* Re: [PATCH v2] kbuild: Check for unknown options with cc-option usage in Kconfig and clang
2019-07-25 15:47 [PATCH v2] kbuild: Check for unknown options with cc-option usage in Kconfig and clang Stephen Boyd
@ 2019-07-25 16:41 ` Doug Anderson
2019-07-25 17:01 ` Nathan Chancellor
2019-07-29 10:02 ` Masahiro Yamada
2 siblings, 0 replies; 6+ messages in thread
From: Doug Anderson @ 2019-07-25 16:41 UTC (permalink / raw)
To: Stephen Boyd
Cc: Masahiro Yamada, Michal Marek, LKML, Linux Kbuild mailing list,
clang-built-linux, Peter Smith, Nathan Chancellor,
Nick Desaulniers
Hi,
On Thu, Jul 25, 2019 at 8:47 AM Stephen Boyd <swboyd@chromium.org> wrote:
>
> If the particular version of clang a user has doesn't enable
> -Werror=unknown-warning-option by default, even though it is the
> default[1], then make sure to pass the option to the Kconfig cc-option
> command so that testing options from Kconfig files works properly.
> Otherwise, depending on the default values setup in the clang toolchain
> we will silently assume options such as -Wmaybe-uninitialized are
> supported by clang, when they really aren't.
>
> A compilation issue only started happening for me once commit
> 589834b3a009 ("kbuild: Add -Werror=unknown-warning-option to
> CLANG_FLAGS") was applied on top of commit b303c6df80c9 ("kbuild:
> compute false-positive -Wmaybe-uninitialized cases in Kconfig"). This
> leads kbuild to try and test for the existence of the
> -Wmaybe-uninitialized flag with the cc-option command in
> scripts/Kconfig.include, and it doesn't see an error returned from the
> option test so it sets the config value to Y. Then the Makefile tries to
> pass the unknown option on the command line and
> -Werror=unknown-warning-option catches the invalid option and breaks the
> build. Before commit 589834b3a009 ("kbuild: Add
> -Werror=unknown-warning-option to CLANG_FLAGS") the build works fine,
> but any cc-option test of a warning option in Kconfig files silently
> evaluates to true, even if the warning option flag isn't supported on
> clang.
>
> Note: This doesn't change cc-option usages in Makefiles because those
> use a different rule that includes KBUILD_CFLAGS by default (see the
> __cc-option command in scripts/Kbuild.incluide). The KBUILD_CFLAGS
> variable already has the -Werror=unknown-warning-option flag set. Thanks
> to Doug for pointing out the different rule.
>
> [1] https://clang.llvm.org/docs/DiagnosticsReference.html#wunknown-warning-option
> Cc: Peter Smith <peter.smith@linaro.org>
> Cc: Nathan Chancellor <natechancellor@gmail.com>
> Cc: Nick Desaulniers <ndesaulniers@google.com>
> Cc: Douglas Anderson <dianders@chromium.org>
> Reviewed-by: Nathan Chancellor <natechancellor@gmail.com>
> Signed-off-by: Stephen Boyd <swboyd@chromium.org>
Fixes: 589834b3a009 ("kbuild: Add -Werror=unknown-warning-option to
CLANG_FLAGS")
Reviewed-by: Douglas Anderson <dianders@chromium.org>
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH v2] kbuild: Check for unknown options with cc-option usage in Kconfig and clang
2019-07-25 15:47 [PATCH v2] kbuild: Check for unknown options with cc-option usage in Kconfig and clang Stephen Boyd
2019-07-25 16:41 ` Doug Anderson
@ 2019-07-25 17:01 ` Nathan Chancellor
2019-07-29 10:02 ` Masahiro Yamada
2 siblings, 0 replies; 6+ messages in thread
From: Nathan Chancellor @ 2019-07-25 17:01 UTC (permalink / raw)
To: Stephen Boyd
Cc: Masahiro Yamada, Michal Marek, linux-kernel, linux-kbuild,
clang-built-linux, Peter Smith, Nick Desaulniers,
Douglas Anderson
On Thu, Jul 25, 2019 at 08:47:30AM -0700, Stephen Boyd wrote:
> If the particular version of clang a user has doesn't enable
> -Werror=unknown-warning-option by default, even though it is the
> default[1], then make sure to pass the option to the Kconfig cc-option
> command so that testing options from Kconfig files works properly.
> Otherwise, depending on the default values setup in the clang toolchain
> we will silently assume options such as -Wmaybe-uninitialized are
> supported by clang, when they really aren't.
>
> A compilation issue only started happening for me once commit
> 589834b3a009 ("kbuild: Add -Werror=unknown-warning-option to
> CLANG_FLAGS") was applied on top of commit b303c6df80c9 ("kbuild:
> compute false-positive -Wmaybe-uninitialized cases in Kconfig"). This
> leads kbuild to try and test for the existence of the
> -Wmaybe-uninitialized flag with the cc-option command in
> scripts/Kconfig.include, and it doesn't see an error returned from the
> option test so it sets the config value to Y. Then the Makefile tries to
> pass the unknown option on the command line and
> -Werror=unknown-warning-option catches the invalid option and breaks the
> build. Before commit 589834b3a009 ("kbuild: Add
> -Werror=unknown-warning-option to CLANG_FLAGS") the build works fine,
> but any cc-option test of a warning option in Kconfig files silently
> evaluates to true, even if the warning option flag isn't supported on
> clang.
>
> Note: This doesn't change cc-option usages in Makefiles because those
> use a different rule that includes KBUILD_CFLAGS by default (see the
> __cc-option command in scripts/Kbuild.incluide). The KBUILD_CFLAGS
> variable already has the -Werror=unknown-warning-option flag set. Thanks
> to Doug for pointing out the different rule.
>
> [1] https://clang.llvm.org/docs/DiagnosticsReference.html#wunknown-warning-option
> Cc: Peter Smith <peter.smith@linaro.org>
> Cc: Nathan Chancellor <natechancellor@gmail.com>
> Cc: Nick Desaulniers <ndesaulniers@google.com>
> Cc: Douglas Anderson <dianders@chromium.org>
> Reviewed-by: Nathan Chancellor <natechancellor@gmail.com>
> Signed-off-by: Stephen Boyd <swboyd@chromium.org>
Commit message wording looks better to me, thanks for the change!
I think we might also want to add:
Cc: stable@vger.kernel.org # 4.19+
in addition to Doug's suggested fixes tag because my patch has been
AUTOSEL'd by Sasha:
https://lore.kernel.org/lkml/20190719040732.17285-44-sashal@kernel.org/
https://git.kernel.org/sashal/linux-stable/c/a28859fa4fea5a834a53d86d51e502012ce09c57
Alternatively, I can just mention that this patch needs to be picked up
in addition to that one when it is formally sent out in a stable review.
Cheers,
Nathan
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH v2] kbuild: Check for unknown options with cc-option usage in Kconfig and clang
2019-07-25 15:47 [PATCH v2] kbuild: Check for unknown options with cc-option usage in Kconfig and clang Stephen Boyd
2019-07-25 16:41 ` Doug Anderson
2019-07-25 17:01 ` Nathan Chancellor
@ 2019-07-29 10:02 ` Masahiro Yamada
2019-07-29 14:23 ` Stephen Boyd
2 siblings, 1 reply; 6+ messages in thread
From: Masahiro Yamada @ 2019-07-29 10:02 UTC (permalink / raw)
To: Stephen Boyd
Cc: Michal Marek, Linux Kernel Mailing List,
Linux Kbuild mailing list, clang-built-linux, Peter Smith,
Nathan Chancellor, Nick Desaulniers, Douglas Anderson
Hi.
On Fri, Jul 26, 2019 at 12:47 AM Stephen Boyd <swboyd@chromium.org> wrote:
>
> If the particular version of clang a user has doesn't enable
> -Werror=unknown-warning-option by default, even though it is the
> default[1], then make sure to pass the option to the Kconfig cc-option
> command so that testing options from Kconfig files works properly.
> Otherwise, depending on the default values setup in the clang toolchain
> we will silently assume options such as -Wmaybe-uninitialized are
> supported by clang, when they really aren't.
>
> A compilation issue only started happening for me once commit
> 589834b3a009 ("kbuild: Add -Werror=unknown-warning-option to
> CLANG_FLAGS") was applied on top of commit b303c6df80c9 ("kbuild:
> compute false-positive -Wmaybe-uninitialized cases in Kconfig"). This
> leads kbuild to try and test for the existence of the
> -Wmaybe-uninitialized flag with the cc-option command in
> scripts/Kconfig.include, and it doesn't see an error returned from the
> option test so it sets the config value to Y. Then the Makefile tries to
> pass the unknown option on the command line and
> -Werror=unknown-warning-option catches the invalid option and breaks the
> build. Before commit 589834b3a009 ("kbuild: Add
> -Werror=unknown-warning-option to CLANG_FLAGS") the build works fine,
> but any cc-option test of a warning option in Kconfig files silently
> evaluates to true, even if the warning option flag isn't supported on
> clang.
>
> Note: This doesn't change cc-option usages in Makefiles because those
> use a different rule that includes KBUILD_CFLAGS by default (see the
> __cc-option command in scripts/Kbuild.incluide). The KBUILD_CFLAGS
> variable already has the -Werror=unknown-warning-option flag set. Thanks
> to Doug for pointing out the different rule.
>
> [1] https://clang.llvm.org/docs/DiagnosticsReference.html#wunknown-warning-option
> Cc: Peter Smith <peter.smith@linaro.org>
> Cc: Nathan Chancellor <natechancellor@gmail.com>
> Cc: Nick Desaulniers <ndesaulniers@google.com>
> Cc: Douglas Anderson <dianders@chromium.org>
> Reviewed-by: Nathan Chancellor <natechancellor@gmail.com>
> Signed-off-by: Stephen Boyd <swboyd@chromium.org>
> ---
Thanks for catching this.
I wonder if we could fix this issue
by one-liner, like this:
diff --git a/scripts/Kconfig.include b/scripts/Kconfig.include
index 8a5c4d645eb1..4bbf4fc163a2 100644
--- a/scripts/Kconfig.include
+++ b/scripts/Kconfig.include
@@ -25,7 +25,7 @@ failure = $(if-success,$(1),n,y)
# $(cc-option,<flag>)
# Return y if the compiler supports <flag>, n otherwise
-cc-option = $(success,$(CC) -Werror $(1) -E -x c /dev/null -o /dev/null)
+cc-option = $(success,$(CC) -Werror $(CLANG_FLAGS) $(1) -E -x c
/dev/null -o /dev/null)
# $(ld-option,<flag>)
# Return y if the linker supports <flag>, n otherwise
This propagates not only -Werror=unknown-warning-option
but also other clang flags to Kconfig.
Currently, we do not pass the target triplet to Kconfig.
This means, cc-option in Kconfig evaluates the given flags
against host-arch instead of target-arch.
The compiler flags are mostly independent of the architecture,
and this is not a big deal, I think.
But, maybe, would it make more sense to pass the other
basic clang flags as well?
To do this, the following is necessary as a prerequisite:
https://patchwork.kernel.org/patch/11063507/
Anyway, uninitialized CLANG_FLAGS is a bug, which must be
back-ported.
Thanks.
> Changes from v1:
> * Reworded commit text a bit
> * Added Reviewed-by tag
>
> Makefile | 5 +++++
> scripts/Kconfig.include | 2 +-
> 2 files changed, 6 insertions(+), 1 deletion(-)
>
> diff --git a/Makefile b/Makefile
> index 9be5834073f8..28177674178a 100644
> --- a/Makefile
> +++ b/Makefile
> @@ -517,6 +517,8 @@ ifdef building_out_of_srctree
> { echo "# this is build directory, ignore it"; echo "*"; } > .gitignore
> endif
>
> +KCONFIG_CC_OPTION_FLAGS := -Werror
> +
> ifneq ($(shell $(CC) --version 2>&1 | head -n 1 | grep clang),)
> ifneq ($(CROSS_COMPILE),)
> CLANG_FLAGS := --target=$(notdir $(CROSS_COMPILE:%-=%))
> @@ -531,11 +533,14 @@ ifeq ($(shell $(AS) --version 2>&1 | head -n 1 | grep clang),)
> CLANG_FLAGS += -no-integrated-as
> endif
> CLANG_FLAGS += -Werror=unknown-warning-option
> +KCONFIG_CC_OPTION_FLAGS += -Werror=unknown-warning-option
> KBUILD_CFLAGS += $(CLANG_FLAGS)
> KBUILD_AFLAGS += $(CLANG_FLAGS)
> export CLANG_FLAGS
> endif
>
> +export KCONFIG_CC_OPTION_FLAGS
> +
> # The expansion should be delayed until arch/$(SRCARCH)/Makefile is included.
> # Some architectures define CROSS_COMPILE in arch/$(SRCARCH)/Makefile.
> # CC_VERSION_TEXT is referenced from Kconfig (so it needs export),
> diff --git a/scripts/Kconfig.include b/scripts/Kconfig.include
> index 8a5c4d645eb1..144e83e7cb81 100644
> --- a/scripts/Kconfig.include
> +++ b/scripts/Kconfig.include
> @@ -25,7 +25,7 @@ failure = $(if-success,$(1),n,y)
>
> # $(cc-option,<flag>)
> # Return y if the compiler supports <flag>, n otherwise
> -cc-option = $(success,$(CC) -Werror $(1) -E -x c /dev/null -o /dev/null)
> +cc-option = $(success,$(CC) $(KCONFIG_CC_OPTION_FLAGS) $(1) -E -x c /dev/null -o /dev/null)
>
> # $(ld-option,<flag>)
> # Return y if the linker supports <flag>, n otherwise
> --
> Sent by a computer through tubes
>
--
Best Regards
Masahiro Yamada
^ permalink raw reply related [flat|nested] 6+ messages in thread