Date: Tue, 23 Aug 2005 13:14:18 +0100 From: Alex Zbyslaw <xfb52@dial.pipex.com> To: =?ISO-8859-1?Q?Jimmy_B=E4ckstr=F6m?= <jimmy.backstrom@crystone.se> Cc: freebsd-questions <freebsd-questions@freebsd.org> Subject: Re: FAILURE - SETFEATURES SET TRANSFER MODE status=41 Message-ID: <430B131A.2060102@dial.pipex.com> In-Reply-To: <430B061A.6090509@crystone.se> References: <430AD74E.8010701@crystone.se> <430AF712.8080809@dial.pipex.com> <430B061A.6090509@crystone.se>
next in thread | previous in thread | raw e-mail | index | archive | help
Jimmy B=E4ckstr=F6m wrote: > > > Alex Zbyslaw wrote: > >> Jimmy B=E4ckstr=F6m wrote: >> >>> Hi list! >>> I've googled and searched the mailing list archives for this but=20 >>> couldn't find anything useful. >>> I just installed freebsd 5.4 on a dell PE 2850 with RAID5 (PERC).=20 >>> The installation goes fine and as far as I can tell it runs ok,=20 >>> haven't really done anything with yet though so I can be wrong, but=20 >>> when booting I get two messages that concerns me: >>> >>> ata2-master: FAILURE - SETFEATURES SET TRANSFER MODE=20 >>> status=3D41<READY,ERROR> error=3D4<ABORTED> >>> ata2-slave: FAILURE - SETFEATURES SET TRANSFER MODE=20 >>> status=3D41<READY,ERROR> error=3D4<ABORTED> >>> >>> and >>> >>> Interrupt storm detected on "irq18: uhci2"; throttling interrupt sour= ce >>> >>> When it comes to the ata2-master/slave errors I understand that they = >>> happen for quite a few people running later versions of freebsd? I=20 >>> haven't found anything on the net that points directly at my=20 >>> hardware specifics, it seems to happen to a lot of different drives. = >>> I'm using 4 seagate harddrives by the way. >>> >>> If this problem has been adressed on the list before, I apologise, I = >>> couldn't find anything about it. I appreciate any pointers and if=20 >>> someone wants me to post dmesg just ask. >>> >> Set up two of these servers recently. >> >> We totally ignored the interrupt storm. We couldn't make it go away=20 >> but it seemed to do no harm. >> >> The ata errors are froeither related to the DRAC virtual devices, or=20 >> the real CD, if you have one. They seem to confuse atapicam. If you = >> take atapicam out of the kernel the messages will go away. Your hard = >> drives are SCSI and have nothing to do with ata. >> > Doh! > Didn't even think about the drives being SCSI. My bad. :) > About atapicam, I can't find it in the kernel config, what line should = > I be looking for? > My bad. That's what I tried to get rid of the error and it turns out,=20 on closer inspection, that it didn't work :-( Should have checked my=20 facts more closely! However, I am convinced that the error is not important. It is=20 certainly not related to your disks (which is what you really care=20 about, right?). My supposition is that it is either to do with the DRAC (do you have=20 one?) or to the chipset used to drive the ATAPI devices. If I look at=20 the boot log: acd0: CDRW <HL-DT-STCD-RW/DVD-ROM GCC-4244N/B101> at ata0-master PIO4 ata2-master: FAILURE - SETFEATURES SET TRANSFER MODE=20 status=3D41<READY,ERROR> error=3D4<ABORTED> ata2-slave: FAILURE - SETFEATURES SET TRANSFER MODE=20 status=3D41<READY,ERROR> error=3D4<ABORTED> acd1: CDROM <VIRTUALCDROM DRIVE/> at ata2-slave BIOSPIO acd1 is the "virtual cd" provided by the DRAC and is ata2-slave. My=20 best guess is that the virtual floppy, also provided by the DRAC, is=20 ata2-master. Since these devices are virtual, I guess that is causing=20 the ata code to barf somehow. However, at least the virtual CD works=20 just fine, and I installed FreeBSD on one of the machines, from a CD=20 that was actually loaded into a Windows host on the same network, and as = far as the install process was concerned the CD was local to the Dell. =20 It was slow, but it was so neat. The machine could, in theory, have=20 been half a globe away and I wouldn't have needed anyone to load a CD=20 for me! You could try taking out atapifd and see if that makes the error go=20 away. But that might stop the virtual floppy from working. (The real=20 floppy should still work as it's not an atapi device). Summary ----------- Despite the errors, all the important devices work just fine (keyboards, = disks, CD, virtual CD, DRAC). I only see the interrupt storm at boot=20 time, and although I don't like it, I'm prepared to ignore it as long as = everything keeps working. --Alex PS If you have the PERC 4e controller, then you can get an interface to=20 it from FreeBSD. See the sysutils/megarc port.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?430B131A.2060102>