[34/34] KVM: selftests: Add a memory region subtest to validate invalid flags

Commit Message

Paolo Bonzini Nov. 5, 2023, 4:30 p.m. UTC

From: Sean Christopherson <[e-mail]>

Add a subtest to set_memory_region_test to verify that KVM rejects invalid
flags and combinations with -EINVAL.  KVM might or might not fail with
EINVAL anyways, but we can at least try.

Signed-off-by: Sean Christopherson <[e-mail]>
Message-Id: <[e-mail]>
Signed-off-by: Paolo Bonzini <[e-mail]>
---
 .../selftests/kvm/set_memory_region_test.c    | 49 +++++++++++++++++++
 1 file changed, 49 insertions(+)

Comments

Anish Moorthy Nov. 9, 2023, 1:08 a.m. UTC | #1

Applying [1] and [2] reveals that this also breaks non-x86 builds- the
MEM_REGION_GPA/SLOT definitions are guarded behind an #ifdef
__x86_64__, while the usages introduced here aren't.

Should

On Sun, Nov 5, 2023 at 8:35 AM Paolo Bonzini <[e-mail]> wrote:
>
> +       test_invalid_memory_region_flags();

be #ifdef'd, perhaps? I'm not quite sure what the intent is.

Side note: I wasn't able to get [2] to apply by copy-pasting the diff
and trying "git apply", and that was after checking out the relevant
commit. Eventually I just did it manually. If anyone can successfully
apply it, please let me know what you did so I can see what I was
doing wrong :)

[1] https://lore.kernel.org/kvm/[e-mail]/
[2] https://lore.kernel.org/kvm/[e-mail]/

Fuad Tabba Nov. 9, 2023, 8:54 a.m. UTC | #2

Hi Anish,

On Thu, Nov 9, 2023 at 1:08 AM Anish Moorthy <[e-mail]> wrote:
>
> Applying [1] and [2] reveals that this also breaks non-x86 builds- the
> MEM_REGION_GPA/SLOT definitions are guarded behind an #ifdef
> __x86_64__, while the usages introduced here aren't.
>
> Should
>
> On Sun, Nov 5, 2023 at 8:35 AM Paolo Bonzini <[e-mail]> wrote:
> >
> > +       test_invalid_memory_region_flags();
>
> be #ifdef'd, perhaps? I'm not quite sure what the intent is.
>
> Side note: I wasn't able to get [2] to apply by copy-pasting the diff
> and trying "git apply", and that was after checking out the relevant
> commit. Eventually I just did it manually. If anyone can successfully
> apply it, please let me know what you did so I can see what I was
> doing wrong :)

For me I applied the whole series as follows:

Checkout kvm-x86-next-2023.11.01 (45b890f7689e) from
https://github.com/kvm-x86/linux.git . Then use b4:
b4 am -o -  [e-mail]  | git am -3

Cheers,
/fuad

>
> [1] https://lore.kernel.org/kvm/[e-mail]/
> [2] https://lore.kernel.org/kvm/[e-mail]/

Mark Brown Nov. 20, 2023, 2:09 p.m. UTC | #3

On Wed, Nov 08, 2023 at 05:08:01PM -0800, Anish Moorthy wrote:
> Applying [1] and [2] reveals that this also breaks non-x86 builds- the
> MEM_REGION_GPA/SLOT definitions are guarded behind an #ifdef
> __x86_64__, while the usages introduced here aren't.
> 
> Should
> 
> On Sun, Nov 5, 2023 at 8:35 AM Paolo Bonzini <[e-mail]> wrote:
> >
> > +       test_invalid_memory_region_flags();
> 
> be #ifdef'd, perhaps? I'm not quite sure what the intent is.

This has been broken in -next for a week now, do we have any progress
on a fix or should we just revert the patch?

Paolo Bonzini Nov. 21, 2023, 5 p.m. UTC | #4

On Mon, Nov 20, 2023 at 3:09 PM Mark Brown <[e-mail]> wrote:
>
> On Wed, Nov 08, 2023 at 05:08:01PM -0800, Anish Moorthy wrote:
> > Applying [1] and [2] reveals that this also breaks non-x86 builds- the
> > MEM_REGION_GPA/SLOT definitions are guarded behind an #ifdef
> > __x86_64__, while the usages introduced here aren't.
> >
> > Should
> >
> > On Sun, Nov 5, 2023 at 8:35 AM Paolo Bonzini <[e-mail]> wrote:
> > >
> > > +       test_invalid_memory_region_flags();
> >
> > be #ifdef'd, perhaps? I'm not quite sure what the intent is.
>
> This has been broken in -next for a week now, do we have any progress
> on a fix or should we just revert the patch?

Sorry, I was away last week. I have now posted a patch.

Paolo

Patch

diff --git a/tools/testing/selftests/kvm/set_memory_region_test.c b/tools/testing/selftests/kvm/set_memory_region_test.c
index 1891774eb6d4..343e807043e1 100644
--- a/tools/testing/selftests/kvm/set_memory_region_test.c
+++ b/tools/testing/selftests/kvm/set_memory_region_test.c
@@ -326,6 +326,53 @@  static void test_zero_memory_regions(void)
 }
 #endif /* __x86_64__ */
 
+static void test_invalid_memory_region_flags(void)
+{
+	uint32_t supported_flags = KVM_MEM_LOG_DIRTY_PAGES;
+	const uint32_t v2_only_flags = KVM_MEM_GUEST_MEMFD;
+	struct kvm_vm *vm;
+	int r, i;
+
+#ifdef __x86_64__
+	supported_flags |= KVM_MEM_READONLY;
+
+	if (kvm_check_cap(KVM_CAP_VM_TYPES) & BIT(KVM_X86_SW_PROTECTED_VM))
+		vm = vm_create_barebones_protected_vm();
+	else
+#endif
+		vm = vm_create_barebones();
+
+	if (kvm_check_cap(KVM_CAP_MEMORY_ATTRIBUTES) & KVM_MEMORY_ATTRIBUTE_PRIVATE)
+		supported_flags |= KVM_MEM_GUEST_MEMFD;
+
+	for (i = 0; i < 32; i++) {
+		if ((supported_flags & BIT(i)) && !(v2_only_flags & BIT(i)))
+			continue;
+
+		r = __vm_set_user_memory_region(vm, MEM_REGION_SLOT, BIT(i),
+						MEM_REGION_GPA, MEM_REGION_SIZE, NULL);
+
+		TEST_ASSERT(r && errno == EINVAL,
+			    "KVM_SET_USER_MEMORY_REGION should have failed on v2 only flag 0x%lx", BIT(i));
+
+		if (supported_flags & BIT(i))
+			continue;
+
+		r = __vm_set_user_memory_region2(vm, MEM_REGION_SLOT, BIT(i),
+						 MEM_REGION_GPA, MEM_REGION_SIZE, NULL, 0, 0);
+		TEST_ASSERT(r && errno == EINVAL,
+			    "KVM_SET_USER_MEMORY_REGION2 should have failed on unsupported flag 0x%lx", BIT(i));
+	}
+
+	if (supported_flags & KVM_MEM_GUEST_MEMFD) {
+		r = __vm_set_user_memory_region2(vm, MEM_REGION_SLOT,
+						 KVM_MEM_LOG_DIRTY_PAGES | KVM_MEM_GUEST_MEMFD,
+						 MEM_REGION_GPA, MEM_REGION_SIZE, NULL, 0, 0);
+		TEST_ASSERT(r && errno == EINVAL,
+			    "KVM_SET_USER_MEMORY_REGION2 should have failed, dirty logging private memory is unsupported");
+	}
+}
+
 /*
  * Test it can be added memory slots up to KVM_CAP_NR_MEMSLOTS, then any
  * tentative to add further slots should fail.
@@ -491,6 +538,8 @@  int main(int argc, char *argv[])
 	test_zero_memory_regions();
 #endif
 
+	test_invalid_memory_region_flags();
+
 	test_add_max_memory_regions();
 
 	if (kvm_has_cap(KVM_CAP_GUEST_MEMFD) &&