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;
MIME-Version: 1.0
References: <20230328111714.2056-1-zhangqing@loongson.cn> <9817aaa043e9f0ed964bd523773447bd64f6e2c0.camel@xry111.site>
 <1c231587-3b70-22ab-d554-ebe3de407909@loongson.cn>
In-Reply-To: <1c231587-3b70-22ab-d554-ebe3de407909@loongson.cn>
From:   Huacai Chen <chenhuacai@kernel.org>
Date:   Wed, 29 Mar 2023 08:22:37 +0800
Message-ID: <CAAhV-H5APsBxC8nNa81t3HXum1EU1hOj4S6UC7xLHD7_BCJd7g@mail.gmail.com>
Subject: Re: [PATCH] LoongArch: Add kernel address sanitizer support
To:     Qing Zhang <zhangqing@loongson.cn>
Cc:     Xi Ruoyao <xry111@xry111.site>,
        Andrey Ryabinin <ryabinin.a.a@gmail.com>,
        Jonathan Corbet <corbet@lwn.net>,
        Andrew Morton <akpm@linux-foundation.org>,
        Alexander Potapenko <glider@google.com>,
        Andrey Konovalov <andreyknvl@gmail.com>,
        Dmitry Vyukov <dvyukov@google.com>,
        Vincenzo Frascino <vincenzo.frascino@arm.com>,
        WANG Xuerui <kernel@xen0n.name>,
        Jiaxun Yang <jiaxun.yang@flygoat.com>,
        kasan-dev@googlegroups.com, linux-doc@vger.kernel.org,
        linux-mm@kvack.org, loongarch@lists.linux.dev,
        linux-kernel@vger.kernel.org, linux-hardening@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Tue, Mar 28, 2023 at 8:15=E2=80=AFPM Qing Zhang <zhangqing@loongson.cn> =
wrote:
>
> Hi, Ruoyao
>
> On 2023/3/28 =E4=B8=8B=E5=8D=887:39, Xi Ruoyao wrote:
> > On Tue, 2023-03-28 at 19:17 +0800, Qing Zhang wrote:
> >
> > /* snip */
> >
> >
> >> -void * __init relocate_kernel(void)
> >> +unsigned long __init relocate_kernel(void)
> >
> > Why we must modify relocate_kernel for KASAN?
>
> When the CONFIG_RANDOMIZE_BASE is enabled, the kernel will be updated to
> a random new address.
> Kasan needs to call kasan_early_init before start_kernel.
> There are two situations:
> 1> After enabling CONFIG_RELOCATABLE, call kasan_early_init.
> 2> After CONFIG_RELOCATABLE is not enabled, call kasan_early_init.
>
> In order to prevent code redundancy and semantic problems caused by
> calling kasan_early_init (before jr a0) at the old PC.
In my opinion, you can call kasan_early_init before relocate_kernel in
head.S, then no redundancy.

Huacai
>
> Thanks,
> -Qing
> >
> >>   {
> >>          unsigned long kernel_length;
> >>          unsigned long random_offset =3D 0;
> >>          void *location_new =3D _text; /* Default to original kernel s=
tart */
> >> -       void *kernel_entry =3D start_kernel; /* Default to original ke=
rnel entry point */
> >>          char *cmdline =3D early_ioremap(fw_arg1, COMMAND_LINE_SIZE); =
/* Boot command line is passed in fw_arg1 */
> >>
> >>          strscpy(boot_command_line, cmdline, COMMAND_LINE_SIZE);
> >> @@ -190,9 +189,6 @@ void * __init relocate_kernel(void)
> >>
> >>                  reloc_offset +=3D random_offset;
> >>
> >> -               /* Return the new kernel's entry point */
> >> -               kernel_entry =3D RELOCATED_KASLR(start_kernel);
> >> -
> >>                  /* The current thread is now within the relocated ker=
nel */
> >>                  __current_thread_info =3D RELOCATED_KASLR(__current_t=
hread_info);
> >>
> >> @@ -204,7 +200,7 @@ void * __init relocate_kernel(void)
> >>
> >>          relocate_absolute(random_offset);
> >>
> >> -       return kernel_entry;
> >> +       return random_offset;
> >
>