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: [RFC] Patch: RAM-based heuristics for ggc-min-heapsize and ggc-min-expand


 > From: Geoff Keating <geoffk@geoffk.org>
 > 
 > > From: "Kaveh R. Ghazi" <ghazi@caip.rutgers.edu>
 > > 
 > > cc1 -E ... foo.c
 > > cc1 -fpreprocessed ... foo.i
 > > as ... foo.s
 > > etc
 > > ...
 > > 
 > > Isn't this output coming from the driver?
 > 
 > Some of it, but not all of it.  For instance, cc1 already prints
 > lines like:
 > 
 > GNU CPP version 3.1 20020420 (prerelease) (cpplib) (Darwin/PowerPC)
 > GNU C version 3.1 20020420 (prerelease) (ppc-darwin)
 >         compiled by GNU C version 3.1 20020420 (prerelease).
 > 
 > > If so, then if we want "-param foo=bar" to appear in the line with the
 > > "cc1 -fpreprocessed", where would this printf live if not in the
 > > driver?  And if it lives in the driver then we must compute the value
 > > there so we can pass it as the "%d" value in the printf statement.
 > 
 > I was thinking of putting it on a separate line.  I'm sure that GCC
 > engineers can cope with copying-and-pasting a few extra parameters in
 > the rare case that they matter.

That string comes out of toplev.c:print_version().  I guess that's
clear enough for me to try and implement should we agree on it.

So now we're back to you wanting it one way and Richard Earnshaw
another.  How do we break the deadlock?

		--Kaveh
--
Kaveh R. Ghazi			ghazi@caip.rutgers.edu


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