rd;
struct vcpu *led;
int handle;
unsigned long frame = 0;
int rc = GNTST_okay;
unsigned int cache_flags;
struct active_grant_entry *act;
struct grant_mapping *mt;
grant_entry_t *sha;
union grant_combo scombo, prev_scombo, new_scombo;
/*
* We bound the number of times we retry CMPXCHG on memory locations that
* we share with a guest OS. The reason is that the guest can modify that
* location at a higher rate than we can read-modify-CMPXCHG, so the guest
* could cause us to livelock. There are a few cases where it is valid for
* the guest to race our updates (e.g., to change the GTF_readonly flag),
* so we allow a few retries before failing.
*/
int retries = 0;
led = current;
ld = led->domain;
if ( unlikely((op->flags & (GNTMAP_device_map|GNTMAP_host_map)) == 0) )
{
gdprintk(XENLOG_INFO, "Bad flags in grant map op (%x).\n", op->flags);
op->status = GNTST_bad_gntref;
return;
}
if ( unlikely((rd = rcu_lock_domain_by_id(op->dom)) == NULL) )
{
gdprintk(XENLOG_INFO, "Could not find domain %d\n", op->dom);
op->status = GNTST_bad_domain;
return;
}
rc = xsm_grant_mapref(ld, rd, op->flags);
if ( rc )
{
rcu_unlock_domain(rd);
op->status = GNTST_permission_denied;
return;
}
if ( unlikely((handle = get_maptrack_handle(ld->grant_table)) == -1) )
{
rcu_unlock_domain(rd);
gdprintk(XENLOG_INFO, "Failed to obtain maptrack handle.\n");
op->status = GNTST_no_device_space;
return;
}
spin_lock(&rd->grant_table->lock);
/* Bounds check on the grant ref */
if ( unlikely(op->ref >= nr_grant_entries(rd->grant_table)))
PIN_FAIL(unlock_out, GNTST_bad_gntref, "Bad ref (%d).\n", op->ref);
act = &active_entry(rd->grant_table, op->ref);
sha = &shared_entry(rd->grant_table, op->ref);
/* If already pinned, check the active domid and avoid refcnt overflow. */
if ( act->pin &&
((act->domid != ld->domain_id) ||
(act->pin & 0x80808080U) != 0) )
PIN_FAIL(unlock_out, GNTST_general_error,
"Bad domain (%d != %d), or risk of counter overflow %08x\n",
act->domid, ld->domain_id, act->pin);
if ( !act->pin ||
(!(op->flags & GNTMAP_readonly) &&
!(act->pin & (GNTPIN_hstw_mask|GNTPIN_devw_mask))) )
{
scombo.word = *(u32 *)&sha->flags;
/*
* This loop attempts to set the access (reading/writing) flags
* in the grant table entry. It tries a cmpxchg on the field
* up to five times, and then fails under the assumption that
* the guest is misbehaving.
*/
for ( ; ; )
{
/* If not already pinned, check the grant domid and type. */
if ( !act->pin &&
(((scombo.shorts.flags & GTF_type_mask) !=
GTF_permit_access) ||
(scombo.shorts.domid != ld->domain_id)) )
PIN_FAIL(unlock_out, GNTST_general_error,
"Bad flags (%x) or dom (%d). (expected dom %d)\n",
scombo.shorts.flags, scombo.shorts.domid,
ld->domain_id);
new_scombo = scombo;
new_scombo.shorts.flags |= GTF_reading;
if ( !(op->flags & GNTMAP_readonly) )
{
new_scombo.shorts.flags |= GTF_writing;
if ( unlikely(scombo.shorts.flags & GTF_readonly) )
PIN_FAIL(unlock_out, GNTST_general_error,
"Attempt to write-pin a r/o grant entry.\n");
}
prev_scombo.word = cmpxchg((u32 *)&sha->flags,
scombo.word, new_scombo.word);
if ( likely(prev_scombo.word == scombo.word) )
break;
if ( retries++ == 4 )
PIN_FAIL(unlock_out, GNTST_general_error,
"Shared grant entry is unstable.\n");
scombo = prev_scombo;
}
if ( !act->pin )
{
act->domid = scombo.shorts.domid;
act->frame = gmfn_to_mfn(rd, sha->frame);
}
}
if ( op->flags & GNTMAP_device_map )
act->pin += (op->flags & GNTMAP_readonly) ?
GNTPIN_devr_inc : GNTPIN_devw_inc;
if ( op->flags & GNTMAP_host_map )
act->pin += (op->flags & GNTMAP_readonly) ?
GNTPIN_hstr_inc : GNTPIN_hstw_inc;
frame = act->frame;
cache_flags = (sha->flags & (GTF_PAT | GTF_PWT | GTF_PCD) );
spin_unlock(&rd->grant_table->lock);
if ( is_iomem_page(frame) )
{
if ( !iomem_access_permitted(rd, frame, frame) )
{
gdprintk(XENLOG_WARNING,
"Iomem mapping not permitted %lx (domain %d)\n",
frame, rd->domain_id);
rc = GNTST_general_error;
goto undo_out;
}
rc = create_grant_host_mapping(
op->host_addr, frame, op->flags, cache_flags);
if ( rc != GNTST_okay )
goto undo_out;
}
else
{
if ( unlikely(!mfn_valid(frame)) ||
unlikely(!(gnttab_host_mapping_get_page_type(op, ld, rd) ?
get_page_and_type(mfn_to_page(frame), rd,
PGT_writable_page) :
get_page(mfn_to_page(frame), rd))) )
{
if ( !rd->is_dying )
gdprintk(XENLOG_WARNING, "Could not pin grant frame %lx\n",
frame);
rc = GNTST_general_error;
goto undo_out;
}
if ( op->flags & GNTMAP_host_map )
{
rc = create_grant_host_mapping(op->host_addr, frame, op->flags, 0);
if ( rc != GNTST_okay )
{
if ( gnttab_host_mapping_get_page_type(op, ld, rd) )
put_page_type(mfn_to_page(frame));
put_page(mfn_to_page(frame));
goto undo_out;
}
if ( op->flags & GNTMAP_device_map )
{
(void)get_page(mfn_to_page(frame), rd);
if ( !(op->flags & GNTMAP_readonly) )
get_page_type(mfn_to_page(frame), PGT_writable_page);
}
}
}
TRACE_1D(TRC_MEM_PAGE_GRANT_MAP, op->dom);
mt = &maptrack_entry(ld->grant_table, handle);
mt->domid = op->dom;
mt->ref = op->ref;
mt->flags = op->flags;
op->dev_bus_addr = (u64)frame << PAGE_SHIFT;
op->handle = handle;
op->status = GNTST_okay;
rcu_unlock_domain(rd);
return;
undo_out:
spin_lock(&rd->grant_table->lock);
act = &active_entry(rd->grant_table, op->ref);
sha = &shared_entry(rd->grant_table, op->ref);
if ( op->flags & GNTMAP_device_map )
act->pin -= (op->flags & GNTMAP_readonly) ?
GNTPIN_devr_inc : GNTPIN_devw_inc;
if ( op->flags & GNTMAP_host_map )
act->pin -= (op->flags & GNTMAP_readonly) ?
GNTPIN_hstr_inc : GNTPIN_hstw_inc;
if ( !(op->flags & GNTMAP_readonly) &&
!(act->pin & (GNTPIN_hstw_mask|GNTPIN_devw_mask)) )
gnttab_clear_flag(_GTF_writing, &sha->flags);
if ( !act->pin )
gnttab_clear_flag(_GTF_reading, &sha->flags);
unlock_out:
spin_unlock(&rd->grant_table->lock);
op->status = rc;
put_maptrack_handle(ld->grant_table, handle);
rcu_unlock_domain(rd);
}
static int physdev_map_pirq(struct physdev_map_pirq *map)
{
struct domain *d;
int vector, pirq, ret = 0;
struct msi_info _msi;
void *map_data = NULL;
if ( !IS_PRIV(current->domain) )
return -EPERM;
if ( !map )
return -EINVAL;
if ( map->domid == DOMID_SELF )
d = rcu_lock_domain(current->domain);
else
d = rcu_lock_domain_by_id(map->domid);
if ( d == NULL )
{
ret = -ESRCH;
goto free_domain;
}
/* Verify or get vector. */
switch ( map->type )
{
case MAP_PIRQ_TYPE_GSI:
if ( map->index < 0 || map->index >= NR_IRQS )
{
dprintk(XENLOG_G_ERR, "dom%d: map invalid irq %d\n",
d->domain_id, map->index);
ret = -EINVAL;
goto free_domain;
}
vector = domain_irq_to_vector(current->domain, map->index);
if ( !vector )
{
dprintk(XENLOG_G_ERR, "dom%d: map irq with no vector %d\n",
d->domain_id, vector);
ret = -EINVAL;
goto free_domain;
}
break;
case MAP_PIRQ_TYPE_MSI:
vector = map->index;
if ( vector == -1 )
vector = assign_irq_vector(AUTO_ASSIGN_IRQ);
if ( vector < 0 || vector >= NR_VECTORS )
{
dprintk(XENLOG_G_ERR, "dom%d: map irq with wrong vector %d\n",
d->domain_id, vector);
ret = -EINVAL;
goto free_domain;
}
_msi.bus = map->bus;
_msi.devfn = map->devfn;
_msi.entry_nr = map->entry_nr;
_msi.table_base = map->table_base;
_msi.vector = vector;
map_data = &_msi;
break;
default:
dprintk(XENLOG_G_ERR, "dom%d: wrong map_pirq type %x\n",
d->domain_id, map->type);
ret = -EINVAL;
goto free_domain;
}
spin_lock(&pcidevs_lock);
/* Verify or get pirq. */
spin_lock(&d->event_lock);
pirq = domain_vector_to_irq(d, vector);
if ( map->pirq < 0 )
{
if ( pirq )
{
dprintk(XENLOG_G_ERR, "dom%d: %d:%d already mapped to %d\n",
d->domain_id, map->index, map->pirq,
pirq);
if ( pirq < 0 )
{
ret = -EBUSY;
goto done;
}
}
else
{
pirq = get_free_pirq(d, map->type, map->index);
if ( pirq < 0 )
{
dprintk(XENLOG_G_ERR, "dom%d: no free pirq\n", d->domain_id);
ret = pirq;
goto done;
}
}
}
else
{
if ( pirq && pirq != map->pirq )
{
dprintk(XENLOG_G_ERR, "dom%d: vector %d conflicts with irq %d\n",
d->domain_id, map->index, map->pirq);
ret = -EEXIST;
goto done;
}
else
pirq = map->pirq;
}
ret = map_domain_pirq(d, pirq, vector, map->type, map_data);
if ( ret == 0 )
map->pirq = pirq;
done:
spin_unlock(&d->event_lock);
spin_unlock(&pcidevs_lock);
if ( (ret != 0) && (map->type == MAP_PIRQ_TYPE_MSI) && (map->index == -1) )
free_irq_vector(vector);
free_domain:
rcu_unlock_domain(d);
return ret;
}