[v5,bpf-next,11/15] libbpf: don't use cxx to test_libpf target

No need to use C++ for test_libbpf target when libbpf is on C and it
can be tested with C, after this change the CXXFLAGS in makefiles can
be avoided, at least in bpf samples, when sysroot is used, passing
same C/LDFLAGS as for lib.

Add "return 0" in test_libbpf to avoid warn, but also remove spaces at
start of the lines to keep same style and avoid warns while apply.

Acked-by: Andrii Nakryiko <andriin@fb.com>
Signed-off-by: Ivan Khoronzhuk <ivan.khoronzhuk@linaro.org>
---
 tools/lib/bpf/Makefile                         | 18 +++++-------------
 .../lib/bpf/{test_libbpf.cpp => test_libbpf.c} | 14 ++++++++------
 2 files changed, 13 insertions(+), 19 deletions(-)
 rename tools/lib/bpf/{test_libbpf.cpp => test_libbpf.c} (61%)

On 10/11, Ivan Khoronzhuk wrote:
> No need to use C++ for test_libbpf target when libbpf is on C and it
> can be tested with C, after this change the CXXFLAGS in makefiles can
> be avoided, at least in bpf samples, when sysroot is used, passing
> same C/LDFLAGS as for lib.
> 
> Add "return 0" in test_libbpf to avoid warn, but also remove spaces at
> start of the lines to keep same style and avoid warns while apply.
Hey, just spotted this patch, not sure how it slipped through.
The c++ test was there to make sure libbpf can be included and
linked against c++ code (i.e. libbpf headers don't have some c++
keywords/etc).

Any particular reason you were not happy with it? Can we revert it
back to c++ and fix your use-case instead? Alternatively, we can just
remove this test if we don't really care about c++.

> Acked-by: Andrii Nakryiko <andriin@fb.com>
> Signed-off-by: Ivan Khoronzhuk <ivan.khoronzhuk@linaro.org>
> ---
>  tools/lib/bpf/Makefile                         | 18 +++++-------------
>  .../lib/bpf/{test_libbpf.cpp => test_libbpf.c} | 14 ++++++++------
>  2 files changed, 13 insertions(+), 19 deletions(-)
>  rename tools/lib/bpf/{test_libbpf.cpp => test_libbpf.c} (61%)
> 
> diff --git a/tools/lib/bpf/Makefile b/tools/lib/bpf/Makefile
> index 1270955e4845..46280b5ad48d 100644
> --- a/tools/lib/bpf/Makefile
> +++ b/tools/lib/bpf/Makefile
> @@ -52,7 +52,7 @@ ifndef VERBOSE
>  endif
>  
>  FEATURE_USER = .libbpf
> -FEATURE_TESTS = libelf libelf-mmap bpf reallocarray cxx
> +FEATURE_TESTS = libelf libelf-mmap bpf reallocarray
>  FEATURE_DISPLAY = libelf bpf
>  
>  INCLUDES = -I. -I$(srctree)/tools/include -I$(srctree)/tools/arch/$(ARCH)/include/uapi -I$(srctree)/tools/include/uapi
> @@ -142,15 +142,7 @@ GLOBAL_SYM_COUNT = $(shell readelf -s --wide $(BPF_IN) | \
>  VERSIONED_SYM_COUNT = $(shell readelf -s --wide $(OUTPUT)libbpf.so | \
>  			      grep -Eo '[^ ]+@LIBBPF_' | cut -d@ -f1 | sort -u | wc -l)
>  
> -CMD_TARGETS = $(LIB_TARGET) $(PC_FILE)
> -
> -CXX_TEST_TARGET = $(OUTPUT)test_libbpf
> -
> -ifeq ($(feature-cxx), 1)
> -	CMD_TARGETS += $(CXX_TEST_TARGET)
> -endif
> -
> -TARGETS = $(CMD_TARGETS)
> +CMD_TARGETS = $(LIB_TARGET) $(PC_FILE) $(OUTPUT)test_libbpf
>  
>  all: fixdep
>  	$(Q)$(MAKE) all_cmd
> @@ -190,8 +182,8 @@ $(OUTPUT)libbpf.so.$(LIBBPF_VERSION): $(BPF_IN)
>  $(OUTPUT)libbpf.a: $(BPF_IN)
>  	$(QUIET_LINK)$(RM) $@; $(AR) rcs $@ $^
>  
> -$(OUTPUT)test_libbpf: test_libbpf.cpp $(OUTPUT)libbpf.a
> -	$(QUIET_LINK)$(CXX) $(INCLUDES) $^ -lelf -o $@
> +$(OUTPUT)test_libbpf: test_libbpf.c $(OUTPUT)libbpf.a
> +	$(QUIET_LINK)$(CC) $(INCLUDES) $^ -lelf -o $@
>  
>  $(OUTPUT)libbpf.pc:
>  	$(QUIET_GEN)sed -e "s|@PREFIX@|$(prefix)|" \
> @@ -266,7 +258,7 @@ config-clean:
>  	$(Q)$(MAKE) -C $(srctree)/tools/build/feature/ clean >/dev/null
>  
>  clean:
> -	$(call QUIET_CLEAN, libbpf) $(RM) $(TARGETS) $(CXX_TEST_TARGET) \
> +	$(call QUIET_CLEAN, libbpf) $(RM) $(CMD_TARGETS) \
>  		*.o *~ *.a *.so *.so.$(LIBBPF_MAJOR_VERSION) .*.d .*.cmd \
>  		*.pc LIBBPF-CFLAGS bpf_helper_defs.h
>  	$(call QUIET_CLEAN, core-gen) $(RM) $(OUTPUT)FEATURE-DUMP.libbpf
> diff --git a/tools/lib/bpf/test_libbpf.cpp b/tools/lib/bpf/test_libbpf.c
> similarity index 61%
> rename from tools/lib/bpf/test_libbpf.cpp
> rename to tools/lib/bpf/test_libbpf.c
> index fc134873bb6d..f0eb2727b766 100644
> --- a/tools/lib/bpf/test_libbpf.cpp
> +++ b/tools/lib/bpf/test_libbpf.c
> @@ -7,12 +7,14 @@
>  
>  int main(int argc, char *argv[])
>  {
> -    /* libbpf.h */
> -    libbpf_set_print(NULL);
> +	/* libbpf.h */
> +	libbpf_set_print(NULL);
>  
> -    /* bpf.h */
> -    bpf_prog_get_fd_by_id(0);
> +	/* bpf.h */
> +	bpf_prog_get_fd_by_id(0);
>  
> -    /* btf.h */
> -    btf__new(NULL, 0);
> +	/* btf.h */
> +	btf__new(NULL, 0);
> +
> +	return 0;
>  }
> -- 
> 2.17.1
>

On Thu, Nov 21, 2019 at 1:42 PM Stanislav Fomichev <sdf@fomichev.me> wrote:
>
> On 10/11, Ivan Khoronzhuk wrote:
> > No need to use C++ for test_libbpf target when libbpf is on C and it
> > can be tested with C, after this change the CXXFLAGS in makefiles can
> > be avoided, at least in bpf samples, when sysroot is used, passing
> > same C/LDFLAGS as for lib.
> >
> > Add "return 0" in test_libbpf to avoid warn, but also remove spaces at
> > start of the lines to keep same style and avoid warns while apply.
> Hey, just spotted this patch, not sure how it slipped through.
> The c++ test was there to make sure libbpf can be included and
> linked against c++ code (i.e. libbpf headers don't have some c++
> keywords/etc).
>
> Any particular reason you were not happy with it? Can we revert it
> back to c++ and fix your use-case instead? Alternatively, we can just
> remove this test if we don't really care about c++.
>

No one seemed to know why we have C++ pieces in pure C library and its
Makefile, so we decide to "fix" this. :)
But I do understand your concern. Would it be possible to instead do
this as a proper selftests test? Do you mind taking a look at that?
Thanks!

(please trim irrelevant parts)
[...]

On 11/21, Andrii Nakryiko wrote:
> On Thu, Nov 21, 2019 at 1:42 PM Stanislav Fomichev <sdf@fomichev.me> wrote:
> >
> > On 10/11, Ivan Khoronzhuk wrote:
> > > No need to use C++ for test_libbpf target when libbpf is on C and it
> > > can be tested with C, after this change the CXXFLAGS in makefiles can
> > > be avoided, at least in bpf samples, when sysroot is used, passing
> > > same C/LDFLAGS as for lib.
> > > Add "return 0" in test_libbpf to avoid warn, but also remove spaces at
> > > start of the lines to keep same style and avoid warns while apply.
> > Hey, just spotted this patch, not sure how it slipped through.
> > The c++ test was there to make sure libbpf can be included and
> > linked against c++ code (i.e. libbpf headers don't have some c++
> > keywords/etc).
> >
> > Any particular reason you were not happy with it? Can we revert it
> > back to c++ and fix your use-case instead? Alternatively, we can just
> > remove this test if we don't really care about c++.
> >
> 
> No one seemed to know why we have C++ pieces in pure C library and its
> Makefile, so we decide to "fix" this. :)
It's surprising, the commit 8c4905b995c6 clearly states the reason
for adding it. Looks like it deserved a real comment in the Makefile :-)

> But I do understand your concern. Would it be possible to instead do
> this as a proper selftests test? Do you mind taking a look at that?
Ack, will move this test_libbpf.c into selftests and convert back to
c++.

On Fri, Nov 22, 2019 at 08:32:11AM -0800, Stanislav Fomichev wrote:
>On 11/21, Andrii Nakryiko wrote:
>> On Thu, Nov 21, 2019 at 1:42 PM Stanislav Fomichev <sdf@fomichev.me> wrote:
>> >
>> > On 10/11, Ivan Khoronzhuk wrote:
>> > > No need to use C++ for test_libbpf target when libbpf is on C and it
>> > > can be tested with C, after this change the CXXFLAGS in makefiles can
>> > > be avoided, at least in bpf samples, when sysroot is used, passing
>> > > same C/LDFLAGS as for lib.
>> > > Add "return 0" in test_libbpf to avoid warn, but also remove spaces at
>> > > start of the lines to keep same style and avoid warns while apply.
>> > Hey, just spotted this patch, not sure how it slipped through.
>> > The c++ test was there to make sure libbpf can be included and
>> > linked against c++ code (i.e. libbpf headers don't have some c++
>> > keywords/etc).
>> >
>> > Any particular reason you were not happy with it? Can we revert it
>> > back to c++ and fix your use-case instead? Alternatively, we can just
>> > remove this test if we don't really care about c++.
>> >
>>
>> No one seemed to know why we have C++ pieces in pure C library and its
>> Makefile, so we decide to "fix" this. :)
>It's surprising, the commit 8c4905b995c6 clearly states the reason
>for adding it. Looks like it deserved a real comment in the Makefile :-)

I dislike changing things like this, but I was asked while review and
it seemed logical enough. The comment could prevent us from doing this.

>
>> But I do understand your concern. Would it be possible to instead do
>> this as a proper selftests test? Do you mind taking a look at that?
>Ack, will move this test_libbpf.c into selftests and convert back to
>c++.

On 11/23, Ivan Khoronzhuk wrote:
> On Fri, Nov 22, 2019 at 08:32:11AM -0800, Stanislav Fomichev wrote:
> > On 11/21, Andrii Nakryiko wrote:
> > > On Thu, Nov 21, 2019 at 1:42 PM Stanislav Fomichev <sdf@fomichev.me> wrote:
> > > >
> > > > On 10/11, Ivan Khoronzhuk wrote:
> > > > > No need to use C++ for test_libbpf target when libbpf is on C and it
> > > > > can be tested with C, after this change the CXXFLAGS in makefiles can
> > > > > be avoided, at least in bpf samples, when sysroot is used, passing
> > > > > same C/LDFLAGS as for lib.
> > > > > Add "return 0" in test_libbpf to avoid warn, but also remove spaces at
> > > > > start of the lines to keep same style and avoid warns while apply.
> > > > Hey, just spotted this patch, not sure how it slipped through.
> > > > The c++ test was there to make sure libbpf can be included and
> > > > linked against c++ code (i.e. libbpf headers don't have some c++
> > > > keywords/etc).
> > > >
> > > > Any particular reason you were not happy with it? Can we revert it
> > > > back to c++ and fix your use-case instead? Alternatively, we can just
> > > > remove this test if we don't really care about c++.
> > > >
> > > 
> > > No one seemed to know why we have C++ pieces in pure C library and its
> > > Makefile, so we decide to "fix" this. :)
> > It's surprising, the commit 8c4905b995c6 clearly states the reason
> > for adding it. Looks like it deserved a real comment in the Makefile :-)
> 
> I dislike changing things like this, but I was asked while review and
> it seemed logical enough. The comment could prevent us from doing this.
No worries, I'll add it back with a comment :-)

> > > But I do understand your concern. Would it be possible to instead do
> > > this as a proper selftests test? Do you mind taking a look at that?
> > Ack, will move this test_libbpf.c into selftests and convert back to
> > c++.
> 
> -- 
> Regards,
> Ivan Khoronzhuk