Re: [PATCH] kbuild: Provide way to actually disable stack protector

[Kees Cook <keescook@chromium.org>]

On Tue, Jun 23, 2020 at 11:33:53AM +0900, Masahiro Yamada wrote:
> On Tue, Jun 23, 2020 at 4:02 AM Kees Cook <keescook@chromium.org> wrote:
> >
> > Some builds of GCC enable stack protector by default. Simply removing
> > the arguments is not sufficient to disable stack protector, as the stack
> > protector for those GCC builds must be explicitly disabled. (Removing the
> > arguments is left as-is just to be sure there are no ordering problems. If
> > -fno-stack-protector ended up _before_ -fstack-protector, it would not
> > disable it: GCC uses whichever -f... comes last on the command line.)
> >
> > Fixes: 20355e5f73a7 ("x86/entry: Exclude low level entry code from sanitizing")
> > Signed-off-by: Kees Cook <keescook@chromium.org>
> > ---
> >  Makefile                          | 4 +++-
> >  arch/Kconfig                      | 3 ---
> >  arch/arm/boot/compressed/Makefile | 4 ++--
> >  arch/x86/entry/Makefile           | 3 +++
> >  4 files changed, 8 insertions(+), 6 deletions(-)
> >
> > diff --git a/Makefile b/Makefile
> > index ac2c61c37a73..b46e91bf0b0e 100644
> > --- a/Makefile
> > +++ b/Makefile
> > @@ -762,7 +762,9 @@ ifneq ($(CONFIG_FRAME_WARN),0)
> >  KBUILD_CFLAGS += -Wframe-larger-than=$(CONFIG_FRAME_WARN)
> >  endif
> >
> > -stackp-flags-$(CONFIG_CC_HAS_STACKPROTECTOR_NONE) := -fno-stack-protector
> > +DISABLE_STACKPROTECTOR := $(call cc-option,-fno-stack-protector)
> > +export DISABLE_STACKPROTECTOR
> > +stackp-flags-y                                    := $(DISABLE_STACKPROTECTOR)
> >  stackp-flags-$(CONFIG_STACKPROTECTOR)             := -fstack-protector
> >  stackp-flags-$(CONFIG_STACKPROTECTOR_STRONG)      := -fstack-protector-strong
> >
> > diff --git a/arch/Kconfig b/arch/Kconfig
> > index 8cc35dc556c7..1ea61290900a 100644
> > --- a/arch/Kconfig
> > +++ b/arch/Kconfig
> > @@ -478,9 +478,6 @@ config HAVE_STACKPROTECTOR
> >           An arch should select this symbol if:
> >           - it has implemented a stack canary (e.g. __stack_chk_guard)
> >
> > -config CC_HAS_STACKPROTECTOR_NONE
> > -       def_bool $(cc-option,-fno-stack-protector)
> > -
> >  config STACKPROTECTOR
> >         bool "Stack Protector buffer overflow detection"
> >         depends on HAVE_STACKPROTECTOR
> > diff --git a/arch/arm/boot/compressed/Makefile b/arch/arm/boot/compressed/Makefile
> > index 00602a6fba04..3693bac525d2 100644
> > --- a/arch/arm/boot/compressed/Makefile
> > +++ b/arch/arm/boot/compressed/Makefile
> > @@ -84,9 +84,9 @@ endif
> >
> >  # -fstack-protector-strong triggers protection checks in this code,
> >  # but it is being used too early to link to meaningful stack_chk logic.
> > -nossp-flags-$(CONFIG_CC_HAS_STACKPROTECTOR_NONE) := -fno-stack-protector
> >  $(foreach o, $(libfdt_objs) atags_to_fdt.o, \
> > -       $(eval CFLAGS_$(o) := -I $(srctree)/scripts/dtc/libfdt $(nossp-flags-y)))
> > +       $(eval CFLAGS_$(o) := -I $(srctree)/scripts/dtc/libfdt \
> > +                             $(DISABLE_STACKPROTECTOR)))
> >
> >  # These were previously generated C files. When you are building the kernel
> >  # with O=, make sure to remove the stale files in the output tree. Otherwise,
> > diff --git a/arch/x86/entry/Makefile b/arch/x86/entry/Makefile
> > index b7a5790d8d63..79902decc3d1 100644
> > --- a/arch/x86/entry/Makefile
> > +++ b/arch/x86/entry/Makefile
> > @@ -10,6 +10,9 @@ KCOV_INSTRUMENT := n
> >  CFLAGS_REMOVE_common.o = $(CC_FLAGS_FTRACE) -fstack-protector -fstack-protector-strong
> >  CFLAGS_REMOVE_syscall_32.o = $(CC_FLAGS_FTRACE) -fstack-protector -fstack-protector-strong
> >  CFLAGS_REMOVE_syscall_64.o = $(CC_FLAGS_FTRACE) -fstack-protector -fstack-protector-strong
> > +CFLAGS_common.o += $(DISABLE_STACKPROTECTOR)
> > +CFLAGS_syscall_32.o += $(DISABLE_STACKPROTECTOR)
> > +CFLAGS_syscall_64.o += $(DISABLE_STACKPROTECTOR)
> 
> There is one more c file in this directory.
> 
> Is it OK to not patch syscall_x32.c ?

Good question. Peter? (It seems all the syscall_*.c files are just a
table, not code -- why do they need any instrumentation changes?)

> 
> 
> >
> >  CFLAGS_syscall_64.o            += $(call cc-option,-Wno-override-init,)
> >  CFLAGS_syscall_32.o            += $(call cc-option,-Wno-override-init,)
> 
> 
> 
> 
> This patch is ugly.
> 
> I'd rather want to fix this by one-liner.

Why not a global export to assist? This isn't the only place it's needed
(see the arm64 chunk...)

> 
> 
> 
> 
> diff --git a/arch/x86/entry/Makefile b/arch/x86/entry/Makefile
> index b7a5790d8d63..0d41eb91aaea 100644
> --- a/arch/x86/entry/Makefile
> +++ b/arch/x86/entry/Makefile
> @@ -11,6 +11,8 @@ CFLAGS_REMOVE_common.o = $(CC_FLAGS_FTRACE)
> -fstack-protector -fstack-protector-
>  CFLAGS_REMOVE_syscall_32.o = $(CC_FLAGS_FTRACE) -fstack-protector
> -fstack-protector-strong
>  CFLAGS_REMOVE_syscall_64.o = $(CC_FLAGS_FTRACE) -fstack-protector
> -fstack-protector-strong
> 
> +ccflags-$(CONFIG_CC_HAS_STACKPROTECTOR_NONE) += -fno-stack-protector
> +

Order matters here -- when is ccflags-y applied?

>  CFLAGS_syscall_64.o            += $(call cc-option,-Wno-override-init,)
>  CFLAGS_syscall_32.o            += $(call cc-option,-Wno-override-init,)
>  obj-y                          := entry_$(BITS).o thunk_$(BITS).o
> syscall_$(BITS).o
> 
> 
> 
> 
> -- 
> Best Regards
> Masahiro Yamada

-- 
Kees Cook