Package: grub-rescue-cdrom Using debian-rescue-cdrom Grub 1.99~rc1-6 using sid 29.03.2011. Target symptom: " error: hd0 cannot get C/H/S values "
I submitted originally as a Debian Installer installation-report. Installation Report reference bug# 607301. Date: Thu, 16 Dec 2010. http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=607301 Having refined the symptoms somewhat and upgraded the level of software, I am now submitting, to [email protected] , not as a Debian di installation-report, where I would not expect any discusion or developers feedback, but for two way questions and discussion on the forum, with a view to progressing my issue. I trust this is not seen as a 'double post'. It is not meant to be. I am always open to help and guidance with the correct process. ENVIRONMENT: debian-6.0.1a-i386-DVD Binary-1 20110322-15:11 Also upgraded to a 'net update' wheezy 28.03.2011 then sid 29.03.2011. HARDWARE:=============== Machine: Motherboard: MS5169 (Micro-Star International) ATX AL9 Mainboard www.msi.com.tw/products/jumpers/5169 ( Yes AL9 ) Processor: AMD K-6-2/400 400.9Mhz Memory: Memtest86+ v4.10 224MB pass:1 errors:0 Chipset: ALLADIN5 (To clarify, Yes, that is a 5.) Bios: American Megatrend AMIBIOS 1996 A5169MS v2.6 121598 51-1215-000000-00111111-071595-41E_43C-MS5169-T99-H DASD pci board: Manufacturer: "For Best Ports Connection" Sil 3114 sata raid bios version 5.3.14 2006. Serial ATA host controller card SA3114-41R 4 port. chip: Silicon Image chip StealVine Sil 3114 CTU. printed on the chip: QG6332.1-1 0644 AD03AX2. printed on the sata pci board: SATA3114-A01 FG-SA3114-4IR-01-BC01 or it might be: FG.SA3114.4IR.01.BC01 I use this card for the fast disk control bus and future proofing, by buying sata and not pata. I am using this SATA controller with only ONE hard disk drive here. No raid at all at the moment. DASD: The only hard disk drive in the machine: 160GB sata ST3160815AS drive 0. I intend to use another disk with Linux Debian MDADM Software raid 1, mirroring, if and when I can get this to boot the succesfull Debian install. BIOS: AMIBIOS HIFLEX setup utility v1.21 1998 American Megatrends. Standard cmos setup: Screen 1. Fast Boot bypasses filesystem checkup ( currently on.) Base memory: 640KB Extd memory 223 MB 229,376KB I Will fully populate memory, if I can overcome this boot issue. Type LBA Block PIO 32Bit Mode Mode Mode Mode IDE Primary master: CDROM Pioneer DVD-RW Model:DVR-110. 4 On IDE Primary slave: Not installed. IDE Secondary master: Not installed. IDE Secondary slave: LiteOnIt.com.tw ATAPI CD-ROM Drive 36x. 4 On Boot Virus Protection 'DISABLED'. Advanced CMOS setup: Screen 2. 1st Boot device: CD-ROM 2nd Boot Device: SCSI ( Works as written in the pci Sata raid Board user manual.) 3rd Boot Device: Disabled. 4th Boot Device: Disabled. Try other Boot Devices?: NO. ( I tried setting to 'YES' and it did not recognised the pci SATA raid Controller board.) Boot to OS/2: NO. Advanced Chip Set Up: Screen 3. (Nothing of interest here.) PCI / Plug and play setup: Screen 5. Plug and play aware O/S: Yes Offboard PCI IDE CARD?: AUTO bios screen during reboot: PCI devices: Slot 4 RAID: IRQ9 Manufacturer: "For Best Ports Connection" END OF HARDWARE:======================== === See my SAVED LOGS from dgi (Debian Graphical Installer) i386 at http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=607301 Date: Thu, 16 Dec 2010. at message #15, which is my third email entry down. Date: Thu, 10 Feb 2011. Here is the lspci excerpt:- lspci -knn: 00:00.0 Host bridge [0600]: ALi Corporation M1541 [10b9:1541] (rev 04) lspci -knn: Subsystem: ALi Corporation M1541 [10b9:1541] lspci -knn: Kernel driver in use: agpgart-ali lspci -knn: 00:01.0 PCI bridge [0604]: ALi Corporation M1541 PCI to AGP Controller [10b9:5243] (rev 04) lspci -knn: 00:02.0 USB Controller [0c03]: ALi Corporation USB 1.1 Controller [10b9:5237] (rev 03) lspci -knn: Kernel driver in use: ohci_hcd lspci -knn: 00:07.0 ISA bridge [0601]: ALi Corporation M1533/M1535/M1543 PCI to ISA Bridge [Aladdin IV/V/V+] [10b9:1533] (rev b4) lspci -knn: Subsystem: ALi Corporation M1533/M1535/M1543 PCI to ISA Bridge [Aladdin IV/V/V+] [10b9:1533] lspci -knn: 00:0e.0 RAID bus controller [0104]: Silicon Image, Inc. SiI 3114 [SATALink/SATARaid] Serial ATA Controller [1095:3114] (rev 02) lspci -knn: Subsystem: Silicon Image, Inc. Device [1095:7114] lspci -knn: Kernel driver in use: sata_sil lspci -knn: 00:0f.0 IDE interface [0101]: ALi Corporation M5229 IDE [10b9:5229] (rev 20) lspci -knn: Kernel driver in use: pata_ali lspci -knn: 00:12.0 VGA compatible controller [0300]: S3 Inc. 86c375 [ViRGE/DX] or 86c385 [ViRGE/GX] [5333:8a01] (rev 01) lspci -knn: Subsystem: S3 Inc. 86c375 [ViRGE/DX] or 86c385 [ViRGE/GX] [5333:8a01] END OF SAVED LOGS:====================== PARTITION TABLE:======================== Booting Debian 6.0.1a DVD-1, Graphical Rescue Mode. With Debian sid 29.03.2011 installed on /dev/sda1 # bash (My note: To use the Command History List arrows.) root@debian:/# /sbin/fdisk -l /dev/sda Disk /dev/sda: 160.0 GB, 160041885696 255 heads. 63 sectors/track, 19457 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes I/O size (minimum/optimal): 512 bytes / 512 bytes Disk identifier 0x0......0 (My note: yes, it has an identifier) Device Boot Start End Blocks Id System /dev/sda1 * 127 19458 155272192 83 Linux /dev/sda2 1 127 1016833 5 Extended Partition 2 does not end on cylinder boundary. /dev/sda5 1 127 1016832 82 Linux swap / Solaris " Partition entries are not in disk order " root@debian:/# (My note: Re " Partition 2 does not end on a cylinder boundary. " OK, my fault. I MANUALLY partitioned differently this time around. Starting with 1GB swap at begining of disk, I then sized sda1 / root with 'MAX'. I believe this is not the problem with some of my other installs I did. I will check when I 'AUTO' partition the WHOLE disk the next time I install. I believe swap space should come first on the disk (quickest to get to) especially on a small (at the moment) RAM machine. If it starts swapping, you want the swap area of disk to be quick to get to by the disk heads. I have always been puzzled, why debian does not do this. Thrash me if I am wrong. I guess I must be wrong, and debian developers must be correct. 30.March.2011 I reinstalled Debian 6.0.1a i386 DVD-1 and Debian-update-6.0.1a-i386 . Note: I have not installed wheezy-testing or sid-unstable at this time, using the 'debian to auto/decide partition settings for the WHOLE disk' option. Then booting Debian 6.0.1a DVD-1, 'Graphical Rescue Mode.' root@debian:/# bash root@debian:/# /sbin/fdisk -l /dev/sda Disk /dev/sda: 160.0 GB, 160041885696 255 heads. 63 sectors/track, 19457 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes I/O size (minimum/optimal): 512 bytes / 512 bytes Disk identifier 0x0......2 (My note: the identifier changed ?) Device Boot Start End Blocks Id System /dev/sda1 * 1 19404 155855872 83 Linux /dev/sda2 19404 19458 432129 5 Extended /dev/sda5 19404 19458 432128 82 Linux swap / Solaris That's better. Note, both installs DO and did have a * Bootable sda1 partition. Despite grml complaining that the previous install had no partition marked bootable. My target error symptom still persists with this latest install and repartioning. Using grub-rescue-cdrom 1.99~rc1-6 Symptom: 'Booting a command list','GNU/Linux' " error: hd0 cannot get C/H/S values " END OF PARTITION TABLE:========================= Base System Installation Checklist: [O] = OK. [E] = Error Debian 6.0.1a, i386 DVD-1 Initial boot: [O] Detect network card: [O] Configure network: [N/A] Not applicable. Detect CD: [O] Load installer modules: [O] Detect hard drives: [O] Partition hard drives: [O] Install base system: [O] Configure Package Manager: Scanning DVD set disk, Binary-1 only. (yes) Clock/timezone setup: [O] User/password setup: [O] Install tasks: [O] Install boot loader: [O] Overall install: [E] Won't auto boot into new installed system, at end of install process. Same symptom with Wheezy weekly build 20110314 and sid 29th March 2011. ===BEGINING OF ISSUE SUMMARY:================= The disk /sda can be SEEN by 5 various programmes, such as; grml-sid-20110316, grub-rescue-pc 1.99~rc1-6, debian-installer-DVD-1-graphical-rescue-mode , Debian 7.0 (Wheezy) installation DVD-1 built 20110314-03:36; d-i 20110313-21:54, and later sid 29.March.2011 rescatux (Rescue A Tux/penguin) 0.22 cdrom_usb_hybrid_i386_2010_12_06, SGD Super_Grub_Disk_hybrid-1.98 S1 Not all the 5 programmes can ACCESS /sda (the single sata 160 GB disk) I cannot boot Using the grub-rescue-cdrom using relevant Linuz... and initrd... Grub 1.99~rc1-6 Symptom: "error: hd0 cannot get C/H/S values, (My target symptom.) Some programmes can ACCESS /sda. So " We have the technology Captain. " ===END OF ISSUE SUMMARY:======== this issue suappears again, further down this report. I had a look at the Debian Squeeze Installation Notes: Stable. Very impresive. It gave me plenty of help, and links, but nothing current that touched on my issue directly, I believe. Chapter 5. Issues to be aware of for squeeze http://www.debian.org/releases/stable/i386/release-notes/ch-information.en.html#ide-pata-transition Table of Contents 5.1. Potential problems 5.1.1. Migration of disk drivers from IDE to PATA subsystem 5.1.2. mdadm metadata format change requires recent Grub --- My note: Jump this paragraph. Debian Releases / Debian “wheezy” Release Information. DEAD Debian DOCUMENT LINK on 25.03.2011. Maybe because "- no release date has been set." under http://www.debian.org/releases/testing/ DEAD * Generic release status page http://release.debian.org/ "The next release of Debian is codenamed "wheezy" — no release date has been set." ===GRML==================== I tried the 'grml GRUB Live Linux' link in the above document notes. grml-daily-sid Standard (110316 i386) DVD. http://daily.grml.org/grml_sid/ dated 17-Mar-2011 01:23 size: 753M http://daily.grml.org/grml_sid/grml_sid_20110316.iso grml-daily-sid 110316 codename grml-live-autobuild [2011-03-16] Boot-options for grml-daily-sid - Debug Mode INIT: version 2.88 booting Running Linux Kernel 2.6.36-grml root@grml ~ # root@grml ~ # ls root@grml ~ # root@grml ~ # / root@grml / # ls bin/ boot/ dev/ etc/ home/ lib/ ** /live ** /media /mnt ... root@grml / # /dev/sda0 (just to see) ls: cannot access /dev/sda0 : No such file or directory. root@grml / # ls /dev/sda1 /dev/sda1 root@grml / # ls /dev/sda2 /dev/sda2 root@grml / # ls /dev/sda5 /dev/sda5 The second e2fsck -nv /dev/sda1 I performed resulted in: /dev/sda1: clean, 133152/9748480 file, 1416072/38963968 blocks . The first e2fsck -nv /dev/sda1 I performed resulted in: root@grml / # mount My note: /dev/sda1 is unmounted.( no line entry in /etc/fstab.) root@grml / # mount /dev/sda1 ... /dev/sda1 on /mnt/sda1 type ext3 (rw) root@grml / # umount /dev/sda1 root@grml / # /sbin/e2fsck -nv /dev/sda1 e2fsck 1.41.12 (17-May-2010) /dev/sda1 was not cleanly unmounted. check forced <<McTech??<< Pass1. Checking inodes, blocks, and sizes Pass2. Checking directory structure Pass3. Checking directory connectivity Pass4. Checking reference counts Pass5. Checking group summary information 113443 inodes used (1.16%) 954 non-contiguous files (0.8%) 166 non-contiguous directories (0.1%) # of inodes with ind/dind/tind blocks: 7293/60/0 1390543 blocks used (3.57%) 0 bad blocks <<McTech<< 1 large file 89420 regular files 9396 directories 12 character device files 25 block device files 2 fifos 419 links 14578 symbolic links (13474 fast symbolic links) 1 socket 113853 files McTech note:Thats all right then. Reminder: sda1 contains, after succesfully installing, (except auto booting at each install or upgrade ) versions of; Debian 6.0.1a, upgrading to wheezy weekly build, unstable-29.03.2011, and previous upgrade steps through wheezy and unstable to experimental-20.02.2011, then experimental-04.03.2011. Currently: Sid. Wait a minute... grml has shuffled things around. root@grml ~ # cat /etc/apt/sources.list.d/debian.list has deb http://cdn.debian.net/debian/ sid main contrib non-free . I had better shut grml down correctly to allow it to restore /etc/apt/sources.list If that is what it does. So no ad-hoc rebooting during a grml session then. Why ? "Important message to read in your /etc/apt/sources.list contents" during grml-ing. I reinstalled to try to reproduce my 'No Boot' symptom with Debian 6.0.1a, then upgraded to testing, then sid, to by pass any grml legacy left around, which grml might Not have left. (Note: This is not a reflection of the excelent grml. My first exposure. It is a reflection on me. I have not read the grml documentation yet. It's got to be a tick in the box for grml that I got this far without rtfm.) Change of direction and environment: REBOOT grml | Menu | Boot from hard disk. Booting from local disk (/sda1 Which still has debian sid 29.03.2011.) Boot failed: press a key to retry... REBOOT grml | Menu | Isolinux prompt boot: forensic ... Setting device /dev/sda to read only mode :done ... ... "Bootoption forensic found important notice Do *NOT* boot with something like 'grml forensic ...' but always use 'forensic ...' instead." ... Finished execution of main grml startup Welcome to grml-quickconfig (My note, McTech: "That's enough to prove grml can find /sda .") Change of direction and environment: REBOOT grml | Boot options for grml-daily-sid | Debug Mode ... INIT version 2.88 booting ... root@grml ~ # ls /dev/sda* /dev/sda /dev/sda1 /dev/sda2 /dev/sda5 Change of direction and environment: http://hdt-project.org Running from the grml disk: REBOOT grml | Addons | Run Hardware Detection Tool. 0.3.6.1 Works ok. Quiet handy. I just saw some extra information regarding the machines BIOS I had not seen or specified in my report before. Wish list: I hope probing for IDE and SATA disk drives will be included in future releases of HDT: PCI Devices (8): Vendor: Silicon Image, Inc. Product: ( My note, I not L in SiI ) , SiI 3114 [SATALink/SATARaid] Serial ATA Controller. Class: RAID bus controller Location: 00:0e.0 PCI ID: 1095:3114[1095:7114] IRQ: 9 Latency 64 Module: unknown Motherboard: Vendor MSI INC. Product: ALADDIN5 Version: 1.0 Serial: 1234567890 ( Hmmm suspicious serial # ) (I guess the motherboard is just not uniquely serial numbered.) I noticed grml HDT 0.3.6.1 defaults, if there IS no number. Chasis: | Asset Tag: ASS-9876543210 for my machine. System: | Serial: SYS-9876543210. Likewise. Bios: American Megatrends, Inc. Version: 0626 ( My note: That's 0626) Release: 07.15.95 ( So that's 15 July 1995. -071595-.) (A different version FORMAT as specified in my entry way above.) ( My note: v2.6 as apposed to 0626. ) My Hardware summary entry way above has: American Megatrend AMIBIOS 1996 A5169MS v2.6 121598 51-1215-000000-00111111-071595-41E_43C-MS5169-T99-H ( So I guess it MAY be 06.2.6 and yet the ALADDIN is 5 as in ALADDIN5.) Not to worry. Just a curiosity. They may have started with ALADDIN0, like GRUB. So in summary, if I have recollect correctly, the disk /sda can be SEEN by 5 various programmes, such as; grml, grub-rescue-pc, debian-installer-DVD-1-graphical-rescue-mode, rescatux (Rescue A Tux/penguin), SGD (Super Grub Disk). Not all the 5 programmes can ACCESS /sda (the single sata 160 GB disk) > I cannot boot Using the grub-rescue-pc using relevant Linuz... and initrd... > Grub 0.95 Symptom: "Error 21: Selected disk does not exist". ( I believe Legacy 0.9X GRUB is now superceeded by grub1.99 for my situation here.) ( My target symptom to follow is:- ) > Grub 1.99~rc1-6 Symptom: "error: hd0 cannot get C/H/S values, Some programmes can ACCESS /sda. So " We have the technology Captain. " As an aside, and I suppose my wish list for grub-rescue-pc. GRUB2 i386. 0) My 'Command List' edits are saved/kept in RAM between boot attempts, but not machine resets. 1) Q: Is there a way to save my own Grub-rescue-disk, edited 'Command List' between machine resets ? I suspect the answers are no. 2) I would like to save to external storage to floppy / usb-thumb-drive, between reboot (failures). 3) To save in memory with a SOFT WARM reboot of the OS, not a HARD COLD reset boot that clears all volatile memory. I do not think Linux has this feature (yet!) 4) or ideally (Joke-on:)to save my edits to the single hard drive.(joke-off :) Not being able to save my edits between power-off's, used to be a pain with device names, before uuid's came along. With uuid's, not being able to save my edits between pc resets is a nightmare. 5) I think I recently read that I could use either device-name or uuid. I will try device-names the next time I use the grub-rescue-pc 'command list'. 1.99~rc1-3 or above. ----------- Change of environment: I booted up the (DI) Debian Install disk,selected Debian Graphical Installer. At Set Up Passwords, I ignored it and selected [GO BACK] into the DI-Main Menu, Selected 'Partition Disks','Manual'. into Screen: 'Partition Disks', 'This is an overiew of your currently configured partitions and mount points...' SCSI3 (0,0,0) (sda) - 160.0 GB ATA ST3160815AS > #1 primary 159.6 GB B ext3 > #5 logical 4425.5MB F swap swap [GO BACK] into the DI-MM (Main Menu). Selected Abort the installation. "Are you sure ?" Yes. The observation being, the flag 'B' for Boot is currently set in partition sda1 according to Main Menu - Help. | Debian 7.0 (Wheezy) installation DVD-1 built 20110314-03:36; d-i 20110313-21:54 . ------------- Change of direction and environment: http://daily.grml.org/grml_sid/grml_sid_20110316.iso grml-daily-sid 110316 codename grml-live-autobuild [2011-03-16] ( -L = do not complain about things irrelevant for Linux.) root@grml ~ # /bin/sfdisk -l -L Disk /dev/sda: 19457 cylinders, 255 heads, 63 sectors/track Warning: extended partiion does not start at a cylinder boundary. DOS and Linux will interpret the contents differently. Units = cylinders of 8225280 bytes, blocks of 1024 bytes, counting from 0 Device Boot Start end #cyls #blocks Id System /dev/sda1 * 0+ 19403- 19404- 155855872 83 Linux /dev/sda2 19403+ 19457- 54- 432129 5 Extended /dev/sda3 0 - 0 0 0 Empty /dev/sda4 0 - 0 0 0 Empty /dev/sda5 19403+ 19457- 54- 432128 82 Linux swap / Solaris ----- Having just installed debian-6.0.1a-i386-DVD Binary-1 20110322-15:11 . I since realised, to get the 6.0.1a update, I should have also downloaded and installed debian-update-6.0.1a-i386-DVD-1.iso 22-Mar-2011 . Rebooting, I received at the end of the bios POST, the following, I then 'net updated' to wheezy-testing and then to sid-unstable 28.03.2011. all three gave symptom: SiI 3114 SATARaid BIOS Version 5.3.14 Silicon Image, Inc. Press <Ctrl+S> or F4 to enter RAID utility 0 ST3160815AS 149.GB No valid device! Press any key to continue... or) Press <Ctrl+S> or F4 to enter RAID utility Searching for Boot Record from SCSI..Not Found Boot Failure Reboot and Select proper Boot device or insert Boot Media in selected Boot device Press any key when ready --- I experimented, by tweeking one my BIOS settings. 'Advanced CMOS Setup','Try Other Boot Devices','NO' >From 'NO' to 'YES' (It still failed to IPL from the scsi disk, with the following.) SiI 3114 SATARaid BIOS Version 5.3.14 Silicon Image, Inc. Press <Ctrl+S> or F4 to enter RAID utility 0 ST3160815AS 149GB No valid device! Press any key to continue... or) Press <Ctrl+S> or F4 to enter RAID utility Searching for Boot Record from SCSI..Not Found Searching for Boot Record from IDE-0..Not Found Searching for Boot Record from Floppy..Not Found Searching for Boot Record from CDROM..Not Found Searching for Boot Record from LS-120..Not Found Searching for Boot Record from Not Found Boot Failure Reboot and Select proper Boot device or insert Boot Media in selected Boot device Press any key when ready I reverted the bios setting: 'Advanced CMOS Setup','Try Other Boot Devices' back to 'NO' TARGET SYMPTOM: Using the net-updated sid-unstable 20110329, grub-rescue-cdrom 1.99~rc1-6 and booting from the GNU/Linux -'Booting a command list' gives error: hd0 cannot get C/H/S values. 29.03.2011 Experimenting with GNU GRUB version 1.99~rc1-6 and editing the default 'GNU/Linux','Booting a command list' echo=all echo 0 setparams 'GNU/Linux' echo 1 set root=(hd0,1) echo2 linux /vmlinuz root=/dev/sda1 echo 3 initrd /initrd.img echo 4 --- booting with F10 gave: " Booting a command list 0 1 2 error: hd0 cannot get C/H/S values. 3 error: you need to load the kernel first. 4 Press any key to continue... " --- So the first error came from the 'set root=(hd0,1) stanza. A very minor wish list entry: echo=all does not work. That would have been nice if it did. Best regards, and thanks to all for the great software. McTech _______________________________________________ Help-grub mailing list [email protected] http://lists.gnu.org/mailman/listinfo/help-grub
