Skip to contentSkip to conversation

tests/qtest/migration-test: Run the dirty ring tests only with the x86 target

PATCH In review Thomas Huth 2 participants 2 messages Aug 1, 2022

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 tests/qtest/migration-test: Run the dirty ring tests only with the x86 target

kvm_dirty_ring_supported() only checks whether the dirty ring support is available on the x86 host, but it ignores whether the target QEMU architecture is x86 or not. Thus the test_vcpu_dirty_limit() test currently fails with the assert((strcmp(arch, "x86_64") == 0)) statement in dirtylimit_start_vm() if the users run e.g. "make check-qtest-aarch64" on their x86 host. Fix it by only executing the tests when we're running with a x86_64 target QEMU binary with KVM.

---
 tests/qtest/migration-test.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)
tests/qtest/migration-test.c +4 -3
index 71595a74fd..f83360e0e0 100644
@@ -2439,6 +2439,7 @@ int main(int argc, char **argv)
 {
     char template[] = "/tmp/migration-test-XXXXXX";
     const bool has_kvm = qtest_has_accel("kvm");
+    const char *arch = qtest_get_arch();
     int ret;
 
     g_test_init(&argc, &argv, NULL);
@@ -2452,7 +2453,7 @@ int main(int argc, char **argv)
      * is touchy due to race conditions on dirty bits (especially on PPC for
      * some reason)
      */
-    if (g_str_equal(qtest_get_arch(), "ppc64") &&
+    if (g_str_equal(arch, "ppc64") &&
         (!has_kvm || access("/sys/module/kvm_hv", F_OK))) {
         g_test_message("Skipping test: kvm_hv not available");
         return g_test_run();
@@ -2462,7 +2463,7 @@ int main(int argc, char **argv)
      * Similar to ppc64, s390x seems to be touchy with TCG, so disable it
      * there until the problems are resolved
      */
-    if (g_str_equal(qtest_get_arch(), "s390x") && !has_kvm) {
+    if (g_str_equal(arch, "s390x") && !has_kvm) {
         g_test_message("Skipping test: s390x host with KVM is required");
         return g_test_run();
     }
@@ -2572,7 +2573,7 @@ int main(int argc, char **argv)
 #endif /* CONFIG_TASN1 */
 #endif /* CONFIG_GNUTLS */
 
-    if (kvm_dirty_ring_supported()) {
+    if (g_str_equal(arch, "x86_64") && has_kvm && kvm_dirty_ring_supported()) {
         qtest_add_func("/migration/dirty_ring",
                        test_precopy_unix_dirty_ring);
         qtest_add_func("/migration/vcpu_dirty_limit",
Signature · 1 line
2.31.1
Reviewed-by on v1 1/1
#1 lore raw
Quoted 9 lines · * Thomas Huth ([email protected]) wrote:

kvm_dirty_ring_supported() only checks whether the dirty ring support is available on the x86 host, but it ignores whether the target QEMU architecture is x86 or not. Thus the test_vcpu_dirty_limit() test currently fails with the assert((strcmp(arch, "x86_64") == 0)) statement in dirtylimit_start_vm() if the users run e.g. "make check-qtest-aarch64" on their x86 host. Fix it by only executing the tests when we're running with a x86_64 target QEMU binary with KVM.

I think that corresponds to a bug Juan mentioned where it was failing with a i386 qemu as well.I guess there should be a more generic way!

Quoted 46 lines
---
 tests/qtest/migration-test.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)
tests/qtest/migration-test.c +4 -3
index 71595a74fd..f83360e0e0 100644
@@ -2439,6 +2439,7 @@ int main(int argc, char **argv)
 {
     char template[] = "/tmp/migration-test-XXXXXX";
     const bool has_kvm = qtest_has_accel("kvm");
+    const char *arch = qtest_get_arch();
     int ret;
 
     g_test_init(&argc, &argv, NULL);
@@ -2452,7 +2453,7 @@ int main(int argc, char **argv)
      * is touchy due to race conditions on dirty bits (especially on PPC for
      * some reason)
      */
-    if (g_str_equal(qtest_get_arch(), "ppc64") &&
+    if (g_str_equal(arch, "ppc64") &&
         (!has_kvm || access("/sys/module/kvm_hv", F_OK))) {
         g_test_message("Skipping test: kvm_hv not available");
         return g_test_run();
@@ -2462,7 +2463,7 @@ int main(int argc, char **argv)
      * Similar to ppc64, s390x seems to be touchy with TCG, so disable it
      * there until the problems are resolved
      */
-    if (g_str_equal(qtest_get_arch(), "s390x") && !has_kvm) {
+    if (g_str_equal(arch, "s390x") && !has_kvm) {
         g_test_message("Skipping test: s390x host with KVM is required");
         return g_test_run();
     }
@@ -2572,7 +2573,7 @@ int main(int argc, char **argv)
 #endif /* CONFIG_TASN1 */
 #endif /* CONFIG_GNUTLS */
 
-    if (kvm_dirty_ring_supported()) {
+    if (g_str_equal(arch, "x86_64") && has_kvm && kvm_dirty_ring_supported()) {
         qtest_add_func("/migration/dirty_ring",
                        test_precopy_unix_dirty_ring);
         qtest_add_func("/migration/vcpu_dirty_limit",
Signature · 1 line
2.31.1
Signature · 1 line
Dr. David Alan Gilbert / [email protected] / Manchester, UK