Re: [PATCH 2/2] KVM: riscv: selftests: Detect supported vm modes

From: "Radim Krčmář" <rkrcmar@ventanamicro.com>
To: "Wu Fei" <atwufei@163.com>, "Andrew Jones" <ajones@ventanamicro.com>
Cc: <wu.fei9@sanechips.com.cn>, <kvm-riscv@lists.infradead.org>,
	"kvm-riscv" <kvm-riscv-bounces@lists.infradead.org>
Subject: Re: [PATCH 2/2] KVM: riscv: selftests: Detect supported vm modes
Date: Fri, 24 Oct 2025 17:59:01 +0200	[thread overview]
Message-ID: <DDQODLNQSLO3.E5HMVVMXWDVN@ventanamicro.com> (raw)
In-Reply-To: <7aa181d9-84fe-41e1-af6e-14a3ad6629ef@163.com>

2025-10-24T21:48:01+08:00, Wu Fei <atwufei@163.com>:
> On 10/22/25 22:26, Radim Krčmář wrote:
>> For tests that that don't care about physical width, we could
>> add:
>> 
>>    VM_MODE_PXXV39_4K
>
> Intel adds VM_MODE_PXXV48_4K because "x86_64 machines are having various 
> physical address width rather than some static values" according to 
> commit 567a9f1e9de, and x86 has to probe the real gpa width on vm 
> creation. RISC-V has 3 fixed gpa widths, so maybe we can use the 
> explicit widths?

Makes sense.

>> It doesn't seem necessary to support more than that, but if our tests
>> want larger virtual address space, we can add V48 and V57 variants.
>
> I think V48/V57 is necessary, if a machine is designed to have SV48 
> guests, it's better to test it.

Right.  We also get information about available guest satp modes from
KVM, so we ought to test that it was correct.

> So is it okay to have the combination of all 9 modes?
> 	P(41, 50, 56) x V(39, 48, 57)
> But only several ones are picked up for a specific machine, just as 
> mentioned above, we won't select P41V39 for Sv48x4 etc.

Sounds good, thanks.

-- 
kvm-riscv mailing list
kvm-riscv@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/kvm-riscv