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]

Re: cpplib: interned representation of identifiers


Zack Weinberg wrote:

> On Sat, Jul 08, 2000 at 10:16:34PM +0200, I wrote:

> > Zack Weinberg wrote:

> > > cpplib already creates a hash table entry for every identifier seen in
> > > the program.  This patch goes one step further, and stores pointers to
> > > those entries in identifier tokens instead of their text.  This
> > > reduces memory consumption and simplifies some interface bits
> > > considerably.

> > Perhaps I'm missing something here.  In ye Ole Days the maximum length
> > of an identifier was 8 characters - certainly not many of them exceed
> > that length nowadays.  Might that explain why I do not see any speedup
> > on alphaev6-unknown-linux-gnu, a thoroughly 8-byte pointer beast ;-)

> I'm not surprised you see no speedup, but this isn't why.

...

> And if you look through the graph a bit more, you discover that the
> place to concentrate on is parse_define.  Full gprof output on
> request.  Also, if you'd care to run the same tests on your system I'd
> much appreciate it.  Instructions on request.

[ Sigh, me and my big mouth ... ]

OK, I'll give it a try; maybe not today anymore, though.

-- 
Toon Moene - mailto:toon@moene.indiv.nluug.nl - phoneto: +31 346 214290
Saturnushof 14, 3738 XG  Maartensdijk, The Netherlands
GNU Fortran 77: http://gcc.gnu.org/onlinedocs/g77_news.html
GNU Fortran 95: http://g95.sourceforge.net/ (under construction)

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