From: Manali Shukla <manali.shukla@amd.com>
To: Muhammad Usama Anjum <usama.anjum@collabora.com>,
kvm@vger.kernel.org, linux-kselftest@vger.kernel.org
Cc: pbonzini@redhat.com, seanjc@google.com, shuah@kernel.org,
nikunj@amd.com, thomas.lendacky@amd.com,
Manali Shukla <manali.shukla@amd.com>
Subject: Re: [PATCH v1 3/3] KVM: selftests: Add a test case for KVM_X86_DISABLE_EXITS_HLT
Date: Mon, 1 Apr 2024 10:58:36 +0530 [thread overview]
Message-ID: <5d667c81-56a8-95bd-4aa1-4df16c42dabf@amd.com> (raw)
In-Reply-To: <7e093e45-9349-4cfd-a0ad-78ae66edda90@collabora.com>
Hi Muhammad Usama Anjum,
Thank you for reviewing my patch.
On 3/30/2024 1:43 AM, Muhammad Usama Anjum wrote:
> On 3/27/24 10:42 AM, Manali Shukla wrote:
>> By default, HLT instruction executed by guest is intercepted by hypervisor.
>> However, KVM_CAP_X86_DISABLE_EXITS capability can be used to not intercept
>> HLT by setting KVM_X86_DISABLE_EXITS_HLT.
>>
>> Add a test case to test KVM_X86_DISABLE_EXITS_HLT functionality.
>>
>> Suggested-by: Sean Christopherson <seanjc@google.com>
>> Signed-off-by: Manali Shukla <manali.shukla@amd.com>
> Thank you for the new test patch. We have been trying to ensure TAP
> conformance for tests which cannot be achieved if new tests aren't using
> TAP already. Please make your test TAP compliant.
As per my understanding about TAP interface, kvm_test_harness.h file includes a MACRO,
which is used to create VM with one vcpu using vm_create_with_one_vcpu(), but
halt_disable_exit_test creates a customized VM with KVM_CAP_X86_DISABLE_EXITS
capability set and different vm_shape parameters to start a VM without in-kernel
APIC support. AFAIU, I won't be able to use KVM_ONE_VCPU_TEST_SUITE MACRO as is.
How do you suggest to proceed with this issue?
>
>> ---
>> tools/testing/selftests/kvm/Makefile | 1 +
>> .../kvm/x86_64/halt_disable_exit_test.c | 113 ++++++++++++++++++
> Add generated object to .gitignore file.
Sure. I will do it.
>
>> 2 files changed, 114 insertions(+)
>> create mode 100644 tools/testing/selftests/kvm/x86_64/halt_disable_exit_test.c
>>
>> diff --git a/tools/testing/selftests/kvm/Makefile b/tools/testing/selftests/kvm/Makefile
>> index c75251d5c97c..9f72abb95d2e 100644
>> --- a/tools/testing/selftests/kvm/Makefile
>> +++ b/tools/testing/selftests/kvm/Makefile
>> @@ -89,6 +89,7 @@ TEST_GEN_PROGS_x86_64 += x86_64/set_sregs_test
>> TEST_GEN_PROGS_x86_64 += x86_64/smaller_maxphyaddr_emulation_test
>> TEST_GEN_PROGS_x86_64 += x86_64/smm_test
>> TEST_GEN_PROGS_x86_64 += x86_64/state_test
>> +TEST_GEN_PROGS_x86_64 += x86_64/halt_disable_exit_test
>> TEST_GEN_PROGS_x86_64 += x86_64/vmx_preemption_timer_test
>> TEST_GEN_PROGS_x86_64 += x86_64/svm_vmcall_test
>> TEST_GEN_PROGS_x86_64 += x86_64/svm_int_ctl_test
>> diff --git a/tools/testing/selftests/kvm/x86_64/halt_disable_exit_test.c b/tools/testing/selftests/kvm/x86_64/halt_disable_exit_test.c
>> new file mode 100644
>> index 000000000000..b7279dd0eaff
>> --- /dev/null
>> +++ b/tools/testing/selftests/kvm/x86_64/halt_disable_exit_test.c
>> @@ -0,0 +1,113 @@
>> +// SPDX-License-Identifier: GPL-2.0-only
>> +/*
>> + * KVM disable halt exit test
>> + *
>> + * Copyright (C) 2024 Advanced Micro Devices, Inc.
>> + */
>> +#include <pthread.h>
>> +#include <signal.h>
>> +#include "kvm_util.h"
>> +#include "svm_util.h"
>> +#include "processor.h"
>> +#include "test_util.h"
>> +
>> +pthread_t task_thread, vcpu_thread;
>> +#define SIG_IPI SIGUSR1
>> +
>> +static void guest_code(uint8_t is_hlt_exec)
>> +{
>> + while (!READ_ONCE(is_hlt_exec))
>> + ;
>> +
>> + safe_halt();
>> + GUEST_DONE();
>> +}
>> +
>> +static void *task_worker(void *arg)
>> +{
>> + uint8_t *is_hlt_exec = (uint8_t *)arg;
>> +
>> + usleep(1000);
>> + WRITE_ONCE(*is_hlt_exec, 1);
>> + pthread_kill(vcpu_thread, SIG_IPI);
>> + return 0;
>> +}
>> +
>> +static void *vcpu_worker(void *arg)
>> +{
>> + int ret;
>> + int sig = -1;
>> + uint8_t *is_hlt_exec = (uint8_t *)arg;
>> + struct kvm_vm *vm;
>> + struct kvm_run *run;
>> + struct kvm_vcpu *vcpu;
>> + struct kvm_signal_mask *sigmask = alloca(offsetof(struct kvm_signal_mask, sigset)
>> + + sizeof(sigset_t));
>> + sigset_t *sigset = (sigset_t *) &sigmask->sigset;
>> +
>> + /* Create a VM without in kernel APIC support */
>> + vm = __vm_create(VM_SHAPE_DEFAULT, 1, 0, false);
>> + vm_enable_cap(vm, KVM_CAP_X86_DISABLE_EXITS, KVM_X86_DISABLE_EXITS_HLT);
>> + vcpu = vm_vcpu_add(vm, 0, guest_code);
>> + vcpu_args_set(vcpu, 1, *is_hlt_exec);
>> +
>> + /*
>> + * SIG_IPI is unblocked atomically while in KVM_RUN. It causes the
>> + * ioctl to return with -EINTR, but it is still pending and we need
>> + * to accept it with the sigwait.
>> + */
>> + sigmask->len = 8;
>> + pthread_sigmask(0, NULL, sigset);
>> + sigdelset(sigset, SIG_IPI);
>> + vcpu_ioctl(vcpu, KVM_SET_SIGNAL_MASK, sigmask);
>> + sigemptyset(sigset);
>> + sigaddset(sigset, SIG_IPI);
>> + run = vcpu->run;
>> +
>> +again:
>> + ret = __vcpu_run(vcpu);
>> + TEST_ASSERT_EQ(errno, EINTR);
>> +
>> + if (ret == -1 && errno == EINTR) {
>> + sigwait(sigset, &sig);
>> + assert(sig == SIG_IPI);
>> + TEST_ASSERT_KVM_EXIT_REASON(vcpu, KVM_EXIT_INTR);
>> + goto again;
>> + }
>> +
>> + if (run->exit_reason == KVM_EXIT_HLT)
>> + TEST_FAIL("Expected KVM_EXIT_INTR, got KVM_EXIT_HLT");
>> +
>> + TEST_ASSERT_KVM_EXIT_REASON(vcpu, KVM_EXIT_IO);
>> + kvm_vm_free(vm);
>> + return 0;
>> +}
>> +
>> +int main(int argc, char *argv[])
>> +{
>> + int ret;
>> + void *retval;
>> + uint8_t is_halt_exec;
>> + sigset_t sigset;
>> +
>> + TEST_REQUIRE(kvm_has_cap(KVM_CAP_X86_DISABLE_EXITS));
>> +
>> + /* Ensure that vCPU threads start with SIG_IPI blocked. */
>> + sigemptyset(&sigset);
>> + sigaddset(&sigset, SIG_IPI);
>> + pthread_sigmask(SIG_BLOCK, &sigset, NULL);
>> +
>> + ret = pthread_create(&vcpu_thread, NULL, vcpu_worker, &is_halt_exec);
>> + TEST_ASSERT(ret == 0, "pthread_create vcpu thread failed errno=%d", errno);
>> +
>> + ret = pthread_create(&task_thread, NULL, task_worker, &is_halt_exec);
>> + TEST_ASSERT(ret == 0, "pthread_create task thread failed errno=%d", errno);
>> +
>> + pthread_join(vcpu_thread, &retval);
>> + TEST_ASSERT(ret == 0, "pthread_join on vcpu thread failed with errno=%d", ret);
>> +
>> + pthread_join(task_thread, &retval);
>> + TEST_ASSERT(ret == 0, "pthread_join on task thread failed with errno=%d", ret);
>> +
>> + return 0;
>> +}
>
- Manali
next prev parent reply other threads:[~2024-04-01 5:28 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-03-27 5:42 [PATCH v1 0/3] Add a test case for KVM_X86_DISABLE_EXIT Manali Shukla
2024-03-27 5:42 ` [PATCH v1 1/3] KVM: selftests: Add safe_halt() and cli() helpers to common code Manali Shukla
2024-03-27 5:42 ` [PATCH v1 2/3] KVM: selftests: Change __vm_create() to create a vm without in-kernel APIC Manali Shukla
2024-03-27 11:27 ` Andrew Jones
2024-03-28 11:12 ` Manali Shukla
2024-03-27 5:42 ` [PATCH v1 3/3] KVM: selftests: Add a test case for KVM_X86_DISABLE_EXITS_HLT Manali Shukla
2024-03-29 20:13 ` Muhammad Usama Anjum
2024-04-01 5:28 ` Manali Shukla [this message]
2024-04-01 6:27 ` Manali Shukla
2024-04-01 9:41 ` Muhammad Usama Anjum
2024-04-01 12:31 ` Manali Shukla
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=5d667c81-56a8-95bd-4aa1-4df16c42dabf@amd.com \
--to=manali.shukla@amd.com \
--cc=kvm@vger.kernel.org \
--cc=linux-kselftest@vger.kernel.org \
--cc=nikunj@amd.com \
--cc=pbonzini@redhat.com \
--cc=seanjc@google.com \
--cc=shuah@kernel.org \
--cc=thomas.lendacky@amd.com \
--cc=usama.anjum@collabora.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