lkml.org 
[lkml]   [2021]   [Apr]   [16]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
Date
SubjectRe: [PATCH 2/2] tools: do not include scripts/Kbuild.include
On Fri, Apr 16, 2021 at 2:56 PM Christian Borntraeger
<borntraeger@de.ibm.com> wrote:
>
>
> On 15.04.21 10:06, Christian Borntraeger wrote:
> >
> > On 15.04.21 09:27, Masahiro Yamada wrote:
> >> Since commit d9f4ff50d2aa ("kbuild: spilt cc-option and friends to
> >> scripts/Makefile.compiler"), some kselftests fail to build.
> >>
> >> The tools/ directory opted out Kbuild, and went in a different
> >> direction. They copy any kind of files to the tools/ directory
> >> in order to do whatever they want to do in their world.
> >>
> >> tools/build/Build.include mimics scripts/Kbuild.include, but some
> >> tool Makefiles included the Kbuild one to import a feature that is
> >> missing in tools/build/Build.include:
> >>
> >> - Commit ec04aa3ae87b ("tools/thermal: tmon: use "-fstack-protector"
> >> only if supported") included scripts/Kbuild.include from
> >> tools/thermal/tmon/Makefile to import the cc-option macro.
> >>
> >> - Commit c2390f16fc5b ("selftests: kvm: fix for compilers that do
> >> not support -no-pie") included scripts/Kbuild.include from
> >> tools/testing/selftests/kvm/Makefile to import the try-run macro.
> >>
> >> - Commit 9cae4ace80ef ("selftests/bpf: do not ignore clang
> >> failures") included scripts/Kbuild.include from
> >> tools/testing/selftests/bpf/Makefile to import the .DELETE_ON_ERROR
> >> target.
> >>
> >> - Commit 0695f8bca93e ("selftests/powerpc: Handle Makefile for
> >> unrecognized option") included scripts/Kbuild.include from
> >> tools/testing/selftests/powerpc/pmu/ebb/Makefile to import the
> >> try-run macro.
> >>
> >> Copy what they want there, and stop including scripts/Kbuild.include
> >> from the tool Makefiles.
> >>
> >> Link: https://lore.kernel.org/lkml/86dadf33-70f7-a5ac-cb8c-64966d2f45a1@linux.ibm.com/
> >> Fixes: d9f4ff50d2aa ("kbuild: spilt cc-option and friends to scripts/Makefile.compiler")
> >> Reported-by: Janosch Frank <frankja@linux.ibm.com>
> >> Reported-by: Christian Borntraeger <borntraeger@de.ibm.com>
> >> Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
> >
> > When applying this on top of d9f4ff50d2aa ("kbuild: spilt cc-option and friends to scripts/Makefile.compiler")
> >
> > I still do get
> >
> > # ==== Test Assertion Failure ====
> > # lib/kvm_util.c:142: vm->fd >= 0
> > # pid=315635 tid=315635 - Invalid argument
> > # 1 0x0000000001002f4b: vm_open at kvm_util.c:142
> > # 2 (inlined by) vm_create at kvm_util.c:258
> > # 3 0x00000000010015ef: test_add_max_memory_regions at set_memory_region_test.c:351
> > # 4 (inlined by) main at set_memory_region_test.c:397
> > # 5 0x000003ff971abb89: ?? ??:0
> > # 6 0x00000000010017ad: .annobin_abi_note.c.hot at crt1.o:?
> > # KVM_CREATE_VM ioctl failed, rc: -1 errno: 22
> > not ok 7 selftests: kvm: set_memory_region_test # exit=254
> >
> > and the testcase compilation does not pickup the pgste option.
>
> What does work is the following:
> diff --git a/tools/testing/selftests/kvm/Makefile b/tools/testing/selftests/kvm/Makefile
> index a6d61f451f88..d9c6d9c2069e 100644
> --- a/tools/testing/selftests/kvm/Makefile
> +++ b/tools/testing/selftests/kvm/Makefile
> @@ -1,5 +1,6 @@
> # SPDX-License-Identifier: GPL-2.0-only
> include ../../../../scripts/Kbuild.include
> +include ../../../../scripts/Makefile.compiler
>
> all:
>
>
> as it does pickup the linker option handling.


Kbuild and the tools are divorced.

They cannot be married unless the tools/
build system is largely refactored.
That will be a tons of works (and
I am not sure if it is welcome).

The Kbuild refactoring should not be bothered by
the tools.
For now, I want them separated from each other.



--
Best Regards

Masahiro Yamada

\
 
 \ /
  Last update: 2021-04-16 14:55    [W:0.041 / U:0.512 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site