From mboxrd@z Thu Jan 1 00:00:00 1970 Message-ID: <541c85407ce34cd8c4beb1bdbe60946e@plan9.bell-labs.com> From: "Russ Cox" To: 9fans@cse.psu.edu Subject: Re: [9fans] Writing (new) drivers MIME-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit Date: Mon, 16 Sep 2002 12:21:13 -0400 Topicbox-Message-UUID: ec443120-eaca-11e9-9e20-41e7f4b1d025 You'll be much happier if you can set up two Plan 9 systems, one as a crash box and one to do your development on. Connect the two with a serial line and ethernet, and then you can load kernels over the ether (set up tftpd and dhcpd on the real machine), boot from your development machine (man kfscmd, look for listen), type at the crash box via the serial line (set up consolefs), and run the debugger over the console (man rdbfs). It sounds like a bit of effort, and it is, but once you try it you'll never want to use a less helpful system. The really nice thing about what I've described is that the crash box has no real persistent state, so you don't have to worry about losing your file system, and being able to keep your editing context while the other machine reboots is really great. I still do one-machine kernel development more often than I'd like, and it's just no fun. Russ