Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Tue, 22 Feb 2022 13:25:58 -0800 (PST)
Subject:     Re: [PATCH] riscv: mm: Remove the copy operation of pmd
In-Reply-To: <20220127100622.11856-1-hanchuanhua@oppo.com>
CC:     Paul Walmsley <paul.walmsley@sifive.com>, aou@eecs.berkeley.edu,
        jszhang@kernel.org, alexandre.ghiti@canonical.com,
        wangkefeng.wang@huawei.com, zhengqi.arch@bytedance.com,
        ebiederm@xmission.com, liushixin2@huawei.com,
        linux-riscv@lists.infradead.org, linux-kernel@vger.kernel.org,
        hanchuanhua@oppo.com
From:   Palmer Dabbelt <palmer@dabbelt.com>
To:     hanchuanhua@oppo.com
Message-ID: <mhng-98c8a68c-ecb2-4db1-ab3e-3a35d5e764d0@palmer-ri-x1c9>
Mime-Version: 1.0 (MHng)
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: 8bit
Precedence: bulk

On Thu, 27 Jan 2022 02:06:22 PST (-0800), hanchuanhua@oppo.com wrote:
> Since all processes share the kernel address space,
> we only need to copy pgd in case of a vmalloc page
> fault exception, the other levels of page tables are
> shared, so the operation of copying pmd is unnecessary.
>
> Signed-off-by: hanchuanhua <hanchuanhua@oppo.com>
> ---
>  arch/riscv/mm/fault.c | 10 +++-------
>  1 file changed, 3 insertions(+), 7 deletions(-)
>
> diff --git a/arch/riscv/mm/fault.c b/arch/riscv/mm/fault.c
> index 4e9efbe..40694f0 100644
> --- a/arch/riscv/mm/fault.c
> +++ b/arch/riscv/mm/fault.c
> @@ -102,9 +102,9 @@ static inline void bad_area(struct pt_regs *regs, struct mm_struct *mm, int code
>  static inline void vmalloc_fault(struct pt_regs *regs, int code, unsigned long addr)
>  {
>         pgd_t *pgd, *pgd_k;
> -       pud_t *pud, *pud_k;
> -       p4d_t *p4d, *p4d_k;
> -       pmd_t *pmd, *pmd_k;
> +       pud_t *pud_k;
> +       p4d_t *p4d_k;
> +       pmd_t *pmd_k;
>         pte_t *pte_k;
>         int index;
>         unsigned long pfn;
> @@ -132,14 +132,12 @@ static inline void vmalloc_fault(struct pt_regs *regs, int code, unsigned long a
>         }
>         set_pgd(pgd, *pgd_k);
>
> -       p4d = p4d_offset(pgd, addr);
>         p4d_k = p4d_offset(pgd_k, addr);
>         if (!p4d_present(*p4d_k)) {
>                 no_context(regs, addr);
>                 return;
>         }
>
> -       pud = pud_offset(p4d, addr);
>         pud_k = pud_offset(p4d_k, addr);
>         if (!pud_present(*pud_k)) {
>                 no_context(regs, addr);
> @@ -150,13 +148,11 @@ static inline void vmalloc_fault(struct pt_regs *regs, int code, unsigned long a
>          * Since the vmalloc area is global, it is unnecessary
>          * to copy individual PTEs
>          */
> -       pmd = pmd_offset(pud, addr);
>         pmd_k = pmd_offset(pud_k, addr);
>         if (!pmd_present(*pmd_k)) {
>                 no_context(regs, addr);
>                 return;
>         }
> -       set_pmd(pmd, *pmd_k);
>
>         /*
>          * Make sure the actual PTE exists as well to

This looks good, but "Signed off by" lines are supposed to have real 
names associated with them and this looks like a handle to me.  There 
also seems to be something broken with the email, as this is causing git 
to blow up, but for a small patch like this that's OK as I can resurrect 
it by hand.

Reviewed-by: Palmer Dabbelt <palmer@rivosinc.com> # aside from the name

Please send a v2 with the proper Signed off by line, as that's not 
something I'm supposed to just fix up myself.

Thanks!