From owner-freebsd-bugs Wed Jan 1 12:18:54 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.4/8.8.4) id MAA21972 for bugs-outgoing; Wed, 1 Jan 1997 12:18:54 -0800 (PST) Received: from irz301.inf.tu-dresden.de (irz301.inf.tu-dresden.de [141.76.1.11]) by freefall.freebsd.org (8.8.4/8.8.4) with SMTP id MAA21947 for ; Wed, 1 Jan 1997 12:18:51 -0800 (PST) Received: from sax.sax.de (sax.sax.de [193.175.26.33]) by irz301.inf.tu-dresden.de (8.6.12/8.6.12-s1) with ESMTP id VAA08469; Wed, 1 Jan 1997 21:18:39 +0100 Received: (from uucp@localhost) by sax.sax.de (8.6.12/8.6.12-s1) with UUCP id VAA24263; Wed, 1 Jan 1997 21:18:38 +0100 Received: (from j@localhost) by uriah.heep.sax.de (8.8.4/8.6.9) id TAA29598; Wed, 1 Jan 1997 19:28:54 +0100 (MET) From: J Wunsch Message-Id: <199701011828.TAA29598@uriah.heep.sax.de> Subject: Re: bin/2347: sysinstall: ppp: recursive call in malloc() To: luigi@labinfo.iet.unipi.it (Luigi Rizzo) Date: Wed, 1 Jan 1997 19:28:54 +0100 (MET) Cc: freebsd-bugs@freefall.freebsd.org, brian@freebsd.org Reply-To: joerg_wunsch@uriah.heep.sax.de (Joerg Wunsch) In-Reply-To: <199701011436.PAA00270@labinfo.iet.unipi.it> from Luigi Rizzo at "Jan 1, 97 03:36:27 pm" X-Phone: +49-351-2012 669 X-PGP-Fingerprint: DC 47 E6 E4 FF A6 E9 8F 93 21 E0 7D F9 12 D6 4E X-Mailer: ELM [version 2.4ME+ PL17 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-bugs@freebsd.org X-Loop: FreeBSD.org Precedence: bulk As Luigi Rizzo wrote: > I am using ppp here (basically 2.1R + phkmalloc) without the above > problem. On passing, iijppp is extremely bad since it does a couple of > malloc()/free() per packet (one for the header, one for the packet > itself). On a 14.4 modem this amounts to at least 10 malloc/free per > second, think what this means on a modem server. > > This could be easily fixed, if someone wants to look at it (I am > afraid bu I don't have the time to work on this, but some of the > people who added NAT might be relatively familiar with iijppp code). > The code is relatively well structured and distinguishes between > different classes of objects, so handling free lists withih iijppp > for each class should be highly efficient. Brian Somers agreed to take over maintenance of iijppp. I'm not sure about your CVS knowledge now, Brian? -- cheers, J"org joerg_wunsch@uriah.heep.sax.de -- http://www.sax.de/~joerg/ -- NIC: JW11-RIPE Never trust an operating system you don't have sources for. ;-)