This is the mail archive of the gcc@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]

INSN attribute to enable/disable alternatives [was Re: GCC 4.3.0: Stage 3]


Hi Ian,

sorry for the late response.  I was on vacation for the last two
weeks.  (Btw. thats also the reason for the S/390 daily build system
being down over the last days.  It is back online now.)

> > The lengthy part is necessary to have attribute getter functions which
> > allow to specify the alternative as additional parameter.  I admit
> > that these are a lot of quite mechanical changes to the genattr stuff
> > but my hope was that these getter functions might be useful in other
> > contexts as well.
> 
> I'm not sure how.

Without having any concrete plans I could imagine that there might be
other special insn attributes added which might be used to influence the
alternative decisions done in reload and co. . Not that the process of
finding the right alternative isn't already complicated enough I think
insn attributes would be a nice interface to enable the back end to
influence that process.

> Right now you can refer to the attribute values for an alternative
> when computing a different attribute for that alternative.  Can you
> use that ability to compute your enabled alternative?

I'm not sure how this can buy me anything here.  What I need is access
to the value of an insn attribute for a *different* alternative.

The only thing I could think of would be an insn attribute defining a
vector of flags saying which alternatives are enabled.  This attribute
would return the whole vector for all the alternatives.  This would
avoid the new getter functions but it is not really a beautiful
approach - right?!

Bye,

-Andreas-


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