Pull xpc-disengage into release branch
[pandora-kernel.git] / arch / ia64 / sn / kernel / xpc_partition.c
index 72ef330..581e113 100644 (file)
@@ -47,13 +47,16 @@ static u64 xpc_sh2_IPI_access3;
 u64 xpc_prot_vec[MAX_NUMNODES];
 
 
-/* this partition's reserved page */
+/* this partition's reserved page pointers */
 struct xpc_rsvd_page *xpc_rsvd_page;
-
-/* this partition's XPC variables (within the reserved page) */
+static u64 *xpc_part_nasids;
+static u64 *xpc_mach_nasids;
 struct xpc_vars *xpc_vars;
 struct xpc_vars_part *xpc_vars_part;
 
+static int xp_nasid_mask_bytes;        /* actual size in bytes of nasid mask */
+static int xp_nasid_mask_words;        /* actual size in words of nasid mask */
+
 
 /*
  * For performance reasons, each entry of xpc_partitions[] is cacheline
@@ -65,20 +68,16 @@ struct xpc_partition xpc_partitions[XP_MAX_PARTITIONS + 1];
 
 
 /*
- * Generic buffer used to store a local copy of the remote partitions
- * reserved page or XPC variables.
+ * Generic buffer used to store a local copy of portions of a remote
+ * partition's reserved page (either its header and part_nasids mask,
+ * or its vars).
  *
  * xpc_discovery runs only once and is a seperate thread that is
  * very likely going to be processing in parallel with receiving
  * interrupts.
  */
-char ____cacheline_aligned
-               xpc_remote_copy_buffer[XPC_RSVD_PAGE_ALIGNED_SIZE];
-
-
-/* systune related variables */
-int xpc_hb_interval = XPC_HB_DEFAULT_INTERVAL;
-int xpc_hb_check_interval = XPC_HB_CHECK_DEFAULT_TIMEOUT;
+char ____cacheline_aligned xpc_remote_copy_buffer[XPC_RP_HEADER_SIZE +
+                                                       XP_NASID_MASK_BYTES];
 
 
 /*
@@ -86,13 +85,16 @@ int xpc_hb_check_interval = XPC_HB_CHECK_DEFAULT_TIMEOUT;
  * for that nasid. This function returns 0 on any error.
  */
 static u64
-xpc_get_rsvd_page_pa(int nasid, u64 buf, u64 buf_size)
+xpc_get_rsvd_page_pa(int nasid)
 {
        bte_result_t bte_res;
        s64 status;
        u64 cookie = 0;
        u64 rp_pa = nasid;      /* seed with nasid */
        u64 len = 0;
+       u64 buf = buf;
+       u64 buf_len = 0;
+       void *buf_base = NULL;
 
 
        while (1) {
@@ -108,13 +110,22 @@ xpc_get_rsvd_page_pa(int nasid, u64 buf, u64 buf_size)
                        break;
                }
 
-               if (len > buf_size) {
-                       dev_err(xpc_part, "len (=0x%016lx) > buf_size\n", len);
-                       status = SALRET_ERROR;
-                       break;
+               if (L1_CACHE_ALIGN(len) > buf_len) {
+                       if (buf_base != NULL) {
+                               kfree(buf_base);
+                       }
+                       buf_len = L1_CACHE_ALIGN(len);
+                       buf = (u64) xpc_kmalloc_cacheline_aligned(buf_len,
+                                                       GFP_KERNEL, &buf_base);
+                       if (buf_base == NULL) {
+                               dev_err(xpc_part, "unable to kmalloc "
+                                       "len=0x%016lx\n", buf_len);
+                               status = SALRET_ERROR;
+                               break;
+                       }
                }
 
-               bte_res = xp_bte_copy(rp_pa, ia64_tpa(buf), buf_size,
+               bte_res = xp_bte_copy(rp_pa, ia64_tpa(buf), buf_len,
                                        (BTE_NOTIFY | BTE_WACQUIRE), NULL);
                if (bte_res != BTE_SUCCESS) {
                        dev_dbg(xpc_part, "xp_bte_copy failed %i\n", bte_res);
@@ -123,6 +134,10 @@ xpc_get_rsvd_page_pa(int nasid, u64 buf, u64 buf_size)
                }
        }
 
+       if (buf_base != NULL) {
+               kfree(buf_base);
+       }
+
        if (status != SALRET_OK) {
                rp_pa = 0;
        }
@@ -141,15 +156,15 @@ xpc_rsvd_page_init(void)
 {
        struct xpc_rsvd_page *rp;
        AMO_t *amos_page;
-       u64 rp_pa, next_cl, nasid_array = 0;
+       u64 rp_pa, nasid_array = 0;
        int i, ret;
 
 
        /* get the local reserved page's address */
 
-       rp_pa = xpc_get_rsvd_page_pa(cnodeid_to_nasid(0),
-                                       (u64) xpc_remote_copy_buffer,
-                                               XPC_RSVD_PAGE_ALIGNED_SIZE);
+       preempt_disable();
+       rp_pa = xpc_get_rsvd_page_pa(cpuid_to_nasid(smp_processor_id()));
+       preempt_enable();
        if (rp_pa == 0) {
                dev_err(xpc_part, "SAL failed to locate the reserved page\n");
                return NULL;
@@ -164,12 +179,19 @@ xpc_rsvd_page_init(void)
 
        rp->version = XPC_RP_VERSION;
 
-       /*
-        * Place the XPC variables on the cache line following the
-        * reserved page structure.
-        */
-       next_cl = (u64) rp + XPC_RSVD_PAGE_ALIGNED_SIZE;
-       xpc_vars = (struct xpc_vars *) next_cl;
+       /* establish the actual sizes of the nasid masks */
+       if (rp->SAL_version == 1) {
+               /* SAL_version 1 didn't set the nasids_size field */
+               rp->nasids_size = 128;
+       }
+       xp_nasid_mask_bytes = rp->nasids_size;
+       xp_nasid_mask_words = xp_nasid_mask_bytes / 8;
+
+       /* setup the pointers to the various items in the reserved page */
+       xpc_part_nasids = XPC_RP_PART_NASIDS(rp);
+       xpc_mach_nasids = XPC_RP_MACH_NASIDS(rp);
+       xpc_vars = XPC_RP_VARS(rp);
+       xpc_vars_part = XPC_RP_VARS_PART(rp);
 
        /*
         * Before clearing xpc_vars, see if a page of AMOs had been previously
@@ -221,33 +243,32 @@ xpc_rsvd_page_init(void)
                amos_page = (AMO_t *) TO_AMO((u64) amos_page);
        }
 
+       /* clear xpc_vars */
        memset(xpc_vars, 0, sizeof(struct xpc_vars));
 
-       /*
-        * Place the XPC per partition specific variables on the cache line
-        * following the XPC variables structure.
-        */
-       next_cl += XPC_VARS_ALIGNED_SIZE;
-       memset((u64 *) next_cl, 0, sizeof(struct xpc_vars_part) *
-                                                       XP_MAX_PARTITIONS);
-       xpc_vars_part = (struct xpc_vars_part *) next_cl;
-       xpc_vars->vars_part_pa = __pa(next_cl);
-
        xpc_vars->version = XPC_V_VERSION;
        xpc_vars->act_nasid = cpuid_to_nasid(0);
        xpc_vars->act_phys_cpuid = cpu_physical_id(0);
+       xpc_vars->vars_part_pa = __pa(xpc_vars_part);
+       xpc_vars->amos_page_pa = ia64_tpa((u64) amos_page);
        xpc_vars->amos_page = amos_page;  /* save for next load of XPC */
 
 
-       /*
-        * Initialize the activation related AMO variables.
-        */
-       xpc_vars->act_amos = xpc_IPI_init(XP_MAX_PARTITIONS);
-       for (i = 1; i < XP_NASID_MASK_WORDS; i++) {
-               xpc_IPI_init(i + XP_MAX_PARTITIONS);
+       /* clear xpc_vars_part */
+       memset((u64 *) xpc_vars_part, 0, sizeof(struct xpc_vars_part) *
+                                                       XP_MAX_PARTITIONS);
+
+       /* initialize the activate IRQ related AMO variables */
+       for (i = 0; i < xp_nasid_mask_words; i++) {
+               (void) xpc_IPI_init(XPC_ACTIVATE_IRQ_AMOS + i);
        }
-       /* export AMO page's physical address to other partitions */
-       xpc_vars->amos_page_pa = ia64_tpa((u64) xpc_vars->amos_page);
+
+       /* initialize the engaged remote partitions related AMO variables */
+       (void) xpc_IPI_init(XPC_ENGAGED_PARTITIONS_AMO);
+       (void) xpc_IPI_init(XPC_DISENGAGE_REQUEST_AMO);
+
+       /* timestamp of when reserved page was setup by XPC */
+       rp->stamp = CURRENT_TIME;
 
        /*
         * This signifies to the remote partition that our reserved
@@ -387,6 +408,11 @@ xpc_check_remote_hb(void)
        remote_vars = (struct xpc_vars *) xpc_remote_copy_buffer;
 
        for (partid = 1; partid < XP_MAX_PARTITIONS; partid++) {
+
+               if (xpc_exiting) {
+                       break;
+               }
+
                if (partid == sn_partition_id) {
                        continue;
                }
@@ -401,7 +427,7 @@ xpc_check_remote_hb(void)
                /* pull the remote_hb cache line */
                bres = xp_bte_copy(part->remote_vars_pa,
                                        ia64_tpa((u64) remote_vars),
-                                       XPC_VARS_ALIGNED_SIZE,
+                                       XPC_RP_VARS_SIZE,
                                        (BTE_NOTIFY | BTE_WACQUIRE), NULL);
                if (bres != BTE_SUCCESS) {
                        XPC_DEACTIVATE_PARTITION(part,
@@ -417,7 +443,7 @@ xpc_check_remote_hb(void)
 
                if (((remote_vars->heartbeat == part->last_heartbeat) &&
                        (remote_vars->kdb_status == 0)) ||
-                            !XPC_HB_ALLOWED(sn_partition_id, remote_vars)) {
+                            !xpc_hb_allowed(sn_partition_id, remote_vars)) {
 
                        XPC_DEACTIVATE_PARTITION(part, xpcNoHeartbeat);
                        continue;
@@ -429,31 +455,31 @@ xpc_check_remote_hb(void)
 
 
 /*
- * Get a copy of the remote partition's rsvd page.
+ * Get a copy of a portion of the remote partition's rsvd page.
  *
  * remote_rp points to a buffer that is cacheline aligned for BTE copies and
- * assumed to be of size XPC_RSVD_PAGE_ALIGNED_SIZE.
+ * is large enough to contain a copy of their reserved page header and
+ * part_nasids mask.
  */
 static enum xpc_retval
 xpc_get_remote_rp(int nasid, u64 *discovered_nasids,
-               struct xpc_rsvd_page *remote_rp, u64 *remote_rsvd_page_pa)
+               struct xpc_rsvd_page *remote_rp, u64 *remote_rp_pa)
 {
        int bres, i;
 
 
        /* get the reserved page's physical address */
 
-       *remote_rsvd_page_pa = xpc_get_rsvd_page_pa(nasid, (u64) remote_rp,
-                                               XPC_RSVD_PAGE_ALIGNED_SIZE);
-       if (*remote_rsvd_page_pa == 0) {
+       *remote_rp_pa = xpc_get_rsvd_page_pa(nasid);
+       if (*remote_rp_pa == 0) {
                return xpcNoRsvdPageAddr;
        }
 
 
-       /* pull over the reserved page structure */
+       /* pull over the reserved page header and part_nasids mask */
 
-       bres = xp_bte_copy(*remote_rsvd_page_pa, ia64_tpa((u64) remote_rp),
-                               XPC_RSVD_PAGE_ALIGNED_SIZE,
+       bres = xp_bte_copy(*remote_rp_pa, ia64_tpa((u64) remote_rp),
+                               XPC_RP_HEADER_SIZE + xp_nasid_mask_bytes,
                                (BTE_NOTIFY | BTE_WACQUIRE), NULL);
        if (bres != BTE_SUCCESS) {
                return xpc_map_bte_errors(bres);
@@ -461,8 +487,11 @@ xpc_get_remote_rp(int nasid, u64 *discovered_nasids,
 
 
        if (discovered_nasids != NULL) {
-               for (i = 0; i < XP_NASID_MASK_WORDS; i++) {
-                       discovered_nasids[i] |= remote_rp->part_nasids[i];
+               u64 *remote_part_nasids = XPC_RP_PART_NASIDS(remote_rp);
+
+
+               for (i = 0; i < xp_nasid_mask_words; i++) {
+                       discovered_nasids[i] |= remote_part_nasids[i];
                }
        }
 
@@ -489,10 +518,10 @@ xpc_get_remote_rp(int nasid, u64 *discovered_nasids,
 
 
 /*
- * Get a copy of the remote partition's XPC variables.
+ * Get a copy of the remote partition's XPC variables from the reserved page.
  *
  * remote_vars points to a buffer that is cacheline aligned for BTE copies and
- * assumed to be of size XPC_VARS_ALIGNED_SIZE.
+ * assumed to be of size XPC_RP_VARS_SIZE.
  */
 static enum xpc_retval
 xpc_get_remote_vars(u64 remote_vars_pa, struct xpc_vars *remote_vars)
@@ -508,7 +537,7 @@ xpc_get_remote_vars(u64 remote_vars_pa, struct xpc_vars *remote_vars)
        /* pull over the cross partition variables */
 
        bres = xp_bte_copy(remote_vars_pa, ia64_tpa((u64) remote_vars),
-                               XPC_VARS_ALIGNED_SIZE,
+                               XPC_RP_VARS_SIZE,
                                (BTE_NOTIFY | BTE_WACQUIRE), NULL);
        if (bres != BTE_SUCCESS) {
                return xpc_map_bte_errors(bres);
@@ -524,7 +553,56 @@ xpc_get_remote_vars(u64 remote_vars_pa, struct xpc_vars *remote_vars)
 
 
 /*
- * Prior code has determine the nasid which generated an IPI.  Inspect
+ * Update the remote partition's info.
+ */
+static void
+xpc_update_partition_info(struct xpc_partition *part, u8 remote_rp_version,
+               struct timespec *remote_rp_stamp, u64 remote_rp_pa,
+               u64 remote_vars_pa, struct xpc_vars *remote_vars)
+{
+       part->remote_rp_version = remote_rp_version;
+       dev_dbg(xpc_part, "  remote_rp_version = 0x%016lx\n",
+               part->remote_rp_version);
+
+       part->remote_rp_stamp = *remote_rp_stamp;
+       dev_dbg(xpc_part, "  remote_rp_stamp (tv_sec = 0x%lx tv_nsec = 0x%lx\n",
+               part->remote_rp_stamp.tv_sec, part->remote_rp_stamp.tv_nsec);
+
+       part->remote_rp_pa = remote_rp_pa;
+       dev_dbg(xpc_part, "  remote_rp_pa = 0x%016lx\n", part->remote_rp_pa);
+
+       part->remote_vars_pa = remote_vars_pa;
+       dev_dbg(xpc_part, "  remote_vars_pa = 0x%016lx\n",
+               part->remote_vars_pa);
+
+       part->last_heartbeat = remote_vars->heartbeat;
+       dev_dbg(xpc_part, "  last_heartbeat = 0x%016lx\n",
+               part->last_heartbeat);
+
+       part->remote_vars_part_pa = remote_vars->vars_part_pa;
+       dev_dbg(xpc_part, "  remote_vars_part_pa = 0x%016lx\n",
+               part->remote_vars_part_pa);
+
+       part->remote_act_nasid = remote_vars->act_nasid;
+       dev_dbg(xpc_part, "  remote_act_nasid = 0x%x\n",
+               part->remote_act_nasid);
+
+       part->remote_act_phys_cpuid = remote_vars->act_phys_cpuid;
+       dev_dbg(xpc_part, "  remote_act_phys_cpuid = 0x%x\n",
+               part->remote_act_phys_cpuid);
+
+       part->remote_amos_page_pa = remote_vars->amos_page_pa;
+       dev_dbg(xpc_part, "  remote_amos_page_pa = 0x%lx\n",
+               part->remote_amos_page_pa);
+
+       part->remote_vars_version = remote_vars->version;
+       dev_dbg(xpc_part, "  remote_vars_version = 0x%x\n",
+               part->remote_vars_version);
+}
+
+
+/*
+ * Prior code has determined the nasid which generated an IPI.  Inspect
  * that nasid to determine if its partition needs to be activated or
  * deactivated.
  *
@@ -542,8 +620,12 @@ xpc_identify_act_IRQ_req(int nasid)
 {
        struct xpc_rsvd_page *remote_rp;
        struct xpc_vars *remote_vars;
-       u64 remote_rsvd_page_pa;
+       u64 remote_rp_pa;
        u64 remote_vars_pa;
+       int remote_rp_version;
+       int reactivate = 0;
+       int stamp_diff;
+       struct timespec remote_rp_stamp = { 0, 0 };
        partid_t partid;
        struct xpc_partition *part;
        enum xpc_retval ret;
@@ -553,7 +635,7 @@ xpc_identify_act_IRQ_req(int nasid)
 
        remote_rp = (struct xpc_rsvd_page *) xpc_remote_copy_buffer;
 
-       ret = xpc_get_remote_rp(nasid, NULL, remote_rp, &remote_rsvd_page_pa);
+       ret = xpc_get_remote_rp(nasid, NULL, remote_rp, &remote_rp_pa);
        if (ret != xpcSuccess) {
                dev_warn(xpc_part, "unable to get reserved page from nasid %d, "
                        "which sent interrupt, reason=%d\n", nasid, ret);
@@ -561,6 +643,10 @@ xpc_identify_act_IRQ_req(int nasid)
        }
 
        remote_vars_pa = remote_rp->vars_pa;
+       remote_rp_version = remote_rp->version;
+       if (XPC_SUPPORTS_RP_STAMP(remote_rp_version)) {
+               remote_rp_stamp = remote_rp->stamp;
+       }
        partid = remote_rp->partid;
        part = &xpc_partitions[partid];
 
@@ -586,44 +672,117 @@ xpc_identify_act_IRQ_req(int nasid)
                "%ld:0x%lx\n", (int) nasid, (int) partid, part->act_IRQ_rcvd,
                remote_vars->heartbeat, remote_vars->heartbeating_to_mask);
 
+       if (xpc_partition_disengaged(part) &&
+                                       part->act_state == XPC_P_INACTIVE) {
 
-       if (part->act_state == XPC_P_INACTIVE) {
+               xpc_update_partition_info(part, remote_rp_version,
+                                       &remote_rp_stamp, remote_rp_pa,
+                                       remote_vars_pa, remote_vars);
 
-               part->remote_rp_pa = remote_rsvd_page_pa;
-               dev_dbg(xpc_part, "  remote_rp_pa = 0x%016lx\n",
-                       part->remote_rp_pa);
+               if (XPC_SUPPORTS_DISENGAGE_REQUEST(part->remote_vars_version)) {
+                       if (xpc_partition_disengage_requested(1UL << partid)) {
+                               /*
+                                * Other side is waiting on us to disengage,
+                                * even though we already have.
+                                */
+                               return;
+                       }
+               } else {
+                       /* other side doesn't support disengage requests */
+                       xpc_clear_partition_disengage_request(1UL << partid);
+               }
 
-               part->remote_vars_pa = remote_vars_pa;
-               dev_dbg(xpc_part, "  remote_vars_pa = 0x%016lx\n",
-                       part->remote_vars_pa);
+               xpc_activate_partition(part);
+               return;
+       }
 
-               part->last_heartbeat = remote_vars->heartbeat;
-               dev_dbg(xpc_part, "  last_heartbeat = 0x%016lx\n",
-                       part->last_heartbeat);
+       DBUG_ON(part->remote_rp_version == 0);
+       DBUG_ON(part->remote_vars_version == 0);
+
+       if (!XPC_SUPPORTS_RP_STAMP(part->remote_rp_version)) {
+               DBUG_ON(XPC_SUPPORTS_DISENGAGE_REQUEST(part->
+                                                       remote_vars_version));
+
+               if (!XPC_SUPPORTS_RP_STAMP(remote_rp_version)) {
+                       DBUG_ON(XPC_SUPPORTS_DISENGAGE_REQUEST(remote_vars->
+                                                               version));
+                       /* see if the other side rebooted */
+                       if (part->remote_amos_page_pa ==
+                               remote_vars->amos_page_pa &&
+                                       xpc_hb_allowed(sn_partition_id,
+                                                               remote_vars)) {
+                               /* doesn't look that way, so ignore the IPI */
+                               return;
+                       }
+               }
 
-               part->remote_vars_part_pa = remote_vars->vars_part_pa;
-               dev_dbg(xpc_part, "  remote_vars_part_pa = 0x%016lx\n",
-                       part->remote_vars_part_pa);
+               /*
+                * Other side rebooted and previous XPC didn't support the
+                * disengage request, so we don't need to do anything special.
+                */
 
-               part->remote_act_nasid = remote_vars->act_nasid;
-               dev_dbg(xpc_part, "  remote_act_nasid = 0x%x\n",
-                       part->remote_act_nasid);
+               xpc_update_partition_info(part, remote_rp_version,
+                                               &remote_rp_stamp, remote_rp_pa,
+                                               remote_vars_pa, remote_vars);
+               part->reactivate_nasid = nasid;
+               XPC_DEACTIVATE_PARTITION(part, xpcReactivating);
+               return;
+       }
 
-               part->remote_act_phys_cpuid = remote_vars->act_phys_cpuid;
-               dev_dbg(xpc_part, "  remote_act_phys_cpuid = 0x%x\n",
-                       part->remote_act_phys_cpuid);
+       DBUG_ON(!XPC_SUPPORTS_DISENGAGE_REQUEST(part->remote_vars_version));
 
-               part->remote_amos_page_pa = remote_vars->amos_page_pa;
-               dev_dbg(xpc_part, "  remote_amos_page_pa = 0x%lx\n",
-                       part->remote_amos_page_pa);
+       if (!XPC_SUPPORTS_RP_STAMP(remote_rp_version)) {
+               DBUG_ON(!XPC_SUPPORTS_DISENGAGE_REQUEST(remote_vars->version));
 
-               xpc_activate_partition(part);
+               /*
+                * Other side rebooted and previous XPC did support the
+                * disengage request, but the new one doesn't.
+                */
+
+               xpc_clear_partition_engaged(1UL << partid);
+               xpc_clear_partition_disengage_request(1UL << partid);
 
-       } else if (part->remote_amos_page_pa != remote_vars->amos_page_pa ||
-                       !XPC_HB_ALLOWED(sn_partition_id, remote_vars)) {
+               xpc_update_partition_info(part, remote_rp_version,
+                                               &remote_rp_stamp, remote_rp_pa,
+                                               remote_vars_pa, remote_vars);
+               reactivate = 1;
+
+       } else {
+               DBUG_ON(!XPC_SUPPORTS_DISENGAGE_REQUEST(remote_vars->version));
 
+               stamp_diff = xpc_compare_stamps(&part->remote_rp_stamp,
+                                                       &remote_rp_stamp);
+               if (stamp_diff != 0) {
+                       DBUG_ON(stamp_diff >= 0);
+
+                       /*
+                        * Other side rebooted and the previous XPC did support
+                        * the disengage request, as does the new one.
+                        */
+
+                       DBUG_ON(xpc_partition_engaged(1UL << partid));
+                       DBUG_ON(xpc_partition_disengage_requested(1UL <<
+                                                               partid));
+
+                       xpc_update_partition_info(part, remote_rp_version,
+                                               &remote_rp_stamp, remote_rp_pa,
+                                               remote_vars_pa, remote_vars);
+                       reactivate = 1;
+               }
+       }
+
+       if (!xpc_partition_disengaged(part)) {
+               /* still waiting on other side to disengage from us */
+               return;
+       }
+
+       if (reactivate) {
                part->reactivate_nasid = nasid;
                XPC_DEACTIVATE_PARTITION(part, xpcReactivating);
+
+       } else if (XPC_SUPPORTS_DISENGAGE_REQUEST(part->remote_vars_version) &&
+                       xpc_partition_disengage_requested(1UL << partid)) {
+               XPC_DEACTIVATE_PARTITION(part, xpcOtherGoingDown);
        }
 }
 
@@ -643,14 +802,17 @@ xpc_identify_act_IRQ_sender(void)
        u64 nasid;                      /* remote nasid */
        int n_IRQs_detected = 0;
        AMO_t *act_amos;
-       struct xpc_rsvd_page *rp = (struct xpc_rsvd_page *) xpc_rsvd_page;
 
 
-       act_amos = xpc_vars->act_amos;
+       act_amos = xpc_vars->amos_page + XPC_ACTIVATE_IRQ_AMOS;
 
 
        /* scan through act AMO variable looking for non-zero entries */
-       for (word = 0; word < XP_NASID_MASK_WORDS; word++) {
+       for (word = 0; word < xp_nasid_mask_words; word++) {
+
+               if (xpc_exiting) {
+                       break;
+               }
 
                nasid_mask = xpc_IPI_receive(&act_amos[word]);
                if (nasid_mask == 0) {
@@ -668,7 +830,7 @@ xpc_identify_act_IRQ_sender(void)
                 * remote nasid in our reserved pages machine mask.
                 * This is used in the event of module reload.
                 */
-               rp->mach_nasids[word] |= nasid_mask;
+               xpc_mach_nasids[word] |= nasid_mask;
 
 
                /* locate the nasid(s) which sent interrupts */
@@ -687,6 +849,55 @@ xpc_identify_act_IRQ_sender(void)
 }
 
 
+/*
+ * See if the other side has responded to a partition disengage request
+ * from us.
+ */
+int
+xpc_partition_disengaged(struct xpc_partition *part)
+{
+       partid_t partid = XPC_PARTID(part);
+       int disengaged;
+
+
+       disengaged = (xpc_partition_engaged(1UL << partid) == 0);
+       if (part->disengage_request_timeout) {
+               if (!disengaged) {
+                       if (jiffies < part->disengage_request_timeout) {
+                               /* timelimit hasn't been reached yet */
+                               return 0;
+                       }
+
+                       /*
+                        * Other side hasn't responded to our disengage
+                        * request in a timely fashion, so assume it's dead.
+                        */
+
+                       xpc_clear_partition_engaged(1UL << partid);
+                       disengaged = 1;
+               }
+               part->disengage_request_timeout = 0;
+
+               /* cancel the timer function, provided it's not us */
+               if (!in_interrupt()) {
+                       del_singleshot_timer_sync(&part->
+                                                     disengage_request_timer);
+               }
+
+               DBUG_ON(part->act_state != XPC_P_DEACTIVATING &&
+                                       part->act_state != XPC_P_INACTIVE);
+               if (part->act_state != XPC_P_INACTIVE) {
+                       xpc_wakeup_channel_mgr(part);
+               }
+
+               if (XPC_SUPPORTS_DISENGAGE_REQUEST(part->remote_vars_version)) {
+                       xpc_cancel_partition_disengage_request(part);
+               }
+       }
+       return disengaged;
+}
+
+
 /*
  * Mark specified partition as active.
  */
@@ -721,7 +932,6 @@ xpc_deactivate_partition(const int line, struct xpc_partition *part,
                                enum xpc_retval reason)
 {
        unsigned long irq_flags;
-       partid_t partid = XPC_PARTID(part);
 
 
        spin_lock_irqsave(&part->act_lock, irq_flags);
@@ -749,17 +959,27 @@ xpc_deactivate_partition(const int line, struct xpc_partition *part,
 
        spin_unlock_irqrestore(&part->act_lock, irq_flags);
 
-       XPC_DISALLOW_HB(partid, xpc_vars);
+       if (XPC_SUPPORTS_DISENGAGE_REQUEST(part->remote_vars_version)) {
+               xpc_request_partition_disengage(part);
+               xpc_IPI_send_disengage(part);
 
-       dev_dbg(xpc_part, "bringing partition %d down, reason = %d\n", partid,
-               reason);
+               /* set a timelimit on the disengage request */
+               part->disengage_request_timeout = jiffies +
+                                       (xpc_disengage_request_timelimit * HZ);
+               part->disengage_request_timer.expires =
+                                       part->disengage_request_timeout;
+               add_timer(&part->disengage_request_timer);
+       }
+
+       dev_dbg(xpc_part, "bringing partition %d down, reason = %d\n",
+               XPC_PARTID(part), reason);
 
-       xpc_partition_down(part, reason);
+       xpc_partition_going_down(part, reason);
 }
 
 
 /*
- * Mark specified partition as active.
+ * Mark specified partition as inactive.
  */
 void
 xpc_mark_partition_inactive(struct xpc_partition *part)
@@ -792,9 +1012,10 @@ xpc_discovery(void)
        void *remote_rp_base;
        struct xpc_rsvd_page *remote_rp;
        struct xpc_vars *remote_vars;
-       u64 remote_rsvd_page_pa;
+       u64 remote_rp_pa;
        u64 remote_vars_pa;
        int region;
+       int region_size;
        int max_regions;
        int nasid;
        struct xpc_rsvd_page *rp;
@@ -804,7 +1025,8 @@ xpc_discovery(void)
        enum xpc_retval ret;
 
 
-       remote_rp = xpc_kmalloc_cacheline_aligned(XPC_RSVD_PAGE_ALIGNED_SIZE,
+       remote_rp = xpc_kmalloc_cacheline_aligned(XPC_RP_HEADER_SIZE +
+                                               xp_nasid_mask_bytes,
                                                GFP_KERNEL, &remote_rp_base);
        if (remote_rp == NULL) {
                return;
@@ -812,13 +1034,13 @@ xpc_discovery(void)
        remote_vars = (struct xpc_vars *) remote_rp;
 
 
-       discovered_nasids = kmalloc(sizeof(u64) * XP_NASID_MASK_WORDS,
+       discovered_nasids = kmalloc(sizeof(u64) * xp_nasid_mask_words,
                                                        GFP_KERNEL);
        if (discovered_nasids == NULL) {
                kfree(remote_rp_base);
                return;
        }
-       memset(discovered_nasids, 0, sizeof(u64) * XP_NASID_MASK_WORDS);
+       memset(discovered_nasids, 0, sizeof(u64) * xp_nasid_mask_words);
 
        rp = (struct xpc_rsvd_page *) xpc_rsvd_page;
 
@@ -827,11 +1049,19 @@ xpc_discovery(void)
         * nodes that can comprise an access protection grouping. The access
         * protection is in regards to memory, IOI and IPI.
         */
-//>>> move the next two #defines into either include/asm-ia64/sn/arch.h or
-//>>> include/asm-ia64/sn/addrs.h
-#define SH1_MAX_REGIONS                64
-#define SH2_MAX_REGIONS                256
-       max_regions = is_shub2() ? SH2_MAX_REGIONS : SH1_MAX_REGIONS;
+       max_regions = 64;
+       region_size = sn_region_size;
+
+       switch (region_size) {
+       case 128:
+               max_regions *= 2;
+       case 64:
+               max_regions *= 2;
+       case 32:
+               max_regions *= 2;
+               region_size = 16;
+               DBUG_ON(!is_shub2());
+       }
 
        for (region = 0; region < max_regions; region++) {
 
@@ -841,8 +1071,8 @@ xpc_discovery(void)
 
                dev_dbg(xpc_part, "searching region %d\n", region);
 
-               for (nasid = (region * sn_region_size * 2);
-                    nasid < ((region + 1) * sn_region_size * 2);
+               for (nasid = (region * region_size * 2);
+                    nasid < ((region + 1) * region_size * 2);
                     nasid += 2) {
 
                        if ((volatile int) xpc_exiting) {
@@ -852,14 +1082,14 @@ xpc_discovery(void)
                        dev_dbg(xpc_part, "checking nasid %d\n", nasid);
 
 
-                       if (XPC_NASID_IN_ARRAY(nasid, rp->part_nasids)) {
+                       if (XPC_NASID_IN_ARRAY(nasid, xpc_part_nasids)) {
                                dev_dbg(xpc_part, "PROM indicates Nasid %d is "
                                        "part of the local partition; skipping "
                                        "region\n", nasid);
                                break;
                        }
 
-                       if (!(XPC_NASID_IN_ARRAY(nasid, rp->mach_nasids))) {
+                       if (!(XPC_NASID_IN_ARRAY(nasid, xpc_mach_nasids))) {
                                dev_dbg(xpc_part, "PROM indicates Nasid %d was "
                                        "not on Numa-Link network at reset\n",
                                        nasid);
@@ -877,7 +1107,7 @@ xpc_discovery(void)
                        /* pull over the reserved page structure */
 
                        ret = xpc_get_remote_rp(nasid, discovered_nasids,
-                                             remote_rp, &remote_rsvd_page_pa);
+                                             remote_rp, &remote_rp_pa);
                        if (ret != xpcSuccess) {
                                dev_dbg(xpc_part, "unable to get reserved page "
                                        "from nasid %d, reason=%d\n", nasid,
@@ -948,6 +1178,13 @@ xpc_discovery(void)
                                remote_vars->act_nasid,
                                remote_vars->act_phys_cpuid);
 
+                       if (XPC_SUPPORTS_DISENGAGE_REQUEST(remote_vars->
+                                                               version)) {
+                               part->remote_amos_page_pa =
+                                               remote_vars->amos_page_pa;
+                               xpc_mark_partition_disengaged(part);
+                               xpc_cancel_partition_disengage_request(part);
+                       }
                        xpc_IPI_send_activate(remote_vars);
                }
        }
@@ -974,12 +1211,12 @@ xpc_initiate_partid_to_nasids(partid_t partid, void *nasid_mask)
                return xpcPartitionDown;
        }
 
-       part_nasid_pa = part->remote_rp_pa +
-               (u64) &((struct xpc_rsvd_page *) 0)->part_nasids;
+       memset(nasid_mask, 0, XP_NASID_MASK_BYTES);
+
+       part_nasid_pa = (u64) XPC_RP_PART_NASIDS(part->remote_rp_pa);
 
        bte_res = xp_bte_copy(part_nasid_pa, ia64_tpa((u64) nasid_mask),
-                               L1_CACHE_ALIGN(XP_NASID_MASK_BYTES),
-                               (BTE_NOTIFY | BTE_WACQUIRE), NULL);
+                       xp_nasid_mask_bytes, (BTE_NOTIFY | BTE_WACQUIRE), NULL);
 
        return xpc_map_bte_errors(bte_res);
 }