[Qemu-devel] [RFC PATCH v0] numa: API to lookup NUMA node by address

[Bharata B Rao <bharata@linux.vnet.ibm.com>]

Keep track of start and end address of each NUMA node in numa_info
structure so that lookup of node by address becomes easier. Add
an API numa_get_node() to lookup a node by address.

This is needed by sPAPR PowerPC to support
ibm,dynamic-reconfiguration-memory device tree node which is needed for
memory hotplug.

Signed-off-by: Bharata B Rao <bharata@linux.vnet.ibm.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
---
This patch was earlier posted as part of sPAPR hotplug patchset here:
https://lists.gnu.org/archive/html/qemu-ppc/2015-04/msg00204.html

 include/sysemu/numa.h |  3 +++
 numa.c                | 61 +++++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 64 insertions(+)

diff --git a/include/sysemu/numa.h b/include/sysemu/numa.h
index 6523b4d..19c0ba3 100644
--- a/include/sysemu/numa.h
+++ b/include/sysemu/numa.h
@@ -15,11 +15,14 @@ typedef struct node_info {
     DECLARE_BITMAP(node_cpu, MAX_CPUMASK_BITS);
     struct HostMemoryBackend *node_memdev;
     bool present;
+    ram_addr_t mem_start;
+    ram_addr_t mem_end;
 } NodeInfo;
 extern NodeInfo numa_info[MAX_NODES];
 void parse_numa_opts(MachineClass *mc);
 void numa_post_machine_init(void);
 void query_numa_node_mem(uint64_t node_mem[]);
 extern QemuOptsList qemu_numa_opts;
+uint32_t numa_get_node(ram_addr_t addr, Error **errp);
 
 #endif
diff --git a/numa.c b/numa.c
index c975fb2..fdf333b 100644
--- a/numa.c
+++ b/numa.c
@@ -53,6 +53,63 @@ static int max_numa_nodeid; /* Highest specified NUMA node ID, plus one.
 int nb_numa_nodes;
 NodeInfo numa_info[MAX_NODES];
 
+/*
+ * Given an address, return the index of the NUMA node to which the
+ * address belongs to.
+ */
+uint32_t numa_get_node(ram_addr_t addr, Error **errp)
+{
+    uint32_t i;
+    MemoryDeviceInfoList *info_list = NULL;
+    MemoryDeviceInfoList **prev = &info_list;
+    MemoryDeviceInfoList *info;
+
+    for (i = 0; i < nb_numa_nodes; i++) {
+        if (addr >= numa_info[i].mem_start && addr < numa_info[i].mem_end) {
+            return i;
+        }
+    }
+
+    /*
+     * If this @addr falls under cold or hotplugged memory regions,
+     * check there too.
+     */
+    qmp_pc_dimm_device_list(qdev_get_machine(), &prev);
+    for (info = info_list; info; info = info->next) {
+        MemoryDeviceInfo *value = info->value;
+
+        if (value) {
+            switch (value->kind) {
+            case MEMORY_DEVICE_INFO_KIND_DIMM:
+                if (addr >= value->dimm->addr &&
+                        addr < (value->dimm->addr + value->dimm->size)) {
+                    qapi_free_MemoryDeviceInfoList(info_list);
+                    return value->dimm->node;
+                }
+                break;
+            default:
+                break;
+            }
+        }
+    }
+    qapi_free_MemoryDeviceInfoList(info_list);
+    error_setg(errp, "Address 0x" RAM_ADDR_FMT " doesn't belong to any "
+                "NUMA node", addr);
+
+    return -1;
+}
+
+static void numa_set_mem_address(int nodenr)
+{
+    if (nodenr) {
+        numa_info[nodenr].mem_start = numa_info[nodenr-1].mem_end;
+    } else {
+        numa_info[nodenr].mem_start = 0;
+    }
+    numa_info[nodenr].mem_end = numa_info[nodenr].mem_start +
+                                   numa_info[nodenr].node_mem;
+}
+
 static void numa_node_parse(NumaNodeOptions *node, QemuOpts *opts, Error **errp)
 {
     uint16_t nodenr;
@@ -276,6 +333,10 @@ void parse_numa_opts(MachineClass *mc)
         }
 
         for (i = 0; i < nb_numa_nodes; i++) {
+            numa_set_mem_address(i);
+        }
+
+        for (i = 0; i < nb_numa_nodes; i++) {
             if (!bitmap_empty(numa_info[i].node_cpu, MAX_CPUMASK_BITS)) {
                 break;
             }
-- 
2.1.0