From owner-freebsd-questions Tue Jan 9 22:13:07 1996 Return-Path: owner-questions Received: (from root@localhost) by freefall.freebsd.org (8.7.3/8.7.3) id WAA11014 for questions-outgoing; Tue, 9 Jan 1996 22:13:07 -0800 (PST) Received: from hoover.stanford.edu (hoover.Stanford.EDU [36.33.0.99]) by freefall.freebsd.org (8.7.3/8.7.3) with ESMTP id WAA10996 for ; Tue, 9 Jan 1996 22:12:55 -0800 (PST) Received: from HOOVER.STANFORD.EDU by HOOVER.STANFORD.EDU (PMDF V4.3-10 #13307) id <01HZTL1RW8TY00AB2F@HOOVER.STANFORD.EDU>; Tue, 09 Jan 1996 22:13:47 -0800 (PST) Date: Tue, 09 Jan 1996 22:13:47 -0800 (PST) From: Annelise Anderson Subject: Device Problems To: questions@FreeBSD.org Message-id: <01HZTL1RW8U000AB2F@HOOVER.STANFORD.EDU> X-VMS-To: IN%"questions@FreeBSD.org" X-VMS-Cc: ANDRSN MIME-version: 1.0 Content-type: TEXT/PLAIN; CHARSET=US-ASCII Content-transfer-encoding: 7BIT Sender: owner-questions@FreeBSD.org Precedence: bulk I compiled a new kernel with ProAudio Spectrum 16 and soundblaster stuff in it. The PAS factory default is irq 7; Soundblaster part of card uses irq 5. These are also the irq's used by my two parallel ports, so I used "conflicts" and the options to allow conflicts etc. in the kernel. I did this quite carefully, reading all the documents as I went along and even changing the defaults in one of the *.h files before compiling. The new kernel booted just fine. I downloaded some *.au files and tried cat test.au > /dev/pcaudio and got "device not configured." Tried /dev/speaker with the same result. With cat test.au > dev/nca0, it sent the file to the printer. I disabled lpt0 and lpt1 with userconfig, i.e., booting kernel -c. I tried some of the other devices associated with sound (pas0, nca4, sb0); the test.au file overwrote the devices. I tried MAKEDEV but it didn't work; I couldn't find information about the major and minor numbers for the devices I wanted to recreate in any documentation, so couldn't create them by hand with mknod. So I recompiled the kernel, hoping the devices would be created in this process. Instead, the old devices (and the overwritten ones) are still there, but there is a whole slew of new devices associated with sound. Not wanting to overwrite the device files again without a backup, I created /dev/bak and tried copying (while in X) all the device files in /dev to /dev/bak. I got some messages on wd* devices "device not configured" (the hard drives are SCSI) and eventually things stopped without exiting. Some new devices may have been created in this process. The only entry in /dev/bak was var. This I deleted and tried copying just one file, /dev/zero, to /dev/bak. This filled up the file system to 109%. I finally found the offending file and deleted it. I tried to copy just the new device /dev/audio0 (to which /dev/audio is now symbolically linked) to /dev/bak, but got a message cp: /dev/audio0: Interrupted system call. Unquestionably I would have preferred to have this system backed up long ago, but I am stilled puzzled about how to run the Colorado tape drive with ft. The system is 2.10 from the cdrom. No doubt some time I will find this very funny, as you may well now. Meanwhile I'm puzzled about not being able to make a copy of a device file and getting them overwritten. Any explanations of what I did wrong and what's going on would be appreciated. Annelise