From owner-freebsd-current@FreeBSD.ORG Mon Apr 21 10:07:17 2008 Return-Path: Delivered-To: current@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id A22A61065671; Mon, 21 Apr 2008 10:07:17 +0000 (UTC) (envelope-from marck@rinet.ru) Received: from woozle.rinet.ru (woozle.rinet.ru [195.54.192.68]) by mx1.freebsd.org (Postfix) with ESMTP id 2E2568FC20; Mon, 21 Apr 2008 10:07:16 +0000 (UTC) (envelope-from marck@rinet.ru) Received: from localhost (localhost [127.0.0.1]) by woozle.rinet.ru (8.14.2/8.14.1) with ESMTP id m3LA7FwC038975; Mon, 21 Apr 2008 14:07:15 +0400 (MSD) (envelope-from marck@rinet.ru) Date: Mon, 21 Apr 2008 14:07:15 +0400 (MSD) From: Dmitry Morozovsky To: Scott Long In-Reply-To: <20080420201013.M56317@woozle.rinet.ru> Message-ID: <20080421140600.E72747@woozle.rinet.ru> References: 9060000000212025383 <1415691208445504@webmail12.yandex.ru> <48077B1C.5070608@samsco.org> <20080420185448.X56317@woozle.rinet.ru> <480B5DA0.6030404@samsco.org> <20080420192735.G56317@woozle.rinet.ru> <20080420193212.Y56317@woozle.rinet.ru> <20080420194331.Y56317@woozle.rinet.ru> <480B6664.9040602@samsco.org> <20080420201013.M56317@woozle.rinet.ru> X-NCC-RegID: ru.rinet X-OpenPGP-Key-ID: 6B691B03 MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0 (woozle.rinet.ru [0.0.0.0]); Mon, 21 Apr 2008 14:07:15 +0400 (MSD) Cc: "Andrey V. Elsukov" , emaste@freebsd.org, current@freebsd.org, sos@freebsd.org Subject: Re: Adaptec 1420SA support? X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 21 Apr 2008 10:07:17 -0000 On Sun, 20 Apr 2008, Dmitry Morozovsky wrote: DM> SL> This is just yet another item in the long list of problems with the ata DM> SL> driver. As a workaround, try the following patch: DM> SL> DM> SL> --- ata-raid.c DM> SL> +++ ata-raid.c DM> SL> @@ -146,7 +146,7 @@ DM> SL> rdp->disk->d_mediasize = (off_t)rdp->total_sectors * DEV_BSIZE; DM> SL> rdp->disk->d_fwsectors = rdp->sectors; DM> SL> rdp->disk->d_fwheads = rdp->heads; DM> SL> - rdp->disk->d_maxsize = 128 * DEV_BSIZE; DM> SL> + rdp->disk->d_maxsize = 32768; DM> SL> rdp->disk->d_drv1 = rdp; DM> SL> rdp->disk->d_unit = rdp->lun; DM> SL> /* we support flushing cache if all components support it */ DM> DM> I thought about just the same quick'n'dirty fix, but you're, as usual, quickier DM> ;-) DM> DM> Here are the results: DM> DM> # bsdlabel -Bw ar0 DM> # fdisk -s ar0 DM> /dev/ar0: 9709 cyl 255 hd 63 sec DM> Part Start Size Type Flags DM> 4: 0 50000 0xa5 0x80 DM> # bsdlabel -R ar0 /dev/stdin DM> a: 256M 16 4.2BSD 0 0 DM> b: 4G * swap 0 0 DM> d: 4G * 4.2BSD 0.0 DM> e: 8G * 4.2BSD 0 0 DM> # bsdlabel ar0 DM> # /dev/ar0: DM> 8 partitions: DM> # size offset fstype [fsize bsize bps/cpg] DM> a: 524288 16 4.2BSD 0 0 0 DM> b: 8388608 524304 swap DM> c: 155985920 0 unused 0 0 # "raw" part, don't DM> edit DM> d: 8388608 8912912 4.2BSD 0 0 0 DM> e: 16777216 17301520 4.2BSD 0 0 0 DM> # newfs /dev/ar0a DM> /dev/ar0a: 256.0MB (524288 sectors) block size 16384, fragment size 2048 DM> using 4 cylinder groups of 64.02MB, 4097 blks, 8256 inodes. DM> super-block backups (for fsck -b #) at: DM> 160, 131264, 262368, 393472 DM> GEOM: Reconfigure ar0a, start 8192 length 268435456 end 268443647 DM> GEOM: Reconfigure ar0b, start 268443648 length 4294967296 end 4563410943 DM> GEOM: Reconfigure ar0c, start 0 length 79864791040 end 79864791039 DM> GEOM: Reconfigure ar0d, start 4563410944 length 4294967296 end 8858378239 DM> GEOM: Reconfigure ar0e, start 8858378240 length 8589934592 end 17448312831 DM> # DM> DM> Seems to behave well so far. Will test further. At least after simulating drive loss (atacotrol detach, atacontrol attach) I can't rebuild ar0: marck@moleskin:~# atacontrol status ar0 ar0: ATA RAID1 status: DEGRADED subdisks: 0 ad16 ONLINE 1 ad18 ONLINE marck@moleskin:~# atacontrol rebuild ar0 atacontrol: ioctl(IOCATARAIDREBUILD): Input/output error marck@moleskin:~# Or, should I wipe out ar label from the second disk to emulate disk replacement? Sincerely, D.Marck [DM5020, MCK-RIPE, DM3-RIPN] [ FreeBSD committer: marck@FreeBSD.org ] ------------------------------------------------------------------------ *** Dmitry Morozovsky --- D.Marck --- Wild Woozle --- marck@rinet.ru *** ------------------------------------------------------------------------