kvm.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH] KVM: selftests: Print a message if /dev/kvm is missing
@ 2021-05-07 19:05 David Matlack
  2021-05-07 20:14 ` Andrew Jones
  0 siblings, 1 reply; 5+ messages in thread
From: David Matlack @ 2021-05-07 19:05 UTC (permalink / raw)
  To: kvm; +Cc: Paolo Bonzini, Andrew Jones, Aaron Lewis, David Matlack

If a KVM selftest is run on a machine without /dev/kvm, it will exit
silently. Make it easy to tell what's happening by printing an error
message.

Opportunistically consolidate all codepaths that open /dev/kvm into a
single function so they all print the same message.

This slightly changes the semantics of vm_is_unrestricted_guest() by
changing a TEST_ASSERT() to exit(KSFT_SKIP). However
vm_is_unrestricted_guest() is only called in one place
(x86_64/mmio_warning_test.c) and that is to determine if the test should
be skipped or not.

Signed-off-by: David Matlack <dmatlack@google.com>
---
 .../testing/selftests/kvm/include/kvm_util.h  |  1 +
 tools/testing/selftests/kvm/lib/kvm_util.c    | 35 +++++++++++--------
 .../selftests/kvm/lib/x86_64/processor.c      | 16 +++------
 .../kvm/x86_64/get_msr_index_features.c       |  8 ++---
 4 files changed, 28 insertions(+), 32 deletions(-)

diff --git a/tools/testing/selftests/kvm/include/kvm_util.h b/tools/testing/selftests/kvm/include/kvm_util.h
index a8f022794ce3..84982eb02b29 100644
--- a/tools/testing/selftests/kvm/include/kvm_util.h
+++ b/tools/testing/selftests/kvm/include/kvm_util.h
@@ -77,6 +77,7 @@ struct vm_guest_mode_params {
 };
 extern const struct vm_guest_mode_params vm_guest_mode_params[];
 
+int open_kvm_dev_path_or_exit(void);
 int kvm_check_cap(long cap);
 int vm_enable_cap(struct kvm_vm *vm, struct kvm_enable_cap *cap);
 int vcpu_enable_cap(struct kvm_vm *vm, uint32_t vcpu_id,
diff --git a/tools/testing/selftests/kvm/lib/kvm_util.c b/tools/testing/selftests/kvm/lib/kvm_util.c
index fc83f6c5902d..bb7dc65d7fb5 100644
--- a/tools/testing/selftests/kvm/lib/kvm_util.c
+++ b/tools/testing/selftests/kvm/lib/kvm_util.c
@@ -31,6 +31,23 @@ static void *align(void *x, size_t size)
 	return (void *) (((size_t) x + mask) & ~mask);
 }
 
+/* Open KVM_DEV_PATH if available, otherwise exit the entire program.
+ *
+ * Return:
+ *   The opened file descriptor of /dev/kvm.
+ */
+int open_kvm_dev_path_or_exit(void) {
+  int fd;
+
+  fd = open(KVM_DEV_PATH, O_RDONLY);
+  if (fd < 0) {
+    print_skip("%s not available", KVM_DEV_PATH);
+    exit(KSFT_SKIP);
+  }
+
+  return fd;
+}
+
 /*
  * Capability
  *
@@ -52,10 +69,7 @@ int kvm_check_cap(long cap)
 	int ret;
 	int kvm_fd;
 
-	kvm_fd = open(KVM_DEV_PATH, O_RDONLY);
-	if (kvm_fd < 0)
-		exit(KSFT_SKIP);
-
+	kvm_fd = open_kvm_dev_path_or_exit();
 	ret = ioctl(kvm_fd, KVM_CHECK_EXTENSION, cap);
 	TEST_ASSERT(ret != -1, "KVM_CHECK_EXTENSION IOCTL failed,\n"
 		"  rc: %i errno: %i", ret, errno);
@@ -128,9 +142,7 @@ void vm_enable_dirty_ring(struct kvm_vm *vm, uint32_t ring_size)
 
 static void vm_open(struct kvm_vm *vm, int perm)
 {
-	vm->kvm_fd = open(KVM_DEV_PATH, perm);
-	if (vm->kvm_fd < 0)
-		exit(KSFT_SKIP);
+	vm->kvm_fd = open_kvm_dev_path_or_exit();
 
 	if (!kvm_check_cap(KVM_CAP_IMMEDIATE_EXIT)) {
 		print_skip("immediate_exit not available");
@@ -925,9 +937,7 @@ static int vcpu_mmap_sz(void)
 {
 	int dev_fd, ret;
 
-	dev_fd = open(KVM_DEV_PATH, O_RDONLY);
-	if (dev_fd < 0)
-		exit(KSFT_SKIP);
+        dev_fd = open_kvm_dev_path_or_exit();
 
 	ret = ioctl(dev_fd, KVM_GET_VCPU_MMAP_SIZE, NULL);
 	TEST_ASSERT(ret >= sizeof(struct kvm_run),
@@ -2015,10 +2025,7 @@ bool vm_is_unrestricted_guest(struct kvm_vm *vm)
 
 	if (vm == NULL) {
 		/* Ensure that the KVM vendor-specific module is loaded. */
-		f = fopen(KVM_DEV_PATH, "r");
-		TEST_ASSERT(f != NULL, "Error in opening KVM dev file: %d",
-			    errno);
-		fclose(f);
+                close(open_kvm_dev_path_or_exit());
 	}
 
 	f = fopen("/sys/module/kvm_intel/parameters/unrestricted_guest", "r");
diff --git a/tools/testing/selftests/kvm/lib/x86_64/processor.c b/tools/testing/selftests/kvm/lib/x86_64/processor.c
index a8906e60a108..efe235044421 100644
--- a/tools/testing/selftests/kvm/lib/x86_64/processor.c
+++ b/tools/testing/selftests/kvm/lib/x86_64/processor.c
@@ -657,9 +657,7 @@ struct kvm_cpuid2 *kvm_get_supported_cpuid(void)
 		return cpuid;
 
 	cpuid = allocate_kvm_cpuid2();
-	kvm_fd = open(KVM_DEV_PATH, O_RDONLY);
-	if (kvm_fd < 0)
-		exit(KSFT_SKIP);
+	kvm_fd = open_kvm_dev_path_or_exit();
 
 	ret = ioctl(kvm_fd, KVM_GET_SUPPORTED_CPUID, cpuid);
 	TEST_ASSERT(ret == 0, "KVM_GET_SUPPORTED_CPUID failed %d %d\n",
@@ -691,9 +689,7 @@ uint64_t kvm_get_feature_msr(uint64_t msr_index)
 
 	buffer.header.nmsrs = 1;
 	buffer.entry.index = msr_index;
-	kvm_fd = open(KVM_DEV_PATH, O_RDONLY);
-	if (kvm_fd < 0)
-		exit(KSFT_SKIP);
+	kvm_fd = open_kvm_dev_path_or_exit();
 
 	r = ioctl(kvm_fd, KVM_GET_MSRS, &buffer.header);
 	TEST_ASSERT(r == 1, "KVM_GET_MSRS IOCTL failed,\n"
@@ -986,9 +982,7 @@ struct kvm_msr_list *kvm_get_msr_index_list(void)
 	struct kvm_msr_list *list;
 	int nmsrs, r, kvm_fd;
 
-	kvm_fd = open(KVM_DEV_PATH, O_RDONLY);
-	if (kvm_fd < 0)
-		exit(KSFT_SKIP);
+	kvm_fd = open_kvm_dev_path_or_exit();
 
 	nmsrs = kvm_get_num_msrs_fd(kvm_fd);
 	list = malloc(sizeof(*list) + nmsrs * sizeof(list->indices[0]));
@@ -1312,9 +1306,7 @@ struct kvm_cpuid2 *kvm_get_supported_hv_cpuid(void)
 		return cpuid;
 
 	cpuid = allocate_kvm_cpuid2();
-	kvm_fd = open(KVM_DEV_PATH, O_RDONLY);
-	if (kvm_fd < 0)
-		exit(KSFT_SKIP);
+	kvm_fd = open_kvm_dev_path_or_exit();
 
 	ret = ioctl(kvm_fd, KVM_GET_SUPPORTED_HV_CPUID, cpuid);
 	TEST_ASSERT(ret == 0, "KVM_GET_SUPPORTED_HV_CPUID failed %d %d\n",
diff --git a/tools/testing/selftests/kvm/x86_64/get_msr_index_features.c b/tools/testing/selftests/kvm/x86_64/get_msr_index_features.c
index cb953df4d7d0..8aed0db1331d 100644
--- a/tools/testing/selftests/kvm/x86_64/get_msr_index_features.c
+++ b/tools/testing/selftests/kvm/x86_64/get_msr_index_features.c
@@ -37,9 +37,7 @@ static void test_get_msr_index(void)
 	int old_res, res, kvm_fd, r;
 	struct kvm_msr_list *list;
 
-	kvm_fd = open(KVM_DEV_PATH, O_RDONLY);
-	if (kvm_fd < 0)
-		exit(KSFT_SKIP);
+	kvm_fd = open_kvm_dev_path_or_exit();
 
 	old_res = kvm_num_index_msrs(kvm_fd, 0);
 	TEST_ASSERT(old_res != 0, "Expecting nmsrs to be > 0");
@@ -101,9 +99,7 @@ static void test_get_msr_feature(void)
 	int res, old_res, i, kvm_fd;
 	struct kvm_msr_list *feature_list;
 
-	kvm_fd = open(KVM_DEV_PATH, O_RDONLY);
-	if (kvm_fd < 0)
-		exit(KSFT_SKIP);
+	kvm_fd = open_kvm_dev_path_or_exit();
 
 	old_res = kvm_num_feature_msrs(kvm_fd, 0);
 	TEST_ASSERT(old_res != 0, "Expecting nmsrs to be > 0");
-- 
2.31.1.607.g51e8a6a459-goog


^ permalink raw reply related	[flat|nested] 5+ messages in thread

* Re: [PATCH] KVM: selftests: Print a message if /dev/kvm is missing
  2021-05-07 19:05 [PATCH] KVM: selftests: Print a message if /dev/kvm is missing David Matlack
@ 2021-05-07 20:14 ` Andrew Jones
  2021-05-07 20:51   ` David Matlack
  0 siblings, 1 reply; 5+ messages in thread
From: Andrew Jones @ 2021-05-07 20:14 UTC (permalink / raw)
  To: David Matlack; +Cc: kvm, Paolo Bonzini, Aaron Lewis

On Fri, May 07, 2021 at 07:05:59PM +0000, David Matlack wrote:
> If a KVM selftest is run on a machine without /dev/kvm, it will exit
> silently. Make it easy to tell what's happening by printing an error
> message.
> 
> Opportunistically consolidate all codepaths that open /dev/kvm into a
> single function so they all print the same message.
> 
> This slightly changes the semantics of vm_is_unrestricted_guest() by
> changing a TEST_ASSERT() to exit(KSFT_SKIP). However
> vm_is_unrestricted_guest() is only called in one place
> (x86_64/mmio_warning_test.c) and that is to determine if the test should
> be skipped or not.
> 
> Signed-off-by: David Matlack <dmatlack@google.com>
> ---
>  .../testing/selftests/kvm/include/kvm_util.h  |  1 +
>  tools/testing/selftests/kvm/lib/kvm_util.c    | 35 +++++++++++--------
>  .../selftests/kvm/lib/x86_64/processor.c      | 16 +++------
>  .../kvm/x86_64/get_msr_index_features.c       |  8 ++---
>  4 files changed, 28 insertions(+), 32 deletions(-)
> 
> diff --git a/tools/testing/selftests/kvm/include/kvm_util.h b/tools/testing/selftests/kvm/include/kvm_util.h
> index a8f022794ce3..84982eb02b29 100644
> --- a/tools/testing/selftests/kvm/include/kvm_util.h
> +++ b/tools/testing/selftests/kvm/include/kvm_util.h
> @@ -77,6 +77,7 @@ struct vm_guest_mode_params {
>  };
>  extern const struct vm_guest_mode_params vm_guest_mode_params[];
>  
> +int open_kvm_dev_path_or_exit(void);
>  int kvm_check_cap(long cap);
>  int vm_enable_cap(struct kvm_vm *vm, struct kvm_enable_cap *cap);
>  int vcpu_enable_cap(struct kvm_vm *vm, uint32_t vcpu_id,
> diff --git a/tools/testing/selftests/kvm/lib/kvm_util.c b/tools/testing/selftests/kvm/lib/kvm_util.c
> index fc83f6c5902d..bb7dc65d7fb5 100644
> --- a/tools/testing/selftests/kvm/lib/kvm_util.c
> +++ b/tools/testing/selftests/kvm/lib/kvm_util.c
> @@ -31,6 +31,23 @@ static void *align(void *x, size_t size)
>  	return (void *) (((size_t) x + mask) & ~mask);
>  }
>  
> +/* Open KVM_DEV_PATH if available, otherwise exit the entire program.
> + *
> + * Return:
> + *   The opened file descriptor of /dev/kvm.
> + */
> +int open_kvm_dev_path_or_exit(void) {
> +  int fd;
> +
> +  fd = open(KVM_DEV_PATH, O_RDONLY);
> +  if (fd < 0) {
> +    print_skip("%s not available", KVM_DEV_PATH);
> +    exit(KSFT_SKIP);
> +  }
> +
> +  return fd;
> +}

Style issues in the function above '{' and 2 spaces vs. 1 tab.

> +
>  /*
>   * Capability
>   *
> @@ -52,10 +69,7 @@ int kvm_check_cap(long cap)
>  	int ret;
>  	int kvm_fd;
>  
> -	kvm_fd = open(KVM_DEV_PATH, O_RDONLY);
> -	if (kvm_fd < 0)
> -		exit(KSFT_SKIP);
> -
> +	kvm_fd = open_kvm_dev_path_or_exit();
>  	ret = ioctl(kvm_fd, KVM_CHECK_EXTENSION, cap);
>  	TEST_ASSERT(ret != -1, "KVM_CHECK_EXTENSION IOCTL failed,\n"
>  		"  rc: %i errno: %i", ret, errno);
> @@ -128,9 +142,7 @@ void vm_enable_dirty_ring(struct kvm_vm *vm, uint32_t ring_size)
>  
>  static void vm_open(struct kvm_vm *vm, int perm)
>  {
> -	vm->kvm_fd = open(KVM_DEV_PATH, perm);

I don't think we should change this one, otherwise the user provided
perms are ignored.

> -	if (vm->kvm_fd < 0)
> -		exit(KSFT_SKIP);
> +	vm->kvm_fd = open_kvm_dev_path_or_exit();
>  
>  	if (!kvm_check_cap(KVM_CAP_IMMEDIATE_EXIT)) {
>  		print_skip("immediate_exit not available");
> @@ -925,9 +937,7 @@ static int vcpu_mmap_sz(void)
>  {
>  	int dev_fd, ret;
>  
> -	dev_fd = open(KVM_DEV_PATH, O_RDONLY);
> -	if (dev_fd < 0)
> -		exit(KSFT_SKIP);
> +        dev_fd = open_kvm_dev_path_or_exit();

spaces vs. tab here

>  
>  	ret = ioctl(dev_fd, KVM_GET_VCPU_MMAP_SIZE, NULL);
>  	TEST_ASSERT(ret >= sizeof(struct kvm_run),
> @@ -2015,10 +2025,7 @@ bool vm_is_unrestricted_guest(struct kvm_vm *vm)
>  
>  	if (vm == NULL) {
>  		/* Ensure that the KVM vendor-specific module is loaded. */
> -		f = fopen(KVM_DEV_PATH, "r");
> -		TEST_ASSERT(f != NULL, "Error in opening KVM dev file: %d",
> -			    errno);
> -		fclose(f);
> +                close(open_kvm_dev_path_or_exit());

spaces

>  	}
>  
>  	f = fopen("/sys/module/kvm_intel/parameters/unrestricted_guest", "r");
> diff --git a/tools/testing/selftests/kvm/lib/x86_64/processor.c b/tools/testing/selftests/kvm/lib/x86_64/processor.c
> index a8906e60a108..efe235044421 100644
> --- a/tools/testing/selftests/kvm/lib/x86_64/processor.c
> +++ b/tools/testing/selftests/kvm/lib/x86_64/processor.c
> @@ -657,9 +657,7 @@ struct kvm_cpuid2 *kvm_get_supported_cpuid(void)
>  		return cpuid;
>  
>  	cpuid = allocate_kvm_cpuid2();
> -	kvm_fd = open(KVM_DEV_PATH, O_RDONLY);
> -	if (kvm_fd < 0)
> -		exit(KSFT_SKIP);
> +	kvm_fd = open_kvm_dev_path_or_exit();
>  
>  	ret = ioctl(kvm_fd, KVM_GET_SUPPORTED_CPUID, cpuid);
>  	TEST_ASSERT(ret == 0, "KVM_GET_SUPPORTED_CPUID failed %d %d\n",
> @@ -691,9 +689,7 @@ uint64_t kvm_get_feature_msr(uint64_t msr_index)
>  
>  	buffer.header.nmsrs = 1;
>  	buffer.entry.index = msr_index;
> -	kvm_fd = open(KVM_DEV_PATH, O_RDONLY);
> -	if (kvm_fd < 0)
> -		exit(KSFT_SKIP);
> +	kvm_fd = open_kvm_dev_path_or_exit();
>  
>  	r = ioctl(kvm_fd, KVM_GET_MSRS, &buffer.header);
>  	TEST_ASSERT(r == 1, "KVM_GET_MSRS IOCTL failed,\n"
> @@ -986,9 +982,7 @@ struct kvm_msr_list *kvm_get_msr_index_list(void)
>  	struct kvm_msr_list *list;
>  	int nmsrs, r, kvm_fd;
>  
> -	kvm_fd = open(KVM_DEV_PATH, O_RDONLY);
> -	if (kvm_fd < 0)
> -		exit(KSFT_SKIP);
> +	kvm_fd = open_kvm_dev_path_or_exit();
>  
>  	nmsrs = kvm_get_num_msrs_fd(kvm_fd);
>  	list = malloc(sizeof(*list) + nmsrs * sizeof(list->indices[0]));
> @@ -1312,9 +1306,7 @@ struct kvm_cpuid2 *kvm_get_supported_hv_cpuid(void)
>  		return cpuid;
>  
>  	cpuid = allocate_kvm_cpuid2();
> -	kvm_fd = open(KVM_DEV_PATH, O_RDONLY);
> -	if (kvm_fd < 0)
> -		exit(KSFT_SKIP);
> +	kvm_fd = open_kvm_dev_path_or_exit();
>  
>  	ret = ioctl(kvm_fd, KVM_GET_SUPPORTED_HV_CPUID, cpuid);
>  	TEST_ASSERT(ret == 0, "KVM_GET_SUPPORTED_HV_CPUID failed %d %d\n",
> diff --git a/tools/testing/selftests/kvm/x86_64/get_msr_index_features.c b/tools/testing/selftests/kvm/x86_64/get_msr_index_features.c
> index cb953df4d7d0..8aed0db1331d 100644
> --- a/tools/testing/selftests/kvm/x86_64/get_msr_index_features.c
> +++ b/tools/testing/selftests/kvm/x86_64/get_msr_index_features.c
> @@ -37,9 +37,7 @@ static void test_get_msr_index(void)
>  	int old_res, res, kvm_fd, r;
>  	struct kvm_msr_list *list;
>  
> -	kvm_fd = open(KVM_DEV_PATH, O_RDONLY);
> -	if (kvm_fd < 0)
> -		exit(KSFT_SKIP);
> +	kvm_fd = open_kvm_dev_path_or_exit();
>  
>  	old_res = kvm_num_index_msrs(kvm_fd, 0);
>  	TEST_ASSERT(old_res != 0, "Expecting nmsrs to be > 0");
> @@ -101,9 +99,7 @@ static void test_get_msr_feature(void)
>  	int res, old_res, i, kvm_fd;
>  	struct kvm_msr_list *feature_list;
>  
> -	kvm_fd = open(KVM_DEV_PATH, O_RDONLY);
> -	if (kvm_fd < 0)
> -		exit(KSFT_SKIP);
> +	kvm_fd = open_kvm_dev_path_or_exit();
>  
>  	old_res = kvm_num_feature_msrs(kvm_fd, 0);
>  	TEST_ASSERT(old_res != 0, "Expecting nmsrs to be > 0");
> -- 
> 2.31.1.607.g51e8a6a459-goog
> 

Thanks,
drew


^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH] KVM: selftests: Print a message if /dev/kvm is missing
  2021-05-07 20:14 ` Andrew Jones
@ 2021-05-07 20:51   ` David Matlack
  2021-05-10  6:32     ` Andrew Jones
  0 siblings, 1 reply; 5+ messages in thread
From: David Matlack @ 2021-05-07 20:51 UTC (permalink / raw)
  To: Andrew Jones; +Cc: kvm list, Paolo Bonzini, Aaron Lewis

On Fri, May 7, 2021 at 1:14 PM Andrew Jones <drjones@redhat.com> wrote:
>
> On Fri, May 07, 2021 at 07:05:59PM +0000, David Matlack wrote:
> > If a KVM selftest is run on a machine without /dev/kvm, it will exit
> > silently. Make it easy to tell what's happening by printing an error
> > message.
> >
> > Opportunistically consolidate all codepaths that open /dev/kvm into a
> > single function so they all print the same message.
> >
> > This slightly changes the semantics of vm_is_unrestricted_guest() by
> > changing a TEST_ASSERT() to exit(KSFT_SKIP). However
> > vm_is_unrestricted_guest() is only called in one place
> > (x86_64/mmio_warning_test.c) and that is to determine if the test should
> > be skipped or not.
> >
> > Signed-off-by: David Matlack <dmatlack@google.com>
> > ---
> >  .../testing/selftests/kvm/include/kvm_util.h  |  1 +
> >  tools/testing/selftests/kvm/lib/kvm_util.c    | 35 +++++++++++--------
> >  .../selftests/kvm/lib/x86_64/processor.c      | 16 +++------
> >  .../kvm/x86_64/get_msr_index_features.c       |  8 ++---
> >  4 files changed, 28 insertions(+), 32 deletions(-)
> >
> > diff --git a/tools/testing/selftests/kvm/include/kvm_util.h b/tools/testing/selftests/kvm/include/kvm_util.h
> > index a8f022794ce3..84982eb02b29 100644
> > --- a/tools/testing/selftests/kvm/include/kvm_util.h
> > +++ b/tools/testing/selftests/kvm/include/kvm_util.h
> > @@ -77,6 +77,7 @@ struct vm_guest_mode_params {
> >  };
> >  extern const struct vm_guest_mode_params vm_guest_mode_params[];
> >
> > +int open_kvm_dev_path_or_exit(void);
> >  int kvm_check_cap(long cap);
> >  int vm_enable_cap(struct kvm_vm *vm, struct kvm_enable_cap *cap);
> >  int vcpu_enable_cap(struct kvm_vm *vm, uint32_t vcpu_id,
> > diff --git a/tools/testing/selftests/kvm/lib/kvm_util.c b/tools/testing/selftests/kvm/lib/kvm_util.c
> > index fc83f6c5902d..bb7dc65d7fb5 100644
> > --- a/tools/testing/selftests/kvm/lib/kvm_util.c
> > +++ b/tools/testing/selftests/kvm/lib/kvm_util.c
> > @@ -31,6 +31,23 @@ static void *align(void *x, size_t size)
> >       return (void *) (((size_t) x + mask) & ~mask);
> >  }
> >
> > +/* Open KVM_DEV_PATH if available, otherwise exit the entire program.
> > + *
> > + * Return:
> > + *   The opened file descriptor of /dev/kvm.
> > + */
> > +int open_kvm_dev_path_or_exit(void) {
> > +  int fd;
> > +
> > +  fd = open(KVM_DEV_PATH, O_RDONLY);
> > +  if (fd < 0) {
> > +    print_skip("%s not available", KVM_DEV_PATH);
> > +    exit(KSFT_SKIP);
> > +  }
> > +
> > +  return fd;
> > +}
>
> Style issues in the function above '{' and 2 spaces vs. 1 tab.

Well that's sure embarrassing! Thanks for catching.

>
> > +
> >  /*
> >   * Capability
> >   *
> > @@ -52,10 +69,7 @@ int kvm_check_cap(long cap)
> >       int ret;
> >       int kvm_fd;
> >
> > -     kvm_fd = open(KVM_DEV_PATH, O_RDONLY);
> > -     if (kvm_fd < 0)
> > -             exit(KSFT_SKIP);
> > -
> > +     kvm_fd = open_kvm_dev_path_or_exit();
> >       ret = ioctl(kvm_fd, KVM_CHECK_EXTENSION, cap);
> >       TEST_ASSERT(ret != -1, "KVM_CHECK_EXTENSION IOCTL failed,\n"
> >               "  rc: %i errno: %i", ret, errno);
> > @@ -128,9 +142,7 @@ void vm_enable_dirty_ring(struct kvm_vm *vm, uint32_t ring_size)
> >
> >  static void vm_open(struct kvm_vm *vm, int perm)
> >  {
> > -     vm->kvm_fd = open(KVM_DEV_PATH, perm);
>
> I don't think we should change this one, otherwise the user provided
> perms are ignored.

Good catch. I don't see any reason to exclude this case, but we do need
to pass `perm` down to open_kvm_dev_path_or_exit().


>
> > -     if (vm->kvm_fd < 0)
> > -             exit(KSFT_SKIP);
> > +     vm->kvm_fd = open_kvm_dev_path_or_exit();
> >
> >       if (!kvm_check_cap(KVM_CAP_IMMEDIATE_EXIT)) {
> >               print_skip("immediate_exit not available");
> > @@ -925,9 +937,7 @@ static int vcpu_mmap_sz(void)
> >  {
> >       int dev_fd, ret;
> >
> > -     dev_fd = open(KVM_DEV_PATH, O_RDONLY);
> > -     if (dev_fd < 0)
> > -             exit(KSFT_SKIP);
> > +        dev_fd = open_kvm_dev_path_or_exit();
>
> spaces vs. tab here
>
> >
> >       ret = ioctl(dev_fd, KVM_GET_VCPU_MMAP_SIZE, NULL);
> >       TEST_ASSERT(ret >= sizeof(struct kvm_run),
> > @@ -2015,10 +2025,7 @@ bool vm_is_unrestricted_guest(struct kvm_vm *vm)
> >
> >       if (vm == NULL) {
> >               /* Ensure that the KVM vendor-specific module is loaded. */
> > -             f = fopen(KVM_DEV_PATH, "r");
> > -             TEST_ASSERT(f != NULL, "Error in opening KVM dev file: %d",
> > -                         errno);
> > -             fclose(f);
> > +                close(open_kvm_dev_path_or_exit());
>
> spaces
>
> >       }
> >
> >       f = fopen("/sys/module/kvm_intel/parameters/unrestricted_guest", "r");
> > diff --git a/tools/testing/selftests/kvm/lib/x86_64/processor.c b/tools/testing/selftests/kvm/lib/x86_64/processor.c
> > index a8906e60a108..efe235044421 100644
> > --- a/tools/testing/selftests/kvm/lib/x86_64/processor.c
> > +++ b/tools/testing/selftests/kvm/lib/x86_64/processor.c
> > @@ -657,9 +657,7 @@ struct kvm_cpuid2 *kvm_get_supported_cpuid(void)
> >               return cpuid;
> >
> >       cpuid = allocate_kvm_cpuid2();
> > -     kvm_fd = open(KVM_DEV_PATH, O_RDONLY);
> > -     if (kvm_fd < 0)
> > -             exit(KSFT_SKIP);
> > +     kvm_fd = open_kvm_dev_path_or_exit();
> >
> >       ret = ioctl(kvm_fd, KVM_GET_SUPPORTED_CPUID, cpuid);
> >       TEST_ASSERT(ret == 0, "KVM_GET_SUPPORTED_CPUID failed %d %d\n",
> > @@ -691,9 +689,7 @@ uint64_t kvm_get_feature_msr(uint64_t msr_index)
> >
> >       buffer.header.nmsrs = 1;
> >       buffer.entry.index = msr_index;
> > -     kvm_fd = open(KVM_DEV_PATH, O_RDONLY);
> > -     if (kvm_fd < 0)
> > -             exit(KSFT_SKIP);
> > +     kvm_fd = open_kvm_dev_path_or_exit();
> >
> >       r = ioctl(kvm_fd, KVM_GET_MSRS, &buffer.header);
> >       TEST_ASSERT(r == 1, "KVM_GET_MSRS IOCTL failed,\n"
> > @@ -986,9 +982,7 @@ struct kvm_msr_list *kvm_get_msr_index_list(void)
> >       struct kvm_msr_list *list;
> >       int nmsrs, r, kvm_fd;
> >
> > -     kvm_fd = open(KVM_DEV_PATH, O_RDONLY);
> > -     if (kvm_fd < 0)
> > -             exit(KSFT_SKIP);
> > +     kvm_fd = open_kvm_dev_path_or_exit();
> >
> >       nmsrs = kvm_get_num_msrs_fd(kvm_fd);
> >       list = malloc(sizeof(*list) + nmsrs * sizeof(list->indices[0]));
> > @@ -1312,9 +1306,7 @@ struct kvm_cpuid2 *kvm_get_supported_hv_cpuid(void)
> >               return cpuid;
> >
> >       cpuid = allocate_kvm_cpuid2();
> > -     kvm_fd = open(KVM_DEV_PATH, O_RDONLY);
> > -     if (kvm_fd < 0)
> > -             exit(KSFT_SKIP);
> > +     kvm_fd = open_kvm_dev_path_or_exit();
> >
> >       ret = ioctl(kvm_fd, KVM_GET_SUPPORTED_HV_CPUID, cpuid);
> >       TEST_ASSERT(ret == 0, "KVM_GET_SUPPORTED_HV_CPUID failed %d %d\n",
> > diff --git a/tools/testing/selftests/kvm/x86_64/get_msr_index_features.c b/tools/testing/selftests/kvm/x86_64/get_msr_index_features.c
> > index cb953df4d7d0..8aed0db1331d 100644
> > --- a/tools/testing/selftests/kvm/x86_64/get_msr_index_features.c
> > +++ b/tools/testing/selftests/kvm/x86_64/get_msr_index_features.c
> > @@ -37,9 +37,7 @@ static void test_get_msr_index(void)
> >       int old_res, res, kvm_fd, r;
> >       struct kvm_msr_list *list;
> >
> > -     kvm_fd = open(KVM_DEV_PATH, O_RDONLY);
> > -     if (kvm_fd < 0)
> > -             exit(KSFT_SKIP);
> > +     kvm_fd = open_kvm_dev_path_or_exit();
> >
> >       old_res = kvm_num_index_msrs(kvm_fd, 0);
> >       TEST_ASSERT(old_res != 0, "Expecting nmsrs to be > 0");
> > @@ -101,9 +99,7 @@ static void test_get_msr_feature(void)
> >       int res, old_res, i, kvm_fd;
> >       struct kvm_msr_list *feature_list;
> >
> > -     kvm_fd = open(KVM_DEV_PATH, O_RDONLY);
> > -     if (kvm_fd < 0)
> > -             exit(KSFT_SKIP);
> > +     kvm_fd = open_kvm_dev_path_or_exit();
> >
> >       old_res = kvm_num_feature_msrs(kvm_fd, 0);
> >       TEST_ASSERT(old_res != 0, "Expecting nmsrs to be > 0");
> > --
> > 2.31.1.607.g51e8a6a459-goog
> >
>
> Thanks,
> drew
>

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH] KVM: selftests: Print a message if /dev/kvm is missing
  2021-05-07 20:51   ` David Matlack
@ 2021-05-10  6:32     ` Andrew Jones
  2021-05-10 17:23       ` David Matlack
  0 siblings, 1 reply; 5+ messages in thread
From: Andrew Jones @ 2021-05-10  6:32 UTC (permalink / raw)
  To: David Matlack; +Cc: kvm list, Paolo Bonzini, Aaron Lewis

On Fri, May 07, 2021 at 01:51:45PM -0700, David Matlack wrote:
> > >  static void vm_open(struct kvm_vm *vm, int perm)
> > >  {
> > > -     vm->kvm_fd = open(KVM_DEV_PATH, perm);
> >
> > I don't think we should change this one, otherwise the user provided
> > perms are ignored.
> 
> Good catch. I don't see any reason to exclude this case, but we do need
> to pass `perm` down to open_kvm_dev_path_or_exit().
>

I've reviewed v2 and gave it an r-b, since I don't have overly strong
opinion about this, but I actually liked that open_kvm_dev_path_or_exit()
didn't take any arguments. To handle this case I would have either left
it open coded, like it was, or created something like

int _open_kvm_dev_path_or_exit(int flags)
{
   int fd = open(KVM_DEV_PATH, flags);
   if (fd < 0)
     ...exit skip...
   return fd;
}

int open_kvm_dev_path_or_exit(void)
{
  return _open_kvm_dev_path_or_exit(O_RDONLY);
}

Thanks,
drew


^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH] KVM: selftests: Print a message if /dev/kvm is missing
  2021-05-10  6:32     ` Andrew Jones
@ 2021-05-10 17:23       ` David Matlack
  0 siblings, 0 replies; 5+ messages in thread
From: David Matlack @ 2021-05-10 17:23 UTC (permalink / raw)
  To: Andrew Jones; +Cc: kvm list, Paolo Bonzini, Aaron Lewis

On Sun, May 9, 2021 at 11:32 PM Andrew Jones <drjones@redhat.com> wrote:
>
> On Fri, May 07, 2021 at 01:51:45PM -0700, David Matlack wrote:
> > > >  static void vm_open(struct kvm_vm *vm, int perm)
> > > >  {
> > > > -     vm->kvm_fd = open(KVM_DEV_PATH, perm);
> > >
> > > I don't think we should change this one, otherwise the user provided
> > > perms are ignored.
> >
> > Good catch. I don't see any reason to exclude this case, but we do need
> > to pass `perm` down to open_kvm_dev_path_or_exit().
> >
>
> I've reviewed v2 and gave it an r-b, since I don't have overly strong
> opinion about this, but I actually liked that open_kvm_dev_path_or_exit()
> didn't take any arguments. To handle this case I would have either left
> it open coded, like it was, or created something like
>
> int _open_kvm_dev_path_or_exit(int flags)
> {
>    int fd = open(KVM_DEV_PATH, flags);
>    if (fd < 0)
>      ...exit skip...
>    return fd;
> }
>
> int open_kvm_dev_path_or_exit(void)
> {
>   return _open_kvm_dev_path_or_exit(O_RDONLY);
> }

I agree so long as hiding O_RDONLY does not decrease code readability.
I could not find any place in KVM where the R/W permissions on
/dev/kvm played a role, so I'm inclined to agree it's better to hide
the permissions.

I'll send another version with this change along with the comment fix.

Thanks.

>
> Thanks,
> drew
>

^ permalink raw reply	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2021-05-10 17:24 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-05-07 19:05 [PATCH] KVM: selftests: Print a message if /dev/kvm is missing David Matlack
2021-05-07 20:14 ` Andrew Jones
2021-05-07 20:51   ` David Matlack
2021-05-10  6:32     ` Andrew Jones
2021-05-10 17:23       ` David Matlack

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).