[v2] numa: hmat: require parent cache description before the next level one

Commit Message

Igor Mammedov Oct. 6, 2020, 3 p.m. UTC

Spec[1] defines 0 - 3 level memory side cache, however QEMU
CLI allows to specify an intermediate cache level without
specifying previous level. Such option(s) silently ignored
when building HMAT table, which leads to incomplete cache
information.
Make sure that previous level exists and error out
if it hasn't been provided.

1) ACPI 6.2A 5.2.27.5 Memory Side Cache Information Structure

Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1842877
Signed-off-by: Igor Mammedov <[e-mail]>
---
v2:
  - fix build inssue on MacOS
---
 hw/core/numa.c | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

Patch

diff --git a/hw/core/numa.c b/hw/core/numa.c
index 7d5d413001..7c4dd4e68e 100644
--- a/hw/core/numa.c
+++ b/hw/core/numa.c
@@ -424,7 +424,13 @@  void parse_numa_hmat_cache(MachineState *ms, NumaHmatCacheOptions *node,
     }
 
     if ((node->level > 1) &&
-        ms->numa_state->hmat_cache[node->node_id][node->level - 1] &&
+        ms->numa_state->hmat_cache[node->node_id][node->level - 1] == NULL) {
+        error_setg(errp, "Cache level=%u shall be defined first",
+                   node->level - 1);
+        return;
+    }
+
+    if ((node->level > 1) &&
         (node->size <=
             ms->numa_state->hmat_cache[node->node_id][node->level - 1]->size)) {
         error_setg(errp, "Invalid size=%" PRIu64 ", the size of level=%" PRIu8