Date: Sat, 11 Apr 2015 20:57:52 +0200 (CEST)
From: Thomas Gleixner <tglx@linutronix.de>
To: Linus Torvalds <torvalds@linux-foundation.org>
cc: Ingo Molnar <mingo@kernel.org>, Denys Vlasenko <dvlasenk@redhat.com>,
        Borislav Petkov <bp@alien8.de>, Tim Chen <tim.c.chen@linux.intel.com>,
        Andy Lutomirski <luto@amacapital.net>, Jason Low <jason.low2@hp.com>,
        Brian Gerst <brgerst@gmail.com>,
        Aswin Chandramouleeswaran <aswin@hp.com>,
        "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>,
        Davidlohr Bueso <dave@stgolabs.net>,
        Peter Zijlstra <a.p.zijlstra@chello.nl>,
        "H. Peter Anvin" <hpa@zytor.com>, LKML <linux-kernel@vger.kernel.org>,
        Peter Zijlstra <peterz@infradead.org>
Subject: Re: [PATCH] x86: Turn off GCC branch probability heuristics
In-Reply-To: <CA+55aFygq02Y9FATg9JOrLkM97a__DNi_LoMd+FmX9n3a-JoBA@mail.gmail.com>
Message-ID: <alpine.DEB.2.11.1504112038590.3845@nanos>
References: <20150409175652.GI6464@linux.vnet.ibm.com> <CA+55aFzXMDjQQ7jTjsPdh1RikXfgV7OCd-+13cz06MOmDBA33w@mail.gmail.com> <CA+55aFwZWi6ecDmVsMBQJTrgrW3GD2DaRtpiOspe=5amR1=dNg@mail.gmail.com> <20150409183926.GM6464@linux.vnet.ibm.com> <20150410090051.GA28549@gmail.com>
 <20150410091252.GA27630@gmail.com> <20150410092152.GA21332@gmail.com> <20150410111427.GA30477@gmail.com> <20150410112748.GB30477@gmail.com> <20150410120846.GA17101@gmail.com> <20150411092021.GA9478@gmail.com>
 <CA+55aFygq02Y9FATg9JOrLkM97a__DNi_LoMd+FmX9n3a-JoBA@mail.gmail.com>
User-Agent: Alpine 2.11 (DEB 23 2013-08-11)
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1770
Lines: 60

On Sat, 11 Apr 2015, Linus Torvalds wrote:
> On Apr 11, 2015 2:20 AM, "Ingo Molnar" <mingo@kernel.org> wrote:
> >
> > Another thing caught my attention (and I'm hijacking the RCU thread
> > again): GCC's notion of how to place branches seems somewhat random,
> > and rather bloaty.
> >
> > So I tried the experiment below on an x86 defconfig, turning off GCC's
> > branch heuristics, and it's rather surprising:
> >
> >      text           data     bss      dec         filename
> >  12566447        1617840 1089536 15273823         vmlinux.fguess-branch-probability
> >  11923593        1617840 1089536 14630969         vmlinux.fno-guess-branch-probability
> >
> > That's an 5.4% code size improvement!
> 
> Ugh. That's much larger than I would have expected. Is it just because
> gcc ends up turning
> 
>          if (a)
>                b
>          c
> 
> into
> 
>          if (a) goto out-of-line
>    return:
>          c
>          ..
>    out-of-line:
>          c
>          goto return;
> 
> a lot? Still, 5% sounds insanely big.
> 
> How much of that 5% comes from code alignment? Or was this on *top* of
> the 1-byte alignment testt?

I thinks its just the no-guess one:

   text	       data       dec  patch           reduction

7563475	    1781048  10302987   

7192973	    1780024   9931461  no-guess	           -4.8%

7354819	    1781048    958464  align-1		   -2.7%

7192973	    1780024   9931461  no-guess + align-1  -4.8%

So with the no-guess applied the align-1 does not matter anymore.

Thanks,

	tglx
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/