Imported Upstream version 2.3.1upstream/2.3.1

Change-Id: I2161522ea1d7ff10cd1d697609d473243c05e1df

1 files changed, 86 insertions, 19 deletions

diff --git a/numa.c b/numa.c
index afd28666b..c975fb268 100644
--- a/numa.c
+++ b/numa.c
@@ -22,7 +22,7 @@
  * THE SOFTWARE.
  */
 
-#include "sysemu/sysemu.h"
+#include "sysemu/numa.h"
 #include "exec/cpu-common.h"
 #include "qemu/bitmap.h"
 #include "qom/cpu.h"
@@ -36,6 +36,8 @@
 #include "sysemu/hostmem.h"
 #include "qmp-commands.h"
 #include "hw/mem/pc-dimm.h"
+#include "qemu/option.h"
+#include "qemu/config-file.h"
 
 QemuOptsList qemu_numa_opts = {
     .name = "numa",
@@ -45,6 +47,11 @@ QemuOptsList qemu_numa_opts = {
 };
 
 static int have_memdevs = -1;
+static int max_numa_nodeid; /* Highest specified NUMA node ID, plus one.
+                             * For all nodes, nodeid < max_numa_nodeid
+                             */
+int nb_numa_nodes;
+NodeInfo numa_info[MAX_NODES];
 
 static void numa_node_parse(NumaNodeOptions *node, QemuOpts *opts, Error **errp)
 {
@@ -59,7 +66,7 @@ static void numa_node_parse(NumaNodeOptions *node, QemuOpts *opts, Error **errp)
 
     if (nodenr >= MAX_NODES) {
         error_setg(errp, "Max number of NUMA nodes reached: %"
-                   PRIu16 "\n", nodenr);
+                   PRIu16 "", nodenr);
         return;
     }
 
@@ -69,16 +76,18 @@ static void numa_node_parse(NumaNodeOptions *node, QemuOpts *opts, Error **errp)
     }
 
     for (cpus = node->cpus; cpus; cpus = cpus->next) {
-        if (cpus->value > MAX_CPUMASK_BITS) {
-            error_setg(errp, "CPU number %" PRIu16 " is bigger than %d",
-                       cpus->value, MAX_CPUMASK_BITS);
+        if (cpus->value >= max_cpus) {
+            error_setg(errp,
+                       "CPU index (%" PRIu16 ")"
+                       " should be smaller than maxcpus (%d)",
+                       cpus->value, max_cpus);
             return;
         }
         bitmap_set(numa_info[nodenr].node_cpu, cpus->value, 1);
     }
 
     if (node->has_mem && node->has_memdev) {
-        error_setg(errp, "qemu: cannot specify both mem= and memdev=\n");
+        error_setg(errp, "qemu: cannot specify both mem= and memdev=");
         return;
     }
 
@@ -87,7 +96,7 @@ static void numa_node_parse(NumaNodeOptions *node, QemuOpts *opts, Error **errp)
     }
     if (node->has_memdev != have_memdevs) {
         error_setg(errp, "qemu: memdev option must be specified for either "
-                   "all or no nodes\n");
+                   "all or no nodes");
         return;
     }
 
@@ -116,7 +125,7 @@ static void numa_node_parse(NumaNodeOptions *node, QemuOpts *opts, Error **errp)
     max_numa_nodeid = MAX(max_numa_nodeid, nodenr + 1);
 }
 
-int numa_init_func(QemuOpts *opts, void *opaque)
+static int parse_numa(QemuOpts *opts, void *opaque)
 {
     NumaOptions *object = NULL;
     Error *err = NULL;
@@ -146,8 +155,7 @@ int numa_init_func(QemuOpts *opts, void *opaque)
     return 0;
 
 error:
-    qerror_report_err(err);
-    error_free(err);
+    error_report_err(err);
 
     if (object) {
         QapiDeallocVisitor *dv = qapi_dealloc_visitor_new();
@@ -159,9 +167,59 @@ error:
     return -1;
 }
 
-void set_numa_nodes(void)
+static char *enumerate_cpus(unsigned long *cpus, int max_cpus)
+{
+    int cpu;
+    bool first = true;
+    GString *s = g_string_new(NULL);
+
+    for (cpu = find_first_bit(cpus, max_cpus);
+        cpu < max_cpus;
+        cpu = find_next_bit(cpus, max_cpus, cpu + 1)) {
+        g_string_append_printf(s, "%s%d", first ? "" : " ", cpu);
+        first = false;
+    }
+    return g_string_free(s, FALSE);
+}
+
+static void validate_numa_cpus(void)
 {
     int i;
+    DECLARE_BITMAP(seen_cpus, MAX_CPUMASK_BITS);
+
+    bitmap_zero(seen_cpus, MAX_CPUMASK_BITS);
+    for (i = 0; i < nb_numa_nodes; i++) {
+        if (bitmap_intersects(seen_cpus, numa_info[i].node_cpu,
+                              MAX_CPUMASK_BITS)) {
+            bitmap_and(seen_cpus, seen_cpus,
+                       numa_info[i].node_cpu, MAX_CPUMASK_BITS);
+            error_report("CPU(s) present in multiple NUMA nodes: %s",
+                         enumerate_cpus(seen_cpus, max_cpus));;
+            exit(EXIT_FAILURE);
+        }
+        bitmap_or(seen_cpus, seen_cpus,
+                  numa_info[i].node_cpu, MAX_CPUMASK_BITS);
+    }
+
+    if (!bitmap_full(seen_cpus, max_cpus)) {
+        char *msg;
+        bitmap_complement(seen_cpus, seen_cpus, max_cpus);
+        msg = enumerate_cpus(seen_cpus, max_cpus);
+        error_report("warning: CPU(s) not present in any NUMA nodes: %s", msg);
+        error_report("warning: All CPU(s) up to maxcpus should be described "
+                     "in NUMA config");
+        g_free(msg);
+    }
+}
+
+void parse_numa_opts(MachineClass *mc)
+{
+    int i;
+
+    if (qemu_opts_foreach(qemu_find_opts("numa"), parse_numa,
+                          NULL, 1) != 0) {
+        exit(1);
+    }
 
     assert(max_numa_nodeid <= MAX_NODES);
 
@@ -222,19 +280,29 @@ void set_numa_nodes(void)
                 break;
             }
         }
-        /* assigning the VCPUs round-robin is easier to implement, guest OSes
-         * must cope with this anyway, because there are BIOSes out there in
-         * real machines which also use this scheme.
+        /* Historically VCPUs were assigned in round-robin order to NUMA
+         * nodes. However it causes issues with guest not handling it nice
+         * in case where cores/threads from a multicore CPU appear on
+         * different nodes. So allow boards to override default distribution
+         * rule grouping VCPUs by socket so that VCPUs from the same socket
+         * would be on the same node.
          */
         if (i == nb_numa_nodes) {
             for (i = 0; i < max_cpus; i++) {
-                set_bit(i, numa_info[i % nb_numa_nodes].node_cpu);
+                unsigned node_id = i % nb_numa_nodes;
+                if (mc->cpu_index_to_socket_id) {
+                    node_id = mc->cpu_index_to_socket_id(i) % nb_numa_nodes;
+                }
+
+                set_bit(i, numa_info[node_id].node_cpu);
             }
         }
+
+        validate_numa_cpus();
     }
 }
 
-void set_numa_modes(void)
+void numa_post_machine_init(void)
 {
     CPUState *cpu;
     int i;
@@ -262,8 +330,7 @@ static void allocate_system_memory_nonnuma(MemoryRegion *mr, Object *owner,
          * regular RAM allocation.
          */
         if (err) {
-            qerror_report_err(err);
-            error_free(err);
+            error_report_err(err);
             memory_region_init_ram(mr, owner, name, ram_size, &error_abort);
         }
 #else
@@ -298,7 +365,7 @@ void memory_region_allocate_system_memory(MemoryRegion *mr, Object *owner,
         }
         MemoryRegion *seg = host_memory_backend_get_memory(backend, &local_err);
         if (local_err) {
-            qerror_report_err(local_err);
+            error_report_err(local_err);
             exit(1);
         }