This is the mail archive of the libstdc++@gcc.gnu.org mailing list for the libstdc++ 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] optional fast iterators for set/map in libstdc++

From: Paolo Carlini <paolo dot carlini at oracle dot com>
To: Paolo Bonzini <bonzini at gnu dot org>
Cc: "janis187 at us dot ibm dot com" <janis187 at us dot ibm dot com>, "libstdc++ at gcc dot gnu dot org" <libstdc++ at gcc dot gnu dot org>, "gcc-patches at gcc dot gnu dot org" <gcc-patches at gcc dot gnu dot org>
Date: Fri, 8 May 2009 09:50:10 +0200
Subject: Re: [RFC] optional fast iterators for set/map in libstdc++
References: <1241744347.6152.26.camel@janis-laptop> <4A03CC71.1040509@gnu.org>

...
implemented using only 2 bits rather than 16 bytes of overhead (and then you can in principle get rid of the parent pointer, and end up actually saving memory).

Thanks Paolo. Then I also suggest exploring other ways to implement the same basic idea.

I'm not sure about the speed benefits of threaded
trees, but +33% in the overhead is quite dramatic.

I agree. Yesterday I also noticed it just before going to sleep and figured out that, in any case, this can't be enabled unconditionally, but was too tired to type ;)

Paolo

Follow-Ups:
- Re: [RFC] optional fast iterators for set/map in libstdc++
  - From: Paolo Carlini

References:
- [RFC] optional fast iterators for set/map in libstdc++
  - From: Janis Johnson
- Re: [RFC] optional fast iterators for set/map in libstdc++
  - From: Paolo Bonzini

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