Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!cs.utexas.edu!uunet!convex!schnoebe@convex.COM From: schnoebe@convex.COM (Eric Schnoebelen) Newsgroups: news.sysadmin Subject: Re: does cnews correct the inode problem? Keywords: Cnews inode ialloc s5ialloc Message-ID: <2112@convex.UUCP> Date: 16 Oct 89 17:47:50 GMT References: <200981@hrc.UUCP> <1989Oct16.043634.3051@twwells.com> Sender: news@convex.UUCP Reply-To: eric@cirr.com (Eric Schnoebelen) Organization: Central Iowa (Model) Railroad, Dallas, Tx. Lines: 33 In article <1989Oct16.043634.3051@twwells.com> bill@twwells.com (T. William Wells) writes: -In article <200981@hrc.UUCP> dan@hrc.UUCP (Dan Troxel) writes: -: I have had it! The alloca/inode problem has brought my system down -: 3 times today. I need to correct it. Does cnews correct the problem? - -No. It is a kernel bug. Yes, the problem is a kernel bug, but it has been my experience that C news doesn't aggravate the bug. I say this because I have a patched version of ialloc.c ( s5ialloc.c ) running on my Microport System V/AT system that tells me when the bug would manifest itself, and I haven't seen any messages from it since I started running C news. ( of course I had set my inode count very high while running B news to help keep the problem at arms length, along with scripts to change init states and run fsck on the news partition. ) -: If not, what do I need to do to correct the problem. I don't have the -: source code, but I do have the object code that I can re-link. - -What kind of system do you have? I just posted a fix for Interactive -2.0.2. I have a fix for Microport SysV/3.0e. I have had other -information sent to me as well, which, once I get it organized, I'll -post. AT&T has also issued fixes for their systems, and Interactive claims to have fixed it with their 2.0.2 release. Comment to them at 386ixsupp@isc780c.isc.com if they haven't. Otherwise, beat on your vendor for them to fix the problem. ------ Eric Schnoebelen eric@egsner.cirr.com "My other computer is a Convex" schneobe@convex.com