Date: Thu, 14 Dec 2000 12:24:33 -0800 (PST) From: Matthew Jacob <mjacob@feral.com> To: current@freebsd.org Subject: Old Business: followup about dirty 'modules' directory Message-ID: <Pine.LNX.4.21.0012141217200.28426-100000@zeppo.feral.com> In-Reply-To: <388F6AB0.8B806F3B@scc.nl>
next in thread | previous in thread | raw e-mail | index | archive | help
A month or so ago I queried/complained about src/sys/modules getting corrupted with architecture specific derived files such that I could no longer share between i386 && alpha. Part of this issue had to do with having some idiocy on my part, but part of it *seems* to have to do with if you change the default /usr/obj to somewhere else. The setup I have here for all of my varied test machines is: private/per-machine /usr/src (CVS to local cvsup'd copy, ergo, out of date) shared, NFS mounted sys, mounted as /tstsys, and thence a loopback mount to a local directory for /tstsys/compile, CVS to freefall (ergo, could be kept up to the minute for all platforms) What was happening is that just one machine seemed to be polluting /tstsys/modules with derived files (.ko's, .depends, etc.). Big PITA. I queried and everyone said, "huh?", so I shut up and simply loopback mounted /tstsys/modules to a local directory for that machine as well. I finally decided that this was stupid, and looked into it more. It appears that this was the only machine that had a MAKEOBJDIRPREFIX set. Somehow this seemed to interfere with the dance that goes on with building modules for each kernel. I haven't tracked it further than this, but thought I should mention it in case anyone else has stubber their toe on this one. -matt To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.LNX.4.21.0012141217200.28426-100000>