Content-Type: text/plain;
        charset=utf-8
Mime-Version: 1.0 (Mac OS X Mail 16.0 \(3696.120.41.1.1\))
Subject: Re: [PATCH v1] riscv: support arch_has_hw_pte_young()
From:   Jessica Clarke <jrtc27@jrtc27.com>
In-Reply-To: <20230130104900.muwgikgkvkn2byna@orel>
Date:   Mon, 30 Jan 2023 17:27:18 +0000
Cc:     Anup Patel <apatel@ventanamicro.com>, Jinyu Tang <tjytimi@163.com>,
        Palmer Dabbelt <palmer@rivosinc.com>,
        Paul Walmsley <paul.walmsley@sifive.com>,
        Palmer Dabbelt <palmer@dabbelt.com>, yuzhao@google.com,
        Conor Dooley <conor.dooley@microchip.com>,
        Guo Ren <guoren@kernel.org>, tongtiangen@huawei.com,
        Anup Patel <anup@brainfault.org>,
        Andrew Morton <akpm@linux-foundation.org>, falcon@tinylab.org,
        linux-riscv <linux-riscv@lists.infradead.org>,
        linux-kernel@vger.kernel.org
Content-Transfer-Encoding: quoted-printable
Message-Id: <EBC27F74-6AB4-43B3-90B9-107C05467167@jrtc27.com>
References: <20230129064956.143664-1-tjytimi@163.com>
 <CAK9=C2VaxNezv+M=s=ZSA-YDf_17+hcom_z1=_j2YkRHg-yg7Q@mail.gmail.com>
 <20230130104900.muwgikgkvkn2byna@orel>
To:     Andrew Jones <ajones@ventanamicro.com>
Precedence: bulk

On 30 Jan 2023, at 10:49, Andrew Jones <ajones@ventanamicro.com> wrote:
>=20
> On Mon, Jan 30, 2023 at 03:55:55PM +0530, Anup Patel wrote:
>> On Sun, Jan 29, 2023 at 12:21 PM Jinyu Tang <tjytimi@163.com> wrote:
>>>=20
>>> The arch_has_hw_pte_young() is false for riscv by default. If it's
>>> false, page table walk is almost skipped for MGLRU reclaim. And it
>>> will also cause useless step in __wp_page_copy_user().
>>>=20
>>> RISC-V Privileged Book says that riscv have two schemes to manage A
>>> and D bit.
>>>=20
>>> So add a config for selecting, the default is true. For simple
>>> implementation riscv CPU which just generate page fault, unselect =
it.
>>=20
>> I totally disagree with this approach.
>>=20
>> Almost all existing RISC-V platforms don't have HW support
>> PTE.A and PTE.D updates.
>>=20
>> We want the same kernel image to run HW with/without PTE.A
>> and PTE.D updates so kconfig based approach is not going to
>> fly.
>>=20
>>>=20
>>> Signed-off-by: Jinyu Tang <tjytimi@163.com>
>>> ---
>>> arch/riscv/Kconfig               | 10 ++++++++++
>>> arch/riscv/include/asm/pgtable.h |  7 +++++++
>>> 2 files changed, 17 insertions(+)
>>>=20
>>> diff --git a/arch/riscv/Kconfig b/arch/riscv/Kconfig
>>> index e2b656043abf..17c82885549c 100644
>>> --- a/arch/riscv/Kconfig
>>> +++ b/arch/riscv/Kconfig
>>> @@ -180,6 +180,16 @@ config PAGE_OFFSET
>>>        default 0x80000000 if 64BIT && !MMU
>>>        default 0xff60000000000000 if 64BIT
>>>=20
>>> +config ARCH_HAS_HARDWARE_PTE_YOUNG
>>> +       bool "Hardware Set PTE Access Bit"
>>> +       default y
>>> +       help
>>> +         Select if hardware set A bit when PTE is accessed. The =
default is
>>> +         'Y', because most RISC-V CPU hardware can manage A and D =
bit.
>>> +         But RISC-V may have simple implementation that do not =
support
>>> +         hardware set A bit but only generate page fault, for that =
case just
>>> +         unselect it.
>>> +
>>> config KASAN_SHADOW_OFFSET
>>>        hex
>>>        depends on KASAN_GENERIC
>>> diff --git a/arch/riscv/include/asm/pgtable.h =
b/arch/riscv/include/asm/pgtable.h
>>> index 4eba9a98d0e3..1db54ab4e1ba 100644
>>> --- a/arch/riscv/include/asm/pgtable.h
>>> +++ b/arch/riscv/include/asm/pgtable.h
>>> @@ -532,6 +532,13 @@ static inline int ptep_clear_flush_young(struct =
vm_area_struct *vma,
>>>         */
>>>        return ptep_test_and_clear_young(vma, address, ptep);
>>> }
>>> +#ifdef CONFIG_ARCH_HAS_HARDWARE_PTE_YOUNG
>>=20
>>> +#define arch_has_hw_pte_young arch_has_hw_pte_young
>>> +static inline bool arch_has_hw_pte_young(void)
>>> +{
>>> +       return true;
>>=20
>> Drop the kconfig option ARCH_HAS_HARDWARE_PTE_YOUNG
>> and instead use code patching to return true only when Svadu
>> ISA extension is available in DT ISA string.
>=20
> Indeed. I should have checked if there was an extension for this
> first. It crossed my mind that we should only be enabling features
> when the extensions are present, but looking at the privileged manual
> isn't sufficient to learn about the Svadu extension. I should have
> checked https://wiki.riscv.org/display/HOME/Specification+Status
>=20
> Anyway, I retract my r-b and agree with Anup.

Svadu is a bit of a mess, for years it=E2=80=99s been legal to implement
hardware A/D tracking and such implementations exist (it=E2=80=99s what =
QEMU
has done for many years, and I know of an FPGA-based implementation
that does it too), yet RVA20S64 outlaws that by requiring what it calls
Ssptead and Svadu gets introduced to re-allow that behaviour gated
behind a CSR bit.

Jess