Bug#760885: partitionmanager: HTML report produced for operations applied gives incorrect KDE version
Package: partitionmanager Version: 1.0.2-1 Severity: minor Dear Maintainer, This is a silly and (presumably) cosmetic bug: if you use KDE Partition Manager to make some changes to a hard drive, and then, when it has successfully completed, you click the 'Open in external browser' button (first clicking 'Details ' if necessary to show the 'Open ...' button) to cause an HTML report to be generated detailing the changes, the report shows the KDE Partition Manager version correctly (1.0.2), but reports the KDE version as '4.4.3', instead of the '4.8.4' that it actually is. This is despite KDE Partition Manager being fully aware elsewhere of the correct KDE version, as shown if you do 'Help | About', in which case the KDE version is shown correctly as 4.8.4. I have screenshots illustrating this bug, which I'll try to attach to the report. The bug is repeatable at will (on every run). There is no apparent adverse effect on the requested disk operation, so this is just a cosmetic bug, albeit a rather strange one. -- System Information: Debian Release: 7.6 APT prefers stable-updates APT policy: (500, 'stable-updates'), (500, 'stable') Architecture: i386 (i686) Kernel: Linux 3.2.0-4-686-pae (SMP w/2 CPU cores) Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Versions of packages partitionmanager depends on: ii kdebase-runtime4:4.8.4-2 ii libblkid1 2.20.1-5.3 ii libc6 2.13-38+deb7u3 ii libgcc11:4.7.2-5 ii libkdecore54:4.8.4-4+deb7u1 ii libkdeui5 4:4.8.4-4+deb7u1 ii libkio54:4.8.4-4+deb7u1 ii libparted0debian1 2.3-12 ii libqtcore4 4:4.8.2+dfsg-11 ii libqtgui4 4:4.8.2+dfsg-11 ii libstdc++6 4.7.2-5 ii libuuid1 2.20.1-5.3 partitionmanager recommends no packages. Versions of packages partitionmanager suggests: ii dosfstools 3.0.13-1 pn hfsplusnone pn hfsutils none pn jfsutils none pn ntfsprogs none pn reiser4progs none pn reiserfsprogs none pn xfsprogs none -- no debconf information More A quick look through the package source code shows the following code fragment in file src/util/report.cpp : s += tableLine(i18n(Date:), KGlobal::locale()-formatDateTime(KDateTime::currentLocalDateTime())); s += tableLine(i18n(Program version:), KGlobal::mainComponent().aboutData()-version()); s += tableLine(i18n(LibParted version:), ped_get_version()); s += tableLine(i18n(KDE version:), KDE_VERSION_STRING); but there are no other references to the variable KDE_VERSION_STRING anywhere in any of the source package files, and I'm not a KDE programmer so don't know whether this is a standard data item in a KDE development environment. FWIW, the system in which I experience this bug is a Wheezy VM, hosted by Virtualbox running on a Squeeze host - which of course is KDE 4.4.3 . -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#759610: Small correction (typo/braino)
Aaaargh ... when I wrote : bug 754580 filed against package pkg-kde-extras by Colin Watson that should have been filed against partitionmanager (I had altogether too many tabs open at the time). Also, it's worth noting that 754580 is marked closed/done, with comments that both an NMU of a (minimally, to support parted 3.1) tweaked 1.0.3 has been prepared (by Colin W), and that upstream 1.1.0 is being packaged (by a new maintainer). It's still unclear (to me) whether or not 1.1.0 addresses 4KiB sector partition alignment issues. Cheers Nick -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#759697: partitionmanager: HTML report produced after disk operation gives incorrect KDE version
Package: partitionmanager Version: 1.0.2-1 Severity: minor This is a silly and peculiar cosmetic bug: if you use KDE Partition Manager to make some changes to a hard drive, and then, when it has successfully completed, you view the operation report in HTML format by clicking the 'Open in external browser' button (first clicking 'Details ' if necessary to show the 'Open ...' button), the report shows the KDE Partition Manager version correctly (1.0.2), but reports the KDE version as '4.4.3', instead of the '4.8.4' that it actually is. This is despite KDE Partition Manager being fully aware elsewhere of the correct KDE version, as shown if you do 'Help | About', in which case the KDE version is correctly shown as 4.8.4. A quick look through the package source code shows the following code fragment in file src/util/report.cpp : s += tableLine(i18n(Date:), KGlobal::locale()-formatDateTime(KDateTime::currentLocalDateTime())); s += tableLine(i18n(Program version:), KGlobal::mainComponent().aboutData()-version()); s += tableLine(i18n(LibParted version:), ped_get_version()); s += tableLine(i18n(KDE version:), KDE_VERSION_STRING); but there are no other references to the variable KDE_VERSION_STRING anywhere in any of the source package files, and I'm not a KDE programmer so don't know whether this is a standard data item (external environment variable ?) in a KDE development environment. This bug is being reported against the Wheezy version of partitionmanager, but exactly the same fault is observed using the Squeeze version - perhaps unsurprising given that both Debian releases have the same partitionmanager version. Squeeze KDE is 4.4.5, and that in Wheezy is 4.8.4, so the reporting of KDE version 4.4.3 is mysterious - perhaps it was on the package maintainer's development workstation at the time the package was built, and has somehow become hard-coded ? There is no obvious adverse effect on the requested disk operation, so this is just a cosmetic bug, albeit a rather strange one. -- System Information: Debian Release: 7.6 APT prefers stable-updates APT policy: (500, 'stable-updates'), (500, 'stable') Architecture: i386 (i686) Kernel: Linux 3.2.0-4-686-pae (SMP w/2 CPU cores) Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Versions of packages partitionmanager depends on: ii kdebase-runtime4:4.8.4-2 ii libblkid1 2.20.1-5.3 ii libc6 2.13-38+deb7u3 ii libgcc11:4.7.2-5 ii libkdecore54:4.8.4-4+deb7u1 ii libkdeui5 4:4.8.4-4+deb7u1 ii libkio54:4.8.4-4+deb7u1 ii libparted0debian1 2.3-12 ii libqtcore4 4:4.8.2+dfsg-11 ii libqtgui4 4:4.8.2+dfsg-11 ii libstdc++6 4.7.2-5 ii libuuid1 2.20.1-5.3 partitionmanager recommends no packages. Versions of packages partitionmanager suggests: ii dosfstools 3.0.13-1 pn hfsplusnone pn hfsutils none pn jfsutils none pn ntfsprogs none pn reiser4progs none pn reiserfsprogs none pn xfsprogs none -- no debconf information -- System Information: Debian Release: 7.6 APT prefers stable-updates APT policy: (500, 'stable-updates'), (500, 'stable') Architecture: i386 (i686) Kernel: Linux 3.2.0-4-686-pae (SMP w/2 CPU cores) Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Versions of packages partitionmanager depends on: ii kdebase-runtime4:4.8.4-2 ii libblkid1 2.20.1-5.3 ii libc6 2.13-38+deb7u3 ii libgcc11:4.7.2-5 ii libkdecore54:4.8.4-4+deb7u1 ii libkdeui5 4:4.8.4-4+deb7u1 ii libkio54:4.8.4-4+deb7u1 ii libparted0debian1 2.3-12 ii libqtcore4 4:4.8.2+dfsg-11 ii libqtgui4 4:4.8.2+dfsg-11 ii libstdc++6 4.7.2-5 ii libuuid1 2.20.1-5.3 partitionmanager recommends no packages. Versions of packages partitionmanager suggests: ii dosfstools 3.0.13-1 pn hfsplusnone pn hfsutils none pn jfsutils none pn ntfsprogs none pn reiser4progs none pn reiserfsprogs none pn xfsprogs none -- no debconf information -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#759610: partitionmanager: partition alignment attribute missing from 'new partition' dialog
Package: partitionmanager Version: 1.0.2-1 Severity: important Tags: upstream KDE Partition Manager 1.0.2 lacks a feature which it seems is needed with modern large hard drives - specifically the ability to specify partition alignment when creating new partitions. More accurately, the feature is needed when creating partitions on drives with 4KB physical sectors but which still report a logical sector size of 512 bytes (so-called first-generation advanced format drives) which is a feature of most modern drives over about 1TB in size. On such drives, if a partition is created starting at a logical sector which is itself not at the beginning of a 4KB physical sector then disk transfers to/from such a partition may require 2 physical sectors to be transferred when 1 would have otherwise been sufficient, depending on filesystem cluster size, access pattern (sequential/random) and transfer direction (read/write), potentially resulting in serious performance problems. See [1] [2] [3] [4] [5]. The situation is complex, and I may not have understood it completely :-) It is compounded by an apparent difference of opinion among drive manufacturers as to whether sector numbering should begin at 0 (traditional) or 1 (easier for 4K-drive firmware ?) - see the LKML thread at [6]. At any rate the general advice seems to be that if partitions are aligned to MiB boundaries on advanced format drives then the performance problems will be avoided. This means the first partition on the drive will start at sector 2048, rather than the traditional sector 63 - which will be fine unless the drive needs to be shared (dual-booted) with a Windows=XP installation (in which case I haven't been able to figure out what is recommended). Since at least 2012, GParted has had a partition alignment widget (a dropdown) on the 'new partition' dialog, offering a choice of cylinder, MiB, or 'none' for the alignment attribute (I assume 'none' means logical sector alignment). Unless partitionmanager is somehow automagically detecting the need for 4KiB-style alignment and silently doing the Right Thing, then I suggest it needs enhancing to offer a similar alignment choice. I note there is a new upstream partitionmanager version 1.1.0, at a new homepage [7], where it is stated the changes include support for GPT partition tables, and some nice GUI improvements in the area of partition widget design (paraphrasing), but there are no screen shots, and I'm still attempting to understand the source code to see whether or not partition alignment issues are addressed. If the new version *does* address the problem then bug 754580 filed against package pkg-kde-extras by Colin Watson [8] (a request for version 1.1.0 to be packaged) will probably be sufficient for Jessie+ releases. [1] http://hothardware.com/Reviews/WDs-1TB-Caviar-Green-w-Advanced-Format-Windows-XP-Users-Pay-Attention/?page=2 [2] http://en.wikipedia.org/wiki/4_KB_sector_alignment#512e [3] http://johannes-bauer.com/linux/wdc/?menuid=3 [4] http://www.ibm.com/developerworks/library/l-4kb-sector-disks/ [5] https://ata.wiki.kernel.org/index.php/ATA_4_KiB_sector_issues [6] https://lkml.org/lkml/2009/2/25/444 [7] https://stikonas.eu/wordpress/?p=8 [8] https://bugs.debian.org/754580 Sorry for the length of this bug report - it seemed helpful to try to give a full statement of the problem. -- System Information: Debian Release: 7.6 APT prefers stable-updates APT policy: (500, 'stable-updates'), (500, 'stable') Architecture: i386 (i686) Kernel: Linux 3.2.0-4-686-pae (SMP w/2 CPU cores) Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Versions of packages partitionmanager depends on: ii kdebase-runtime4:4.8.4-2 ii libblkid1 2.20.1-5.3 ii libc6 2.13-38+deb7u3 ii libgcc11:4.7.2-5 ii libkdecore54:4.8.4-4+deb7u1 ii libkdeui5 4:4.8.4-4+deb7u1 ii libkio54:4.8.4-4+deb7u1 ii libparted0debian1 2.3-12 ii libqtcore4 4:4.8.2+dfsg-11 ii libqtgui4 4:4.8.2+dfsg-11 ii libstdc++6 4.7.2-5 ii libuuid1 2.20.1-5.3 partitionmanager recommends no packages. Versions of packages partitionmanager suggests: ii dosfstools 3.0.13-1 pn hfsplusnone pn hfsutils none pn jfsutils none pn ntfsprogs none pn reiser4progs none pn reiserfsprogs none pn xfsprogs none -- no debconf information -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#759610: Small correction (typo/braino)
Aaaargh ... when I wrote : bug 754580 filed against package pkg-kde-extras by Colin Watson that should have been filed against partitionmanager (I had altogether too many tabs open at the time). Also, it's worth noting that 754580 is marked closed/done, with comments that both an NMU of a (minimally, to support parted 3.1) tweaked 1.0.3 has been prepared (by Colin W), and that upstream 1.1.0 is being packaged (by a new maintainer). It's still unclear (to me) whether or not 1.1.0 addresses 4KiB sector partition alignment issues. Cheers Nick -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#759610: More info: warning from partitionmanager 1.0.2 on squeeze
I've just tried creating a test partition on the only advanced format hard drive I have, which is a new Toshiba DT01ACA200 attached to a system still running squeeze (but with the 3.2.0 kernel from backports), with partitionmanager 1.0.2-1 (identical functionality presumably to the wheezy version). Creating a partition leaving the free space before at its default value of 0MiB results in a partition that begins at sector 63. The HTML operation report produced at the end shows that ext4 filesystem creation produces the following warning: Command: mkfs.ext4 /dev/sdb1 mke2fs 1.41.12 (17-May-2010) /dev/sdb1 alignment is offset by 512 bytes. This may result in very poor performance, (re)-partitioning suggested. Filesystem label= OS type: Linux Block size=4096 (log=2) Fragment size=4096 (log=2) Stride=1 blocks, Stripe width=0 blocks i.e. the partition is reported misaligned by 1 logical sector. If I repeat the test but specify a free space before of 1MiB, then the new partition begins at sector 16065 (see below), and filesystem creation reports : Command: mkfs.ext4 /dev/sdb1 mke2fs 1.41.12 (17-May-2010) /dev/sdb1 alignment is offset by 3584 bytes. This may result in very poor performance, (re)-partitioning suggested. Filesystem label= OS type: Linux Block size=4096 (log=2) Fragment size=4096 (log=2) Stride=1 blocks, Stripe width=0 blocks i.e. the partition is reported misaligned by 7 logical sectors. The seemingly bizarre start sector 16065 would correspond in virtual (BIOS- style) CHS disk geometry to 255heads x 63sectors, i.e. 1 cylinder, so it's possible that partitionmanager has a further bug where it works in cylinders even when claiming to work in MiB :) It is not possible to use the free space before widget to enter a quantity in any unit other than MiB, so it appears impossible to create a properly aligned partition. Cheers, Nick -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#686632: Packaging Fault: kgpg should depend on some version of gpg
Package: kgpg Version: 4:4.4.5-1+squeeze1 Severity: minor Content-Type: text/plain; charset=us-ascii MIME-Version: 1.0 Content-Transfer-Encoding: 7bit From: Nick Boyce n...@glimmer.adsl24.co.uk To: Debian Bug Tracking System sub...@bugs.debian.org Subject: Packaging Fault: kgpg should depend on some version of gpg Message-ID: 20120904035835.2770.23363.report...@shimmer.glimmer.adsl24.co.uk X-Mailer: reportbug 4.12.6 Date: Tue, 04 Sep 2012 04:58:35 +0100 Package: kgpg Version: 4.4.5-1+squeeze1 Severity: minor kgpg is designed as a KDE4 GUI wrapper for the commandline GnuPG crypto toolkit, and as such does not itself contain any OpenPGP facilities. It is therefore useless unless it is installed alongside some version of GnuPG, and so it would surely be helpful to users - if only for documentary reasons - if the kgpg package depended on either gnupg or gnupg2 (I don't think gpgme contains enough functionality). This would of course also provide for automatic installation of the supporting GnuPG software on any system which doesn't already have it. I acknowledge that current Debian systems are likely to have GnuPG already installed because it seems to be a base package these days, presumably because it's necessary in order to check package signatures during package installation by APT. I note that Ubuntu's current kgpg package also lacks the gnupg dependency, having apparently dropped it around the time of the change from KDE3 to KDE4. I can confirm that on Debian Sarge / KDE 3.3.2 the kgpg package had a gnupg dependency (I still have such a system) so maybe the dependency was dropped by Debian at the KDE3/4 transition, perhaps on the grounds that GnuPG will always be present. This bug is being submitted in case the ommission is inadvertent. -- System Information: Debian Release: 6.0.5 APT prefers stable-updates APT policy: (500, 'stable-updates'), (500, 'stable') Architecture: amd64 (x86_64) Kernel: Linux 2.6.32-5-amd64 (SMP w/4 CPU cores) Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash -- System Information: Debian Release: 6.0.5 APT prefers stable-updates APT policy: (500, 'stable-updates'), (500, 'stable') Architecture: amd64 (x86_64) Kernel: Linux 2.6.32-5-amd64 (SMP w/4 CPU cores) Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Versions of packages kgpg depends on: ii kdebase-runtime 4:4.4.5-1 runtime components from the offici ii kdepim-runtime4:4.4.7-1 Runtime components for akonadi-kde ii libc6 2.11.3-3 Embedded GNU C Library: Shared lib ii libkabc4 4:4.4.5-2 library for handling address book ii libkdecore5 4:4.4.5-2+squeeze3 the KDE Platform Core Library ii libkdeui5 4:4.4.5-2+squeeze3 the KDE Platform User Interface Li ii libkio5 4:4.4.5-2+squeeze3 the Network-enabled File Managemen ii libkpimutils4 4:4.4.5-2 library for dealing with email add ii libqt4-dbus 4:4.6.3-4+squeeze1 Qt 4 D-Bus module ii libqtcore44:4.6.3-4+squeeze1 Qt 4 core module ii libqtgui4 4:4.6.3-4+squeeze1 Qt 4 GUI module ii libsolid4 4:4.4.5-2+squeeze3 Solid Library for KDE Platform ii libstdc++64.4.5-8The GNU Standard C++ Library v3 kgpg recommends no packages. kgpg suggests no packages. -- no debconf information -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#465207: Query: shouldn't the --unrar switch itself should be distributable ?
While I understand the problem with the unrar code itself, it seems to me that simply enabling the --unrar parameter of clamscan should not entail incorporating or distributing any unrar code at all - the code to parse the --unrar parameter and call the non-free unrar binary if specified surely belongs to ClamAV alone ? In which case the ClamAV binaries could remain free and distributable under Debian policy, while individual sysadmins could make their own decision about whether to install the non-free unrar binary package, and use the --unrar switch to request that clamscan call it. If I've stated this correctly, please consider enabling the functionality of the --unrar parameter. Thanks Nick Boyce -- Science is the poetry of reality -- Richard Dawkins -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#292724: Installer Fails To Install Grub - /boot/grub/stage1 not read correctly
On Sun, 30 Jan 2005 17:52:39 +0100, Frans Pop wrote: On Saturday 29 January 2005 03:11, Nick Boyce wrote: ... there is no /boot directory at all - but there _is_ one on /target - i.e. the file /target/boot/grub/stage1 *does* exist. Is the installer simply passing the wrong --root-directory parameter to grub-install ? [...] No: grub installation is run inside a chroot on /target. Ah - thanks - I should have thought of that. (1) I notice that the above partition table print from fdisk shows my two Linux partitions (7 8) have type=FAT16 (type 06) despite their just having been formatted by the installer as ext3 and swap That looks more likely. We've seen before that the partition types are not always set correctly in edge cases. Suggest you try fixing this with fdisk from vt2 after hardware detection I've just done this, and it fixed the problem :-) I switched to VC2 when the prompt about not finding a DHCP server appeared, then ran cfdisk, and changed the partition types to 83 and 82 respectively. When I exited cfdisk I was told I *had* to reboot to have the partition changes recognised (any change to the MBR I guess). After a reboot and restart of the installation from the start, the Grub installation step completed (to the root partition boot sector, /dev/hda7) straight away, no problem. Then, I wiped reinstalled again, and this time selected the LILO bootloader, and that installed without problem too. (2) I created all the partitions using Win2K Disk Manager, at a time We advise to never use non-linux tools for partitioning. Well I understand what you're saying, but I was a bit nervous of trying the Sarge partitioner - I tried the Ubuntu Warty Warthog installer 3 weeks ago on this same box, and it trashed my MBR. Maybe I used it badly, but I asked it to create a 2nd primary partition for its root (Win2K was already installed), and specified create the new partition at the start of the available space. The effect seemed to be to reorder the partition table entries (or something) such that Win2K wouldn't boot any more - it couldn't find its boot partition, even though the partition was still there and the contents were visible using the Win2K Recovery Console. I assume the Ubuntu partitioner is closely related to Sarge's ? Also, I used the same plan with several other boxes at work which dual-boot WinNT Workstation and Woody : I used WinNT Disk Manager to create the Linux partitions, and the Woody installer didn't have any problem at all with them. There may be a regression in Sarge here You might try again later next week using a _daily_ CD image. I will - I'm happy to keep this box available for a while to test Sarge installations if that helps iron out any other installer issues. I've made a backup copy of the MBR now, so I'm thinking of deleting the entire extended partition, and installing Sarge into a new 2nd primary partition, to see if there's any difference from the above experience. I'll add any new relevant info to this bug report, but I guess you could close this bug now if that's The Right Thing. Thanks for your help. Nick Boyce -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#292724: Installer Fails To Install Grub - /boot/grub/stage1 not read correctly
Package: installation-reports Debian-installer-version: 2005-01-22 sarge-i386-netinst.iso RC2 from http://cdimage.debian.org/pub/cdimage-testing/sarge_d-i/i386/rc2/sarge-i386-netinst.iso uname -a: not possible Date: 2005-01-27 Method: booted from network installer CD (on CDRW media) Machine: home-built, ABIT NF7-S V2 motherboard (not overclocked) Processor: AMD Athlon XP 2600 (Barton core) Memory: 512Mb Corsair (DDR) Root Device: primary master IDE /dev/hda7 - intended, anyway :-) Root Size/partition table: Output from fdisk on VC2: # fdisk -l /dev/ide/host0/bus0/target0/lun0/disc Disk /dev/ide/host0/bus0/target0/lun0/disc: 41.1 GB, 41110142976 bytes 255 heads, 63 sectors/track, 4998 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Device Boot Start End Blocks Id System /dev/ide/host0/bus0/target0/lun0/part1 * 11275 102414067 HPFS/NTFS /dev/ide/host0/bus0/target0/lun0/part212764717 27647865f W95 Ext'd (LBA) /dev/ide/host0/bus0/target0/lun0/part512761536 20964517 HPFS/NTFS /dev/ide/host0/bus0/target0/lun0/part615372058 4192933+ b W95 FAT32 /dev/ide/host0/bus0/target0/lun0/part720592580 4192933+ 6 FAT16 /dev/ide/host0/bus0/target0/lun0/part825812645 5220816 FAT16 *All* partitions created using Win2K Disk Manager. Output of lspci and lspci -n: not possible Base System Installation Checklist: [O] = OK, [E] = Error (please elaborate below), [ ] = didn't try it Initial boot worked:[O] Configure network HW: [O] Config network: [ ] Detect CD: [O] Load installer modules: [O] Detect hard drives: [O] Partition hard drives: [ ] Create file systems:[O] Mount partitions: [O] Install base system:[O] Install boot loader:[E] Reboot: [ ] Comments/Problems: Installation proceeded well and without incident until it's time to install the bootloader. I requested installation of Grub on the root partition (not the MBR), and this failed with the message (on VC3) : file /boot/grub/stage1 not read correctly I may be misunderstanding the situation, but digging about with ls on VC2 revealed that there is indeed no such file - in fact there is no /boot directory at all - but there _is_ one on /target - i.e. the file /target/boot/grub/stage1 *does* exist. Is the installer simply passing the wrong --root-directory parameter to grub-install ? That would be hard to believe, because surely a gazillion other people would have noticed already ... Two other possible factors : (1) I notice that the above partition table print from fdisk shows my two Linux partitions (7 8) have type=FAT16 (type 06) despite their just having been formatted by the installer as ext3 and swap respectively ... Huh !? (2) I created all the partitions using Win2K Disk Manager, at a time when I had the BIOS hard drive access mode set to CHS. This didn't seem to faze Win2K at all (it ignores the BIOS ?), but I've since changed the setting to LBA, though this hasn't helped here. Might it have caused screwy partition table values ? Also: FWIW, I can't install LILO either - it hangs at 75% on VC1, with no useful diagnostics on VC3 (the last two messages on VC4 are about something called os-prober probing my root swap partitions, 7 and 8). This brings my testing of the Sarge installer to a shuddering halt ... cluesticks gratefully received. I'll keep this machine available for any experiments you may want me to try (I have another Woody machine to carry on using). Here's the installer log for the Grub installation failure : cut == [ many lines snipped ] Reading Package Lists... Building Dependency Tree... e2fsprogs is already the newest version. 0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded. Reading Package Lists... Building Dependency Tree... Suggested packages: grub-doc grubconf The following NEW packages will be installed: grub 0 upgraded, 1 newly installed, 0 to remove and 0 not upgraded. Need to get 0B/350kB of archives. After unpacking 750kB of additional disk space will be used. Selecting previously deselected package grub. (Reading database ... 9461 files and directories currently installed.) Unpacking grub (from .../grub_0.95+cvs20040624-10_i386.deb) ... Setting up grub (0.95+cvs20040624-10) ... Probing devices to guess BIOS drives. This may take a long time. The file /boot/grub/stage1 not read correctly. cut == And here's an ls -l of / (for the installer) and /target (etc.) : cut == # ls -l / drwxr-xr-x2 root root 1500 Jan 27 22:41 bin dr-xr-xr-x 10 root root 2048 Nov 21 22:10 cdrom drwxr-xr-x1 root root0 Jan 1 1970 dev drwxr-xr-x 10 root
Bug#292278: Installer Fails To Read Kernel deb From ASUS E616P2 DVDROM
Package: installation-reports Debian-installer-version: 2005-01-22 sarge-i386-netinst.iso RC2 from http://cdimage.debian.org/pub/cdimage-testing/sarge_d-i/i386/rc2/sarge-i386-netinst.iso uname -a: not possible Date: 2005-01-23 Method: booted from network installer CD (on CDRW media) Machine: home-built, ABIT NF7-S V2 motherboard (not overclocked) Processor: AMD Athlon XP 2600 (barton core) Memory: 512Mb Corsair (DDR) Root Device: primary master IDE /dev/hda7 - intended, anyway :-) Root Size/partition table: output from DOS utility MBRtool follows : Geometry values (from BIOS!) for this disk : (C/H/S) - 1023/254/63 Partition Table Information ACT TYPE START-C/H/SENDC/H/S LBA-start LBA-length Entry 1: 128 07 0111023 254 63 63 20482812 Entry 2: 00F 1023 011023 254 63 2048287555295730 Entry 3: 000 0000000 0 Entry 4: 000 0000000 0 Entry 1 is a 10Gb primary NTFS partition holding Win2K Entry 2 is a 26Gb extended partition containing logical drive 1 (partition /dev/hda5) : 2Gb NTFS data logical drive 2 (partition /dev/hda6) : 4Gb FAT32 data logical drive 3 (partition /dev/hda7) : 4Gb for Linux root logical drive 4 (partition /dev/hda8) : 512Mb for Linux swap (and a lot of free space) *All* partitions created using Win2K Disk Manager. Output of lspci and lspci -n: not possible Base System Installation Checklist: [O] = OK, [E] = Error (please elaborate below), [ ] = didn't try it Initial boot worked:[O] Configure network HW: [O] Config network: [ ] Detect CD: [O] Load installer modules: [O] Detect hard drives: [O] Partition hard drives: [ ] Create file systems:[O] Mount partitions: [O] Install base system:[E] Install boot loader:[ ] Reboot: [ ] Comments/Problems: The installation proceeded very nicely until the failure. Network configuration was skipped when the DHCP request timed out (not plugged into any LAN yet). I used the manual partitioning option to simply tell the installer to use the precreated (see above) hda7 and hda8 for ext3 root and swap respectively, allowing it to format them. However, installation of the base system failed repeatably with the error short read in buffer_copy error while trying to install the 2.4 kernel deb (see log extract below). Googling indicated this was probably an I/O fault, but the CD, CDROM drive and all other hardware are brand new. The CDROM drive is in fact an Asus E616P2 DVDROM drive. Further googling showed up some reports of other users having problems with this drive model in unusual circumstances, and also a report on the Linux Kernel ML about the ide-cd driver hanging with this drive model : http://marc.theaimsgroup.com/?l=linux-kernelm=109916114213228w=2 I had noticed that this step : Detecting hardware to find CDROM drives Loading module 'ide-detect' for 'Linux IDE detection' spent a *long* time (3 or 4 mins) at 81% complete ... Checking the Asus website revealed a firmware update V1.07, dated 17.Jan.2005 is available : http://www.asus.com.tw/support/download/item.aspx?ModelName=DVD-E616P2Type=AllSLanguage=en-us My drive came with firmware V1.03, which the above web-page says was supposed to Fix Play MP3 Cause Delay issue in Linux OS, so in the hope that the 1.07 release improved other relevant things I downloaded it and flashed it into the drive. After this, a repeat installation completed the install base system step with out incident - so I guess you can close this bug report as Fixed by vendor DVDROM drive firmware update. [ In fact I then had a subsequent installation failure (failure to install either of the Grub or Lilo bootloaders) which will be the subject of my next Sarge installation report bug submission. ] Here's the installer log for the base system installation failure : cut == [ many lines snipped ] Selecting previously deselected package initrd-tools. Unpacking initrd-tools (from .../initrd-tools_0.1.74_all.deb) ... Setting up dash (0.5.1-3) ... Setting up cramfsprogs (1.1-6) ... Setting up initrd-tools (0.1.74) ... Reading Package Lists... Building Dependency Tree... Suggested packages: lilo grub kernel-doc-2.4.27 kernel-source-2.4.27 kernel-pcmcia-modules-2.4.27-1-386 The following NEW packages will be installed: kernel-image-2.4.27-1-386 0 upgraded, 1 newly installed, 0 to remove and 0 not upgraded. Need to get 0B/11.0MB of archives. After unpacking 28.4MB of additional disk space will be used. Selecting previously deselected package kernel-image-2.4.27-1-386. (Reading database ... 7544 files and directories currently installed.) Unpacking kernel-image-2.4.27-1-386 (from .../kernel-image-2.4.27-1-386_2.4.27-2_i386.deb) ... dpkg-deb (subprocess): error in buffer_read(stream): failed to write to pipe in copy: Input/output error dpkg-deb:
Bug#292278: Installer Fails To Read Kernel deb From ASUS E616P2 DVDROM
I meant to add two further comments to the original report : 1) The Win2K installer CD had no problem at all with the same hardware, even when I was using the earlier firmware. 2) I guess there's an underlying problem with the kernel ide-cd driver - maybe I should submit another bug report against that. Nick Boyce Bristol, UK -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]