This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]
Other format:	[Raw text]

Re: dataflow branch merging plans.

From: "Vladimir N. Makarov" <vmakarov at redhat dot com>
To: Steven Bosscher <stevenb dot gcc at gmail dot com>
Cc: Bernd Schmidt <bernds_cb1 at t-online dot de>, Kenneth Zadeck <zadeck at naturalbridge dot com>, gcc-patches <gcc-patches at gcc dot gnu dot org>, "Park, Seongbae" <seongbae dot park at gmail dot com>, "Bonzini, Paolo" <bonzini at gnu dot org>, Serge Belyshev <belyshev at depni dot sinp dot msu dot ru>, richard dot earnshaw at arm dot com, echristo at apple dot com, "Pinski, Andrew" <andrew_pinski at playstation dot sony dot com>, "Weigand, Ulrich" <Ulrich dot Weigand at de dot ibm dot com>, Ian Lance Taylor <iant at google dot com>, "Edelsohn, David" <dje at watson dot ibm dot com>, "Berlin, Daniel" <dberlin at dberlin dot org>
Date: Thu, 24 May 2007 20:22:14 -0400
Subject: Re: dataflow branch merging plans.
References: <46543F49.8060104@naturalbridge.com> <46557380.9060105@t-online.de> <571f6b510705241344u750c6499r27942cda78f40e32@mail.gmail.com>

Steven Bosscher wrote:

On 5/24/07, Bernd Schmidt <bernds_cb1@t-online.de> wrote:

Kenneth Zadeck wrote:
> I believe that the dataflow branch is now ready to merge into the
> mainline.  We have fixed almost all of the performance problems
> associated with it.  While there are still some left, we feel
> confident that these can be addressed during the rest of stage I and
> during stage II.

Vlad's last benchmark run still showed up to 11% compile time
regression, didn't it?

No, 8% in ia64, and somewhere between 5% and 6% on the other targets
Vlad's tested.

Yes, 11% is the worst dataflow branch compile time regression which is achieved on SPEC2000 on Itanium. 8% is on SPECINT on Itanium.

  Note that the scheduler on ia64
is already dog slow even on the trunk. The DFA for Itanium is the
problem here.

Scheduler is slow on Itanium. That is true. But the DFA (used for fast pipeline hazard recognizing) was a big step forward. It sped up all Itanium compiler (not only scheduler) up to 2 times and give about 3% improvement on SPECINT2000 (which is much harder then get the same improvement on SPECFP2000). You can read more details about it in my artcile for the 1st GCC summit.

I am not saying that there is no better solution. I encorauge you to write the scheduler (and insn bundling) faster and better without the DFA. Although I should say that usage of DFA is common practise for Itanium schedulers. For example, Open64 got this a few years later than GCC.

By the way, if you switch off scheduler (-fno-schedule-insns and -fno-schedule-insns2) the datafalow branch compilation speed degradation on Itanium will be not 11% but almost 14% on SPECFP2000.

Steven, I really value your help in rewriting old pipileine descriptions to the DFA ones but please be more accurate when you jump the conclusion like "the DFA for Itanium is the problem here" or saying on IRC that most interesting work on the DFA was done by Zack. That is not true.

Follow-Ups:
- Re: dataflow branch merging plans.
  - From: Steven Bosscher

References:
- dataflow branch merging plans.
  - From: Kenneth Zadeck
- Re: dataflow branch merging plans.
  - From: Bernd Schmidt
- Re: dataflow branch merging plans.
  - From: Steven Bosscher

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]