From owner-freebsd-questions Thu Mar 16 08:41:33 1995 Return-Path: questions-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id IAA12572 for questions-outgoing; Thu, 16 Mar 1995 08:41:33 -0800 Received: from gaudi.diatel.upm.es (gaudi.diatel.upm.es [138.100.49.10]) by freefall.cdrom.com (8.6.10/8.6.6) with SMTP id IAA12564 for ; Thu, 16 Mar 1995 08:41:19 -0800 Received: by gaudi.diatel.upm.es (4.1/SMI-4.1) Thu, 16 Mar 95 17:40:33 +0100 X400-Received: by mta diatel.upm in /PRMD=/ADMD=/C=/; Relayed; Thu, 16 Mar 1995 17:40:31 UTC+0100 X400-Received: by /PRMD=iris/ADMD=mensatex/C=es/; Relayed; Thu, 16 Mar 1995 17:40:31 UTC+0100 Date: Thu, 16 Mar 1995 17:40:31 UTC+0100 X400-Originator: jmrueda@diatel.upm.es X400-Recipients: non-disclosure:; X400-Content-Type: P2-1984 (2) X400-Mts-Identifier: [/PRMD=iris/ADMD=mensatex/C=es/;950316174031] Content-Identifier: 635 From: Javier Martin Rueda To: questions@FreeBSD.org Message-Id: <635*/S=jmrueda/OU=diatel/O=upm/PRMD=iris/ADMD=mensatex/C=es/@MHS> Subject: Input/output errors with Buslogic Bt946C Mime-Version: 1.0 (Generated by Ean X.400 to MIME gateway) Sender: questions-owner@FreeBSD.org Precedence: bulk This is me again, but now with a different problem. I repeat the hardware config, just in case: Motherboard: 486DX2-66 PCI+VESA, Opti chipset PCI SCSI adapter: Buslogic Bt946C SCSI disk: Conner CFP1060S PCI Ethernet card: SMC EtherPower 8432BT DSP 550: serial card with two 16550A ports. The problem is that I get occasional "Input/output error" when accessing the hard disk. It happens with files but once it happened when accessing the swap area, and the corresponding process was killed by a SEGV after the kernel printed an error message about not being able to retrieve a page from disk. For example, if I recompile from scratch a new kernel, there's a chance that I may get one or two of those errors during the process. Sometimes, it does the whole thing without a single error. The errors don't seem to be located in a particular spot of the hard disk, but instead they happen randomly (though rarely, at least). I've read the entire disk several times and didn't get any error. I used "dd if=/dev/rsd0d of=/dev/null bs=8k count=129680" (it's a 1 Gb disk). Also, I've put the disk under heavy activity (with several finds and several bonnies concurrently), and it might report one or several errors, or it might do everything without a single error. Again, no repeatable pattern. Finally, the kernel does not print any error message. In other systems running FreeBSD, if I got a hard disk error, something appeared in the console in a bright color saying the track, cylinder, sector, etc. But here, just a message from the user process saying "Input/Output error". Shouldn't the device driver print something if it got some physical error from the controller? The disk is brand new. Before attempting to exchange it by a new one, I'm trying to figure out if this is due to some problem other than a physical defect on the disk. Any idea? Again, thanks for reading this.