Received-SPF: pass (google.com: domain of linux-kernel+bounces-47363-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.48.161 as permitted sender) client-ip=147.75.48.161;
Date: Wed, 31 Jan 2024 15:20:40 -0800
In-Reply-To: <20231102155111.28821-4-guang.zeng@intel.com>
Precedence: bulk
Mime-Version: 1.0
References: <20231102155111.28821-1-guang.zeng@intel.com> <20231102155111.28821-4-guang.zeng@intel.com>
Message-ID: <ZbrVyB1Pj9KrH-Cl@google.com>
Subject: Re: [RFC PATCH v1 3/8] KVM: selftests: Add virt_arch_ucall_prealloc()
 arch specific implementation
From: Sean Christopherson <seanjc@google.com>
To: Zeng Guang <guang.zeng@intel.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>, Shuah Khan <shuah@kernel.org>, Marc Zyngier <maz@kernel.org>, 
	Oliver Upton <oliver.upton@linux.dev>, James Morse <james.morse@arm.com>, 
	Suzuki K Poulose <suzuki.poulose@arm.com>, Zenghui Yu <yuzenghui@huawei.com>, 
	Anup Patel <anup@brainfault.org>, Atish Patra <atishp@atishpatra.org>, 
	David Hildenbrand <david@redhat.com>, kvm@vger.kernel.org, linux-kselftest@vger.kernel.org, 
	linux-kernel@vger.kernel.org, linux-arm-kernel@lists.infradead.org, 
	kvmarm@lists.linux.dev, kvm-riscv@lists.infradead.org, 
	linux-riscv@lists.infradead.org
Content-Type: text/plain; charset="us-ascii"

On Thu, Nov 02, 2023, Zeng Guang wrote:
> diff --git a/tools/testing/selftests/kvm/include/kvm_util_base.h b/tools/testing/selftests/kvm/include/kvm_util_base.h
> index a18db6a7b3cf..dbaa2cf83c1c 100644
> --- a/tools/testing/selftests/kvm/include/kvm_util_base.h
> +++ b/tools/testing/selftests/kvm/include/kvm_util_base.h
> @@ -917,6 +917,23 @@ static inline void virt_dump(FILE *stream, struct kvm_vm *vm, uint8_t indent)
>  	virt_arch_dump(stream, vm, indent);
>  }
>  
> +/*
> + * Virtual UCALL memory pre-processing
> + *
> + * Input Args:
> + *   ucall_gva - Guest virtual address point to memory of ucall pool
> + *
> + * Output Args: None
> + *
> + * Return:
> + *   Processed guest virtual address point to memory of ucall pool
> + */

Please omit the massive comments, they are yet another misguided remnant in
selftests that we are purging.

> diff --git a/tools/testing/selftests/kvm/lib/x86_64/processor.c b/tools/testing/selftests/kvm/lib/x86_64/processor.c
> index 6f4295a13d00..525b714ee13c 100644
> --- a/tools/testing/selftests/kvm/lib/x86_64/processor.c
> +++ b/tools/testing/selftests/kvm/lib/x86_64/processor.c
> @@ -388,6 +388,18 @@ void virt_arch_dump(FILE *stream, struct kvm_vm *vm, uint8_t indent)
>  	}
>  }
>  
> +void *virt_arch_ucall_prealloc(uint64_t ucall_gva)
> +{
> +	unsigned short desc_cs;
> +
> +	asm volatile ("mov %%cs,%0" :  "=r" (desc_cs));

Strictly speaking, CS.DPL is not the source of truth for CPL, SS.DPL is.  But
that's probably a moot point, because I again think this is a hack that shows the
overall approach isn't maintainable.

Can you post the actual usage of userspace selftests, i.e. the "full" series?
It's really hard to build a mental model of how this all fits together without
seeing the actual usage.