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;
MIME-Version: 1.0
In-Reply-To: <CAKv+Gu87DmGLG7N8X8+V7LppXm72SJUmCcEpaehCE5Rt0Z4FNw@mail.gmail.com>
References: <20180627160604.8154-1-ard.biesheuvel@linaro.org>
 <20180627160604.8154-3-ard.biesheuvel@linaro.org> <20180628085003.GA2494@hirez.programming.kicks-ass.net>
 <CAKv+Gu87DmGLG7N8X8+V7LppXm72SJUmCcEpaehCE5Rt0Z4FNw@mail.gmail.com>
From:   Ard Biesheuvel <ard.biesheuvel@linaro.org>
Date:   Thu, 28 Jun 2018 11:04:45 +0200
Message-ID: <CAKv+Gu-xPR6UEiP0NT0zYXKRYXqwFCXP=nf0cC87PZdDYBZB4A@mail.gmail.com>
Subject: Re: [PATCH 2/5] kernel/jump_label: implement generic support for
 relative references
To:     Peter Zijlstra <peterz@infradead.org>
Cc:     Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        linux-arm-kernel <linux-arm-kernel@lists.infradead.org>,
        "the arch/x86 maintainers" <x86@kernel.org>,
        Catalin Marinas <catalin.marinas@arm.com>,
        Will Deacon <will.deacon@arm.com>,
        Thomas Gleixner <tglx@linutronix.de>,
        Ingo Molnar <mingo@redhat.com>, Arnd Bergmann <arnd@arndb.de>,
        Steven Rostedt <rostedt@goodmis.org>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On 28 June 2018 at 11:02, Ard Biesheuvel <ard.biesheuvel@linaro.org> wrote:
> On 28 June 2018 at 10:50, Peter Zijlstra <peterz@infradead.org> wrote:
>> On Wed, Jun 27, 2018 at 06:06:01PM +0200, Ard Biesheuvel wrote:
>>> diff --git a/include/linux/jump_label.h b/include/linux/jump_label.h
>>> index 86ec0652d3b1..aa203dffe72c 100644
>>> --- a/include/linux/jump_label.h
>>> +++ b/include/linux/jump_label.h
>>> @@ -121,6 +121,32 @@ struct static_key {
>>>  #include <asm/jump_label.h>
>>>
>>>  #ifndef __ASSEMBLY__
>>> +#ifdef CONFIG_HAVE_ARCH_JUMP_LABEL_RELATIVE
>>> +
>>> +struct jump_entry {
>>> +     int code;
>>> +     int target;
>>> +     int key;
>>> +};
>>
>> I much prefer you use 'u32' there.
>>
>
> Actually, they are signed so that would be s32. But yeah, I can change that.
>
>>
>>> +static void jump_label_swap(void *a, void *b, int size)
>>> +{
>>> +     long delta = (unsigned long)a - (unsigned long)b;
>>> +     struct jump_entry *jea = a;
>>> +     struct jump_entry *jeb = b;
>>> +     struct jump_entry tmp = *jea;
>>> +
>>> +     jea->code       = jeb->code - delta;
>>> +     jea->target     = jeb->target - delta;
>>> +     jea->key        = jeb->key - delta;
>>> +
>>> +     jeb->code       = tmp.code + delta;
>>> +     jeb->target     = tmp.target + delta;
>>> +     jeb->key        = tmp.key + delta;
>>> +}
>>> +
>>>  static void
>>>  jump_label_sort_entries(struct jump_entry *start, struct jump_entry *stop)
>>>  {
>>> @@ -56,7 +72,9 @@ jump_label_sort_entries(struct jump_entry *start, struct jump_entry *stop)
>>>
>>>       size = (((unsigned long)stop - (unsigned long)start)
>>>                                       / sizeof(struct jump_entry));
>>> -     sort(start, size, sizeof(struct jump_entry), jump_label_cmp, NULL);
>>> +     sort(start, size, sizeof(struct jump_entry), jump_label_cmp,
>>> +          IS_ENABLED(CONFIG_HAVE_ARCH_JUMP_LABEL_RELATIVE) ? jump_label_swap
>>> +                                                           : NULL);
>>>  }
>>
>> That will result in jump_label_swap being an unused symbol for some
>> compile options.
>>
>
> No, and isn't that the point of IS_ENABLED()? The compiler sees a
> reference to jump_label_swap(), so it won't complain about it being
> unused.
>
>> Would it not be much nicer to write that like:
>>
>> static void jump_label_swap(void *a, void *b, int size)
>> {
>>         struct jump_entry *jea = a, *jeb = b;
>>
>> #ifdef CONFIG_HAVE_ARCH_JUMP_LABEL_RELATIVE
>>         long delta = a - b;
>>
>>         jea->code += delta;
>>         jea->target += delta;
>>         jea->key += delta;
>>
>>         jeb->code -= delta;
>>         jeb->target -= delta;
>>         jeb->key -= delta;
>> #else
>>
>>         swap(*jea, *jeb);
>> }
>>
>> And then unconditionally use jump_label_swap().
>
> Meh. I thought IS_ENABLED() was preferred over #ifdef, no? That way,
> the compiler always sees the code, and simply discards it without
> complaining if it ends up left unused.

... and it means the sort() routine will unconditionally perform an
indirect function call even if the arch does not require it.