[Devel] [PATCH vz10 1/3] ms/KVM: selftests: Forcefully override ARCH from x86_64 to x86

[Devel] [PATCH vz10 1/3] ms/KVM: selftests: Forcefully override ARCH from x86_64 to x86

[Eva Kurchatova]

From: Sean Christopherson <seanjc@google.com>

Forcefully override ARCH from x86_64 to x86 to handle the scenario where
the user specifies ARCH=x86_64 on the command line.

Fixes: 9af04539d474 ("KVM: selftests: Override ARCH for x86_64 instead of using ARCH_DIR")
Cc: stable@vger.kernel.org
Reported-by: David Matlack <dmatlack@google.com>
Closes: https://lore.kernel.org/all/20250724213130.3374922-1-dmatlack@google.com
Link: https://lore.kernel.org/r/20251007223057.368082-1-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
(cherry picked from commit 17e5a9b77716564540d81f0c1e6082d28cf305c9)

https://virtuozzo.atlassian.net/browse/VSTOR-139672
Feature: fix selftests
Signed-off-by: Eva Kurchatova <eva.kurchatova@virtuozzo.com>
---
 tools/testing/selftests/kvm/Makefile | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/testing/selftests/kvm/Makefile b/tools/testing/selftests/kvm/Makefile
index 20af35a91d6f..4e85f3634e2e 100644
--- a/tools/testing/selftests/kvm/Makefile
+++ b/tools/testing/selftests/kvm/Makefile
@@ -6,7 +6,7 @@ ARCH            ?= $(SUBARCH)
 ifeq ($(ARCH),$(filter $(ARCH),arm64 s390 riscv x86 x86_64))
 # Top-level selftests allows ARCH=x86_64 :-(
 ifeq ($(ARCH),x86_64)
-	ARCH := x86
+	override ARCH := x86
 endif
 include Makefile.kvm
 else
-- 
2.55.0

_______________________________________________
Devel mailing list
Devel@openvz.org
https://lists.openvz.org/mailman/listinfo/devel

[Devel] [PATCH vz10 2/3] selftests: cgroup: give the O_TMPFILE open in get_temp_fd() a mode

[Eva Kurchatova]

O_TMPFILE, like O_CREAT, needs the third argument. Without it glibc
refuses the call at compile time as soon as fortification is on:

  In function 'open',
      inlined from 'get_temp_fd' at cgroup_util.c:24:9:
  /usr/include/bits/fcntl2.h:50:11: error: call to '__open_missing_mode'
    declared with attribute error: open with O_CREAT or O_TMPFILE in
    second argument needs 3 arguments

cgroup/Makefile builds with "-Wall -pthread" and the fortify checks
take effect only once the compiler optimises, so the cgroup target
builds. kvm includes ../cgroup/lib/libcgroup.mk and does
build with -O2, so the shared object fails there and takes the whole
kvm target with it. Building with -D_FORTIFY_SOURCE=3, as distributions
commonly do, therefore loses every kvm test rather than anything in
cgroup.

Fixes: 844934d16dbc ("selftests/cgroup: move get_temp_fd() to the shared cgroup lib")
https://virtuozzo.atlassian.net/browse/VSTOR-139672
Feature: fix selftests
Signed-off-by: Eva Kurchatova <eva.kurchatova@virtuozzo.com>
---
 tools/testing/selftests/cgroup/lib/cgroup_util.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/testing/selftests/cgroup/lib/cgroup_util.c b/tools/testing/selftests/cgroup/lib/cgroup_util.c
index 9a677fdf0bf6..8f996b54849e 100644
--- a/tools/testing/selftests/cgroup/lib/cgroup_util.c
+++ b/tools/testing/selftests/cgroup/lib/cgroup_util.c
@@ -21,7 +21,7 @@
 
 int get_temp_fd(void)
 {
-	return open(".", O_TMPFILE | O_RDWR | O_EXCL);
+	return open(".", O_TMPFILE | O_RDWR | O_EXCL, 0600);
 }
 
 /* Returns read len on success, or -errno on failure. */
-- 
2.55.0

_______________________________________________
Devel mailing list
Devel@openvz.org
https://lists.openvz.org/mailman/listinfo/devel

[Devel] [PATCH vz10 3/3] selftests: kvm: raise the timeout the suite gives each test

[Eva Kurchatova]

kvm/settings asks for 120 seconds per test, which mmu_stress_test cannot
meet on any machine. It writes a 128gb guest address space by default,
and the size is settable only on the command line, which the kselftest
runner does not pass:

  not ok 1 selftests: kvm: mmu_stress_test # TIMEOUT 300 seconds

It is not stuck, it is long, and how long depends on how many vCPUs it
gets, which is 3/4 of the cpus of the machine. Measured to completion,
all passing:

  40 cpus, 30 vCPUs    171s
   4 cpus,  3 vCPUs    436s
   2 cpus,  1 vCPU    1094s

Ask for 1800 seconds, which leaves room above the slowest of those for a
machine that is also busy. The cost is that the limit applies to every
test in the suite, so a test that really does hang now takes that long
to be killed; nothing else in the suite runs for more than a few tens of
seconds today.

A timeout given on the runner command line replaces this value
entirely, so it has to be at least as large.

https://virtuozzo.atlassian.net/browse/VSTOR-139672
Feature: fix vz selftests
Signed-off-by: Eva Kurchatova <eva.kurchatova@virtuozzo.com>
---
 tools/testing/selftests/kvm/settings | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/testing/selftests/kvm/settings b/tools/testing/selftests/kvm/settings
index 6091b45d226b..abc5648b59ab 100644
--- a/tools/testing/selftests/kvm/settings
+++ b/tools/testing/selftests/kvm/settings
@@ -1 +1 @@
-timeout=120
+timeout=1800
-- 
2.55.0

_______________________________________________
Devel mailing list
Devel@openvz.org
https://lists.openvz.org/mailman/listinfo/devel