Return to BSD News archive
Path: sserve!manuel.anu.edu.au!munnari.oz.au!news.Hawaii.Edu!ames!agate!howland.reston.ans.net!spool.mu.edu!uunet!pipex!sunic!isgate!krafla!adamd From: adamd@rhi.hi.is (Adam David) Newsgroups: comp.os.386bsd.questions Subject: Re: sigseg and sigbus errors and cache problems Message-ID: <6495@krafla.rhi.hi.is> Date: 7 Mar 93 03:19:38 GMT References: <1993Feb17.193715.3658@spang.Camosun.BC.CA> <1993Feb24.144055.13545@cm.cf.ac.uk> Sender: news@rhi.hi.is Lines: 17 Nntp-Posting-Host: hengill.rhi.hi.is paul@isl.cf.ac.uk (Paul) writes: >I've been having lots of traps when doing long compiles and I noticed a >few articles recently that suggest that it's because of cache problems. [speculation about CPU external cache deleted] The problem is with the FS buffer cache, it gets corrupted. This is less evident after applying patches 1-82 but the problem is still there. During compiles the most common symptom is that programs get killed by signal 11 (segmentation violation) for instance when cpp is reading in a header file. The header file now seems to contain binary data, but after a reboot the file almost always contains the correct (textual) contents. -- Adam David (adam@veda.is)