Processed: Re: Bug#373594: installation-report: daily snapshot on compaq armada m700
Processing commands for [EMAIL PROTECTED]: > reassign 373594 partman-base Bug#373594: parted_devices causes delays when probing non-existing floppy device Bug reassigned from package `partman-base' to `partman-base'. > End of message, stopping processing here. Please contact me if you need assistance. Debian bug tracking system administrator (administrator, Debian Bugs database) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Processed: Re: Bug#373594: installation-report: daily snapshot on compaq armada m700
Processing commands for [EMAIL PROTECTED]: > reassign 373594 partmen-base Bug#373594: installation-report: daily snapshot on compaq armada m700 Warning: Unknown package 'partmen-base' Bug reassigned from package `installation-reports' to `partmen-base'. > retitle 373594 parted_devices causes delays when probing non-existing floppy > device Bug#373594: installation-report: daily snapshot on compaq armada m700 Warning: Unknown package 'partmen-base' Changed Bug title. > End of message, stopping processing here. Please contact me if you need assistance. Debian bug tracking system administrator (administrator, Debian Bugs database) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#373594: installation-report: daily snapshot on compaq armada m700
On Thu, Jun 22, 2006 at 02:54:40PM +0200, Frans Pop wrote: > Could you restart the installer I can now confirm that the end_request IO errors concede with the waiting period before the partitioning screen comes up. There also lengthy delays while "Scanning disks" with the IO errors being logged again. > and when you get to the partitioning > screen switch to VT2 and run parted_devices manually? Does it pause then > and do you get the same errors in syslog? Yes, it does. Confirmed. Greetings Marc -- - Marc Haber | "I don't trust Computers. They | Mailadresse im Header Mannheim, Germany | lose things."Winona Ryder | Fon: *49 621 72739834 Nordisch by Nature | How to make an American Quilt | Fax: *49 621 72739835 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#373594: installation-report: daily snapshot on compaq armada m700
On Thu, Jun 22, 2006 at 02:54:40PM +0200, Frans Pop wrote: > Marc: > Could you restart the installer and when you get to the partitioning > screen switch to VT2 and run parted_devices manually? Does it pause then > and do you get the same errors in syslog? I could, but it is rather painful to get access to the machine again, which has been re-windosized in the mean time. I'd like to avoid going through the hassles again if possible. > Shortly after that the errors also happen during initialization of > partman, which was apparently canceled the first time and restarted. I remember trying to use LVM first, which was not possible due to the LVM bug in d-i that was present that time. Could that be the cause for the cancelation of partman? Greetings Marc -- - Marc Haber | "I don't trust Computers. They | Mailadresse im Header Mannheim, Germany | lose things."Winona Ryder | Fon: *49 621 72739834 Nordisch by Nature | How to make an American Quilt | Fax: *49 621 72739835 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#373594: installation-report: daily snapshot on compaq armada m700
On Thursday 22 June 2006 10:26, Michael Prokop wrote: > IMHO this looks like an userland application is trying to access > /dev/fd0 several times (as the 'end_request: I/O error...' is > present several times). This usually happens when scanning for > disks/partitions/... Might this be the case, Frans Looking at the log again, it seems that the kernel really thinks there's floppy drive, so IMHO you can't really blame the installer for trying to access it...: Jun 14 12:11:12 hw-detect: Detected module 'floppy' for 'Linux Floppy' Jun 14 12:11:12 kernel: Floppy drive(s): fd0 is 1.44M Jun 14 12:11:12 hw-detect: insmod /lib/modules/2.6.16-2-486/kernel/drivers/block/floppy.ko Jun 14 12:11:12 kernel: FDC 0 is a post-1991 82077 The place in the log where the errors first happen is during the last hardware detection phase, just before partitioning. The installer calls the disk-detect script [1], which in turn calls parted_devices. Jun 14 12:12:18 main-menu[2115]: INFO: Menu item 'disk-detect' selected Jun 14 12:12:59 kernel: end_request: I/O error, dev fd0, sector 0 Jun 14 12:13:33 kernel: end_request: I/O error, dev fd0, sector 0 Marc: Could you restart the installer and when you get to the partitioning screen switch to VT2 and run parted_devices manually? Does it pause then and do you get the same errors in syslog? Shortly after that the errors also happen during initialization of partman, which was apparently canceled the first time and restarted. The partman init scripts also call parted_devices. Jun 14 12:13:35 main-menu[2115]: INFO: Menu item 'partman-base' selected Jun 14 12:14:11 kernel: end_request: I/O error, dev fd0, sector 0 Jun 14 12:14:45 kernel: end_request: I/O error, dev fd0, sector 0 Jun 14 12:15:14 main-menu[2115]: INFO: Menu item 'partman-base' succeeded but requested to be left unconfigured. [...] Jun 14 12:15:18 main-menu[2115]: INFO: Menu item 'partman-base' selected Jun 14 12:15:54 kernel: end_request: I/O error, dev fd0, sector 0 Jun 14 12:16:28 kernel: end_request: I/O error, dev fd0, sector 0 Cheers, FJP [1]http://svn.debian.org/wsvn/d-i/trunk/packages/hw-detect/disk-detect.sh?op=file&rev=0&sc=0 pgpH6CVPm1iIl.pgp Description: PGP signature
Bug#373594: installation-report: daily snapshot on compaq armada m700
On Thu, Jun 22, 2006 at 10:26:01AM +0200, Michael Prokop wrote: > Marc, does the same behaviour (hanging for ~2 min) occur on harddisk > installation or is it just at d-i? It does only happen at installation time. The installed system boots flawlessly (if that's what you mean). Greetings Marc -- - Marc Haber | "I don't trust Computers. They | Mailadresse im Header Mannheim, Germany | lose things."Winona Ryder | Fon: *49 621 72739834 Nordisch by Nature | How to make an American Quilt | Fax: *49 621 72739835 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#373594: installation-report: daily snapshot on compaq armada m700
[Mika, the last paragraph might be of interest of you] On Wed, Jun 21, 2006 at 12:09:33AM +0200, Frans Pop wrote: > On Tuesday 20 June 2006 23:32, Marc Haber wrote: > > No, the notebook does not have a floppy drive. It has a floppy LED > > though. > > Huh? Then how the hell does the kernel manage to make the floppy light > light up? I suspect that your hardware does support a floppy device to > some extend and that that confuses the kernel, causing it to retry and > thus the delay. Yes. Looks the same way like standard desktops which all support floppy drives but do not have one connected. > Probably the laptop supports a floppy unit in the cdrom/2nd_battery bay? Most probably. > I'd say this is a kernel bug. Anyway, something we cannot solve in the > installer so if you want to follow up on this, I suggest you do so > separately (probably with upstream kernel people). Actually, this sounds like an issue I have with the grml live CD on some systems where the behavior can be fixed by calling the special option "noautoconfig" when booting. Maybe you can ask Mika Prokop about what this option actually does. Ccing him on this issue. Greetings Marc -- - Marc Haber | "I don't trust Computers. They | Mailadresse im Header Mannheim, Germany | lose things."Winona Ryder | Fon: *49 621 72739834 Nordisch by Nature | How to make an American Quilt | Fax: *49 621 72739835 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#373594: installation-report: daily snapshot on compaq armada m700
* Marc Haber <[EMAIL PROTECTED]> [20060622 07:15]: > On Wed, Jun 21, 2006 at 12:09:33AM +0200, Frans Pop wrote: > > On Tuesday 20 June 2006 23:32, Marc Haber wrote: > > > No, the notebook does not have a floppy drive. It has a floppy LED > > > though. > > Huh? Then how the hell does the kernel manage to make the floppy light > > light up? I suspect that your hardware does support a floppy device to > > some extend and that that confuses the kernel, causing it to retry and > > thus the delay. > Yes. Looks the same way like standard desktops which all support > floppy drives but do not have one connected. ACK > > I'd say this is a kernel bug. Anyway, something we cannot solve in the > > installer so if you want to follow up on this, I suggest you do so > > separately (probably with upstream kernel people). > Actually, this sounds like an issue I have with the grml live CD on > some systems where the behavior can be fixed by calling the special > option "noautoconfig" when booting. Maybe you can ask Mika Prokop > about what this option actually does. Ccing him on this issue. noautoconfig disables a check while booting in userland at the grml system. What I'm doing there is: * check for existence of /dev/fd0; if it exits: * run 'if dd if=/dev/fd0 of=/dev/null bs=512 count=1', if successful: * run blkid to identify devices with a specific label (GRMLCFG) noautoconfig disables this check at all. Booting grml on a system with a fd-connector but no fd present at all provides the following syslog entry: end_request: I/O error, dev fd0, sector 0 which I could also find in #373594. Marc, does the same behaviour (hanging for ~2 min) occur on harddisk installation or is it just at d-i? IMHO this looks like an userland application is trying to access /dev/fd0 several times (as the 'end_request: I/O error...' is present several times). This usually happens when scanning for disks/partitions/... Might this be the case, Frans? HTH && regards, -mika- -- ,'"`. http://www.michael-prokop.at/ ( grml.org -ยป Linux Live-CD for texttool-users and sysadmins `._,' http://www.grml.org/ pgpy3Vm0MV8GW.pgp Description: PGP signature
Bug#373594: installation-report: daily snapshot on compaq armada m700
On Thu, Jun 22, 2006 at 01:24:50PM +0200, Frans Pop wrote: > What is the reason you have taken the BTS off the address list? Mistake on my part, and missing reply-to in your original answer. Usually, the reply-to is there, which in turn promptet my mistake. Greetings Marc -- - Marc Haber | "I don't trust Computers. They | Mailadresse im Header Mannheim, Germany | lose things."Winona Ryder | Fon: *49 621 72739834 Nordisch by Nature | How to make an American Quilt | Fax: *49 621 72739835 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#373594: installation-report: daily snapshot on compaq armada m700
Marc, What is the reason you have taken the BTS off the address list? To be honest, I'm not very interested in following up on this if it is not tracked through the BTS. Please make sure that the follow-ups so far get added to the original installation report. Cheers, Frans On Thursday 22 June 2006 07:03, you wrote: > [Mika, the last paragraph might be of interest of you] > > On Wed, Jun 21, 2006 at 12:09:33AM +0200, Frans Pop wrote: > > On Tuesday 20 June 2006 23:32, Marc Haber wrote: > > > No, the notebook does not have a floppy drive. It has a floppy LED > > > though. > > > > Huh? Then how the hell does the kernel manage to make the floppy > > light light up? I suspect that your hardware does support a floppy > > device to some extend and that that confuses the kernel, causing it > > to retry and thus the delay. > > Yes. Looks the same way like standard desktops which all support > floppy drives but do not have one connected. > > > Probably the laptop supports a floppy unit in the cdrom/2nd_battery > > bay? > > Most probably. > > > I'd say this is a kernel bug. Anyway, something we cannot solve in > > the installer so if you want to follow up on this, I suggest you do > > so separately (probably with upstream kernel people). > > Actually, this sounds like an issue I have with the grml live CD on > some systems where the behavior can be fixed by calling the special > option "noautoconfig" when booting. Maybe you can ask Mika Prokop > about what this option actually does. Ccing him on this issue. > > Greetings > Marc pgpZIjx0QxjEV.pgp Description: PGP signature
Bug#373594: installation-report: daily snapshot on compaq armada m700
On Tuesday 20 June 2006 23:32, Marc Haber wrote: > No, the notebook does not have a floppy drive. It has a floppy LED > though. Huh? Then how the hell does the kernel manage to make the floppy light light up? I suspect that your hardware does support a floppy device to some extend and that that confuses the kernel, causing it to retry and thus the delay. Probably the laptop supports a floppy unit in the cdrom/2nd_battery bay? I'd say this is a kernel bug. Anyway, something we cannot solve in the installer so if you want to follow up on this, I suggest you do so separately (probably with upstream kernel people). -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#373594: installation-report: daily snapshot on compaq armada m700
On Tue, Jun 20, 2006 at 08:19:10PM +0200, Frans Pop wrote: > On Wed, 14 Jun 2006 15:13:26 +0200 Marc Haber wrote: > > Before starting the partitioner, system halts for like two minutes on a > > blank blue screen with the floppy light on > > That must be during driver loading. Strange that it takes so long. > That probably coincided with: > Jun 14 12:12:59 kernel: end_request: I/O error, dev fd0, sector 0 > Jun 14 12:13:33 kernel: end_request: I/O error, dev fd0, sector 0 > Jun 14 12:13:35 main-menu[2115]: DEBUG: resolver (libtextwrap1): package > doesn't exist (ignored) > Jun 14 12:13:35 main-menu[2115]: INFO: Menu item 'partman-base' selected > Jun 14 12:13:36 kernel: JFS: nTxBlock = 4543, nTxLock = 36350 > Jun 14 12:13:36 kernel: SGI XFS with ACLs, security attributes, realtime, > large block numbers, no debug enabled > Jun 14 12:13:36 kernel: SGI XFS Quota Management subsystem > Jun 14 12:14:11 kernel: end_request: I/O error, dev fd0, sector 0 > Jun 14 12:14:45 kernel: end_request: I/O error, dev fd0, sector 0 > > Not sure why this is, especially as you actually do have a floppy drive. No, the notebook does not have a floppy drive. It has a floppy LED though. Greetings Marc -- - Marc Haber | "I don't trust Computers. They | Mailadresse im Header Mannheim, Germany | lose things."Winona Ryder | Fon: *49 621 72739834 Nordisch by Nature | How to make an American Quilt | Fax: *49 621 72739835 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#373594: installation-report: daily snapshot on compaq armada m700
On Wed, 14 Jun 2006 15:13:26 +0200 Marc Haber wrote: > Kudos. That installer job is very well done. Thank you, always nice to hear :-) > Comments/Problems: > "Germany" is still not present in the list of countries offered in the > initial dialog. A lot more obscure countries are. If you select English as language the installer only shows countries that have English as an official language, same for other languages. > Before starting the partitioner, system halts for like two minutes on a > blank blue screen with the floppy light on That must be during driver loading. Strange that it takes so long. That probably coincided with: Jun 14 12:12:59 kernel: end_request: I/O error, dev fd0, sector 0 Jun 14 12:13:33 kernel: end_request: I/O error, dev fd0, sector 0 Jun 14 12:13:35 main-menu[2115]: DEBUG: resolver (libtextwrap1): package doesn't exist (ignored) Jun 14 12:13:35 main-menu[2115]: INFO: Menu item 'partman-base' selected Jun 14 12:13:36 kernel: JFS: nTxBlock = 4543, nTxLock = 36350 Jun 14 12:13:36 kernel: SGI XFS with ACLs, security attributes, realtime, large block numbers, no debug enabled Jun 14 12:13:36 kernel: SGI XFS Quota Management subsystem Jun 14 12:14:11 kernel: end_request: I/O error, dev fd0, sector 0 Jun 14 12:14:45 kernel: end_request: I/O error, dev fd0, sector 0 Not sure why this is, especially as you actually do have a floppy drive. On most systems this goes quite smoothly. > "Erase entire disk and use LVM" does not make any changes to the > partitioning table Was due to an error in new lvm2 package; already fixed. > Why is /tmp not a tmpfs on a "separate /home, /usr, /var, /tmp" layout? Not sure. D-I team: any reactions to that? > popcon seems to install even if answer was "no", but popularity-contest > is purged on the final system. It is being installed when the question is shown as the question is part of the popcon package; it is purged based on the answer. That's a choice made by Joey when he (re)implemented popcon support. IIRC Woody used to leave popcon installed after answering no, but both solutions can be argued. > x11-common shows an upgrade warning even if we are not updating but > installing anew. That irritates us too: http://bugs.debian.org/372077 > After installation, quite a few mailboxes, including Debian-exim, > identd and the user created during installation, are created in > /var/mail, but with zero size and owned by root, inhibiting mail > delivery. That is serious. Looks like a bug in add-user. I get the following myself: -- 1 root mail 0 2006-06-20 13:41 Debian-exim -- 1 root mail 0 2006-06-20 13:55 fjp -- 1 root mail 0 2006-06-20 13:41 identd -- 1 root mail 0 2006-06-20 13:41 statd Note also the lack of permissions on the files... Thanks for your report. pgpgOwYuzD9f9.pgp Description: PGP signature
