SYS-OSS/FRET-qemu
4
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

numa: Skip invalidation of cluster and NUMA node boundary for qtest

Browse Source 
There are warning messages printed from tests/qtest/numa-test.c,
to complain the CPU cluster and NUMA node boundary is broken. Since
the broken boundary is expected, we don't want to see the warning
messages.

  # cd /home/gavin/sandbox/qemu.main/build
  # MALLOC_PERTURB_=255 QTEST_QEMU_BINARY=./qemu-system-aarch64           \
    G_TEST_DBUS_DAEMON=../tests/dbus-vmstate-daemon.sh                    \
    QTEST_QEMU_IMG=./qemu-img                                             \
    QTEST_QEMU_STORAGE_DAEMON_BINARY=./storage-daemon/qemu-storage-daemon \
    tests/qtest/numa-test --tap -k
      :
    qemu-system-aarch64: warning: CPU-0 and CPU-4 in socket-0-cluster-0   \
    have been associated with node-0 and node-1 respectively.             \
    It can cause OSes like Linux to misbehave
      :

Skip the invalidation of CPU cluster and NUMA node boundary when
qtest is enabled, to avoid the warning messages.

Fixes: a494fdb715 ("numa: Validate cluster and NUMA node boundary if required")
Signed-off-by: Gavin Shan <gshan@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
This commit is contained in:
Gavin Shan 2023-07-21 20:44:37 +10:00 committed by Peter Maydell
parent 3fab43d2f6
commit 393d5c5bc7
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
hw/core/machine.c
View File

@ -1312,7 +1312,7 @@ static void validate_cpu_cluster_to_numa_boundary(MachineState *ms)
const CPUArchId *cpus = possible_cpus->cpus; const CPUArchId *cpus = possible_cpus->cpus;
int i, j; int i, j;
if (state->num_nodes <= 1 || possible_cpus->len <= 1) { if (qtest_enabled() || state->num_nodes <= 1 || possible_cpus->len <= 1) {
return; return;
} }