[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [PATCH v3 04/15] argo: init, destroy and soft-reset, with enable command line opt



On 01/07/2019 02:42 AM, Christopher Clark wrote:
> Initialises basic data structures and performs teardown of argo state
> for domain shutdown.
> 
> Inclusion of the Argo implementation is dependent on CONFIG_ARGO.
> 
> Introduces a new Xen command line parameter 'argo': bool to enable/disable
> the argo hypercall. Defaults to disabled.
> 
> New headers:
>   public/argo.h: with definions of addresses and ring structure, including
>   indexes for atomic update for communication between domain and hypervisor.
> 
>   xen/argo.h: to expose the hooks for integration into domain lifecycle:
>     argo_init: per-domain init of argo data structures for domain_create.
>     argo_destroy: teardown for domain_destroy and the error exit
>                   path of domain_create.
>     argo_soft_reset: reset of domain state for domain_soft_reset.
> 
> Adds two new fields to struct domain:
>     rwlock_t argo_lock;
>     struct argo_domain *argo;
> 
> In accordance with recent work on _domain_destroy, argo_destroy is
> idempotent. It will tear down: all rings registered by this domain, all
> rings where this domain is the single sender (ie. specified partner,
> non-wildcard rings), and all pending notifications where this domain is
> awaiting signal about available space in the rings of other domains.
> 
> A count will be maintained of the number of rings that a domain has
> registered in order to limit it below the fixed maximum limit defined here.
> 
> The software license on the public header is the BSD license, standard
> procedure for the public Xen headers. The public header was originally
> posted under a GPL license at: [1]:
> https://lists.xenproject.org/archives/html/xen-devel/2013-05/msg02710.html
> 
> The following ACK by Lars Kurth is to confirm that only people being
> employees of Citrix contributed to the header files in the series posted at
> [1] and that thus the copyright of the files in question is fully owned by
> Citrix. The ACK also confirms that Citrix is happy for the header files to
> be published under a BSD license in this series (which is based on [1]).
> 
> Signed-off-by: Christopher Clark <christopher.clark6@xxxxxxxxxxxxxx>
> Acked-by: Lars Kurth <lars.kurth@xxxxxxxxxx>

Other than an indentation issue in domain_rings_remove_all, this LGTM.

Reviewed-by: Ross Philipson <ross.philipson@xxxxxxxxxx>

> ---
> v2 rewrite locking explanation comment
> v2 header copyright line now includes 2019
> v2 self: use ring_info backpointer in pending_ent to maintain npending
> v2 self: rename all_rings_remove_info to domain_rings_remove_all
> v2 feedback Jan: drop cookie, implement teardown
> v2 self: add npending to track number of pending entries per ring
> v2 self: amend comment on locking; drop section comments
> v2 cookie_eq: test low bits first and use likely on high bits
> v2 self: OVERHAUL
> v2 self: s/argo_pending_ent/pending_ent/g
> v2 self: drop pending_remove_ent, inline at single call site
> v1 feedback Roger, Jan: drop argo prefix on static functions
> v2 #4 Lars: add Acked-by and details to commit message.
> v2 feedback #9 Jan: document argo boot opt in xen-command-line.markdown
> v2 bugfix: xsm use in soft-reset prior to introduction
> v2 feedback #9 Jan: drop 'message' from do_argo_message_op
> v1 #5 feedback Paul: init/destroy unsigned, brackets and whitespace fixes
> v1 #5 feedback Paul: Use mfn_eq for comparing mfns.
> v1 #5 feedback Paul: init/destroy : use currd
> v1 #6 (#5) feedback Jan: init/destroy: s/ENOSYS/EOPNOTSUPP/
> v1 #6 feedback Paul: Folded patch 6 into patch 5.
> v1 #6 feedback Jan: drop opt_argo_enabled initializer
> v1 $6 feedback Jan: s/ENOSYS/EOPNOTSUPP/g and drop useless dprintk
> v1. #5 feedback Paul: change the license on public header to BSD
> - ack from Lars at Citrix.
> v1. self, Jan: drop unnecessary xen include from sched.h
> v1. self, Jan: drop inclusion of public argo.h in private one
> v1. self, Jan: add include of public argo.h to argo.c
> v1. self, Jan: drop fwd decl of argo_domain in priv header
> v1. Paul/self/Jan: add data structures to xlat.lst and compat/argo.h to 
> Makefile
> v1. self: removed allocation of event channel since switching to VIRQ
> v1. self: drop types.h include from private argo.h
> v1: reorder public argo include position
> v1: #13 feedback Jan: public namespace: prefix with xen
> v1: self: rename pending ent "id" to "domain_id"
> v1: self: add domain_cookie to ent struct
> v1. #15 feedback Jan: make cmd unsigned
> v1. #15 feedback Jan: make i loop variable unsigned
> v1: self: adjust dprintks in init, destroy
> v1: #18 feedback Jan: meld max ring count limit
> v1: self: use type not struct in public defn, affects compat gen header
> v1: feedback #15 Jan: handle upper-halves of hypercall args
> v1: add comment explaining the 'magic' field
> v1: self + Jan feedback: implement soft reset
> v1: feedback #13 Roger: use ASSERT_UNREACHABLE
> 
>  docs/misc/xen-command-line.pandoc |  11 +
>  xen/common/argo.c                 | 461 
> +++++++++++++++++++++++++++++++++++++-
>  xen/common/domain.c               |  20 ++
>  xen/include/Makefile              |   1 +
>  xen/include/public/argo.h         |  59 +++++
>  xen/include/xen/argo.h            |  23 ++
>  xen/include/xen/sched.h           |   6 +
>  xen/include/xlat.lst              |   2 +
>  8 files changed, 582 insertions(+), 1 deletion(-)
>  create mode 100644 xen/include/public/argo.h
>  create mode 100644 xen/include/xen/argo.h
> 
> diff --git a/docs/misc/xen-command-line.pandoc 
> b/docs/misc/xen-command-line.pandoc
> index a755a67..aea13eb 100644
> --- a/docs/misc/xen-command-line.pandoc
> +++ b/docs/misc/xen-command-line.pandoc
> @@ -182,6 +182,17 @@ Permit Xen to use "Always Running APIC Timer" support on 
> compatible hardware
>  in combination with cpuidle.  This option is only expected to be useful for
>  developers wishing Xen to fall back to older timing methods on newer 
> hardware.
>  
> +### argo
> +> `= <boolean>`
> +
> +> Default: `false`
> +
> +Enable the Argo hypervisor-mediated interdomain communication mechanism.
> +
> +This allows domains access to the Argo hypercall, which supports registration
> +of memory rings with the hypervisor to receive messages, sending messages to
> +other domains by hypercall and querying the ring status of other domains.
> +
>  ### asid (x86)
>  > `= <boolean>`
>  
> diff --git a/xen/common/argo.c b/xen/common/argo.c
> index 6f782f7..86195d3 100644
> --- a/xen/common/argo.c
> +++ b/xen/common/argo.c
> @@ -17,7 +17,177 @@
>   */
>  
>  #include <xen/errno.h>
> +#include <xen/sched.h>
> +#include <xen/domain.h>
> +#include <xen/argo.h>
> +#include <xen/event.h>
> +#include <xen/domain_page.h>
>  #include <xen/guest_access.h>
> +#include <xen/time.h>
> +#include <public/argo.h>
> +
> +DEFINE_XEN_GUEST_HANDLE(xen_argo_addr_t);
> +DEFINE_XEN_GUEST_HANDLE(xen_argo_ring_t);
> +
> +/* Xen command line option to enable argo */
> +static bool __read_mostly opt_argo_enabled;
> +boolean_param("argo", opt_argo_enabled);
> +
> +typedef struct argo_ring_id
> +{
> +    uint32_t port;
> +    domid_t partner_id;
> +    domid_t domain_id;
> +} argo_ring_id;
> +
> +/* Data about a domain's own ring that it has registered */
> +struct argo_ring_info
> +{
> +    /* next node in the hash, protected by L2 */
> +    struct hlist_node node;
> +    /* this ring's id, protected by L2 */
> +    struct argo_ring_id id;
> +    /* L3 */
> +    spinlock_t lock;
> +    /* length of the ring, protected by L3 */
> +    uint32_t len;
> +    /* number of pages in the ring, protected by L3 */
> +    uint32_t npage;
> +    /* number of pages translated into mfns, protected by L3 */
> +    uint32_t nmfns;
> +    /* cached tx pointer location, protected by L3 */
> +    uint32_t tx_ptr;
> +    /* mapped ring pages protected by L3 */
> +    uint8_t **mfn_mapping;
> +    /* list of mfns of guest ring, protected by L3 */
> +    mfn_t *mfns;
> +    /* list of struct pending_ent for this ring, protected by L3 */
> +    struct hlist_head pending;
> +    /* number of pending entries queued for this ring, protected by L3 */
> +    uint32_t npending;
> +};
> +
> +/* Data about a single-sender ring, held by the sender (partner) domain */
> +struct argo_send_info
> +{
> +    /* next node in the hash, protected by Lsend */
> +    struct hlist_node node;
> +    /* this ring's id, protected by Lsend */
> +    struct argo_ring_id id;
> +};
> +
> +/* A space-available notification that is awaiting sufficient space */
> +struct pending_ent
> +{
> +    /* List node within argo_ring_info's pending list */
> +    struct hlist_node node;
> +    /*
> +     * List node within argo_domain's wildcard_pend_list. Only used if the
> +     * ring is one with a wildcard partner (ie. that any domain may send to)
> +     * to enable cancelling signals on wildcard rings on domain destroy.
> +     */
> +    struct hlist_node wildcard_node;
> +    /*
> +     * Pointer to the ring_info that this ent pertains to. Used to ensure 
> that
> +     * ring_info->npending is decremented when ents for wildcard rings are
> +     * cancelled for domain destroy.
> +     * Caution: Must hold the correct locks before accessing ring_info via 
> this.
> +     */
> +    struct argo_ring_info *ring_info;
> +    /* domain to be notified when space is available */
> +    domid_t domain_id;
> +    uint16_t pad;
> +    /* minimum ring space available that this signal is waiting upon */
> +    uint32_t len;
> +};
> +
> +/*
> + * The value of the argo element in a struct domain is
> + * protected by the global lock argo_lock: L1
> + */
> +#define ARGO_HTABLE_SIZE 32
> +struct argo_domain
> +{
> +    /* L2 */
> +    rwlock_t lock;
> +    /*
> +     * Hash table of argo_ring_info about rings this domain has registered.
> +     * Protected by L2.
> +     */
> +    struct hlist_head ring_hash[ARGO_HTABLE_SIZE];
> +    /* Counter of rings registered by this domain. Protected by L2. */
> +    uint32_t ring_count;
> +
> +    /* Lsend */
> +    spinlock_t send_lock;
> +    /*
> +     * Hash table of argo_send_info about rings other domains have registered
> +     * for this domain to send to. Single partner, non-wildcard rings.
> +     * Protected by Lsend.
> +     */
> +    struct hlist_head send_hash[ARGO_HTABLE_SIZE];
> +
> +    /* Lwildcard */
> +    spinlock_t wildcard_lock;
> +    /*
> +     * List of pending space-available signals for this domain about wildcard
> +     * rings registered by other domains. Protected by Lwildcard.
> +     */
> +    struct hlist_head wildcard_pend_list;
> +};
> +
> +/*
> + * Locking is organized as follows:
> + *
> + * Terminology: R(<lock>) means taking a read lock on the specified lock;
> + *              W(<lock>) means taking a write lock on it.
> + *
> + * L1 : The global lock: argo_lock
> + * Protects the argo elements of all struct domain *d in the system.
> + * It does not protect any of the elements of d->argo, only their
> + * addresses.
> + *
> + * By extension since the destruction of a domain with a non-NULL
> + * d->argo will need to free the d->argo pointer, holding W(L1)
> + * guarantees that no domains pointers that argo is interested in
> + * become invalid whilst this lock is held.
> + */
> +
> +static DEFINE_RWLOCK(argo_lock); /* L1 */
> +
> +/*
> + * L2 : The per-domain ring hash lock: d->argo->lock
> + * Holding a read lock on L2 protects the ring hash table and
> + * the elements in the hash_table d->argo->ring_hash, and
> + * the node and id fields in struct argo_ring_info in the
> + * hash table.
> + * Holding a write lock on L2 protects all of the elements of
> + * struct argo_ring_info.
> + *
> + * To take L2 you must already have R(L1). W(L1) implies W(L2) and L3.
> + *
> + * L3 : The ringinfo lock: argo_ring_info *ringinfo; ringinfo->lock
> + * Protects all the fields within the argo_ring_info, aside from the ones 
> that
> + * L2 already protects: node, id, lock.
> + *
> + * To aquire L3 you must already have R(L2). W(L2) implies L3.
> + *
> + * Lsend : The per-domain single-sender partner rings lock: 
> d->argo->send_lock
> + * Protects the per-domain send hash table : d->argo->send_hash
> + * and the elements in the hash table, and the node and id fields
> + * in struct argo_send_info in the hash table.
> + *
> + * To take Lsend, you must already have R(L1). W(L1) implies Lsend.
> + * Do not attempt to acquire a L2 on any domain after taking and while
> + * holding a Lsend lock -- acquire the L2 (if one is needed) beforehand.
> + *
> + * Lwildcard : The per-domain wildcard pending list lock: 
> d->argo->wildcard_lock
> + * Protects the per-domain list of outstanding signals for space availability
> + * on wildcard rings.
> + *
> + * To take Lwildcard, you must already have R(L1). W(L1) implies Lwildcard.
> + * No other locks are acquired after obtaining Lwildcard.
> + */
>  
>  /* Change this to #define ARGO_DEBUG here to enable more debug messages */
>  #undef ARGO_DEBUG
> @@ -28,10 +198,299 @@
>  #define argo_dprintk(format, ... ) ((void)0)
>  #endif
>  
> +static void
> +ring_unmap(struct argo_ring_info *ring_info)
> +{
> +    unsigned int i;
> +
> +    if ( !ring_info->mfn_mapping )
> +        return;
> +
> +    for ( i = 0; i < ring_info->nmfns; i++ )
> +    {
> +        if ( !ring_info->mfn_mapping[i] )
> +            continue;
> +        if ( ring_info->mfns )
> +            argo_dprintk(XENLOG_ERR "argo: unmapping page %"PRI_mfn" from 
> %p\n",
> +                         mfn_x(ring_info->mfns[i]),
> +                         ring_info->mfn_mapping[i]);
> +        unmap_domain_page_global(ring_info->mfn_mapping[i]);
> +        ring_info->mfn_mapping[i] = NULL;
> +    }
> +}
> +
> +static void
> +wildcard_pending_list_remove(domid_t domain_id, struct pending_ent *ent)
> +{
> +    struct domain *d = get_domain_by_id(domain_id);
> +    if ( !d )
> +        return;
> +
> +    if ( d->argo )
> +    {
> +        spin_lock(&d->argo->wildcard_lock);
> +        hlist_del(&ent->wildcard_node);
> +        spin_unlock(&d->argo->wildcard_lock);
> +    }
> +    put_domain(d);
> +}
> +
> +static void
> +pending_remove_all(struct argo_ring_info *ring_info)
> +{
> +    struct hlist_node *node, *next;
> +    struct pending_ent *ent;
> +
> +    hlist_for_each_entry_safe(ent, node, next, &ring_info->pending, node)
> +    {
> +        if ( ring_info->id.partner_id == XEN_ARGO_DOMID_ANY )
> +            wildcard_pending_list_remove(ent->domain_id, ent);
> +        hlist_del(&ent->node);
> +        xfree(ent);
> +    }
> +    ring_info->npending = 0;
> +}
> +
> +static void
> +wildcard_rings_pending_remove(struct domain *d)
> +{
> +    struct hlist_node *node, *next;
> +    struct pending_ent *ent;
> +
> +    ASSERT(rw_is_write_locked(&argo_lock));
> +
> +    hlist_for_each_entry_safe(ent, node, next, &d->argo->wildcard_pend_list,
> +                              node)
> +    {
> +        hlist_del(&ent->node);
> +        ent->ring_info->npending--;
> +        hlist_del(&ent->wildcard_node);
> +        xfree(ent);
> +    }
> +}
> +
> +static void
> +ring_remove_mfns(const struct domain *d, struct argo_ring_info *ring_info)
> +{
> +    unsigned int i;
> +
> +    ASSERT(rw_is_write_locked(&d->argo->lock) ||
> +           rw_is_write_locked(&argo_lock));
> +
> +    if ( !ring_info->mfns )
> +        return;
> +
> +    if ( !ring_info->mfn_mapping )
> +    {
> +        ASSERT_UNREACHABLE();
> +        return;
> +    }
> +
> +    ring_unmap(ring_info);
> +
> +    for ( i = 0; i < ring_info->nmfns; i++ )
> +        if ( !mfn_eq(ring_info->mfns[i], INVALID_MFN) )
> +            put_page_and_type(mfn_to_page(ring_info->mfns[i]));
> +
> +    xfree(ring_info->mfns);
> +    ring_info->mfns = NULL;
> +    ring_info->npage = 0;
> +    xfree(ring_info->mfn_mapping);
> +    ring_info->mfn_mapping = NULL;
> +    ring_info->nmfns = 0;
> +}
> +
> +static void
> +ring_remove_info(struct domain *d, struct argo_ring_info *ring_info)
> +{
> +    ASSERT(rw_is_write_locked(&d->argo->lock) ||
> +           rw_is_write_locked(&argo_lock));
> +
> +    pending_remove_all(ring_info);
> +    hlist_del(&ring_info->node);
> +    ring_remove_mfns(d, ring_info);
> +    xfree(ring_info);
> +}
> +
> +static void
> +domain_rings_remove_all(struct domain *d)
> +{
> +    unsigned int i;
> +
> +    for ( i = 0; i < ARGO_HTABLE_SIZE; ++i )
> +    {
> +        struct hlist_node *node, *next;
> +        struct argo_ring_info *ring_info;
> +
> +        hlist_for_each_entry_safe(ring_info, node, next,
> +                                  &d->argo->ring_hash[i], node)
> +            ring_remove_info(d, ring_info);
> +    }
> +    d->argo->ring_count = 0;
> +}
> +
> +/*
> + * Tear down all rings of other domains where src_d domain is the partner.
> + * (ie. it is the single domain that can send to those rings.)
> + * This will also cancel any pending notifications about those rings.
> + */
> +static void
> +partner_rings_remove(struct domain *src_d)
> +{
> +    unsigned int i;
> +
> +    ASSERT(rw_is_write_locked(&argo_lock));
> +
> +    for ( i = 0; i < ARGO_HTABLE_SIZE; ++i )
> +    {
> +        struct hlist_node *node, *next;
> +        struct argo_send_info *send_info;
> +
> +        hlist_for_each_entry_safe(send_info, node, next,
> +                                  &src_d->argo->send_hash[i], node)
> +        {
> +            struct argo_ring_info *ring_info;
> +            struct domain *dst_d;
> +
> +            dst_d = get_domain_by_id(send_info->id.domain_id);
> +            if ( dst_d )
> +            {
> +                ring_info = ring_find_info(dst_d, &send_info->id);
> +                if ( ring_info )
> +                {
> +                    ring_remove_info(dst_d, ring_info);
> +                    dst_d->argo->ring_count--;
> +                }
> +
> +                put_domain(dst_d);
> +            }
> +
> +            hlist_del(&send_info->node);
> +            xfree(send_info);
> +        }
> +    }
> +}
> +
>  long
>  do_argo_op(unsigned int cmd, XEN_GUEST_HANDLE_PARAM(void) arg1,
>             XEN_GUEST_HANDLE_PARAM(void) arg2, unsigned long arg3,
>             unsigned long arg4)
>  {
> -    return -ENOSYS;
> +    struct domain *currd = current->domain;
> +    long rc = -EFAULT;
> +
> +    argo_dprintk("->do_argo_op(%u,%p,%p,%d,%d)\n", cmd,
> +                 (void *)arg1.p, (void *)arg2.p, (int) arg3, (int) arg4);
> +
> +    if ( unlikely(!opt_argo_enabled) )
> +    {
> +        rc = -EOPNOTSUPP;
> +        return rc;
> +    }
> +
> +    domain_lock(currd);
> +
> +    switch (cmd)
> +    {
> +    default:
> +        rc = -EOPNOTSUPP;
> +        break;
> +    }
> +
> +    domain_unlock(currd);
> +
> +    argo_dprintk("<-do_argo_op(%u)=%ld\n", cmd, rc);
> +
> +    return rc;
> +}
> +
> +static void
> +argo_domain_init(struct argo_domain *argo)
> +{
> +    unsigned int i;
> +
> +    rwlock_init(&argo->lock);
> +    spin_lock_init(&argo->send_lock);
> +    spin_lock_init(&argo->wildcard_lock);
> +    argo->ring_count = 0;
> +
> +    for ( i = 0; i < ARGO_HTABLE_SIZE; ++i )
> +    {
> +        INIT_HLIST_HEAD(&argo->ring_hash[i]);
> +        INIT_HLIST_HEAD(&argo->send_hash[i]);
> +    }
> +    INIT_HLIST_HEAD(&argo->wildcard_pend_list);
> +}
> +
> +int
> +argo_init(struct domain *d)
> +{
> +    struct argo_domain *argo;
> +
> +    if ( !opt_argo_enabled )
> +    {
> +        argo_dprintk("argo disabled, domid: %d\n", d->domain_id);
> +        return 0;
> +    }
> +
> +    argo_dprintk("init: domid: %d\n", d->domain_id);
> +
> +    argo = xmalloc(struct argo_domain);
> +    if ( !argo )
> +        return -ENOMEM;
> +
> +    write_lock(&argo_lock);
> +
> +    argo_domain_init(argo);
> +
> +    d->argo = argo;
> +
> +    write_unlock(&argo_lock);
> +
> +    return 0;
> +}
> +
> +void
> +argo_destroy(struct domain *d)
> +{
> +    BUG_ON(!d->is_dying);
> +
> +    write_lock(&argo_lock);
> +
> +    argo_dprintk("destroy: domid %d d->argo=%p\n", d->domain_id, d->argo);
> +
> +    if ( d->argo )
> +    {
> +        domain_rings_remove_all(d);
> +        partner_rings_remove(d);
> +        wildcard_rings_pending_remove(d);
> +        xfree(d->argo);
> +        d->argo = NULL;
> +    }
> +    write_unlock(&argo_lock);
> +}
> +
> +void
> +argo_soft_reset(struct domain *d)
> +{
> +    write_lock(&argo_lock);
> +
> +    argo_dprintk("soft reset d=%d d->argo=%p\n", d->domain_id, d->argo);
> +
> +    if ( d->argo )
> +    {
> +        domain_rings_remove_all(d);
> +        partner_rings_remove(d);
> +        wildcard_rings_pending_remove(d);
> +
> +        if ( !opt_argo_enabled )
> +        {
> +            xfree(d->argo);
> +            d->argo = NULL;
> +        }
> +        else
> +            argo_domain_init(d->argo);
> +    }
> +
> +    write_unlock(&argo_lock);
>  }
> diff --git a/xen/common/domain.c b/xen/common/domain.c
> index c623dae..9596840 100644
> --- a/xen/common/domain.c
> +++ b/xen/common/domain.c
> @@ -32,6 +32,7 @@
>  #include <xen/grant_table.h>
>  #include <xen/xenoprof.h>
>  #include <xen/irq.h>
> +#include <xen/argo.h>
>  #include <asm/debugger.h>
>  #include <asm/p2m.h>
>  #include <asm/processor.h>
> @@ -277,6 +278,10 @@ static void _domain_destroy(struct domain *d)
>  
>      xfree(d->pbuf);
>  
> +#ifdef CONFIG_ARGO
> +    argo_destroy(d);
> +#endif
> +
>      rangeset_domain_destroy(d);
>  
>      free_cpumask_var(d->dirty_cpumask);
> @@ -376,6 +381,9 @@ struct domain *domain_create(domid_t domid,
>      spin_lock_init(&d->hypercall_deadlock_mutex);
>      INIT_PAGE_LIST_HEAD(&d->page_list);
>      INIT_PAGE_LIST_HEAD(&d->xenpage_list);
> +#ifdef CONFIG_ARGO
> +    rwlock_init(&d->argo_lock);
> +#endif
>  
>      spin_lock_init(&d->node_affinity_lock);
>      d->node_affinity = NODE_MASK_ALL;
> @@ -445,6 +453,11 @@ struct domain *domain_create(domid_t domid,
>              goto fail;
>          init_status |= INIT_gnttab;
>  
> +#ifdef CONFIG_ARGO
> +        if ( (err = argo_init(d)) != 0 )
> +            goto fail;
> +#endif
> +
>          err = -ENOMEM;
>  
>          d->pbuf = xzalloc_array(char, DOMAIN_PBUF_SIZE);
> @@ -717,6 +730,9 @@ int domain_kill(struct domain *d)
>          if ( d->is_dying != DOMDYING_alive )
>              return domain_kill(d);
>          d->is_dying = DOMDYING_dying;
> +#ifdef CONFIG_ARGO
> +        argo_destroy(d);
> +#endif
>          evtchn_destroy(d);
>          gnttab_release_mappings(d);
>          tmem_destroy(d->tmem_client);
> @@ -1175,6 +1191,10 @@ int domain_soft_reset(struct domain *d)
>  
>      grant_table_warn_active_grants(d);
>  
> +#ifdef CONFIG_ARGO
> +    argo_soft_reset(d);
> +#endif
> +
>      for_each_vcpu ( d, v )
>      {
>          set_xen_guest_handle(runstate_guest(v), NULL);
> diff --git a/xen/include/Makefile b/xen/include/Makefile
> index f7895e4..3d14532 100644
> --- a/xen/include/Makefile
> +++ b/xen/include/Makefile
> @@ -5,6 +5,7 @@ ifneq ($(CONFIG_COMPAT),)
>  compat-arch-$(CONFIG_X86) := x86_32
>  
>  headers-y := \
> +    compat/argo.h \
>      compat/callback.h \
>      compat/elfnote.h \
>      compat/event_channel.h \
> diff --git a/xen/include/public/argo.h b/xen/include/public/argo.h
> new file mode 100644
> index 0000000..4818684
> --- /dev/null
> +++ b/xen/include/public/argo.h
> @@ -0,0 +1,59 @@
> +/******************************************************************************
> + * Argo : Hypervisor-Mediated data eXchange
> + *
> + * Derived from v4v, the version 2 of v2v.
> + *
> + * Copyright (c) 2010, Citrix Systems
> + * Copyright (c) 2018-2019, BAE Systems
> + *
> + * Permission is hereby granted, free of charge, to any person obtaining a 
> copy
> + * of this software and associated documentation files (the "Software"), to
> + * deal in the Software without restriction, including without limitation the
> + * rights to use, copy, modify, merge, publish, distribute, sublicense, 
> and/or
> + * sell copies of the Software, and to permit persons to whom the Software is
> + * furnished to do so, subject to the following conditions:
> + *
> + * The above copyright notice and this permission notice shall be included in
> + * all copies or substantial portions of the Software.
> + *
> + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
> + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
> + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL 
> THE
> + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
> + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
> + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
> + * DEALINGS IN THE SOFTWARE.
> + *
> + */
> +
> +#ifndef __XEN_PUBLIC_ARGO_H__
> +#define __XEN_PUBLIC_ARGO_H__
> +
> +#include "xen.h"
> +
> +typedef struct xen_argo_addr
> +{
> +    uint32_t port;
> +    domid_t domain_id;
> +    uint16_t pad;
> +} xen_argo_addr_t;
> +
> +typedef struct xen_argo_ring
> +{
> +    /* Guests should use atomic operations to access rx_ptr */
> +    uint32_t rx_ptr;
> +    /* Guests should use atomic operations to access tx_ptr */
> +    uint32_t tx_ptr;
> +    /*
> +     * Header space reserved for later use. Align the start of the ring to a
> +     * multiple of the message slot size.
> +     */
> +    uint8_t reserved[56];
> +#if defined(__STDC_VERSION__) && __STDC_VERSION__ >= 199901L
> +    uint8_t ring[];
> +#elif defined(__GNUC__)
> +    uint8_t ring[0];
> +#endif
> +} xen_argo_ring_t;
> +
> +#endif
> diff --git a/xen/include/xen/argo.h b/xen/include/xen/argo.h
> new file mode 100644
> index 0000000..29d32a9
> --- /dev/null
> +++ b/xen/include/xen/argo.h
> @@ -0,0 +1,23 @@
> +/******************************************************************************
> + * Argo : Hypervisor-Mediated data eXchange
> + *
> + * Copyright (c) 2018, BAE Systems
> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> + * GNU General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public License
> + * along with this program; if not, write to the Free Software
> + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
> + */
> +
> +#ifndef __XEN_ARGO_H__
> +#define __XEN_ARGO_H__
> +
> +int argo_init(struct domain *d);
> +void argo_destroy(struct domain *d);
> +void argo_soft_reset(struct domain *d);
> +
> +#endif
> diff --git a/xen/include/xen/sched.h b/xen/include/xen/sched.h
> index 4956a77..20418e7 100644
> --- a/xen/include/xen/sched.h
> +++ b/xen/include/xen/sched.h
> @@ -490,6 +490,12 @@ struct domain
>          unsigned int guest_request_enabled       : 1;
>          unsigned int guest_request_sync          : 1;
>      } monitor;
> +
> +#ifdef CONFIG_ARGO
> +    /* Argo interdomain communication support */
> +    rwlock_t argo_lock;
> +    struct argo_domain *argo;
> +#endif
>  };
>  
>  /* Protect updates/reads (resp.) of domain_list and domain_hash. */
> diff --git a/xen/include/xlat.lst b/xen/include/xlat.lst
> index 5273320..9f616e4 100644
> --- a/xen/include/xlat.lst
> +++ b/xen/include/xlat.lst
> @@ -148,3 +148,5 @@
>  ?    flask_setenforce                xsm/flask_op.h
>  !    flask_sid_context               xsm/flask_op.h
>  ?    flask_transition                xsm/flask_op.h
> +?    argo_addr                       argo.h
> +?    argo_ring                       argo.h
> 


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel

 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.