[PATCH 00/12] ARM: use adr_l/ldr_l macros for PC-relative references

[Ard Biesheuvel]

On Wed, 16 Sep 2020 at 02:31, Nick Desaulniers <ndesaulniers at google.com> wrote:
>
> On Tue, Sep 15, 2020 at 2:30 PM Ard Biesheuvel <ardb at kernel.org> wrote:
> >
> > Excellent, thanks for testing. Do you have any coverage for Thumb2
> > builds as well? (CONFIG_THUMB2_KERNEL=y)
>
> Ah, right, manually testing ARCH=arm defconfig with
> CONFIG_THUMB2_KERNEL enabled via menuconfig:
>
> Builds and boots for clang.
>
> (Also needed https://lore.kernel.org/lkml/20200915225751.274531-1-ndesaulniers@google.com/T/#u
> for an unrelated issue).
>
> For GCC, I observe:
>
> arch/arm/vfp/vfphw.o: in function `vfp_support_entry':
> (.text+0xa): relocation truncated to fit: R_ARM_THM_JUMP19 against
> symbol `vfp_kmode_exception' defined in .text.unlikely section in
> arch/arm/vfp/vfpmodule.o
>

Interesting. And this is using ld.bfd ?

> There doesn't seem to be a config option to work around that for now.
> Though it's not caused by your series; if I drop your series, I can
> still reproduce.
>
> Is there a different config I should be testing rather than
> defconfig+manual testing, like one of the existing configs? Looks like
> only milbeaut_m10v_defconfig enable THUMB2 by default.  I should add
> that to my CI and kernelCI for coverage with clang.
> https://github.com/ClangBuiltLinux/continuous-integration/issues/94#issuecomment-693030376
>
> milbeaut_m10v_defconfig
> builds with your series for clang.  Looks like that config may be
> currently broken for GCC?
> Looks like it doesn't boot with Clang, so I'll need to debug that.
> Again, both of the two past sentences are regardless of your series.
> So your series still LGTM.

Cheers.

I usually build multi_v7_defconfig with/wihout CONFIG_LPAE x
with/without CONFIG_THUMB2_KERNEL (LPAE affects the size of
dma_addr_t, and uses a different page table format, so it is a useful
bit to flick in testing)