Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1332272 > unrolled thread

Re: [PATCH v2] ARM: mm: mark section-aligned portion of rodata NX

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  Re: [PATCH v2] ARM: mm: mark section-aligned portion of rodata NX Russell King - ARM Linux <linux@arm.linux.org.uk> - 2016-02-11 19:10 +0100
    Re: [PATCH v2] ARM: mm: mark section-aligned portion of rodata NX Arnd Bergmann <arnd@arndb.de> - 2016-02-12 10:50 +0100
      Re: [PATCH v2] ARM: mm: mark section-aligned portion of rodata NX Geert Uytterhoeven <geert@linux-m68k.org> - 2016-02-15 09:50 +0100

#1332272 — Re: [PATCH v2] ARM: mm: mark section-aligned portion of rodata NX

On Fri, Jan 22, 2016 at 01:19:03PM -0800, Kees Cook wrote:
> On Tue, Dec 8, 2015 at 10:38 AM, Kees Cook <keescook@chromium.org> wrote:
> > On Mon, Dec 7, 2015 at 11:47 PM, Ard Biesheuvel
> > <ard.biesheuvel@linaro.org> wrote:
> >> On 7 December 2015 at 23:35, Kees Cook <keescook@chromium.org> wrote:
> >>>  /*
> >>> + * Without CONFIG_DEBUG_ALIGN_RODATA, __start_rodata_section_aligned will
> >>> + * be the first section-aligned location after __start_rodata. Otherwise,
> >>> + * it will be equal to __start_rodata.
> >>> + */
> >>> +__start_rodata_section_aligned = ALIGN(__start_rodata, 1 << SECTION_SHIFT);
...

I'm afraid this causes build errors on two configurations:

./arch/arm/kernel/vmlinux.lds:701: undefined symbol `SECTION_SHIFT' referenced in expression

which I think is down to the new __start_rodata_section_aligned line.

Olof's builder errored out with this for both allnoconfig and
lpc18xx_defconfig builds.  Both probably due to CONFIG_MMU being
unset, and hence no definition for SECTION_SHIFT.

I think you need to make this conditional on CONFIG_MMU.

-- 
RMK's Patch system: http://www.arm.linux.org.uk/developer/patches/
FTTC broadband for 0.8mile line: currently at 9.6Mbps down 400kbps up
according to speedtest.net.

[toc] | [next] | [standalone]

#1332635

On Thursday 11 February 2016 18:00:32 Russell King - ARM Linux wrote:
> On Fri, Jan 22, 2016 at 01:19:03PM -0800, Kees Cook wrote:
> > On Tue, Dec 8, 2015 at 10:38 AM, Kees Cook <keescook@chromium.org> wrote:
> > > On Mon, Dec 7, 2015 at 11:47 PM, Ard Biesheuvel
> > > <ard.biesheuvel@linaro.org> wrote:
> > >> On 7 December 2015 at 23:35, Kees Cook <keescook@chromium.org> wrote:
> > >>>  /*
> > >>> + * Without CONFIG_DEBUG_ALIGN_RODATA, __start_rodata_section_aligned will
> > >>> + * be the first section-aligned location after __start_rodata. Otherwise,
> > >>> + * it will be equal to __start_rodata.
> > >>> + */
> > >>> +__start_rodata_section_aligned = ALIGN(__start_rodata, 1 << SECTION_SHIFT);
> ...
> 
> I'm afraid this causes build errors on two configurations:
> 
> ./arch/arm/kernel/vmlinux.lds:701: undefined symbol `SECTION_SHIFT' referenced in expression
> 
> which I think is down to the new __start_rodata_section_aligned line.
> 
> Olof's builder errored out with this for both allnoconfig and
> lpc18xx_defconfig builds.  Both probably due to CONFIG_MMU being
> unset, and hence no definition for SECTION_SHIFT.
> 
> I think you need to make this conditional on CONFIG_MMU.

I've hit another problem with this patch, when using XIP_KERNEL, I now see

arch/arm/mm/built-in.o:(.data+0x4bc): undefined reference to `__start_rodata_section_aligned'

in linux-next, apparently because of the combination with "ARM: 8513/1: xip: Move
XIP linking to a separate file".

	Arnd

[toc] | [prev] | [next] | [standalone]

#1334280

On Fri, Feb 12, 2016 at 10:45 AM, Arnd Bergmann <arnd@arndb.de> wrote:
> On Thursday 11 February 2016 18:00:32 Russell King - ARM Linux wrote:
>> On Fri, Jan 22, 2016 at 01:19:03PM -0800, Kees Cook wrote:
>> > On Tue, Dec 8, 2015 at 10:38 AM, Kees Cook <keescook@chromium.org> wrote:
>> > > On Mon, Dec 7, 2015 at 11:47 PM, Ard Biesheuvel
>> > > <ard.biesheuvel@linaro.org> wrote:
>> > >> On 7 December 2015 at 23:35, Kees Cook <keescook@chromium.org> wrote:
>> > >>>  /*
>> > >>> + * Without CONFIG_DEBUG_ALIGN_RODATA, __start_rodata_section_aligned will
>> > >>> + * be the first section-aligned location after __start_rodata. Otherwise,
>> > >>> + * it will be equal to __start_rodata.
>> > >>> + */
>> > >>> +__start_rodata_section_aligned = ALIGN(__start_rodata, 1 << SECTION_SHIFT);
>> ...
>>
>> I'm afraid this causes build errors on two configurations:
>>
>> ./arch/arm/kernel/vmlinux.lds:701: undefined symbol `SECTION_SHIFT' referenced in expression
>>
>> which I think is down to the new __start_rodata_section_aligned line.
>>
>> Olof's builder errored out with this for both allnoconfig and
>> lpc18xx_defconfig builds.  Both probably due to CONFIG_MMU being
>> unset, and hence no definition for SECTION_SHIFT.
>>
>> I think you need to make this conditional on CONFIG_MMU.
>
> I've hit another problem with this patch, when using XIP_KERNEL, I now see
>
> arch/arm/mm/built-in.o:(.data+0x4bc): undefined reference to `__start_rodata_section_aligned'
>
> in linux-next, apparently because of the combination with "ARM: 8513/1: xip: Move
> XIP linking to a separate file".

Is this due to the copy not taking into account the recent changes to
vmlinux.lds.S?

"git show 538bf4694898b19e" with copy-rename detection shows:

copy from arch/arm/kernel/vmlinux.lds.S
copy to arch/arm/kernel/vmlinux-xip.lds.S
index a6e395c53a48e045..6f59ef290289e078 100644
--- a/arch/arm/kernel/vmlinux.lds.S
+++ b/arch/arm/kernel/vmlinux-xip.lds.S

[...]

@@ -117,7 +114,7 @@ SECTIONS
                        ARM_CPU_KEEP(PROC_INFO)
        }

-#ifdef CONFIG_DEBUG_ALIGN_RODATA
+#ifdef CONFIG_DEBUG_RODATA
        . = ALIGN(1<<SECTION_SHIFT);
 #endif
        RO_DATA(PAGE_SIZE)

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web