Received-SPF: pass (google.com: domain of linux-kernel+bounces-104417-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45d1:ec00::1 as permitted sender) client-ip=2604:1380:45d1:ec00::1;
Precedence: bulk
MIME-Version: 1.0
References: <20240315063154.696633-1-liuyuntao12@huawei.com>
 <CAMj1kXH+FaddHV5--kqB_wVgw_M682MvchPB1BoCuDuA6vVyvg@mail.gmail.com>
 <d7a1996f-f558-4cea-b8ca-eb071d02c6d4@huawei.com> <CAMj1kXHP-M7hTsCmorgzvDiPB1jRXHciJrXjJR82zmkxUTD-5Q@mail.gmail.com>
 <b3e3f664-f8ad-41d7-b8d5-3c83818ff489@huawei.com> <d91cf190-e583-422f-884d-79552b674f72@app.fastmail.com>
 <baed1e49-d854-4517-8219-2a5bb0b3cd1b@huawei.com> <CAMj1kXG5dHsdGn0_hSv2vPyMJbe1zOf4HLprLUSKJ+Dy2e_DzQ@mail.gmail.com>
 <31f3a3fe-b16f-47a9-8c41-80fc0a236d52@huawei.com> <CAMj1kXEgQLizmzUwvw79DQ9zkq2L5ed5576MzmYX6uHA5QUh1Q@mail.gmail.com>
 <dc3dad0c-d444-4ea0-8d6f-0a0d850d948d@huawei.com>
In-Reply-To: <dc3dad0c-d444-4ea0-8d6f-0a0d850d948d@huawei.com>
From: Ard Biesheuvel <ardb@kernel.org>
Date: Fri, 15 Mar 2024 13:58:14 +0100
Message-ID: <CAMj1kXEkd-MAN8x88DPTnkZzojg8TRbTgzujO+DZCpo58mZxaA@mail.gmail.com>
Subject: Re: [PATCH-next v3] arm32: enable HAVE_LD_DEAD_CODE_DATA_ELIMINATION
To: "liuyuntao (F)" <liuyuntao12@huawei.com>
Cc: Arnd Bergmann <arnd@arndb.de>, linux-arm-kernel@lists.infradead.org, 
	linux-kernel@vger.kernel.org, Geert Uytterhoeven <geert@linux-m68k.org>, 
	Russell King <linux@armlinux.org.uk>, Andrew Davis <afd@ti.com>, 
	Andrew Morton <akpm@linux-foundation.org>, 
	"Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>, 
	Geert Uytterhoeven <geert+renesas@glider.be>, Jonathan Corbet <corbet@lwn.net>, 
	Mike Rapoport <rppt@kernel.org>, Rob Herring <robh@kernel.org>, Thomas Gleixner <tglx@linutronix.de>, 
	Linus Walleij <linus.walleij@linaro.org>, Fangrui Song <maskray@google.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Fri, 15 Mar 2024 at 13:52, liuyuntao (F) <liuyuntao12@huawei.com> wrote:
>
>
>
> On 2024/3/15 20:46, Ard Biesheuvel wrote:
> > On Fri, 15 Mar 2024 at 13:22, liuyuntao (F) <liuyuntao12@huawei.com> wr=
ote:
> >>
> >>
> >>
> >> On 2024/3/15 20:13, Ard Biesheuvel wrote:
> >>> On Fri, 15 Mar 2024 at 13:10, liuyuntao (F) <liuyuntao12@huawei.com> =
wrote:
> >>>>
> >>>>
> >>>>
> >>>> On 2024/3/15 19:51, Arnd Bergmann wrote:
> >>>>> On Fri, Mar 15, 2024, at 12:48, liuyuntao (F) wrote:
> >>>>>> On 2024/3/15 18:46, Ard Biesheuvel wrote:
> >>>>>>> @@ -28,6 +28,7 @@
> >>>>>>>                                        -DEFI_HAVE_MEMCHR -DEFI_HA=
VE_STRRCHR \
> >>>>>>>                                        -DEFI_HAVE_STRCMP -fno-bui=
ltin -fpic \
> >>>>>>>                                        $(call cc-option,-mno-sing=
le-pic-base)
> >>>>>>> +cflags-$(CONFIG_ARM)           :=3D $(filter-out -fdata-sections=
, $(CFLAGS-y))
> >>>>>>>      cflags-$(CONFIG_RISCV)         +=3D -fpic -DNO_ALTERNATIVE -=
mno-relax
> >>>>>>>      cflags-$(CONFIG_LOONGARCH)     +=3D -fpie
> >>>>>>
> >>>>>> Another error:
> >>>>>> cannot initialize array of =E2=80=98short unsigned int=E2=80=99 fr=
om a string literal
> >>>>>> with type array of =E2=80=98unsigned int=E2=80=99
> >>>>>>      17 | static const efi_char16_t shim_MokSBState_name[] =3D L"M=
okSBStateRT";
> >>>>>
> >>>>> The line sets cflags-y to an empty string.
> >>>>>
> >>>>> It probably should have been
> >>>>>
> >>>>> cflags-$(CONFIG_ARM)           :=3D $(filter-out -fdata-sections, $=
(cflags-y))
> >>>>>
> >>>>> (small cflags instead of CFLAGS).
> >>>>>
> >>>
> >>> Oops indeed.
> >>>
> >>>> Oh, my mistake, after fixed this bug, still this error:
> >>>>> arm-linux-gnueabi-ld: warning: orphan section `.data.efi_loglevel' =
from `drivers/firmware/efi/libstub/printk.stub.o' being placed in section `=
data.efi_loglevel'
> >>>>> arm-linux-gnueabi-ld: warning: orphan section `.data.screen_info_gu=
id' from `drivers/firmware/efi/libstub/screen_info.stub.o' being placed in =
section `.data.screen_info_guid'
> >>>>> arm-linux-gnueabi-ld: warning: orphan section `.data.cpu_state_guid=
' from `drivers/firmware/efi/libstub/arm32-stub.stub.o' being placed in sec=
tion `.data.cpu_state_guid'
> >>>>> arm-linux-gnueabi-ld: warning: orphan section `.data.efi_nokaslr' f=
rom `drivers/firmware/efi/libstub/efi-stub-helper.stub.o' being placed in s=
ection `.data.efi_nokaslr'
> >>>>
> >>>> The -fdata-sections option is present in the KBUILD_CFLAGS_KERNEL
> >>>> variable; altering cflags-y only affect the KBUILD_CFLAGS variable.
> >>>> It seems that the -fdata-sections option needs to be removed from th=
e
> >>>> KBUILD_CFLAGS_KERNEL variable.
> >>>
> >>> Indeed.
> >>>
> >>> So
> >>>
> >>> KBUILD_CFLAGS_KERNEL :=3D $(filter-out -fdata-sections,$(KBUILD_CFLAG=
S_KERNEL))
> >>>
> >>> (assuming I got it right this time without build testing)
> >>
> >> Um, it not worked,too, same error.
> >
> > Aren't you missing a 'L' ?
> >
> >>> --- a/drivers/firmware/efi/libstub/Makefile
> >>> +++ b/drivers/firmware/efi/libstub/Makefile
> >>> @@ -28,6 +28,7 @@ cflags-$(CONFIG_ARM)          +=3D -DEFI_HAVE_STRLE=
N -DEFI_HAVE_STRNLEN \
> >>>                                     -DEFI_HAVE_MEMCHR -DEFI_HAVE_STRR=
CHR \
> >>>
> >
> >>> ifdef CONFIG_LD_DEAD_CODE_DATA_ELIMINATION
> >>> KBUILD_CFLAGS_KERNEL +=3D -ffunction-sections -fdata-sections
> >>> KBUILD_RUSTFLAGS_KERNEL +=3D -Zfunction-sections=3Dy
> >>> LDFLAGS_vmlinux +=3D --gc-sections
> >>> endif
> >>
> >>                                    -DEFI_HAVE_STRCMP -fno-builtin -fpi=
c \
> >>>                                     $(call cc-option,-mno-single-pic-=
base)
> >>> +KBUILD_CFLAGS_KERNE            :=3D $(filter-out -fdata-sections, $(=
KBUILD_CFLAGS_KERNE))
> >>>   cflags-$(CONFIG_RISCV)         +=3D -fpic -DNO_ALTERNATIVE -mno-rel=
ax
> >>>   cflags-$(CONFIG_LOONGARCH)     +=3D -fpie
> >>
> >>
> Sorry, I was too careless.
> How about reset KBUILD_CFLAGS_KERNEL?
>

No, that is not the solution.

Either avoid setting -fdata-sections altogether, or filter it out.


> > +# KBUILD_CFLAGS_KERNEL is is set by CONFIG_LD_DEAD_CODE_DATA_ELIMINATI=
ON,
> > +# not need here, just reset it.
> > +KBUILD_CFLAGS_KERNEL :=3D
> > +
> >  # non-x86 reuses KBUILD_CFLAGS, x86 does not
> >  cflags-y                       :=3D $(KBUILD_CFLAGS)
>