[POWERPC] Support ibm,dynamic-reconfiguration-memory nodes

For PAPR partitions with large amounts of memory, the firmware has an
alternative, more compact representation for the information about the
memory in the partition and its NUMA associativity information.  This
adds the code to the kernel to parse this alternative representation.

The other part of this patch is telling the firmware that we can
handle the alternative representation.  There is however a subtlety
here, because the firmware will invoke a reboot if the memory
representation we request is different from the representation that
firmware is currently using.  This is because firmware can't change
the representation on the fly.  Further, some firmware versions used
on POWER5+ machines have a bug where this reboot leaves the machine
with an altered value of load-base, which will prevent any kernel
booting until it is reset to the normal value (0x4000).  Because of
this bug, we do NOT set fake_elf.rpanote.new_mem_def = 1, and thus we
do not request the new representation on POWER5+ and earlier machines.
We do request the new representation on POWER6, which uses the
ibm,client-architecture-support call.

Signed-off-by: Paul Mackerras <paulus@samba.org>

3 files changed, 121 insertions, 1 deletions

diff --git a/arch/powerpc/kernel/prom.c b/arch/powerpc/kernel/prom.c
index c18dbe77fdc2..1fc732a552db 100644
--- a/arch/powerpc/kernel/prom.c
+++ b/arch/powerpc/kernel/prom.c
@@ -804,6 +804,56 @@ static unsigned long __init dt_mem_next_cell(int s, cell_t **cellp)
         return of_read_ulong(p, s);
 }
 
+#ifdef CONFIG_PPC_PSERIES
+/*
+ * Interpret the ibm,dynamic-memory property in the
+ * /ibm,dynamic-reconfiguration-memory node.
+ * This contains a list of memory blocks along with NUMA affinity
+ * information.
+ */
+static int __init early_init_dt_scan_drconf_memory(unsigned long node)
+{
+        cell_t *dm, *ls;
+        unsigned long l, n;
+        unsigned long base, size, lmb_size, flags;
+
+        ls = (cell_t *)of_get_flat_dt_prop(node, "ibm,lmb-size", &l);
+        if (ls == NULL || l < dt_root_size_cells * sizeof(cell_t))
+                return 0;
+        lmb_size = dt_mem_next_cell(dt_root_size_cells, &ls);
+
+        dm = (cell_t *)of_get_flat_dt_prop(node, "ibm,dynamic-memory", &l);
+        if (dm == NULL || l < sizeof(cell_t))
+                return 0;
+
+        n = *dm++;      /* number of entries */
+        if (l < (n * (dt_root_addr_cells + 4) + 1) * sizeof(cell_t))
+                return 0;
+
+        for (; n != 0; --n) {
+                base = dt_mem_next_cell(dt_root_addr_cells, &dm);
+                flags = dm[3];
+                /* skip DRC index, pad, assoc. list index, flags */
+                dm += 4;
+                /* skip this block if the reserved bit is set in flags (0x80)
+                   or if the block is not assigned to this partition (0x8) */
+                if ((flags & 0x80) || !(flags & 0x8))
+                        continue;
+                size = lmb_size;
+                if (iommu_is_off) {
+                        if (base >= 0x80000000ul)
+                                continue;
+                        if ((base + size) > 0x80000000ul)
+                                size = 0x80000000ul - base;
+                }
+                lmb_add(base, size);
+        }
+        lmb_dump_all();
+        return 0;
+}
+#else
+#define early_init_dt_scan_drconf_memory(node)  0
+#endif /* CONFIG_PPC_PSERIES */
 
 static int __init early_init_dt_scan_memory(unsigned long node,
                                             const char *uname, int depth, void *data)
@@ -812,6 +862,11 @@ static int __init early_init_dt_scan_memory(unsigned long node,
         cell_t *reg, *endp;
         unsigned long l;
 
+        /* Look for the ibm,dynamic-reconfiguration-memory node */
+        if (depth == 1 &&
+            strcmp(uname, "ibm,dynamic-reconfiguration-memory") == 0)
+                return early_init_dt_scan_drconf_memory(node);
+
         /* We are scanning "memory" nodes only */
         if (type == NULL) {
                 /*
diff --git a/arch/powerpc/kernel/prom_init.c b/arch/powerpc/kernel/prom_init.c
index 46cf32670ddb..520ef42f642e 100644
--- a/arch/powerpc/kernel/prom_init.c
+++ b/arch/powerpc/kernel/prom_init.c
@@ -679,7 +679,7 @@ static unsigned char ibm_architecture_vec[] = {
         /* option vector 5: PAPR/OF options */
         3 - 2,                          /* length */
         0,                              /* don't ignore, don't halt */
-        OV5_LPAR | OV5_SPLPAR | OV5_LARGE_PAGES,
+        OV5_LPAR | OV5_SPLPAR | OV5_LARGE_PAGES | OV5_DRCONF_MEMORY,
 };
 
 /* Old method - ELF header with PT_NOTE sections */
diff --git a/arch/powerpc/mm/numa.c b/arch/powerpc/mm/numa.c
index 9da01dc8cfd9..262790910ff2 100644
--- a/arch/powerpc/mm/numa.c
+++ b/arch/powerpc/mm/numa.c
@@ -295,6 +295,63 @@ static unsigned long __init numa_enforce_memory_limit(unsigned long start,
         return lmb_end_of_DRAM() - start;
 }
 
+/*
+ * Extract NUMA information from the ibm,dynamic-reconfiguration-memory
+ * node.  This assumes n_mem_{addr,size}_cells have been set.
+ */
+static void __init parse_drconf_memory(struct device_node *memory)
+{
+        const unsigned int *lm, *dm, *aa;
+        unsigned int ls, ld, la;
+        unsigned int n, aam, aalen;
+        unsigned long lmb_size, size;
+        int nid, default_nid = 0;
+        unsigned int start, ai, flags;
+
+        lm = get_property(memory, "ibm,lmb-size", &ls);
+        dm = get_property(memory, "ibm,dynamic-memory", &ld);
+        aa = get_property(memory, "ibm,associativity-lookup-arrays", &la);
+        if (!lm || !dm || !aa ||
+            ls < sizeof(unsigned int) || ld < sizeof(unsigned int) ||
+            la < 2 * sizeof(unsigned int))
+                return;
+
+        lmb_size = read_n_cells(n_mem_size_cells, &lm);
+        n = *dm++;              /* number of LMBs */
+        aam = *aa++;            /* number of associativity lists */
+        aalen = *aa++;          /* length of each associativity list */
+        if (ld < (n * (n_mem_addr_cells + 4) + 1) * sizeof(unsigned int) ||
+            la < (aam * aalen + 2) * sizeof(unsigned int))
+                return;
+
+        for (; n != 0; --n) {
+                start = read_n_cells(n_mem_addr_cells, &dm);
+                ai = dm[2];
+                flags = dm[3];
+                dm += 4;
+                /* 0x80 == reserved, 0x8 = assigned to us */
+                if ((flags & 0x80) || !(flags & 0x8))
+                        continue;
+                nid = default_nid;
+                /* flags & 0x40 means associativity index is invalid */
+                if (min_common_depth > 0 && min_common_depth <= aalen &&
+                    (flags & 0x40) == 0 && ai < aam) {
+                        /* this is like of_node_to_nid_single */
+                        nid = aa[ai * aalen + min_common_depth - 1];
+                        if (nid == 0xffff || nid >= MAX_NUMNODES)
+                                nid = default_nid;
+                }
+                node_set_online(nid);
+
+                size = numa_enforce_memory_limit(start, lmb_size);
+                if (!size)
+                        continue;
+
+                add_active_range(nid, start >> PAGE_SHIFT,
+                                 (start >> PAGE_SHIFT) + (size >> PAGE_SHIFT));
+        }
+}
+
 static int __init parse_numa_properties(void)
 {
         struct device_node *cpu = NULL;
@@ -385,6 +442,14 @@ new_range:
                         goto new_range;
         }
 
+        /*
+         * Now do the same thing for each LMB listed in the ibm,dynamic-memory
+         * property in the ibm,dynamic-reconfiguration-memory node.
+         */
+        memory = of_find_node_by_path("/ibm,dynamic-reconfiguration-memory");
+        if (memory)
+                parse_drconf_memory(memory);
+
         return 0;
 }