public inbox for gcc-help@gcc.gnu.org
 help / color / mirror / Atom feed
From: "John (Eljay) Love-Jensen" <eljay@adobe.com>
To: "NightStrike" <nightstrike@gmail.com>,
	        "Tom St Denis" <tstdenis@ellipticsemi.com>
Cc: "J.C. Pizarro" <jcpiza@gmail.com>,
	"Galloth" <lordgalloth@gmail.com>,         <gcc-help@gcc.gnu.org>
Subject: RE: reduce compilation times?
Date: Tue, 27 Nov 2007 19:41:00 -0000	[thread overview]
Message-ID: <EDC8DDD212FEB34C884CBB0EE8EC2D9103B25E0A@namailgen.corp.adobe.com> (raw)
In-Reply-To: <b609cb3b0711271121p57206b2br87c6a6c99309255d@mail.gmail.com>

Hi NightStrike,

> Is there a difference in the speed of the resulting program when everything is split into many object files instead of being combined into a single main.c ...?

There may be a negative small performance impact in a resulting program that is split into many object files instead of being combined into a single main.c.  (My expectation is that, overall, the performance impact will be negligible if it is even measurable.)

There may be a few interoperating routines that are strongly negatively impacted by being split into many object files instead of being combined into a single object file.  (My expectation for these particular routines is that they should be heavily optimized, perhaps even being rewritten in lovingly handcrafted assembly -- assuming assembly chops are superior to the optimizing compiler's amazing optimizations.)  If not re-written in hand coded assembly, at least having performance critical routines' code hand-tweaked to allow the GCC optimizer to do it's best to optimize it would be prudent (including using inline functions, and avoiding the anti-patterns that cripple optimization).

The way to assess those routines is through profiling.

GCC does not do "holistic" optimizations (yet).  In contrast, LLVM does "holistic" optimizations.

>... or is the resulting binary identical bit for bit?

No, the resulting binary is not identical bit-for-bit.

It should be identical output for identical input.  (Barring non-compliant or undefined behavior code, of course.)

HTH,
--Eljay

  reply	other threads:[~2007-11-27 19:35 UTC|newest]

Thread overview: 69+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2007-11-27 16:07 J.C. Pizarro
2007-11-27 16:19 ` Brian Dessent
2007-11-27 16:26   ` J.C. Pizarro
     [not found] ` <5abcb5650711270804o171e1facr565beec70314af75@mail.gmail.com>
2007-11-27 16:41   ` J.C. Pizarro
2007-11-27 16:46     ` Tom St Denis
2007-11-27 17:16       ` J.C. Pizarro
2007-11-27 17:46         ` Tom St Denis
2007-11-27 18:26           ` Wesley Smith
2007-11-27 19:35       ` NightStrike
2007-11-27 19:41         ` John (Eljay) Love-Jensen [this message]
2007-11-27 19:49         ` Tom St Denis
2007-11-28  9:19           ` Brian Dessent
2007-11-28 12:07             ` Tom St Denis
2007-11-28 12:35               ` Brian Dessent
2007-11-27 17:44     ` Vladimir Vassilev
     [not found]       ` <998d0e4a0711271310k657b791cy6ad5cc5721105f4c@mail.gmail.com>
2007-11-27 22:30         ` J.C. Pizarro
  -- strict thread matches above, loose matches on Subject: below --
2007-11-28 16:06 J.C. Pizarro
2007-11-28 16:16 ` Tom St Denis
2007-11-28 16:34   ` J.C. Pizarro
2007-11-28 18:18     ` Tom St Denis
2007-11-28 13:56 Duft Markus
2007-11-28 14:35 ` Tom St Denis
2007-11-29  0:23 ` Tim Prince
2007-11-28 13:25 Duft Markus
2007-11-28 13:26 ` Tom St Denis
2007-11-28 12:36 Duft Markus
2007-11-28  7:57 Duft Markus
2007-11-28 12:01 ` J.C. Pizarro
2007-11-28 12:28   ` Tom St Denis
2007-11-28 12:49     ` Fabian Cenedese
2007-11-28 13:03       ` Tom St Denis
2007-11-28 12:52     ` J.C. Pizarro
2007-11-28 13:17       ` Tom St Denis
2007-11-28 13:40         ` J.C. Pizarro
2007-11-28 13:51           ` Tom St Denis
2007-11-28 13:59             ` Tom St Denis
2007-11-28 15:51             ` John (Eljay) Love-Jensen
2007-11-28 13:30       ` Ted Byers
2007-11-28 12:12 ` John (Eljay) Love-Jensen
2007-11-28 12:31   ` J.C. Pizarro
2007-11-28 12:39     ` Tom St Denis
2007-11-28 12:54     ` John (Eljay) Love-Jensen
2007-11-28 12:18 ` Tom St Denis
2007-11-28 13:09   ` Ted Byers
2007-11-27 10:04 mahmoodn
2007-11-27 11:11 ` Andrew Haley
2007-11-27 11:15   ` mahmoodn
2007-11-27 11:30     ` Andrew Haley
2007-11-27 12:20       ` mahmoodn
2007-11-27 12:25         ` John Love-Jensen
2007-11-27 15:27           ` Tim Prince
2007-11-27 14:07         ` Andrew Haley
2007-11-28  9:01           ` mahmoodn
2007-11-28 12:11             ` John (Eljay) Love-Jensen
2007-11-30  9:15               ` mahmoodn
2007-11-30 13:33                 ` mahmoodn
2007-11-27 15:48   ` Sven Eschenberg
2007-11-27 16:27     ` Andrew Haley
2007-11-27 18:51       ` Sven Eschenberg
2007-11-27 19:21         ` Andrew Haley
2007-11-27 20:43           ` Sven Eschenberg
2007-12-01 12:20   ` mahmoodn
2007-12-03 16:14     ` Andrew Haley
2007-12-04 11:23       ` mahmoodn
2007-12-04 12:19         ` Tom Browder
2007-12-05  7:44           ` mahmoodn
2007-12-05 10:24             ` Tom Browder
2007-12-05 10:29               ` mahmoodn
2007-11-27 13:48 ` John Love-Jensen

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=EDC8DDD212FEB34C884CBB0EE8EC2D9103B25E0A@namailgen.corp.adobe.com \
    --to=eljay@adobe.com \
    --cc=gcc-help@gcc.gnu.org \
    --cc=jcpiza@gmail.com \
    --cc=lordgalloth@gmail.com \
    --cc=nightstrike@gmail.com \
    --cc=tstdenis@ellipticsemi.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).