From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S267490AbUHDWi0 (ORCPT ); Wed, 4 Aug 2004 18:38:26 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S267485AbUHDWiZ (ORCPT ); Wed, 4 Aug 2004 18:38:25 -0400 Received: from adsl-67-114-19-185.dsl.pltn13.pacbell.net ([67.114.19.185]:17793 "EHLO bastard.smallmerchant.com") by vger.kernel.org with ESMTP id S267488AbUHDWhU (ORCPT ); Wed, 4 Aug 2004 18:37:20 -0400 Message-ID: <4111651E.1040406@tupshin.com> Date: Wed, 04 Aug 2004 15:37:18 -0700 From: Tupshin Harper1 User-Agent: Mozilla Thunderbird 0.7.1 (Windows/20040626) X-Accept-Language: en-us, en MIME-Version: 1.0 To: Kernel Mailing List Subject: Re: /dev/hdl not showing up because of fix-ide-probe-double-detection patch References: <411013F7.7080800@tupshin.com> In-Reply-To: <411013F7.7080800@tupshin.com> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Tupshin Harper1 wrote: > I have an x86 setup with a large(9) number of ide hard drives. With > 2.6.8-rc2, all of them show up (as a,b,d,e,f,g,h,i,j,k, and l), but on > 2.6.8-rc2-mm1, the last one (hdl) does not show up, even though it's a > slave on the same controller as hdk. Everything about the config is > identical between the two. Is there some default limit or other > restriction in the current mm kernels that's preventing that drive > from showing up? I haven't tried any other recent mm kernels, so I > don't know when this was introduced. > > -Tupshin Well, I found the source of the problem. Dmesg gives a "ignoring undecoded slave" message, which is coming from the fix-ide-probe-double-detection patch. Both /dev/hdk and /dev/hdl are the same model of hard drive, and unfortunately, they both report that they are Model M0000000000000000000, which triggers the double detection removal. Does anybody have a suggestion (besides maintaining my own kernel without this patch)? Is the serial number settable? Is there a boot param I can use? Also, what is in /proc/ide/hd?/identity besides serial number. The two drives have very similar identity files, but they are slightly different. Could that additional info be used to check for duplicates? -Tupshin ---snippet from dmesg--- PDC20276: IDE controller at PCI slot 0000:00:0c.0 ACPI: PCI interrupt 0000:00:0c.0[A] -> GSI 19 (level, low) -> IRQ 19 PDC20276: chipset revision 1 PDC20276: 100% native mode on irq 19 ide4: BM-DMA at 0xdc00-0xdc07, BIOS settings: hdi:pio, hdj:pio ide5: BM-DMA at 0xdc08-0xdc0f, BIOS settings: hdk:pio, hdl:pio hdi: SAMSUNG SV8004H, ATA DISK drive hdj: Maxtor 6Y250P0, ATA DISK drive ide4 at 0xec00-0xec07,0xe802 on irq 19 hdk: Maxtor 4R120L0, ATA DISK drive hdl: Maxtor 4R120L0, ATA DISK drive ide-probe: ignoring undecoded slave ide5 at 0xe400-0xe407,0xe002 on irq 19