Re: [PATCH 4.14 114/165] x86/vdso: Fix vDSO build if a retpoline is emitted

From: Nikola Ciprich
Date: Tue Oct 02 2018 - 04:21:46 EST


Hi Greg and others,

sorry for reporting this so late, but still...

this breaks build on older compilers, since it requires
-mindirect-branch=thunk-inline -mindirect-branch-register even though
retpoline support is disabled in kernel config.. is this expected?

BR

nik


On Mon, Sep 03, 2018 at 06:56:40PM +0200, Greg Kroah-Hartman wrote:
> 4.14-stable review patch. If anyone has any objections, please let me know.
>
> ------------------
>
> From: Andy Lutomirski <luto@xxxxxxxxxx>
>
> commit 2e549b2ee0e358bc758480e716b881f9cabedb6a upstream.
>
> Currently, if the vDSO ends up containing an indirect branch or
> call, GCC will emit the "external thunk" style of retpoline, and it
> will fail to link.
>
> Fix it by building the vDSO with inline retpoline thunks.
>
> I haven't seen any reports of this triggering on an unpatched
> kernel.
>
> Fixes: commit 76b043848fd2 ("x86/retpoline: Add initial retpoline support")
> Signed-off-by: Andy Lutomirski <luto@xxxxxxxxxx>
> Signed-off-by: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
> Acked-by: Matt Rickard <matt@xxxxxxxxxxxxxxx>
> Cc: Borislav Petkov <bp@xxxxxxxxx>
> Cc: Jason Vas Dias <jason.vas.dias@xxxxxxxxx>
> Cc: David Woodhouse <dwmw2@xxxxxxxxxxxxx>
> Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
> Cc: Andi Kleen <ak@xxxxxxxxxxxxxxx>
> Cc: stable@xxxxxxxxxxxxxxx
> Link: https://lkml.kernel.org/r/c76538cd3afbe19c6246c2d1715bc6a60bd63985.1534448381.git.luto@xxxxxxxxxx
> Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
>
> ---
> Makefile | 4 ++++
> arch/x86/entry/vdso/Makefile | 6 ++++--
> 2 files changed, 8 insertions(+), 2 deletions(-)
>
> --- a/Makefile
> +++ b/Makefile
> @@ -490,9 +490,13 @@ KBUILD_AFLAGS += $(CLANG_TARGET) $(CLANG
> endif
>
> RETPOLINE_CFLAGS_GCC := -mindirect-branch=thunk-extern -mindirect-branch-register
> +RETPOLINE_VDSO_CFLAGS_GCC := -mindirect-branch=thunk-inline -mindirect-branch-register
> RETPOLINE_CFLAGS_CLANG := -mretpoline-external-thunk
> +RETPOLINE_VDSO_CFLAGS_CLANG := -mretpoline
> RETPOLINE_CFLAGS := $(call cc-option,$(RETPOLINE_CFLAGS_GCC),$(call cc-option,$(RETPOLINE_CFLAGS_CLANG)))
> +RETPOLINE_VDSO_CFLAGS := $(call cc-option,$(RETPOLINE_VDSO_CFLAGS_GCC),$(call cc-option,$(RETPOLINE_VDSO_CFLAGS_CLANG)))
> export RETPOLINE_CFLAGS
> +export RETPOLINE_VDSO_CFLAGS
>
> ifeq ($(config-targets),1)
> # ===========================================================================
> --- a/arch/x86/entry/vdso/Makefile
> +++ b/arch/x86/entry/vdso/Makefile
> @@ -74,9 +74,9 @@ $(obj)/vdso-image-%.c: $(obj)/vdso%.so.d
> CFL := $(PROFILING) -mcmodel=small -fPIC -O2 -fasynchronous-unwind-tables -m64 \
> $(filter -g%,$(KBUILD_CFLAGS)) $(call cc-option, -fno-stack-protector) \
> -fno-omit-frame-pointer -foptimize-sibling-calls \
> - -DDISABLE_BRANCH_PROFILING -DBUILD_VDSO
> + -DDISABLE_BRANCH_PROFILING -DBUILD_VDSO $(RETPOLINE_VDSO_CFLAGS)
>
> -$(vobjs): KBUILD_CFLAGS := $(filter-out $(GCC_PLUGINS_CFLAGS),$(KBUILD_CFLAGS)) $(CFL)
> +$(vobjs): KBUILD_CFLAGS := $(filter-out $(GCC_PLUGINS_CFLAGS) $(RETPOLINE_CFLAGS),$(KBUILD_CFLAGS)) $(CFL)
>
> #
> # vDSO code runs in userspace and -pg doesn't help with profiling anyway.
> @@ -147,11 +147,13 @@ KBUILD_CFLAGS_32 := $(filter-out -mcmode
> KBUILD_CFLAGS_32 := $(filter-out -fno-pic,$(KBUILD_CFLAGS_32))
> KBUILD_CFLAGS_32 := $(filter-out -mfentry,$(KBUILD_CFLAGS_32))
> KBUILD_CFLAGS_32 := $(filter-out $(GCC_PLUGINS_CFLAGS),$(KBUILD_CFLAGS_32))
> +KBUILD_CFLAGS_32 := $(filter-out $(RETPOLINE_CFLAGS),$(KBUILD_CFLAGS_32))
> KBUILD_CFLAGS_32 += -m32 -msoft-float -mregparm=0 -fpic
> KBUILD_CFLAGS_32 += $(call cc-option, -fno-stack-protector)
> KBUILD_CFLAGS_32 += $(call cc-option, -foptimize-sibling-calls)
> KBUILD_CFLAGS_32 += -fno-omit-frame-pointer
> KBUILD_CFLAGS_32 += -DDISABLE_BRANCH_PROFILING
> +KBUILD_CFLAGS_32 += $(RETPOLINE_VDSO_CFLAGS)
> $(obj)/vdso32.so.dbg: KBUILD_CFLAGS = $(KBUILD_CFLAGS_32)
>
> $(obj)/vdso32.so.dbg: FORCE \
>
>

--
-------------------------------------
Ing. Nikola CIPRICH
LinuxBox.cz, s.r.o.
28.rijna 168, 709 00 Ostrava

tel.: +420 591 166 214
fax: +420 596 621 273
mobil: +420 777 093 799
www.linuxbox.cz

mobil servis: +420 737 238 656
email servis: servis@xxxxxxxxxxx
-------------------------------------