Re: [PATCH v2 11/11] KVM: riscv: selftests: Add get-reg-list test

From: Haibo Xu <xiaobo55x@gmail.com>
To: Andrew Jones <ajones@ventanamicro.com>
Cc: Haibo Xu <haibo1.xu@intel.com>,
	maz@kernel.org, oliver.upton@linux.dev, seanjc@google.com,
	Paolo Bonzini <pbonzini@redhat.com>,
	Jonathan Corbet <corbet@lwn.net>,
	Anup Patel <anup@brainfault.org>,
	Atish Patra <atishp@atishpatra.org>,
	Paul Walmsley <paul.walmsley@sifive.com>,
	Palmer Dabbelt <palmer@dabbelt.com>,
	Albert Ou <aou@eecs.berkeley.edu>, Shuah Khan <shuah@kernel.org>,
	James Morse <james.morse@arm.com>,
	Suzuki K Poulose <suzuki.poulose@arm.com>,
	Zenghui Yu <yuzenghui@huawei.com>,
	David Matlack <dmatlack@google.com>,
	Ben Gardon <bgardon@google.com>,
	Vipin Sharma <vipinsh@google.com>,
	Colton Lewis <coltonlewis@google.com>,
	kvm@vger.kernel.org, linux-doc@vger.kernel.org,
	linux-kernel@vger.kernel.org, kvm-riscv@lists.infradead.org,
	linux-riscv@lists.infradead.org, linux-kselftest@vger.kernel.org,
	linux-arm-kernel@lists.infradead.org, kvmarm@lists.linux.dev
Subject: Re: [PATCH v2 11/11] KVM: riscv: selftests: Add get-reg-list test
Date: Mon, 29 May 2023 20:20:58 +0800	[thread overview]
Message-ID: <CAJve8o=wX+Lb84YeGBa9anp+eZnrgvTW6AOVgoeUnHZu7eWswA@mail.gmail.com> (raw)
In-Reply-To: <20230529-7b3e41e6aca55f9e90dd6cd3@orel>

On Mon, May 29, 2023 at 3:08 PM Andrew Jones <ajones@ventanamicro.com> wrote:
>
> On Sat, May 27, 2023 at 12:39:57PM +0800, Haibo Xu wrote:
> > On Fri, May 26, 2023 at 1:18 AM Andrew Jones <ajones@ventanamicro.com> wrote:
> > >
> > > On Thu, May 25, 2023 at 03:38:35PM +0800, Haibo Xu wrote:
> ...
> > > > +     KVM_REG_RISCV | KVM_REG_SIZE_U64 | KVM_REG_RISCV_CORE | KVM_REG_RISCV_CORE_REG(regs.a6),
> > > > +     KVM_REG_RISCV | KVM_REG_SIZE_U64 | KVM_REG_RISCV_CORE | KVM_REG_RISCV_CORE_REG(regs.a7),
> > > > +     KVM_REG_RISCV | KVM_REG_SIZE_U64 | KVM_REG_RISCV_CORE | KVM_REG_RISCV_CORE_REG(regs.s2),
> > > > +     KVM_REG_RISCV | KVM_REG_SIZE_U64 | KVM_REG_RISCV_CORE | KVM_REG_RISCV_CORE_REG(regs.s3),
> > > > +     KVM_REG_RISCV | KVM_REG_SIZE_U64 | KVM_REG_RISCV_CORE | KVM_REG_RISCV_CORE_REG(regs.s4),
> > > > +     KVM_REG_RISCV | KVM_REG_SIZE_U64 | KVM_REG_RISCV_CORE | KVM_REG_RISCV_CORE_REG(regs.s5),
> > > > +     KVM_REG_RISCV | KVM_REG_SIZE_U64 | KVM_REG_RISCV_CORE | KVM_REG_RISCV_CORE_REG(regs.s6),
> > > > +     KVM_REG_RISCV | KVM_REG_SIZE_U64 | KVM_REG_RISCV_CORE | KVM_REG_RISCV_CORE_REG(regs.s7),
> > > > +     KVM_REG_RISCV | KVM_REG_SIZE_U64 | KVM_REG_RISCV_CORE | KVM_REG_RISCV_CORE_REG(regs.s8),
> > > > +     KVM_REG_RISCV | KVM_REG_SIZE_U64 | KVM_REG_RISCV_CORE | KVM_REG_RISCV_CORE_REG(regs.s9),
> > > > +     KVM_REG_RISCV | KVM_REG_SIZE_U64 | KVM_REG_RISCV_CORE | KVM_REG_RISCV_CORE_REG(regs.s10),
> > > > +     KVM_REG_RISCV | KVM_REG_SIZE_U64 | KVM_REG_RISCV_CORE | KVM_REG_RISCV_CORE_REG(regs.s11),
> > > > +     KVM_REG_RISCV | KVM_REG_SIZE_U64 | KVM_REG_RISCV_CORE | KVM_REG_RISCV_CORE_REG(regs.t3),
> > > > +     KVM_REG_RISCV | KVM_REG_SIZE_U64 | KVM_REG_RISCV_CORE | KVM_REG_RISCV_CORE_REG(regs.t4),
> > > > +     KVM_REG_RISCV | KVM_REG_SIZE_U64 | KVM_REG_RISCV_CORE | KVM_REG_RISCV_CORE_REG(regs.t5),
> > > > +     KVM_REG_RISCV | KVM_REG_SIZE_U64 | KVM_REG_RISCV_CORE | KVM_REG_RISCV_CORE_REG(regs.t6),
> > >
> > > ...all the above would just be indices rather than named registers. I
> > > guess that's better for these registers.
> > >
> >
> > You mean to show it as KVM_REG_RISCV_CORE_REG(regs.regs[0]) ...
> > KVM_REG_RISCV_CORE_REG(regs.regs[31])?
> >
>
> I'm OK with these registers using their names in this list, it does look
> better. However the original idea for these lists was that they would be
> generated from print_reg(). In this case, print_reg() is generating them
> with their number instead of name. Either print_reg() could learn how to
> generate their names by handling the offset ranges of each register type,
> e.g.
>
> switch (reg_off) {
> case 10 ... 17:
>    strdup_printf("... KVM_REG_RISCV_CORE_REG(regs.a%d),", reg_off - 10);
>
> or we can use the numbers here in this list, or we can leave it as you
> have it (i.e. done manually).
>

I agree that the print_reg() should print the register name instead of
the register offset.
It may be better to manually encode/decode the register offset to name
case by case
in the print_reg() since we just need to handle it once. Will update it in v3.

Thanks,
Haibo

> Thanks,
> drew