Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Wed, 04 Mar 2020 02:10:29 -0800
User-Agent: K-9 Mail for Android
In-Reply-To: <20200304100437.GM2596@hirez.programming.kicks-ass.net>
References: <20200303204144.GA9913@avx2> <9947D7CB-B9CD-47E0-BC5E-C7FC3A81FC7B@zytor.com> <20200304100437.GM2596@hirez.programming.kicks-ass.net>
MIME-Version: 1.0
Content-Type: text/plain;
 charset=utf-8
Content-Transfer-Encoding: quoted-printable
Subject: Re: [PATCH] x86/acpi: make "asmlinkage" part first thing in the function definition
To:     Peter Zijlstra <peterz@infradead.org>
CC:     Alexey Dobriyan <adobriyan@gmail.com>, rjw@rjwysocki.net,
        lenb@kernel.org, tglx@linutronix.de, mingo@redhat.com,
        bp@alien8.de, luto@kernel.org, x86@kernel.org,
        linux-kernel@vger.kernel.org, linux-acpi@vger.kernel.org
From:   hpa@zytor.com
Message-ID: <E88DA201-307C-4F47-A931-13C3894D2E8F@zytor.com>
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On March 4, 2020 2:04:38 AM PST, Peter Zijlstra <peterz@infradead=2Eorg> wr=
ote:
>On Wed, Mar 04, 2020 at 12:54:09AM -0800, hpa@zytor=2Ecom wrote:
>> On March 3, 2020 12:41:44 PM PST, Alexey Dobriyan
><adobriyan@gmail=2Ecom> wrote:
>> >g++ insists that function declaration must start with extern "C"
>> >(which asmlinkage expands to)=2E
>> >
>> >gcc doesn't care=2E
>> >
>> >Signed-off-by: _Z6Alexeyv <adobriyan@gmail=2Ecom>
>> >---
>> >
>> > arch/x86/kernel/acpi/sleep=2Ec |    2 +-
>> > arch/x86/kernel/acpi/sleep=2Eh |    2 +-
>> > 2 files changed, 2 insertions(+), 2 deletions(-)
>> >
>> >--- a/arch/x86/kernel/acpi/sleep=2Ec
>> >+++ b/arch/x86/kernel/acpi/sleep=2Ec
>> >@@ -43,7 +43,7 @@ unsigned long acpi_get_wakeup_address(void)
>> >  *
>> >  * Wrapper around acpi_enter_sleep_state() to be called by
>assmebly=2E
>> >  */
>> >-acpi_status asmlinkage __visible x86_acpi_enter_sleep_state(u8
>state)
>> >+asmlinkage acpi_status __visible x86_acpi_enter_sleep_state(u8
>state)
>> > {
>> > 	return acpi_enter_sleep_state(state);
>> > }
>> >--- a/arch/x86/kernel/acpi/sleep=2Eh
>> >+++ b/arch/x86/kernel/acpi/sleep=2Eh
>> >@@ -19,4 +19,4 @@ extern void do_suspend_lowlevel(void);
>> >=20
>> > extern int x86_acpi_suspend_lowlevel(void);
>> >=20
>> >-acpi_status asmlinkage x86_acpi_enter_sleep_state(u8 state);
>> >+asmlinkage acpi_status x86_acpi_enter_sleep_state(u8 state);
>>=20
>> Are you building the kernel with C++?!
>
>He is :-) IIRC he's got a whole bunch of patches that removes all the
>C++ keywords from the kernel=2E

I'm genuinely curious: I have heard it says that gcc is something like 10x=
 slower in C++ mode even for what is otherwise basically C code (i=2Ee=2E n=
o templates etc=2E)

Does that match observations?
--=20
Sent from my Android device with K-9 Mail=2E Please excuse my brevity=2E