Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Mon, 05 Sep 2022 16:17:39 +0530
From:   "Naveen N. Rao" <naveen.n.rao@linux.vnet.ibm.com>
Subject: Re: [PATCH v2 05/16] powerpc: Skip objtool from running on
 drivers/crypto/vmx/aesp8-ppc.o
To:     Peter Zijlstra <peterz@infradead.org>,
        Sathvika Vasireddy <sv@linux.ibm.com>
Cc:     aik@ozlabs.ru, chenzhongjin@huawei.com,
        christophe.leroy@csgroup.eu, jpoimboe@redhat.com,
        linux-kernel@vger.kernel.org, linuxppc-dev@lists.ozlabs.org,
        mbenes@suse.cz, mingo@redhat.com, mpe@ellerman.id.au,
        npiggin@gmail.com, rostedt@goodmis.org
References: <20220829055223.24767-1-sv@linux.ibm.com>
        <20220829055223.24767-6-sv@linux.ibm.com>
        <YxEFZXCnCH7F8yku@hirez.programming.kicks-ass.net>
In-Reply-To: <YxEFZXCnCH7F8yku@hirez.programming.kicks-ass.net>
MIME-Version: 1.0
User-Agent: astroid/4d6b06ad (https://github.com/astroidmail/astroid)
Message-Id: <1662374747.rv53oucl9s.naveen@linux.ibm.com>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

Peter Zijlstra wrote:
> On Mon, Aug 29, 2022 at 11:22:12AM +0530, Sathvika Vasireddy wrote:
>> With objtool enabled, below warnings are seen when trying to build:
>>=20
>> drivers/crypto/vmx/aesp8-ppc.o: warning: objtool: aes_p8_set_encrypt_key=
+0x44: unannotated intra-function call
>>=20
>> drivers/crypto/vmx/aesp8-ppc.o: warning: objtool: .text+0x2448: unannota=
ted intra-function call
>>=20
>> drivers/crypto/vmx/aesp8-ppc.o: warning: objtool: .text+0x2d68: unannota=
ted intra-function call
>>=20
>> Skip objtool from running on this file, for two main reasons:
>>=20
>> Since this file comes from OpenSSL, and since it is a perl file
>> which generates a .S file, it may not be the best choice to
>> make too many code changes to such files, unless absolutely
>> necessary.
>>=20
>> Second reason is that, at least as far as the objtool --mcount
>> functionality is concerned, we do not have to run objtool on
>> that file because that file does not have any calls to _mcount.
>>=20
>> Signed-off-by: Sathvika Vasireddy <sv@linux.ibm.com>
>> ---
>>  drivers/crypto/vmx/Makefile | 2 ++
>>  1 file changed, 2 insertions(+)
>>=20
>> diff --git a/drivers/crypto/vmx/Makefile b/drivers/crypto/vmx/Makefile
>> index 2560cfea1dec..7b41f0da6807 100644
>> --- a/drivers/crypto/vmx/Makefile
>> +++ b/drivers/crypto/vmx/Makefile
>> @@ -9,3 +9,5 @@ targets +=3D aesp8-ppc.S ghashp8-ppc.S
>> =20
>>  $(obj)/aesp8-ppc.S $(obj)/ghashp8-ppc.S: $(obj)/%.S: $(src)/%.pl FORCE
>>  	$(call if_changed,perl)
>> +
>> +OBJECT_FILES_NON_STANDARD_aesp8-ppc.o :=3D y
>=20
> Just be aware that the moment you start to run objtool on vmlinux.o this
> annotation becomes unstuck.
>=20
> Specifically things like LTO builds require this because there simply
> isn't an object file before that.

Good point. LTO isn't supported today with ftrace on powerpc, so I guess=20
we are ok with this patch for now.

To add LTO support, I guess we will need to add some annotations to this=20
file later on.


- Naveen