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: [wwwdocs] gcc-4.0/changes.html - Mention SMS as new optimization.


"Giovanni Bajo" <giovannibajo@libero.it> wrote on 02/15/2005 03:28:55 AM:

> Mostafa Hagog <MUSTAFA@il.ibm.com> wrote:
> 
> > This patch adds SMS to gcc-4.0/changes.html as a new optimization.
> > OK to commit?
> 
> 
> I personally don't think this is appropriate. Your explanation is very
> technical, too much for the audience of changes.html (which are GCC 
users and
> not compiler writers). I think you should provide something along 
> the lines of:
> 
> - why SMS makes the code better (what does "better scheduling" mean)

I cannot explain this without technical information, like interleaving 
loop 
iterations and so on.  We can just say that SMS is intended to schedule 
instructions in loops, rather than the traditional scheduling that look at
basic blocks without giving special handling for loops. 

> - which are the targets which are best served by SMS? Will a Pentium1 
gain the
> same benefit as a Pentium 4 from this? And a PPC? What about an embedded
> target?
I will try doing this for a general classification of machines like 
in-order 
vs. out-of-order machines, super-scalar and so on.  mentioning specific 
machines 
makes the description complicated, and I will end-up giving technical 
description 
again.

> - what are the figures on what SMS can achieve, on at least a coupleof 
target?
> Is it 10% or 1% or 50% of run-time speed? You mentioned it is disabled 
because
> it is slow, so you should give some hints on why users should still try 
to use
> it.

OK, I will look in the results that I have and make a summary, but in 
bottom line 
this is not (yet) a win (on SPEC). 

> 
> I think these information are what users would like to know about SMS. I 
am
> sure *I* do not know the answers of the above questions, and I would 
like to.
> The brief explanation of the SMS algorithm is irrelevant: the link 
> to the paper
> will be followed by those interested in details.
> 
> Another thing you could mention is that work on SMS is going on and an 
update
> is expected in GCC 4.1.
OK.

Thanks for the tips, 
Mostafa.

> 
> Giovanni Bajo
> 


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