X-Git-Url: https://git.proxmox.com/?a=blobdiff_plain;f=numa.c;h=3875e1efda3aafb67ce0427cac714bb1bead6f11;hb=2fa23277d58ce7ec527541b3baf52894ded530cc;hp=7151b24d1c827f096d4b683e514b03d190e24b64;hpb=fec035a53fa15c4c8c4e62bfef56a35df4161e38;p=mirror_qemu.git

diff --git a/numa.c b/numa.c
index 7151b24d1c..3875e1efda 100644
--- a/numa.c
+++ b/numa.c
@@ -29,13 +29,14 @@
 #include "qemu/bitmap.h"
 #include "qom/cpu.h"
 #include "qemu/error-report.h"
-#include "include/exec/cpu-common.h" /* for RAM_ADDR_FMT */
-#include "qapi-visit.h"
+#include "qapi/error.h"
 #include "qapi/opts-visitor.h"
+#include "qapi/qapi-commands-misc.h"
+#include "qapi/qapi-visit-misc.h"
 #include "hw/boards.h"
 #include "sysemu/hostmem.h"
-#include "qmp-commands.h"
 #include "hw/mem/pc-dimm.h"
+#include "hw/mem/memory-device.h"
 #include "qemu/option.h"
 #include "qemu/config-file.h"
 #include "qemu/cutils.h"
@@ -55,96 +56,11 @@ int nb_numa_nodes;
 bool have_numa_distance;
 NodeInfo numa_info[MAX_NODES];
 
-void numa_set_mem_node_id(ram_addr_t addr, uint64_t size, uint32_t node)
-{
-    struct numa_addr_range *range;
-
-    /*
-     * Memory-less nodes can come here with 0 size in which case,
-     * there is nothing to do.
-     */
-    if (!size) {
-        return;
-    }
-
-    range = g_malloc0(sizeof(*range));
-    range->mem_start = addr;
-    range->mem_end = addr + size - 1;
-    QLIST_INSERT_HEAD(&numa_info[node].addr, range, entry);
-}
-
-void numa_unset_mem_node_id(ram_addr_t addr, uint64_t size, uint32_t node)
-{
-    struct numa_addr_range *range, *next;
-
-    QLIST_FOREACH_SAFE(range, &numa_info[node].addr, entry, next) {
-        if (addr == range->mem_start && (addr + size - 1) == range->mem_end) {
-            QLIST_REMOVE(range, entry);
-            g_free(range);
-            return;
-        }
-    }
-}
-
-static void numa_set_mem_ranges(void)
-{
-    int i;
-    ram_addr_t mem_start = 0;
-
-    /*
-     * Deduce start address of each node and use it to store
-     * the address range info in numa_info address range list
-     */
-    for (i = 0; i < nb_numa_nodes; i++) {
-        numa_set_mem_node_id(mem_start, numa_info[i].node_mem, i);
-        mem_start += numa_info[i].node_mem;
-    }
-}
-
-/*
- * Check if @addr falls under NUMA @node.
- */
-static bool numa_addr_belongs_to_node(ram_addr_t addr, uint32_t node)
-{
-    struct numa_addr_range *range;
-
-    QLIST_FOREACH(range, &numa_info[node].addr, entry) {
-        if (addr >= range->mem_start && addr <= range->mem_end) {
-            return true;
-        }
-    }
-    return false;
-}
-
-/*
- * 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;
-
-    /* For non NUMA configurations, check if the addr falls under node 0 */
-    if (!nb_numa_nodes) {
-        if (numa_addr_belongs_to_node(addr, 0)) {
-            return 0;
-        }
-    }
-
-    for (i = 0; i < nb_numa_nodes; i++) {
-        if (numa_addr_belongs_to_node(addr, i)) {
-            return i;
-        }
-    }
-
-    error_setg(errp, "Address 0x" RAM_ADDR_FMT " doesn't belong to any "
-                "NUMA node", addr);
-    return -1;
-}
 
 static void parse_numa_node(MachineState *ms, NumaNodeOptions *node,
                             Error **errp)
 {
+    Error *err = NULL;
     uint16_t nodenr;
     uint16List *cpus = NULL;
     MachineClass *mc = MACHINE_GET_CLASS(ms);
@@ -166,9 +82,9 @@ static void parse_numa_node(MachineState *ms, NumaNodeOptions *node,
         return;
     }
 
-    if (!mc->cpu_index_to_instance_props) {
-        error_report("NUMA is not supported by this machine-type");
-        exit(1);
+    if (!mc->cpu_index_to_instance_props || !mc->get_default_cpu_node_id) {
+        error_setg(errp, "NUMA is not supported by this machine-type");
+        return;
     }
     for (cpus = node->cpus; cpus; cpus = cpus->next) {
         CpuInstanceProperties props;
@@ -182,7 +98,11 @@ static void parse_numa_node(MachineState *ms, NumaNodeOptions *node,
         props = mc->cpu_index_to_instance_props(ms, cpus->value);
         props.node_id = nodenr;
         props.has_node_id = true;
-        machine_set_cpu_numa_node(ms, &props, &error_fatal);
+        machine_set_cpu_numa_node(ms, &props, &err);
+        if (err) {
+            error_propagate(errp, err);
+            return;
+        }
     }
 
     if (node->has_mem && node->has_memdev) {
@@ -226,9 +146,8 @@ static void parse_numa_distance(NumaDistOptions *dist, Error **errp)
     uint8_t val = dist->val;
 
     if (src >= MAX_NODES || dst >= MAX_NODES) {
-        error_setg(errp,
-                   "Invalid node %d, max possible could be %d",
-                   MAX(src, dst), MAX_NODES);
+        error_setg(errp, "Parameter '%s' expects an integer between 0 and %d",
+                   src >= MAX_NODES ? "src" : "dst", MAX_NODES - 1);
         return;
     }
 
@@ -255,28 +174,11 @@ static void parse_numa_distance(NumaDistOptions *dist, Error **errp)
     have_numa_distance = true;
 }
 
-static int parse_numa(void *opaque, QemuOpts *opts, Error **errp)
+static
+void set_numa_options(MachineState *ms, NumaOptions *object, Error **errp)
 {
-    NumaOptions *object = NULL;
-    MachineState *ms = opaque;
     Error *err = NULL;
 
-    {
-        Visitor *v = opts_visitor_new(opts);
-        visit_type_NumaOptions(v, NULL, &object, &err);
-        visit_free(v);
-    }
-
-    if (err) {
-        goto end;
-    }
-
-    /* Fix up legacy suffix-less format */
-    if ((object->type == NUMA_OPTIONS_TYPE_NODE) && object->u.node.has_mem) {
-        const char *mem_str = qemu_opt_get(opts, "mem");
-        qemu_strtosz_MiB(mem_str, NULL, &object->u.node.mem);
-    }
-
     switch (object->type) {
     case NUMA_OPTIONS_TYPE_NODE:
         parse_numa_node(ms, &object->u.node, &err);
@@ -309,10 +211,35 @@ static int parse_numa(void *opaque, QemuOpts *opts, Error **errp)
         abort();
     }
 
+end:
+    error_propagate(errp, err);
+}
+
+static int parse_numa(void *opaque, QemuOpts *opts, Error **errp)
+{
+    NumaOptions *object = NULL;
+    MachineState *ms = MACHINE(opaque);
+    Error *err = NULL;
+    Visitor *v = opts_visitor_new(opts);
+
+    visit_type_NumaOptions(v, NULL, &object, &err);
+    visit_free(v);
+    if (err) {
+        goto end;
+    }
+
+    /* Fix up legacy suffix-less format */
+    if ((object->type == NUMA_OPTIONS_TYPE_NODE) && object->u.node.has_mem) {
+        const char *mem_str = qemu_opt_get(opts, "mem");
+        qemu_strtosz_MiB(mem_str, NULL, &object->u.node.mem);
+    }
+
+    set_numa_options(ms, object, &err);
+
 end:
     qapi_free_NumaOptions(object);
     if (err) {
-        error_report_err(err);
+        error_propagate(errp, err);
         return -1;
     }
 
@@ -424,15 +351,11 @@ void numa_default_auto_assign_ram(MachineClass *mc, NodeInfo *nodes,
     nodes[i].node_mem = size - usedmem;
 }
 
-void parse_numa_opts(MachineState *ms)
+void numa_complete_configuration(MachineState *ms)
 {
     int i;
     MachineClass *mc = MACHINE_GET_CLASS(ms);
 
-    if (qemu_opts_foreach(qemu_find_opts("numa"), parse_numa, ms, NULL)) {
-        exit(1);
-    }
-
     /*
      * If memory hotplug is enabled (slots > 0) but without '-numa'
      * options explicitly on CLI, guestes will break.
@@ -449,7 +372,7 @@ void parse_numa_opts(MachineState *ms)
     if (ms->ram_slots > 0 && nb_numa_nodes == 0 &&
         mc->auto_enable_numa_with_memhp) {
             NumaNodeOptions node = { };
-            parse_numa_node(ms, &node, NULL);
+            parse_numa_node(ms, &node, &error_abort);
     }
 
     assert(max_numa_nodeid <= MAX_NODES);
@@ -497,12 +420,6 @@ void parse_numa_opts(MachineState *ms)
             exit(1);
         }
 
-        for (i = 0; i < nb_numa_nodes; i++) {
-            QLIST_INIT(&numa_info[i].addr);
-        }
-
-        numa_set_mem_ranges();
-
         /* QEMU needs at least all unique node pair distances to build
          * the whole NUMA distance table. QEMU treats the distance table
          * as symmetric by default, i.e. distance A->B == distance B->A.
@@ -522,11 +439,25 @@ void parse_numa_opts(MachineState *ms)
             /* Validation succeeded, now fill in any missing distances. */
             complete_init_numa_distance();
         }
-    } else {
-        numa_set_mem_node_id(0, ram_size, 0);
     }
 }
 
+void parse_numa_opts(MachineState *ms)
+{
+    qemu_opts_foreach(qemu_find_opts("numa"), parse_numa, ms, &error_fatal);
+}
+
+void qmp_set_numa_node(NumaOptions *cmd, Error **errp)
+{
+    if (!runstate_check(RUN_STATE_PRECONFIG)) {
+        error_setg(errp, "The command is permitted only in '%s' state",
+                   RunState_str(RUN_STATE_PRECONFIG));
+         return;
+    }
+
+    set_numa_options(MACHINE(qdev_get_machine()), cmd, errp);
+}
+
 void numa_cpu_pre_plug(const CPUArchId *slot, DeviceState *dev, Error **errp)
 {
     int node_id = object_property_get_int(OBJECT(dev), "node-id", &error_abort);
@@ -551,17 +482,19 @@ static void allocate_system_memory_nonnuma(MemoryRegion *mr, Object *owner,
     if (mem_path) {
 #ifdef __linux__
         Error *err = NULL;
-        memory_region_init_ram_from_file(mr, owner, name, ram_size, false,
+        memory_region_init_ram_from_file(mr, owner, name, ram_size, 0, 0,
                                          mem_path, &err);
         if (err) {
             error_report_err(err);
             if (mem_prealloc) {
                 exit(1);
             }
+            error_report("falling back to regular RAM allocation.");
 
             /* Legacy behavior: if allocation failed, fall back to
              * regular RAM allocation.
              */
+            mem_path = NULL;
             memory_region_init_ram_nomigrate(mr, owner, name, ram_size, &error_fatal);
         }
 #else
@@ -593,14 +526,14 @@ void memory_region_allocate_system_memory(MemoryRegion *mr, Object *owner,
         if (!backend) {
             continue;
         }
-        MemoryRegion *seg = host_memory_backend_get_memory(backend,
-                                                           &error_fatal);
+        MemoryRegion *seg = host_memory_backend_get_memory(backend);
 
         if (memory_region_is_mapped(seg)) {
             char *path = object_get_canonical_path_component(OBJECT(backend));
             error_report("memory backend %s is used multiple times. Each "
                          "-numa option must use a different memdev value.",
                          path);
+            g_free(path);
             exit(1);
         }
 
@@ -613,30 +546,33 @@ void memory_region_allocate_system_memory(MemoryRegion *mr, Object *owner,
 
 static void numa_stat_memory_devices(NumaNodeMem node_mem[])
 {
-    MemoryDeviceInfoList *info_list = NULL;
-    MemoryDeviceInfoList **prev = &info_list;
+    MemoryDeviceInfoList *info_list = qmp_memory_device_list();
     MemoryDeviceInfoList *info;
     PCDIMMDeviceInfo     *pcdimm_info;
 
-    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->type) {
-            case MEMORY_DEVICE_INFO_KIND_DIMM: {
+            case MEMORY_DEVICE_INFO_KIND_DIMM:
                 pcdimm_info = value->u.dimm.data;
-                node_mem[pcdimm_info->node].node_mem += pcdimm_info->size;
-                if (pcdimm_info->hotpluggable && pcdimm_info->hotplugged) {
-                    node_mem[pcdimm_info->node].node_plugged_mem +=
-                        pcdimm_info->size;
-                }
                 break;
-            }
+
+            case MEMORY_DEVICE_INFO_KIND_NVDIMM:
+                pcdimm_info = value->u.nvdimm.data;
+                break;
 
             default:
+                pcdimm_info = NULL;
                 break;
             }
+
+            if (pcdimm_info) {
+                node_mem[pcdimm_info->node].node_mem += pcdimm_info->size;
+                node_mem[pcdimm_info->node].node_plugged_mem +=
+                    pcdimm_info->size;
+            }
         }
     }
     qapi_free_MemoryDeviceInfoList(info_list);
@@ -666,7 +602,7 @@ static int query_memdev(Object *obj, void *opaque)
 
         m->value = g_malloc0(sizeof(*m->value));
 
-        m->value->id = object_property_get_str(obj, "id", NULL);
+        m->value->id = object_get_canonical_path_component(obj);
         m->value->has_id = !!m->value->id;
 
         m->value->size = object_property_get_uint(obj, "size",