numa: equally distribute memory on nodes

When there are more nodes than available memory to put the minimum
allowed memory by node, all the memory is put on the last node.

This is because we put (ram_size / nb_numa_nodes) &
~((1 << mc->numa_mem_align_shift) - 1); on each node, and in this
case the value is 0. This is particularly true with pseries,
as the memory must be aligned to 256MB.

To avoid this problem, this patch uses an error diffusion algorithm [1]
to distribute equally the memory on nodes.

We introduce numa_auto_assign_ram() function in MachineClass
to keep compatibility between machine type versions.
The legacy function is used with pseries-2.9, pc-q35-2.9 and
pc-i440fx-2.9 (and previous), the new one with all others.

Example:

qemu-system-ppc64 -S -nographic  -nodefaults -monitor stdio -m 1G -smp 8 \
                  -numa node -numa node -numa node \
                  -numa node -numa node -numa node

Before:

(qemu) info numa
6 nodes
node 0 cpus: 0 6
node 0 size: 0 MB
node 1 cpus: 1 7
node 1 size: 0 MB
node 2 cpus: 2
node 2 size: 0 MB
node 3 cpus: 3
node 3 size: 0 MB
node 4 cpus: 4
node 4 size: 0 MB
node 5 cpus: 5
node 5 size: 1024 MB

After:
(qemu) info numa
6 nodes
node 0 cpus: 0 6
node 0 size: 0 MB
node 1 cpus: 1 7
node 1 size: 256 MB
node 2 cpus: 2
node 2 size: 0 MB
node 3 cpus: 3
node 3 size: 256 MB
node 4 cpus: 4
node 4 size: 256 MB
node 5 cpus: 5
node 5 size: 256 MB

[1] https://en.wikipedia.org/wiki/Error_diffusion

Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Message-Id: <20170502162955.1610-2-lvivier@redhat.com>
Reviewed-by: Eduardo Habkost <ehabkost@redhat.com>
[ehabkost: s/ram_size/size/ at numa_default_auto_assign_ram()]
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>

1 files changed, 38 insertions, 11 deletions

diff --git a/numa.c b/numa.c
index 2b3fc69915..d753687dec 100644
--- a/numa.c
+++ b/numa.c
@@ -407,6 +407,42 @@ static void complete_init_numa_distance(void)
     }
 }
 
+void numa_legacy_auto_assign_ram(MachineClass *mc, NodeInfo *nodes,
+                                 int nb_nodes, ram_addr_t size)
+{
+    int i;
+    uint64_t usedmem = 0;
+
+    /* Align each node according to the alignment
+     * requirements of the machine class
+     */
+
+    for (i = 0; i < nb_nodes - 1; i++) {
+        nodes[i].node_mem = (size / nb_nodes) &
+                            ~((1 << mc->numa_mem_align_shift) - 1);
+        usedmem += nodes[i].node_mem;
+    }
+    nodes[i].node_mem = size - usedmem;
+}
+
+void numa_default_auto_assign_ram(MachineClass *mc, NodeInfo *nodes,
+                                  int nb_nodes, ram_addr_t size)
+{
+    int i;
+    uint64_t usedmem = 0, node_mem;
+    uint64_t granularity = size / nb_nodes;
+    uint64_t propagate = 0;
+
+    for (i = 0; i < nb_nodes - 1; i++) {
+        node_mem = (granularity + propagate) &
+                   ~((1 << mc->numa_mem_align_shift) - 1);
+        propagate = granularity + propagate - node_mem;
+        nodes[i].node_mem = node_mem;
+        usedmem += node_mem;
+    }
+    nodes[i].node_mem = size - usedmem;
+}
+
 void parse_numa_opts(MachineClass *mc)
 {
     int i;
@@ -449,17 +485,8 @@ void parse_numa_opts(MachineClass *mc)
             }
         }
         if (i == nb_numa_nodes) {
-            uint64_t usedmem = 0;
-
-            /* Align each node according to the alignment
-             * requirements of the machine class
-             */
-            for (i = 0; i < nb_numa_nodes - 1; i++) {
-                numa_info[i].node_mem = (ram_size / nb_numa_nodes) &
-                                        ~((1 << mc->numa_mem_align_shift) - 1);
-                usedmem += numa_info[i].node_mem;
-            }
-            numa_info[i].node_mem = ram_size - usedmem;
+            assert(mc->numa_auto_assign_ram);
+            mc->numa_auto_assign_ram(mc, numa_info, nb_numa_nodes, ram_size);
         }
 
         numa_total = 0;