Received-SPF: pass (google.com: domain of linux-kernel+bounces-153669-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.199.223 as permitted sender) client-ip=147.75.199.223;
Precedence: bulk
MIME-Version: 1.0
References: <20240422060556.1226848-1-liuyuntao12@huawei.com>
In-Reply-To: <20240422060556.1226848-1-liuyuntao12@huawei.com>
From: Masahiro Yamada <masahiroy@kernel.org>
Date: Tue, 23 Apr 2024 01:01:55 +0900
Message-ID: <CAK7LNAQgkt6t6UEB+_q15KJb2STVL6oqUo3mFM8EzumFH+-mYw@mail.gmail.com>
Subject: Re: [PATCH] x86: enable HAVE_LD_DEAD_CODE_DATA_ELIMINATION
To: Yuntao Liu <liuyuntao12@huawei.com>
Cc: linux-kernel@vger.kernel.org, linux-kbuild@vger.kernel.org, 
	tglx@linutronix.de, mingo@redhat.com, bp@alien8.de, 
	dave.hansen@linux.intel.com, hpa@zytor.com, nathan@kernel.org, 
	nicolas@fjasle.eu, peterz@infradead.org, jpoimboe@kernel.org, 
	leitao@debian.org, petr.pavlu@suse.com, richard.weiyang@gmail.com, 
	ruanjinjie@huawei.com, ndesaulniers@google.com, jgross@suse.com
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Mon, Apr 22, 2024 at 3:41=E2=80=AFPM Yuntao Liu <liuyuntao12@huawei.com>=
 wrote:
>
> The current x86 architecture does not yet support the
> HAVE_LD_DEAD_CODE_DATA_ELIMINATION feature. x86 is widely used in
> embedded scenarios, and enabling this feature would be beneficial for
> reducing the size of the kernel image.
>
> In order to make this work, we keep the necessary tables by annotating
> them with KEEP, also it requires further changes to linker script to KEEP
> some tables and wildcard compiler generated sections into the right place=
.
>
> Enabling CONFIG_UNWINDER_ORC or CONFIG_MITIGATION_RETPOLINE will enable
> the objtool's --orc and --retpoline parameters, which will alter the
> layout of the binary file, thereby preventing gc-sections from functionin=
g
> properly. Therefore, HAVE_LD_DEAD_CODE_DATA_ELIMINATION should only be
> selected when they are not enabled.
>
> Enabling CONFIG_LTO_CLANG or CONFIG_X86_KERNEL_IBT will use vmlinux.o
> instead of performing the slow LTO link again. This can also prevent
> gc-sections from functioning properly. Therefore, using this optimization
> when CONFIG_LD_DEAD_CODE_DATA_ELIMINATION is not enabled.
>
> The size comparison of zImage is as follows:
> x86_def_defconfig  i386_defconfig    tinyconfig
> 10892288           10826240          607232          no dce
> 10748928           10719744          529408          dce
> 1.3%               0.98%             12.8%           shrink
>
> When using smaller config file, there is a significant reduction in the
> size of the zImage.
> ---

> diff --git a/scripts/link-vmlinux.sh b/scripts/link-vmlinux.sh
> index 7862a8101747..7287b5a9f17d 100755
> --- a/scripts/link-vmlinux.sh
> +++ b/scripts/link-vmlinux.sh
> @@ -60,7 +60,7 @@ vmlinux_link()
>         # skip output file argument
>         shift
>
> -       if is_enabled CONFIG_LTO_CLANG || is_enabled CONFIG_X86_KERNEL_IB=
T; then
> +       if [ is_enabled CONFIG_LTO_CLANG || is_enabled CONFIG_X86_KERNEL_=
IBT ] && [ ! is_enabled CONFIG_LD_DEAD_CODE_DATA_ELIMINATION ]; then
>                 # Use vmlinux.o instead of performing the slow LTO link a=
gain.
>                 objs=3Dvmlinux.o
>                 libs=3D
> --


This is wrong.
You should not put is_enabled inside [ ... ]

(is_enabled CONFIG_LTO_CLANG || is_enabled CONFIG_X86_KERNEL_IBT) && !
is_enabled CONFIG_LD_DEAD_CODE_DATA_ELIMINATION

is still weird.


When CONFIG_LTO_CLANG=3Dy and CONFIG_LD_DEAD_CODE_DATA_ELIMINATION=3Dy,
the result of LTO will be discarded.


--=20
Best Regards
Masahiro Yamada