[v4,bpf-next,1/4] selftests/bpf: compile progs with -D__TARGET_ARCH_$(SRCARCH)
diff mbox series

Message ID 20190711142930.68809-2-iii@linux.ibm.com
State Accepted
Delegated to: BPF Maintainers
Headers show
Series
  • selftests/bpf: fix compiling loop{1,2,3}.c on s390
Related show

Commit Message

Ilya Leoshkevich July 11, 2019, 2:29 p.m. UTC
This opens up the possibility of accessing registers in an
arch-independent way.

Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com>
---
 tools/testing/selftests/bpf/Makefile | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

Comments

Andrii Nakryiko July 12, 2019, 12:53 a.m. UTC | #1
On Thu, Jul 11, 2019 at 7:32 AM Ilya Leoshkevich <iii@linux.ibm.com> wrote:
>
> This opens up the possibility of accessing registers in an
> arch-independent way.
>
> Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com>
> ---
>  tools/testing/selftests/bpf/Makefile | 4 +++-
>  1 file changed, 3 insertions(+), 1 deletion(-)
>
> diff --git a/tools/testing/selftests/bpf/Makefile b/tools/testing/selftests/bpf/Makefile
> index 2620406a53ec..ad84450e4ab8 100644
> --- a/tools/testing/selftests/bpf/Makefile
> +++ b/tools/testing/selftests/bpf/Makefile
> @@ -1,4 +1,5 @@
>  # SPDX-License-Identifier: GPL-2.0
> +include ../../../scripts/Makefile.arch
>
>  LIBDIR := ../../../lib
>  BPFDIR := $(LIBDIR)/bpf
> @@ -138,7 +139,8 @@ CLANG_SYS_INCLUDES := $(shell $(CLANG) -v -E - </dev/null 2>&1 \
>
>  CLANG_FLAGS = -I. -I./include/uapi -I../../../include/uapi \
>               $(CLANG_SYS_INCLUDES) \
> -             -Wno-compare-distinct-pointer-types
> +             -Wno-compare-distinct-pointer-types \
> +             -D__TARGET_ARCH_$(SRCARCH)

samples/bpf/Makefile uses $(ARCH), why does it work for samples?
Should we update samples/bpf/Makefile as well?

>
>  $(OUTPUT)/test_l4lb_noinline.o: CLANG_FLAGS += -fno-inline
>  $(OUTPUT)/test_xdp_noinline.o: CLANG_FLAGS += -fno-inline
> --
> 2.21.0
>
Ilya Leoshkevich July 12, 2019, 8:59 a.m. UTC | #2
> Am 12.07.2019 um 02:53 schrieb Andrii Nakryiko <andrii.nakryiko@gmail.com>:
> 
> On Thu, Jul 11, 2019 at 7:32 AM Ilya Leoshkevich <iii@linux.ibm.com> wrote:
>> 
>> This opens up the possibility of accessing registers in an
>> arch-independent way.
>> 
>> Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com>
>> ---
>> tools/testing/selftests/bpf/Makefile | 4 +++-
>> 1 file changed, 3 insertions(+), 1 deletion(-)
>> 
>> diff --git a/tools/testing/selftests/bpf/Makefile b/tools/testing/selftests/bpf/Makefile
>> index 2620406a53ec..ad84450e4ab8 100644
>> --- a/tools/testing/selftests/bpf/Makefile
>> +++ b/tools/testing/selftests/bpf/Makefile
>> @@ -1,4 +1,5 @@
>> # SPDX-License-Identifier: GPL-2.0
>> +include ../../../scripts/Makefile.arch
>> 
>> LIBDIR := ../../../lib
>> BPFDIR := $(LIBDIR)/bpf
>> @@ -138,7 +139,8 @@ CLANG_SYS_INCLUDES := $(shell $(CLANG) -v -E - </dev/null 2>&1 \
>> 
>> CLANG_FLAGS = -I. -I./include/uapi -I../../../include/uapi \
>>              $(CLANG_SYS_INCLUDES) \
>> -             -Wno-compare-distinct-pointer-types
>> +             -Wno-compare-distinct-pointer-types \
>> +             -D__TARGET_ARCH_$(SRCARCH)
> 
> samples/bpf/Makefile uses $(ARCH), why does it work for samples?
> Should we update samples/bpf/Makefile as well?

I believe that in common cases both are okay, but judging by
linux:Makefile and linux:tools/scripts/Makefile.arch, one could use e.g.
ARCH=i686, and that would be converted to SRCARCH=x86. So IMHO SRCARCH
is safer, and we should change bpf/samples/Makefile. I could send a
patch separately.
Andrii Nakryiko July 12, 2019, 5:28 p.m. UTC | #3
On Fri, Jul 12, 2019 at 2:00 AM Ilya Leoshkevich <iii@linux.ibm.com> wrote:
>
> > Am 12.07.2019 um 02:53 schrieb Andrii Nakryiko <andrii.nakryiko@gmail.com>:
> >
> > On Thu, Jul 11, 2019 at 7:32 AM Ilya Leoshkevich <iii@linux.ibm.com> wrote:
> >>
> >> This opens up the possibility of accessing registers in an
> >> arch-independent way.
> >>
> >> Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com>
> >> ---
> >> tools/testing/selftests/bpf/Makefile | 4 +++-
> >> 1 file changed, 3 insertions(+), 1 deletion(-)
> >>
> >> diff --git a/tools/testing/selftests/bpf/Makefile b/tools/testing/selftests/bpf/Makefile
> >> index 2620406a53ec..ad84450e4ab8 100644
> >> --- a/tools/testing/selftests/bpf/Makefile
> >> +++ b/tools/testing/selftests/bpf/Makefile
> >> @@ -1,4 +1,5 @@
> >> # SPDX-License-Identifier: GPL-2.0
> >> +include ../../../scripts/Makefile.arch
> >>
> >> LIBDIR := ../../../lib
> >> BPFDIR := $(LIBDIR)/bpf
> >> @@ -138,7 +139,8 @@ CLANG_SYS_INCLUDES := $(shell $(CLANG) -v -E - </dev/null 2>&1 \
> >>
> >> CLANG_FLAGS = -I. -I./include/uapi -I../../../include/uapi \
> >>              $(CLANG_SYS_INCLUDES) \
> >> -             -Wno-compare-distinct-pointer-types
> >> +             -Wno-compare-distinct-pointer-types \
> >> +             -D__TARGET_ARCH_$(SRCARCH)
> >
> > samples/bpf/Makefile uses $(ARCH), why does it work for samples?
> > Should we update samples/bpf/Makefile as well?
>
> I believe that in common cases both are okay, but judging by
> linux:Makefile and linux:tools/scripts/Makefile.arch, one could use e.g.
> ARCH=i686, and that would be converted to SRCARCH=x86. So IMHO SRCARCH
> is safer, and we should change bpf/samples/Makefile. I could send a
> patch separately.

Yeah, let's do that then, thanks!

Patch
diff mbox series

diff --git a/tools/testing/selftests/bpf/Makefile b/tools/testing/selftests/bpf/Makefile
index 2620406a53ec..ad84450e4ab8 100644
--- a/tools/testing/selftests/bpf/Makefile
+++ b/tools/testing/selftests/bpf/Makefile
@@ -1,4 +1,5 @@ 
 # SPDX-License-Identifier: GPL-2.0
+include ../../../scripts/Makefile.arch
 
 LIBDIR := ../../../lib
 BPFDIR := $(LIBDIR)/bpf
@@ -138,7 +139,8 @@  CLANG_SYS_INCLUDES := $(shell $(CLANG) -v -E - </dev/null 2>&1 \
 
 CLANG_FLAGS = -I. -I./include/uapi -I../../../include/uapi \
 	      $(CLANG_SYS_INCLUDES) \
-	      -Wno-compare-distinct-pointer-types
+	      -Wno-compare-distinct-pointer-types \
+	      -D__TARGET_ARCH_$(SRCARCH)
 
 $(OUTPUT)/test_l4lb_noinline.o: CLANG_FLAGS += -fno-inline
 $(OUTPUT)/test_xdp_noinline.o: CLANG_FLAGS += -fno-inline