From mboxrd@z Thu Jan  1 00:00:00 1970
From: "Indan Zupancic" <indan@nul.nu>
Subject: RE: [RFC] net: bpf_jit: Two pass JIT and other changes
Date: Mon, 2 Apr 2012 22:03:12 +1000
Message-ID: <5f98ccc12c7887af855ff44700b528af.squirrel@webmail.greenhost.nl>
References: <AE90C24D6B3A694183C094C60CF0A2F6026B6ECF@saturn3.aculab.com>
Mime-Version: 1.0
Content-Type: text/plain;charset=UTF-8
Content-Transfer-Encoding: 8bit
Cc: "Eric Dumazet" <eric.dumazet@gmail.com>, netdev@vger.kernel.org,
	linux-kernel@vger.kernel.org
To: "David Laight" <David.Laight@ACULAB.COM>
Return-path: <linux-kernel-owner@vger.kernel.org>
In-Reply-To: <AE90C24D6B3A694183C094C60CF0A2F6026B6ECF@saturn3.aculab.com>
Sender: linux-kernel-owner@vger.kernel.org
List-Id: netdev.vger.kernel.org

On Mon, April 2, 2012 19:28, David Laight wrote:
>
>> 1) Sadly, gcc isn't always smart enough to inline emit_code().
>>    So turn it into a macro to force gcc to inline it. This saves
>>    about 500 bytes. Tested with gcc 4.6.0. An alternative to the
>>    macro would be to create inline emit_code1(), emit_code2() etc.
>
> Does giving it the __attribute__((always_inline)) help?
> I had to mark a load of small static functions that way
> after a minor change (possibly to the overall size)
> stopped gcc inlining everything.

I did try that first, but it didn't work for some reason. Perhaps
the kernel uses some flag that overrides it when optimising for size.
Maybe plain -Os does that already, I don't know.

Greetings,

Indan