Date: Thu, 15 Apr 1999 01:05:57 -0500 (CDT) From: Anthony Kimball <alk@pobox.com> To: davids@webmaster.com Cc: chat@FreeBSD.ORG Subject: RE: swap-related problems Message-ID: <14101.32737.825891.700729@avalon.east> References: <14101.21892.460225.807747@avalon.east> <000001be86f8$fcd4d2d0$021d85d1@whenever.youwant.to>
next in thread | previous in thread | raw e-mail | index | archive | help
Quoth David Schwartz on Wed, 14 April: : : I'm not sure I'm following you. How are you suggesting a person write a : portable application so that copy-on-write of its code or libraries doesn't : cause it to blow up in an overcommit situation? I'm not -- but at least the parent of a child that dies in this way can detect it. And it won't die because of an operation which the ANSI C standard specifies should be legal and return null was performed, if mmap-backed malloc is used. Which is an improvement. Your comments point out that the issue is much more grey than my earlier postings have painted it: A reliable malloc is insufficient to prevent over-commit death. Okay, so it's not good enough, practically. Let's do better: How about an mmap-backed fork? To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-chat" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?14101.32737.825891.700729>