From: Andi Kleen <ak@muc.de>
Subject: Re: Hotspot. Dynamic compilers "better" than static one?
Date: 1998/05/30
Date: 1998-05-30T00:00:00+00:00 [thread overview]
Message-ID: <m3d8cvfqbk.fsf@fred.muc.de> (raw)
In-Reply-To: 6kpk0h$qmo$1@supernews.com
Roedy Green <roedy@oberon.ark.com> writes:
> Over a decade ago I spent months writing screaming fast code for the
> nucleus of a 32-bit Forth compiler that hand optimised every jump this way.
> The secondary advantage is that the most commonly used code is more likely
> to be pre-fetched or in cache. A static optimiser can't do this, since it
> has no knowledge of which branch is the more likely.
Many modern compilers support profiling feedback. This means you compile
the program, run it to generate the profiling option and compile the
program again with feeding the profiling data into the compiler.
Dynamic compiling has the potential advantage that the code is tuned to
the particular usage pattern of the enduser, but I think for most programs
that does not matter much.
-Andi
next prev parent reply other threads:[~1998-05-30 0:00 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <6knj4m$odp$1@nnrp1.dejanews.com>
1998-05-30 0:00 ` Hotspot. Dynamic compilers "better" than static one? nabbasi
1998-05-30 0:00 ` Roedy Green
1998-05-30 0:00 ` Andi Kleen [this message]
[not found] ` <dewar.896629645@merv>
1998-06-02 0:00 ` Dr Richard A. O'Keefe
1998-06-02 0:00 ` Lieven Marchand
1998-06-01 0:00 ` Norman H. Cohen
1998-06-03 0:00 ` John Volan
1998-06-05 0:00 ` Norman H. Cohen
1998-06-08 0:00 ` John Volan
replies disabled
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox