This is the mail archive of the
mailing list for the GCC project.
Re: [PATCH] Value profile based optimizations, part 1
- From: Nathan Sidwell <nathan at codesourcery dot com>
- To: Jan Hubicka <jh at suse dot cz>
- Cc: Zdenek Dvorak <rakdver at atrey dot karlin dot mff dot cuni dot cz>, gcc-patches at gcc dot gnu dot org
- Date: Mon, 09 Jun 2003 12:55:20 +0100
- Subject: Re: [PATCH] Value profile based optimizations, part 1
- Organization: Codesourcery LLC
- References: <20030426220941.GA14611@atrey.karlin.mff.cuni.cz> <3EABD519.firstname.lastname@example.org> <20030427163514.GA5895@atrey.karlin.mff.cuni.cz> <3EAE8A13.email@example.com> <20030429153741.GA20897@atrey.karlin.mff.cuni.cz> <3EB3BA9E.firstname.lastname@example.org> <20030503215839.GA7007@atrey.karlin.mff.cuni.cz> <3EB51BF4.email@example.com> <20030608204856.GA23446@atrey.karlin.mff.cuni.cz> <3EE45249.firstname.lastname@example.org> <20030609105305.GN31246@kam.mff.cuni.cz>
Jan Hubicka wrote:
Although you have vpt.h and vpt.c, you have inserted code into profile.c
that has intimate knowledge of vpt. The vpt files are small. Would it
not be more sensible to either
1) put all the vpt stuff into profile.h & profile.c
2) put all the vpt stuff into vpt.h & vpt.c
my inclination is for #1?
vpt stuff is going to grow and I would be against putting all that into
profile.c/h contains the routines to actually profile stuff, while vpt
will contain optimizations that does use the data. That split seems to
be pretty clean to me (everything in profile.c so far is about analysis)
Perhaps the interface and/or comment can be clarified to avoid the
ok, makes sense. Zdenek, there needs to be some explanetory comments
about the big picture.
makes sense too.
* flags.h (flag_value_histograms): Declare.
why not flag_value_profile? you've called this value profiling after
all? Histogram is the implementation technique, value profiling is
the desired outcome. A user is more likely to understand what the
-fvalue-profile would be fine with me.
Perhaps we can also think about sanyfing the naming if -profile-arcs
-fbranch-probabilities pair. Perhaps by adding aliased -fprofile-gen
-fprofile-use that is easier to remember.
Nathan Sidwell :: http://www.codesourcery.com :: CodeSourcery LLC
The voices in my head said this was stupid too
email@example.com :: http://www.planetfall.pwp.blueyonder.co.uk