From shenkin@still3.chem.columbia.edu Sun Mar 19 23:35:27 1995 Received: from mailhub.cc.columbia.edu for shenkin@still3.chem.columbia.edu by www.ccl.net (8.6.10/930601.1506) id WAA13290; Sun, 19 Mar 1995 22:49:07 -0500 Received: from still3.chem.columbia.edu by mailhub.cc.columbia.edu with SMTP id AA12796 (5.65c+CU/IDA-1.4.4/HLK for <@mailhub.cc.columbia.edu:chemistry@ccl.net>); Sun, 19 Mar 1995 22:49:05 -0500 Received: by still3.chem.columbia.edu (931110.SGI/930416.SGI.AUTO) for @mailhub.cc.columbia.edu:chemistry@ccl.net id AA27855; Sun, 19 Mar 95 22:49:04 -0500 From: "Peter Shenkin" Message-Id: <9503192249.ZM27853@still3.chem.columbia.edu> Date: Sun, 19 Mar 1995 22:49:02 -0500 In-Reply-To: "Simon Kilvington" "CCL:malloc etc. on DEC Alpha" (Mar 16, 6:13pm) References: <199503161813.SAA27313@vision.soton.ac.uk> X-Mailer: Z-Mail (3.1.0 22feb94 MediaMail) To: "Simon Kilvington" , chemistry@ccl.net Subject: Re: CCL:malloc etc. on DEC Alpha Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 On Mar 16, 6:13pm, Simon Kilvington wrote: > Subject: CCL:malloc etc. on DEC Alpha > Konrad Hinsen wrote: > > > > However, in all cases of malloc()-related crashes I have experienced > > so far (and that is a large number), the code calling malloc() was at > > fault. ..... This is my experience as well. Whether the user error trashes the malloc() arena or not depends essentially on luck. > a (fairly) easy way to check for writing passed the end of malloc > blocks is to replace calls to malloc with calls to a routine which claims > slightly more space than is needed, puts some id characters at the start > and end of the block eg "MALLOC_BLOCK" or something, and returns a ptr to > the first byte after the id.... You can also use the public-domain dbmalloc package, by Conor Cahill. As I understand it, this package does things similar to what Konrad suggests. Conor also has a commercial package called Sentinel which does fancier things as well. I've had great success with dbmalloc over the years -- most recently, about two weeks ago! -P. -- ************************ The secret of life: ************************* *Peter S. Shenkin, Box 768 Havemeyer Hall, Chemistry, Columbia Univ.,* * New York, NY 10027; shenkin@columbia.edu; (212) 854-5143 * ************* If you find a loose thread, don't pull it. *************