From: Sean Christopherson <seanjc@google.com>
To: Maxim Levitsky <mlevitsk@redhat.com>
Cc: kvm@vger.kernel.org, Cathy Avery <cavery@redhat.com>,
Paolo Bonzini <pbonzini@redhat.com>
Subject: Re: [kvm-unit-tests PATCH 11/16] svm: add svm_suported
Date: Thu, 20 Oct 2022 18:21:53 +0000 [thread overview]
Message-ID: <Y1GRwf071rJDqVbh@google.com> (raw)
In-Reply-To: <20221020152404.283980-12-mlevitsk@redhat.com>
s/suported/supported
On Thu, Oct 20, 2022, Maxim Levitsky wrote:
Please provide a changelog.
> Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
> ---
> lib/x86/svm_lib.h | 5 +++++
> x86/svm.c | 2 +-
> 2 files changed, 6 insertions(+), 1 deletion(-)
>
> diff --git a/lib/x86/svm_lib.h b/lib/x86/svm_lib.h
> index 04910281..2d13b066 100644
> --- a/lib/x86/svm_lib.h
> +++ b/lib/x86/svm_lib.h
> @@ -4,6 +4,11 @@
> #include <x86/svm.h>
> #include "processor.h"
>
> +static inline bool svm_supported(void)
> +{
> + return this_cpu_has(X86_FEATURE_SVM);
Why add a wrapper? The only reason NPT and a few others have wrappers is to
play nice with svm_test's "bool (*supported)(void)" hook.
I would rather go the opposite direction and get rid of the wrappers, which IMO
only make it harder to understand what is being checked.
E.g. add a required_feature to the tests and use that for all X86_FEATURE_*
checks instead of adding wrappers. And unless there's a supported helper I'm not
seeing, the .supported hook can go away entirely by adding a dedicated "smp_required"
flag.
We'd probaby want helper macros for SMP vs. non-SMP, e.g.
#define SVM_V1_TEST(name, feature, ...)
{ #name, feature, false, ... }
#define SVM_SMP_V1_TEST(name, feature, ...)
{ #name, feature, true, ... }
diff --git a/x86/svm.c b/x86/svm.c
index 7aa3ebd2..2a412c27 100644
--- a/x86/svm.c
+++ b/x86/svm.c
@@ -170,6 +170,7 @@ test_wanted(const char *name, char *filters[], int filter_count)
int run_svm_tests(int ac, char **av, struct svm_test *svm_tests)
{
+ bool smp_supported = cpu_count() > 1;
int i = 0;
ac--;
@@ -187,7 +188,10 @@ int run_svm_tests(int ac, char **av, struct svm_test *svm_tests)
for (; svm_tests[i].name != NULL; i++) {
if (!test_wanted(svm_tests[i].name, av, ac))
continue;
- if (svm_tests[i].supported && !svm_tests[i].supported())
+ if (svm_tests[i].required_feature &&
+ !this_cpu_has(svm_tests[i].required_feature))
+ continue;
+ if (svm_tests[i].smp_required && !smp_supported)
continue;
if (svm_tests[i].v2 == NULL) {
if (svm_tests[i].on_vcpu) {
diff --git a/x86/svm.h b/x86/svm.h
index 0c40a086..632287ca 100644
--- a/x86/svm.h
+++ b/x86/svm.h
@@ -9,7 +9,8 @@
struct svm_test {
const char *name;
- bool (*supported)(void);
+ u64 required_feature;
+ bool smp_required;
void (*prepare)(struct svm_test *test);
void (*prepare_gif_clear)(struct svm_test *test);
void (*guest_func)(struct svm_test *test);
next prev parent reply other threads:[~2022-10-20 18:22 UTC|newest]
Thread overview: 51+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-10-20 15:23 [kvm-unit-tests PATCH 00/16] kvm-unit-tests: set of fixes and new tests Maxim Levitsky
2022-10-20 15:23 ` [kvm-unit-tests PATCH 01/16] x86: make irq_enable avoid the interrupt shadow Maxim Levitsky
2022-10-20 18:01 ` Sean Christopherson
2022-10-24 12:36 ` Maxim Levitsky
2022-10-24 22:49 ` Sean Christopherson
2022-10-27 10:16 ` Maxim Levitsky
2022-10-27 15:50 ` Sean Christopherson
2022-10-27 17:10 ` Maxim Levitsky
2022-10-20 15:23 ` [kvm-unit-tests PATCH 02/16] x86: add few helper functions for apic local timer Maxim Levitsky
2022-10-20 19:14 ` Sean Christopherson
2022-10-24 12:37 ` Maxim Levitsky
2022-10-24 16:10 ` Sean Christopherson
2022-10-27 10:19 ` Maxim Levitsky
2022-10-27 15:54 ` Sean Christopherson
2022-10-27 17:11 ` Maxim Levitsky
2022-10-20 15:23 ` [kvm-unit-tests PATCH 03/16] svm: use irq_enable instead of sti/nop Maxim Levitsky
2022-10-20 15:23 ` [kvm-unit-tests PATCH 04/16] svm: make svm_intr_intercept_mix_if/gif test a bit more robust Maxim Levitsky
2022-10-20 15:23 ` [kvm-unit-tests PATCH 05/16] svm: use apic_start_timer/apic_stop_timer instead of open coding it Maxim Levitsky
2022-10-20 15:23 ` [kvm-unit-tests PATCH 06/16] x86: Add test for #SMI during interrupt window Maxim Levitsky
2022-10-20 15:23 ` [kvm-unit-tests PATCH 07/16] x86: Add a simple test for SYSENTER instruction Maxim Levitsky
2022-10-20 19:25 ` Sean Christopherson
2022-10-24 12:38 ` Maxim Levitsky
2022-10-20 15:23 ` [kvm-unit-tests PATCH 08/16] svm: add nested shutdown test Maxim Levitsky
2022-10-20 15:26 ` Maxim Levitsky
2022-10-20 19:06 ` Sean Christopherson
2022-10-24 12:39 ` Maxim Levitsky
2022-10-20 15:23 ` [kvm-unit-tests PATCH 09/16] svm: move svm spec definitions to lib/x86/svm.h Maxim Levitsky
2022-10-20 19:08 ` Sean Christopherson
2022-10-20 15:23 ` [kvm-unit-tests PATCH 10/16] svm: move some svm support functions into lib/x86/svm_lib.h Maxim Levitsky
2022-10-20 15:23 ` [kvm-unit-tests PATCH 11/16] svm: add svm_suported Maxim Levitsky
2022-10-20 18:21 ` Sean Christopherson [this message]
2022-10-24 12:40 ` Maxim Levitsky
2022-10-20 15:24 ` [kvm-unit-tests PATCH 12/16] svm: move setup_svm to svm_lib.c Maxim Levitsky
2022-10-20 15:24 ` [kvm-unit-tests PATCH 13/16] svm: move vmcb_ident " Maxim Levitsky
2022-10-20 18:37 ` Sean Christopherson
2022-10-24 12:46 ` Maxim Levitsky
2022-10-20 15:24 ` [kvm-unit-tests PATCH 14/16] svm: rewerite vm entry macros Maxim Levitsky
2022-10-20 18:55 ` Sean Christopherson
2022-10-24 12:45 ` Maxim Levitsky
2022-10-24 19:56 ` Sean Christopherson
2022-10-27 12:07 ` Maxim Levitsky
2022-10-27 19:39 ` Sean Christopherson
2022-10-20 15:24 ` [kvm-unit-tests PATCH 15/16] svm: introduce svm_vcpu Maxim Levitsky
2022-10-20 19:02 ` Sean Christopherson
2022-10-24 12:46 ` Maxim Levitsky
2022-10-20 15:24 ` [kvm-unit-tests PATCH 16/16] add IPI loss stress test Maxim Levitsky
2022-10-20 20:23 ` Sean Christopherson
2022-10-24 12:54 ` Maxim Levitsky
2022-10-24 17:19 ` Sean Christopherson
2022-10-27 11:00 ` Maxim Levitsky
2022-10-27 18:41 ` Sean Christopherson
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=Y1GRwf071rJDqVbh@google.com \
--to=seanjc@google.com \
--cc=cavery@redhat.com \
--cc=kvm@vger.kernel.org \
--cc=mlevitsk@redhat.com \
--cc=pbonzini@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox