Bug#264314: hwclock time incorrectly substituted in tzsetup
Package: base-config Version: 2.40 Severity: normal Hello, While testing the d-i RC1 I've noticed that during the configuration of base system (after reboot) the current time was not correctly substituted into the template of the question about the hardware clock being set to GMT. In my case the line was displayed as: The hardware clock says the time is now Sun Aug. I was able to trace the problem to the tzsetup in base-config. The following patch fixes the substition, so the string shows the complete time: cut here-- --- tzsetup.orig2004-07-19 17:49:26.0 -0400 +++ tzsetup 2004-08-08 06:38:48.0 -0400 @@ -139,7 +139,7 @@ # The sed removes info about drift from the # hwclock output. - db_subst tzconfig/gmt hwtime $(hwclock --show --localtime | sed 's/ .*//') + db_subst tzconfig/gmt hwtime $(hwclock --show --localtime | cut -d ' ' -f 1-5) db_fset tzconfig/gmt seen false db_input high tzconfig/gmt || true fi cut here------ Best regards, Jurij Smakov[EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: new debian-installer images for sparc
On Sun, 30 May 2004, Joshua Kwan wrote: > Folks, > > Here is a mini-ISO image I cooked up with SILO rolled back to 1.3.2. Tom > Callaway suggested that I try this because a lot of new stuff related to > initrds changed in 1.4.x. > > http://june.voxel.net/~joshk/mini.iso > > Tell us what happens! (And wish me luck on finals!) > > -- > Joshua Kwan Hi, I have tested the installation with the above image on Sparcstation 10 using the serial console. The image works fine except for a couple of minor quirks, which do not seem to affect overall functionality (was able to complete the install just fine): 1. After the initial boot it tries to load modules sunhme (which seems to be wrong for my hardware - Sparcstation 10) and sr_mod (which is not included). Following messages appear in /var/log/messages: Using /lib/modules/2.4.26-sparc32/kernel/drivers/net/sunhme.o insmod: init_module: sunhme: No such device insmodsunhme modprobe: failed to load module sunhme insmod: sr_mod.o: no module by that name found insmodsr_mod modprobe: failed to load module sr_mod 2. After the installer components are downloaded, attempt to load sunhme module is done again and, again, it fails. 3. The pseudographics is screwed up after reboot as reported before and as shown at http://www.wooyd.org/d-i/second.png Best regards and good luck with exams, Jurij Smakov[EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#253742: Sparcstation2 netboot problems
Hi Alan, It seems like one explanation for your problems is that all the output goes to the serial console instead of the terminal (for reasons unknown). I have heard at least from one other person on IRC (maybe it were you? ;-) who had very similar problems. Could you please try and hook up to the machine's serial console and see if it works? Best regards, Jurij Smakov[EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#244294: Installation report for Sparcstation 10 (netboot, serial console)
Package: installation-reports INSTALL REPORT Debian-installer-version: http://people.debian.org/~jbailey/d-i/sparc/2004-04-17/sparc32/netboot/boot.img uname -a: Linux snoopy 2.4.24-sparc32 #1 Fri Jan 30 16:04:55 EST 2004 sparc GNU/Linux Date: Sat, 17 Apr 2004 Method: Over serial console, by netbooting from a PC with rarpd and tftpd. Machine: Sparcstation 10 (sun4m). Processor: SuperSPARC TMS390 Memory: 80MB Root Device: SCSI drive, Seagate ST1480N, 437 MB, /dev/sda Root Size/partition table: output of fdisk -l /dev/sda: Disk /dev/sda (Sun disk label): 9 heads, 80 sectors, 1018 cylinders Units = cylinders of 720 * 512 bytes Device FlagStart EndBlocks Id System /dev/sda1 0 8142930401 Boot /dev/sda2 814 923 39240 82 Linux swap /dev/sda3 0 10183664805 Whole disk /dev/sda4 923 1018 34200 83 Linux native Output of lspci: pcilib: Cannot open /sys/bus/pci/devices pcilib: Cannot open /proc/bus/pci lspci: Cannot find any working access method. 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: [O] Detect CD: [O] (no CD is present in the system) Load installer modules: [O] Detect hard drives: [O] Partition hard drives: [O] Create file systems:[O] Mount partitions: [O] Install base system:[E] (see comment [1] below) Install boot loader:[O] Reboot: [E] (see comment [2] below) Comments/Problems: [1] One minor problem which I encountered is triggered, when the chosen mirror is bad (in my case the mirror I have chosen did not carry sparc packages). I was then presented by a screen with an informational message and three options: "Retry", "Change mirror", and "Cancel". When choosing "Cancel", I was returned back to the very same screen (does not seem correct), but I was able to get out using the "Change mirror" option. [2] A more serious problem occured after reboot. Even though I was installing over a serial console, installer failed to recognize it, it seems. It activated the usual 6 virtual consoles, however failed to activate one for a serial port. As a result after the completion of configuration I received a bunch of messages from init about getty respawning too quickly and was left without a working console to login. Booting manually with init=/bin/sh and editing inittab to start an mgetty on a serial console fixed the problem. One more minor quirk which I've noticed is an error message from apt-setup (package base-config). The problem is the code snippet from it around line 326: if [ -n $COUNTRY ]; then db_set apt-setup/country $COUNTRY fi For multi-word country names (like US, for example) this will produce an error: /usr/sbin/apt-setup: line 326: [: United: binary operator expected The correct way is to surround references to $COUNTRY by quotes. I'll file a bug about it for base-config. Best regards, Jurij Smakov[EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#244300: base-config: Unquoted variable in apt-setup
Package: base-config Version: 2.18 Severity: minor Hi, The following code snippet around line 326 of /usr/sbin/apt-setup if [ -n $COUNTRY ]; then db_set apt-setup/country $COUNTRY fi produces an error like /usr/sbin/apt-setup: line 326: [: United: binary operator expected if the COUNTRY variable contains more than one word. It probably needs to be quoted to be treated like a single argument in such a case. It seems like this problem does not affect the db_set statement, i.e. the complete name of the country is put into the debconf database. Best regards, Jurij Smakov[EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- System Information: Debian Release: testing/unstable APT prefers unstable APT policy: (500, 'unstable') Architecture: sparc Kernel: Linux 2.4.24-sparc32 Locale: LANG=C, LC_CTYPE=C Versions of packages base-config depends on: ii adduser 3.52 Add and remove users and groups ii apt 0.5.24 Advanced front-end for dpkg ii aptitude0.2.14.1-2 curses-based apt frontend ii bsdutils1:2.12-6 Basic utilities from 4.4BSD-Lite ii console-data2002.12.04dbs-37 Keymaps, fonts, charset maps, fall ii console-tools 1:0.2.3dbs-50Linux console and font utilities ii debconf 1.4.22 Debian configuration management sy ii debianutils 2.8.1Miscellaneous utilities specific t ii gettext-base0.14.1-2 GNU Internationalization utilities ii passwd 1:4.0.3-26 Change and administer password and -- debconf information excluded -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Serial console install using Russian language is not functional?
[Please CC responses to me, thanks.] Hi, I am trying an installation on a Sparcstation 10 over serial console. Everything goes fine with English as the language, however when I choose Russian (Russia) as the language in the initial screen, the things go pretty bad. All I get is the first screen (screenshot at [0]) and then it just sits there (waited for at least 3 minutes before aborting). Nothing like this happens during an English install. Before filing bugs, I would like to ask a few questions: * Are non-English (in particular, Russian) installs over serial console supported at all? Is Russian translation actually incorporated into the latest boot images? I am using an image built yesterday by Joshua Kwan. * Since I do not see any Russian characters in the window header, what settings should I use for the terminal? The terminal (xterm) I am using is set up with ru_RU.KOI8-R locale, and is known for sure to be able to display Russian (Cyrillic) characters in KOI8-R encoding. Should I use UTF? Tried using xterm in the UTF8 mode (xterm -u8), but result is the same. * Any issues with serial console which might prevent me from seeing the non-English characters? I am using cu to connect, it works completely fine with English install. [0] http://www.wooyd.org/d-i/shot1.jpg Thanks for your excellent work and best regards, Jurij Smakov[EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Sarge install -- still no ACPI
On Fri, 14 May 2004, Mike Oliver wrote: > Update on this: I *can* boot if I leave out "acpi=off" but > substitute "noapic nolapic". I don't know what either of > those options means; I just happened to see another message > where they were being discussed and decided to try them. > > So would someone fill me in on what all these things do? I > know in general terms what ACPI does; for example when I use > the boot options above, if I pull out the AC cable, the screen > dims, whereas it didn't when I had "acpi=off". But I really > don't know what the implications of "noapic nolapic" are. There seem to be a nice page detailing Linux installation and setup on precisely the kind of machine you have [0]. It gives some recipes on ACPI/PCMCIA fixes, so it might be useful. In particular there is a suggestion of booting with 'pci=noacpi nolapic' which might not knock out the ACPI completely. [0] http://twiki.complexfission.com/twiki/bin/view/Computer/CompaqPresario2199US Best regards, Jurij Smakov[EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#250453: Partman is confused about the swap partition on sparc32
Package: partman Severity: normal Hi, I've run into a problem when testing the May 21st sarge-d_i CDs [0] on a sparc32 machine. I have a 4.2 GB hard drive [1], which I divided into a big 4.0 GB partition for root and intended to use the rest as the swap. However, when I choose the option to use the second partition as a swap area, partman gets confused and still displays prompts for mount point, mount options and bootable flag! [2]. If I leave it like that, and choose to continue the installation, I receive an error message that the attempt to mount partition #2 has failed [3]. The problem does not seem to be present if more than two partitions (like /, /home and swap) are created. [0] http://cdimage.debian.org/pub/cdimage-testing/daily/sparc/current/ [1] http://www.wooyd.org/d-i/free.png [2] http://www.wooyd.org/d-i/swap.png [3] http://www.wooyd.org/d-i/fail.png Best regards, Jurij Smakov[EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#250589: May 22nd sarge_d-i CD images on sparc32
Package: installation-reports
Debian-installer-version: May 22nd sarge_d-i images from
http://cdimage.debian.org/pub/cdimage-testing/sarge_d-i/sparc/current/
uname -a: Linux debian 2.4.26-sparc32 #1 Sat Apr 24 00:12:33 EDT 2004 sparc GNU/Linux
Date: May 23rd, 2004
Method: Businesscard and netinst CDs, booted from a SCSI CD-ROM drive,
using serial console.
Machine: Sparcstation 10
Processor: SuperSparc TMS390
Memory: 80MB
Root Device: /dev/sda: SCSI1 (0,3,0) - 4.2 GB FUJITSU MAB3045S SUN4.2G
Root Size/partition table:
Disk /dev/sda (Sun disk label): 16 heads, 135 sectors, 3880 cylinders
Units = cylinders of 2160 * 512 bytes
Device FlagStart EndBlocks Id System
/dev/sda1 090 972001 Boot
/dev/sda290 3616 3808080 83 Linux native
/dev/sda3 0 3880 41904005 Whole disk
/dev/sda4 3616 3880285120 83 Linux native
Output of lspci: pcilib: Cannot open /sys/bus/pci/devices
pcilib: Cannot open /proc/bus/pci
lspci: Cannot find any working access method.
Base System Installation Checklist:
Initial boot worked:[O]
Configure network HW: [O]
Config network: [O]
Detect CD: [O]
Load installer modules: [O]
Detect hard drives: [O]
Partition hard drives: [O/E] [Note 1]
Create file systems:[O]
Mount partitions: [O]
Install base system:[O/E] [Note 2]
Install boot loader:[O/E] [Note 3]
Reboot: [O] [Note 4]
[O] = OK, [E] = Error (please elaborate below), [ ] = didn't try it
Comments/Problems:
1. The partitioner occasionally gets confused about the swap partition.
See bug 250453 for a detailed description. The problem is not
reproducible every time, but usually may be triggered by repeatedly
deleting/recreating the partitions. In the installation above I have
chosen the partition sda4 to be swap (and partman was displaying it as
swap), but as a result it was marked as "Linux native" and no swap
entry made in /etc/fstab.
2. The installer from businesscard CD still tries to install the 64-bit
SMP kernel on a sparc32 machine, even though the corresponding
32-bit image is available. This breaks the installation. The netinst
install is fine. I'll try to investigate the problem by debugging
the base-installer's postinst.
3. I've tried guided partitioning offered by partman and 'All files in
one partition' option, recommended for beginners. As a result, two
partitions, / (4.0GB) and swap were created. When it comes to
installing SILO, it fails, claiming that the second stage boot loader
is above the "magic" 1GB boundary, even though the base system occupies
about 230MB. I am not sure whether this is a problem with SILO
miscalculating the position of /boot/second.b or this file is really
placed too far from the beginning of the disk for no good reason. It
probably should be documented somewhere. Two ways to work around are
a) to make a small /boot partition in the beginning, or b) keep the
size or / (which should be in the beginning of the disk) under 1GB.
4. After reboot the pseudographics in debconf screens is screwed up
(looked fine during the first stage), screenshot is available at
http://www.wooyd.org/d-i/second.png.
Summary:
Netinst CD - usable, completed the install, some minor setbacks.
Businesscard CD - ok up to base system install, then breaks by trying to
install the wrong kernel.
Best regards,
Jurij Smakov[EMAIL PROTECTED]
Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#401524: installation-report: on sparc64 1st partition not useable for raid and X config wrong/not working
On Mon, Dec 04, 2006 at 09:33:45AM +, Andre' Breiler wrote: > > > 2) The X server config was configured with 24bit depth but the > > > suncg6 card supports 8 bit only. Unfortunatly the driver fails to > > > initialize the card (XFree86 on sparc32 worked find with the same card) > > > in a way that the screen switches and the cursor goes blue from white > > > but nothing else happens. I've recently bumped into the default depth problem, it has been fixed in xorg svn. Do you mean that it hangs even after you manually change the DefaultDepth to 8? Could you please attach your /etc/X11/xorg.conf and /var/log/Xorg.0.log files? I only have a cg6 framebuffer in a SparcStation20, and it has been working fine when I've tested it during the weekend. > I'll have a look if I can find out where to file the bug against. The driver package is xserver-xorg-video-suncg6. Best regards, -- Jurij Smakov [EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#401524: xorg problem with cgsix
Hi Andre, Can you please try debugging the hang you're experiencing? Approximately sequence of steps: 1. Run X, wait for it to hang. 2. Log in via ssh, or from another virtual console. 3. Note the process id of the hanged X process. 4. Run gdb. 5. Enter command 'attach ', where pid is process id. 6. Enter 'backtrace' to see where it is looping/hanging (you can send the output of this command to the bug report). It would be even more valuable, if you would rebuild the xorg packages with debugging symbols and install them before debugging. If you don't know how to do that, let me know. Best regards, -- Jurij Smakov [EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#389516: Supported hardware chapter for installation manual
On Tue, Dec 26, 2006 at 08:36:03PM +0100, Frans Pop wrote: > On Tuesday 26 December 2006 19:51, Geert Stappers wrote: > > Op 26-12-2006 om 03:44 schreef Frans Pop: > > > Please check: > > > http://d-i.alioth.debian.org/manual/en.sparc/ch02s01.html#sparc-cpus > > > > The information about memory banks and memory maps is gone, > > please include it again. > > I'll leave that to the author of the original patch. > > Jurij, > > I guess this is about the following text. Should this be included again? > >Memory Configuration > > > Some older Sun workstations, notably the Sun IPX and Sun IPC have Both Sun IPX and Sun IPC are sun4c machines, not supported by Etch. That was the main reason for dropping this paragraph. If someone (not me) will check that it still applies for the sun4m machines, I'll have no objections for including it again. Best regards, -- Jurij Smakov [EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#373621: tagging wontfix
user [EMAIL PROTECTED] usertag 334104 - oh-noes tag 334104 - pending tag 334104 - patch tag 334104 wontfix tag 373621 wontfix thanks My attempt to actually do something about this ancient bug (#334104) has been blocked by Bastian Blank, who considers the proposed patch to violate the patch acceptance policy. As I don't feel like wasting my breath arguing about it, the best thing I can do is tag this bug wontfix, in hope that the debian-installer people will be able to provide some alternative solution. -- Jurij Smakov [EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#373621: Bug#334104: tagging wontfix
On Fri, Jan 05, 2007 at 08:27:25AM +0100, maximilian attems wrote: > tags 334104 -wontfix > tags 334104 pending > thanks > > morning jurij, > > On Thu, Jan 04, 2007 at 10:36:12PM -0800, Jurij Smakov wrote: > > tag 373621 wontfix > > thanks > > > > My attempt to actually do something about this ancient bug (#334104) > > has been blocked by Bastian Blank, who considers the proposed patch to > > violate the patch acceptance policy. As I don't feel like wasting my > > breath arguing about it, the best thing I can do is tag this bug > > wontfix, in hope that the debian-installer people will be able to > > provide some alternative solution. > > well i considered your patch to be a fine distro specific workaround. > now i've gone another way and disabled the dup pci-id's in tulip for i386. > afaik dmfe.ko explodes on parisc. so you could do the inverse for sparc. > > ideally dmfe and tulip shouldn't pretend to work for chipsets they > can't, but that codebase won't get fixed in time for the etch release. Hi, As I said, I don't think the solution of dropping the duplicate PCI IDs in different drivers for different architectures to be optimal. There is always a possibility that someone will insert an arbitrary PCI card into any box with a PCI bus. If currently the user will be slightly inconvenienced, if your proposed solution is implemented, he/she will have to rebuild the kernel to get it to work. Anyway, Frans Pop has recently posted a proposed patch for the installer, which will make it possible to blacklist and arbitrary module from loading, see http://lists.debian.org/debian-boot/2007/01/msg00207.html If this will make it into etch, that will solve the problem. Best regards, -- Jurij Smakov [EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [RFC] Option to blacklist modules for d-i and installed system
On Fri, Jan 05, 2007 at 04:18:06PM +0100, Frans Pop wrote: > We sometimes see that a module that is automatically loaded causes > problems, either because of a bug or because of a driver conflict. > > Therefore, it would be nice to be able to have a generic facility to > blacklist modules for the installer and to have those carried over into > the installed system. > > I propose to implement something like the patch below to rootskel. > Note that debconf cannot be used as the blacklisting needs to happen > _before_ udev is run for the first time. > Patch has been tested and seems to work nicely. Looks good to me. I'd be very happy if something like that would make it into etch. > Current format is "blacklist=module1,module2". > Alternative formats could be "module.blacklist" or "module.off" (possibly > a bit harder to implement). > A last option could be "module.blacklist=yes", which is probably most > consistent with "module.param=value". I think the first way (blacklist=foo,bar) is the most straightforward, but it's your call, of course. Best regards, -- Jurij Smakov [EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Extending accessibility support in D-I for Lenny
On Fri, Feb 16, 2007 at 12:14:04AM +0100, Frans Pop wrote: > On Thursday 15 February 2007 23:53, Gilles Casse wrote: > > Today, in principle using a Speakup enabled kernel + Speechd-Up + > > SpeechDispatcher + eSpeak or in user space, Yasr + emacspeak server + > > eSpeak, a text based dialog might be correctly spoken. > > However, we understand that the speakup patches have never been properly > ported to current 2.6 kernels and that there were security issues with > the speakup patches. That is why speakup support was dropped from the > installer when we dropped support for 2.4 kernels. I've recently recognized that we no longer have a speakup-enabled kernel, and I'm willing to work on the kernel team side to bring it back for Lenny. It looks like Ubuntu has it working with their kernels, so we should be able to do it too. Best regards, -- Jurij Smakov [EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Extending accessibility support in D-I for Lenny
Hi Mario, On Fri, Feb 16, 2007 at 02:23:31PM +0100, Mario Lang wrote: > > I've recently recognized that we no longer have a speakup-enabled > > kernel, and I'm willing to work on the kernel team side to bring it > > back for Lenny. > > Great, I'd be very happy to see a volunteer working > on this, since speakup has a quite large user base in > the blind linux users community... > > Back when I was still maintaining a speakup enabled kernel > in Debian, I bought a hardware speech synthesizer especially > for testing speakup support for it. If you are really going to work on > speakup > in Debian, I can provide that hardware to you for testing purposes since these > days, you can't even buy those things anymore. However, many > people still have them, and would like to continue using them instead > of software speech synthesis. Besides, using hardware speech > synthesis in speakup does enable a very cool feature > for blind users, i.e. being able to review the screen even > after a kernel crash or panic. Thanks for your kind offer. I'm scheduled to move in about a week, so I'll be forced to take a leave from Debian work for a while (in the worst case scenario it will probably take a couple of months). We can discuss the speakup issues when I get back. Thanks, -- Jurij Smakov [EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#262568: [sparc64] [20041019] [netboot] installation failed
On Mon, 18 Oct 2004, Erinn Clark wrote: As requested by the recent ping, here is my follow-up installation report. [skipped] This time I get: Running "/sbin/silo" failed with error code "1". Hi Erinn, This is most probably due to the fact that SILO is not happy when the root partition is in the beginning of the disk and is larger than 1 GB. Possible workarounds are to create a small /boot partition in the beginning or to keep the / partition under 1 GB. Could you please try this and see if it helps? Best regards, Jurij Smakov[EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#262568: [sparc64] [20041019] [netboot] installation failed
Hello again, On Mon, 18 Oct 2004, Erinn Clark wrote: [skipped] Thanks for the tip. As it happens, the only thing I did differently this time around was use XFS instead of ext3 (with the same partitioning scheme that I had before - 700mb /, 1gb+ /home, and 128 swap). It appears to have eaten up the entire partitioning scheme as a result, leaving a 2.2GB XFS partition in its wake. I would have mentioned this, but I didn't think to check on the partition table after being unable to install SILO... Oops, SILO does not work with XFS either (at least it didn't last time I checked). That was reported and partman should not really let the bootable partition to be XFS (or, at least, that should be documented somewhere). Trying to repartition with ext3 lets me use the first 10% of the disk for / and then tells me the rest of the disk is unusable. Trying again with the automatic partitioning tool in partman gives me a usable /boot (at 100mb) and then says the remaining 2.1gb are unusable. That sounds like broken partition table. You can recreate the partition table (aka 'Sun disklabel') by choosing the disk line (the one with disk model and size specification) in partman's main menu and pressing Enter. It will then let you recreate a blank partition table, after that all the space should be accessible for partitioning. Best regards, Jurij Smakov[EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#248963: Sarge sparc netinstall rc2 still fails on Sparc IPX
Tjena Mats, On Tue, 23 Nov 2004, Magnus Hyllander wrote: I just downloaded the new Sarge rc2 sarge-sparc-netinst.iso and tried to install. The install still fails. Now I do not get any error message, instead the kernel boot hangs efter the messages: Mounted devfs on /dev Freeing unused kernel memory: 144k freed After these messages init normaly should be started off the initrd. Since I do not have access to any sun4c hardware, could you try a couple of things: 1. Append a boot argument 'init=/bin/sh' (without the quotes). This should normally just give you a shell prompt in the end of the boot. 2. If you have a working Linux installation on this machine, could you try running the RC2 busybox binary on it? This binary provides the shell and a bunch of other utilities for initrd. You can either extract it from the initrd image on RC2 CD by mounting it, or download it from http://www.wooyd.org/bug_sun4c/. Try something like 'busybox ls' or 'busybox sh' to see if it works. Best regards, Jurij Smakov[EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#282979: Sun Sparcstation 20, Installation of Sarge netinst failed. Don't detect CDROM
On Thu, 25 Nov 2004, Yuri Pimenov wrote: It boots normaly, but after "language/region" dialog the installation stops with message that "CDRom can't be mounted". If I press "Yes", the dialog appears again and so on. My hardware is ok, I can provide the full information about the drive (serial number and so on if needed). Also can run some tests or what you'll say to me to do to find the right description of the error. Hi, This is a known problem, documented in the RC2 errata [0]. It should be fixed by the next upload of linux-kernel-di-sparc. [0] http://www.debian.org/devel/debian-installer/errata Best regards, Jurij Smakov[EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#382590: debian-installer: sparc testing images missing 32-bit isofs
reassign 382590 linux-kernel-di-sparc-2.6 thanks On Sat, 19 Aug 2006, Frans Pop wrote: I would first like to see it confirmed as it seems like a very strange regression to me. Jurij will probably look into this this weekend. He's just back from his holidays. Indeed, the isofs kernel module is not included in any of the kernel udebs. In kernel-wedge it is included in ide-modules, and that's how it is normally pulled in. On sparc32 we are not building ide-modules udeb, so isofs is not available. I plan to include it explicitly in cdrom-core-modules udeb in the next l-k-di-sparc-2.6 upload. Best regards, Jurij Smakov[EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#384342: installation-report: doesn't boot kernel: Illegal Instruction (sparc)
Hi Joost, This is a known problem. So far we could not figure out what triggers it and how it can be fixed. I would appreciate if you could do the following: boot the machine until the failure, then issue the 'ctrace' command at the 'ok' prompt. Then copy the output of that command (serial console might be of great help here) and post it to the bug. Thanks, Jurij Smakov[EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#382990: etch beta 3 on SparcStation 20
Hi Hendrik, Thanks for you report. I don't have much to say about the video card problem, but I've just tried the beta3 on my SS20 and the disks worked fine. In my case the disks are spun-up already on power-up, this setting can be controlled by a jumper on the drive, usually. I will try to change it to the program-controlled spin up and retry the installation. The fact that you cannot mount the CD is expected, due to an oversight the isofs module was not included in the installer. This information is going to be added to beta3 errata (if it wasn't already). The problem is fixed in latest dailies, can you perhaps try them out? Best regards, Jurij Smakov[EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#384549: cassini driver failure
Hi, I'm a bit confused here. The netboot beta3 image you linked to contains 2.6.16 kernel yet in your report you mention 2.6.17. Are you sure you are using beta3 and not a daily image? Images featuring 2.6.17 have appeared only a couple of days ago... Best regards, Jurij Smakov[EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Installation to software RAID
Hi! Does the Debian currently support the installation into a software RAID? Is it possible during the installation to create a software RAID, install the system into it and make it bootable? I don't think that it is in the standard installation manual, but maybe it is possible to arrange with minimum amount of boot floppy hacking? Best regards, J. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Installation to software RAID
On 29 Nov 2000, Adam Di Carlo wrote: [skipped] > It's absolutely not yet supported -- partly because it's not so easy > (esp if root is a RAID device), and partly because potato included > rather old raid-tools. [skipped] Hmmm, from RAID-Howto it seems, that there is no problem with RAID being a root device. Besides, if there is such thing as Live! CD-ROM (which I've never heard about) it would be quite simple to run Debian off it, create RAID, partition it and then start the installer. Booting is a more complicated issue though, but also manageable (I think). Best regards, Jurij Smakov -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#505789: Patch adding the usb-storage module for sparc
Hi,
With valuable help from Frans, I've created a patch which adds
usb-storage module on sparc, so that it is actually possible to load
needed firmware from a USB stick (attached). Please consider applying
it.
Thanks.
--
Jurij Smakov [EMAIL PROTECTED]
Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC
Index: installer/build/pkg-lists/cdrom/sparc.cfg
===
--- installer/build/pkg-lists/cdrom/sparc.cfg (revision 56652)
+++ installer/build/pkg-lists/cdrom/sparc.cfg (working copy)
@@ -4,10 +4,9 @@
sparc-utils-udeb
kbd-chooser
-# Some modules are only available for sparc64, so only include them if they
-# are available.
-ide-modules-${kernel:Version} ?
-usb-modules-${kernel:Version} ?
+ide-modules-${kernel:Version}
+usb-modules-${kernel:Version}
+usb-storage-modules-${kernel:Version}
scsi-common-modules-${kernel:Version}
Index: installer/build/pkg-lists/netboot/sparc.cfg
===
--- installer/build/pkg-lists/netboot/sparc.cfg (revision 56652)
+++ installer/build/pkg-lists/netboot/sparc.cfg (working copy)
@@ -9,5 +9,6 @@
nic-modules-${kernel:Version}
usb-modules-${kernel:Version}
+usb-storage-modules-${kernel:Version}
console-keymaps-at
Index: packages/kernel/linux-kernel-di-sparc-2.6/modules/sparc/usb-storage-modules
===
--- packages/kernel/linux-kernel-di-sparc-2.6/modules/sparc/usb-storage-modules (revision 0)
+++ packages/kernel/linux-kernel-di-sparc-2.6/modules/sparc/usb-storage-modules (revision 0)
@@ -0,0 +1 @@
+#include
Bug#505789: Patch adding the usb-storage module for sparc
On Mon, Nov 17, 2008 at 09:47:29AM -0200, Otavio Salvador wrote:
> Jurij Smakov <[EMAIL PROTECTED]> writes:
>
> > Hi,
> >
> > With valuable help from Frans, I've created a patch which adds
> > usb-storage module on sparc, so that it is actually possible to load
> > needed firmware from a USB stick (attached). Please consider applying
> > it.
> >
> > Thanks.
> > --
> > Jurij Smakov [EMAIL PROTECTED]
> > Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC
> >
> > Index: installer/build/pkg-lists/cdrom/sparc.cfg
> > ===
> > --- installer/build/pkg-lists/cdrom/sparc.cfg (revision 56652)
> > +++ installer/build/pkg-lists/cdrom/sparc.cfg (working copy)
> > @@ -4,10 +4,9 @@
> > sparc-utils-udeb
> > kbd-chooser
> >
> > -# Some modules are only available for sparc64, so only include them if they
> > -# are available.
> > -ide-modules-${kernel:Version} ?
> > -usb-modules-${kernel:Version} ?
> > +ide-modules-${kernel:Version}
> > +usb-modules-${kernel:Version}
>
> Why you've droped the '?' ?
>
> Otherwise it looks good.
Frans suggested doing it, these modules are guaranteed to exist on
sparc there days, and question marks are leftovers from the old times.
Cheers.
--
Jurij Smakov [EMAIL PROTECTED]
Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#505789: Fix confirmation
Hi, I've just confirmed using the daily build netinst that getting firmware file from a USB stick works. Thanks. -- Jurij Smakov [EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#504721: Possible reason for serial console misdetection
Hi, I looked into it, and here's what I get on my SunBlade 1000 when I boot using serial console: ~ # dmesg | grep -i console [0.00] console [earlyprom0] enabled [ 53.918290] Console: colour dummy device 80x25 [ 53.971361] console handover: boot [earlyprom0] -> real [tty0] [ 59.676537] Console: switching to mono PROM 80x34 [ 64.238061] Console: ttyS0 (SAB82532) [ 64.308951] console [ttyS0] enabled ~ # The logic in reopen-console is the following: 1. Look for the 'console handover' line in dmesg output and extract whatever is listed there as a "real" console. If it's successful, then we are done. In this case this line matches, but it contains an incorrect console setting 'tty0', referring to the terminal. As a result of this setting the information is still displayed on the serial console, but you cannot enter information through it, only keyboard input is accepted. 2. If we still don't have a value for the console, look for 'console [...] enabled' line(s), and extract console values from it. In this case two lines would match, producing console values 'earlyprom0' and 'ttyS0'. Function also checks whether corresponding devices in /dev exist, and if this reduces the number of values to a single console value, then this value is chosen. In this case /dev/earlyprom0 does not exist, so we would end up with a single 'ttyS0' value, which is correct. So, the problem is that reopen-console gives preference to the value in 'console handover' line, which is incorrect on sparc (refers to a real terminal console even if one is connecting through serial). If it is obvious, that sparc is unique that way, and it works correctly on other arches, the following (untested) patch might do the trick: --- a/src/sbin/reopen-console 2008-09-21 09:29:17.0 +0100 +++ b/src/sbin/reopen-console 2008-11-23 22:09:56.0 + @@ -12,7 +12,7 @@ sed -n -e 's/.*\] console handover: boot \[.*\] -> real \[\(.*\)\]$/\1/p')" consoles= - if [ -z "$console" ]; then + if [ -z "$console" ] || [ "$(/bin/archdetect)" = "sparc/sparc64" ]; then # Retrieve all enabled consoles from boot log; ignore those # for which no device file exists for cons in $(dmesg -s 65535 | I'll try to ask around and see whether current console detection in reopen-console is correct for other arches. By the way, appending console=ttyS0 at the boot prompt fixes the problem, so can be used as a temporary workaround. Cheers. -- Jurij Smakov [EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Please help solve #504721
Bcc: debian-devel Hi, We have recently recognized that serial console detection in the installer is not working as expected, which resulted in RC bug http://bugs.debian.org/504721. If you have access to a machine through serial console, please provide debugging information which can help with resolving this bug. The following information would be very useful: 1. Basic information about the serial console setup (whether machine is completely headless, etc.) and contents of /proc/cmdline. 2. Output of 'grep -i console /var/log/dmesg' (or 'dmesg | grep -i console' if you booted recently and boot messages have not been purged from dmesg buffer yet). 3. If you have a chance of booting the installer on the machine, output of 'dmesg | grep -i console' from the installer shell both without and with the 'console=' (for example, 'console=ttyS0') specified on the command line. Does adding the option result in any changed in installer appearance? Thanks. -- Jurij Smakov [EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Proposed updated errata file for D-I upcoming RC2
Installer images are missing the sun_esp kernel module > - which is needed to support CD and disc drives connected to an ESP or > - FAS controller. > + The Sparc port has several issues in this release: > + > + Broken serial console support. In some situations > + (see http://bugs.debian.org/504721";>bug #504721 > +for details), the serial console used may not > +be properly re-opened and therefore become non > +responding. Passing 'console=ttyS0' (or whichever > +device the console is attached to) may we used > +as a workaround. > + fails to recognize ESP/FAS SCSI controller: > + installer images are missing the sun_esp > + kernel module which is needed to support CD and disc > + drives connected to an ESP or FAS controller. > + > > > s390: unsupported features -- Jurij Smakov [EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: XVR-500 support in Lenny
On Wed, Oct 22, 2008 at 12:46:16AM +0200, Petr Vyslouzil wrote: > Dave, Julien, thanks for your replies. > > I asked just to make sure I did not miss something obvious (like the > .config file in /boot dir :)). I've been wondering why Debian on Sparc > doesn't have compiled-in drivers for SUN's graphics cards. I can confirm > that sunxvr500 driver works flawlessly even with XVR-600 card (although > lspci shows wrong description), so at least all Sun Blade [12]500 users > would benefit. And I suppose that sunxvr2500 is not any worse. > So, please, could you consider to compile them in? The difference between > OpenBoot ansi terminal and XVR-600 framebuffer is abysmal. [adding debian-boot since it relates to the problems with installer on Sunblade 1000 which I recently reported] Petr, can you please try the images I've built from http://www.wooyd.org/debian/kernel/ They are the same as Debian's stock kernels, except that the CONFIG_FB_XVR500 and CONFIG_FB_XVR2500 are enabled (note that it does not make sense to build framebuffer drivers as modules, because they are used early during boot). I was hoping that enabling them will fix problems with the installer (black-and-white terminal, with invisible menu highlight) on my SunBlade 1000, however there is a PCI id mismatch which makes it not work for me. The XVR-500 framebuffer driver supports PCI ID 3d3d:07a0 (vendor is 3D Labs) while my card is reported as 1901:07a0 (vendor is Intergraph). Since 3D Labs bought Intergraph at some point, there is a chance that this is the same chip, but I will have to rebuild the kernel again to test it. Cheers. > regards, > Petr > >> Petr: >> >> Well, a google search turned up this page >> (http://merkel.debian.org/~jurij/) which has the Debian kernel >> configurations. >> >> The entry for 2.6.26-8 for sparc, "config-2.6.26-1-sparc64.gz >> <http://merkel.debian.org/%7Ejurij/2.6.26-8/sparc/config-2.6.26-1-sparc64.gz>" >> >> shows: >> >> # CONFIG_FB_XVR500 is not set >> # CONFIG_FB_XVR2500 is not set >> >> Does that help? >> >> [I was expecting to find information that tells me where the "config" >> file details are contained on a running system (under /proc, I >> thought), but I haven't found that, yet Perhaps that functionality >> isn't available / compiled into the sparc kernels...] >> >> Cheers, >> Dave > > > -- > To UNSUBSCRIBE, email to [EMAIL PROTECTED] > with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] -- Jurij Smakov [EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: XVR-500 support in Lenny
On Tue, Dec 02, 2008 at 04:22:49AM +0100, Petr Vyslouzil wrote: > Hello Jurij et al. > Thank you for caring. > >> [adding debian-boot since it relates to the problems with installer on >> Sunblade 1000 which I recently reported] >> >> Petr, can you please try the images I've built from >> >> http://www.wooyd.org/debian/kernel/ >> >> They are the same as Debian's stock kernels, except that the >> CONFIG_FB_XVR500 and CONFIG_FB_XVR2500 are enabled (note that it does >> not make sense to build framebuffer drivers as modules, because they >> are used early during boot). > > I've installed linux-image-2.6.26-1-sparc64-smp_2.6.26-10_sparc.deb and > it works OK on my SunBlade 2500 (Silver) with XVR-600 graphics card. The > framebuffer resolution is 1280x768, font size is 8x12 => the screen makes > 160x64 chars. > > Here is hwinfo --gfxcard output: > > 16: PCI 30002.0: 0380 Display controller > [Created at pci.310] > UDI: /org/freedesktop/Hal/devices/pci_3d3d_7a2 > Unique ID: IA2G.qqQQz3Ovxr7 > SysFS ID: /devices/pci0003:00/0003:00:02.0 > SysFS BusID: 0003:00:02.0 > Hardware Class: graphics card > Model: "3DLabs Sun XVR-500 Graphics Accelerator" > Vendor: pci 0x3d3d "3DLabs" > Device: pci 0x07a2 "Sun XVR-500 Graphics Accelerator" > SubVendor: pci 0x3d3d "3DLabs" > SubDevice: pci 0x1047 > Revision: 0x01 > Driver: "e3d" > Memory Range: 0x0400-0x07ff (rw,prefetchable) > Memory Range: 0x0010-0x0011 (rw,non-prefetchable) > Memory Range: 0x0200-0x03ff (rw,prefetchable) > Memory Range: 0x0012-0x0012 (ro,non-prefetchable,disabled) > IRQ: 39 (no events) > Module Alias: "pci:v3D3Dd07A2sv3D3Dsd1047bc03sc80i00" > Config Status: cfg=new, avail=yes, need=no, active=unknown > > As I stated in some previous E-mail, the card is incorrectly detected as > XVR-500. Probably because of missing subsystem-id in pci.ids. > > XVR-500 card in Solaris looks like this: > subsystem-id: 1024 > subsystem-vendor-id: 3d3d > revision-id: 0001 > device-id: 07a2 > vendor-id: 3d3d > > meanwhile XVR-600 looks like this: > subsystem-id: 1047 > subsystem-vendor-id: 3d3d > revision-id: 0001 > device-id: 07a2 > vendor-id: 3d3d Note that PCI IDs for both cards are the same, 3d3d:07a2. As kernel relies only on PCI IDs to determine which driver should handle it, there is no way to distinguish one from another (and, probably, no point, as they should be completely compatible). I guess the difference between the two versions is only the amount of available memory or something like this. > > I'd like to test your linux image on SunBlade 1500 (with XVR-500), but as > it's running Solaris and colleague of mine is working on it all the time, > I thought I'd just make a custom d-i with sunxvr drivers turned on and > try to run d-i with newt or gtk frontend. > I'm gonna try to do it myself, but as I'm a Debian rookie, let's see how > much time it will take. :) > > >> I was hoping that enabling them will fix problems with the installer >> (black-and-white terminal, with invisible menu highlight) on my >> SunBlade 1000, > > Oh yes, this is exactly the same behaviour I had. > >> however there is a PCI id mismatch which makes it not work for me. The >> XVR-500 framebuffer driver supports PCI ID 3d3d:07a0 (vendor is 3D >> Labs) while my card is reported as 1901:07a0 (vendor is Intergraph). > > 1091:07a0 I suppose. Yes, 1091:07a0 is correct. >> Since 3D Labs bought Intergraph at some point, there is a chance that >> this is the same chip, but I will have to rebuild the kernel again to >> test it. > > Are you going to add PCI_DEVICE(0x1091, 0x7a0) to e3d_pci_table struct? Yes, I did, built a new kernel and it worked nicely. I've sent the patch for the XVR-500 driver to get this change included upstream [0] and will push it to our kernel team as soon as it's acked. [0] http://marc.info/?l=linux-sparc&m=122816564110952&w=2 Best regards, -- Jurij Smakov [EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Please enable XVR-500 and XVR-2500 fb drivers for sparc
Package: linux-2.6
Version: 2.6.26-11
Severity: wishlist
We have recently found out that Debian installation UI looks much
better on various SunBlade systems if the framebuffer drivers for
XVR-500 and XVR-2500 cards (CONFIG_FB_XVR500, CONFIG_FB_XVR2500) are
enabled [0,1]. If there is going to be another kernel rebuild before
lenny, it would be great to enable these drivers. As cards are
sparc-specific, it is sufficient to enable them for sparc. Also, it
would be nice to apply the attached simple patch, which adds the PCI
ID of a card I have in my SunBlade 1000 [1], it appears that it's the
same chip (same device ID), but manufactured by a different company.
I've sent the patch to David Miller to be included upstream, however
haven't heard back yet [2].
[0] http://lists.debian.org/debian-boot/2008/12/msg00040.html
[1] http://lists.debian.org/debian-boot/2008/12/msg00042.html
[2] http://marc.info/?l=linux-sparc&m=122816564110952&w=2
--
Jurij Smakov [EMAIL PROTECTED]
Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC
diff -aur a/drivers/video/sunxvr500.c b/drivers/video/sunxvr500.c
--- a/drivers/video/sunxvr500.c 2008-11-20 18:02:37.0 -0500
+++ b/drivers/video/sunxvr500.c 2008-12-01 15:43:04.0 -0500
@@ -396,6 +396,7 @@
static struct pci_device_id e3d_pci_table[] = {
{ PCI_DEVICE(PCI_VENDOR_ID_3DLABS, 0x7a0), },
+ { PCI_DEVICE(PCI_VENDOR_ID_IGRAPH, 0x7a0), },
{ PCI_DEVICE(PCI_VENDOR_ID_3DLABS, 0x7a2), },
{ .vendor = PCI_VENDOR_ID_3DLABS,
.device = PCI_ANY_ID,
diff -aur a/include/linux/pci_ids.h b/include/linux/pci_ids.h
--- a/include/linux/pci_ids.h 2008-11-20 18:02:37.0 -0500
+++ b/include/linux/pci_ids.h 2008-12-01 15:43:04.0 -0500
@@ -930,6 +930,8 @@
#define PCI_DEVICE_ID_SUN_TOMATILLO 0xa801
#define PCI_DEVICE_ID_SUN_CASSINI 0xabba
+#define PCI_VENDOR_ID_IGRAPH 0x1091
+
#define PCI_VENDOR_ID_CMD 0x1095
#define PCI_DEVICE_ID_CMD_643 0x0643
#define PCI_DEVICE_ID_CMD_646 0x0646
Re: XVR-500 support in Lenny
On Tue, Dec 16, 2008 at 12:21:16AM +0100, Petr Vyslouzil wrote: > Otavio Salvador wrote: >> Petr Vyslouzil writes: >> >>> I've installed linux-image-2.6.26-1-sparc64-smp_2.6.26-10_sparc.deb >>> and it works OK on my SunBlade 2500 (Silver) with XVR-600 graphics >>> card. The framebuffer resolution is 1280x768, font size is 8x12 => the >>> screen makes 160x64 chars. >> >> Could you try sparc daily image? it has 2.6.26-11 kernels and is >> suppose to have this enabled. >> > > Well, it doesn't seem so. According to Debian Official Kernel > Configurations[1] the XVR drivers are not compiled in yet. Today I also > tried the latest d-i snapshot[2] and it indeed came without the driver. > Are there any issues that prevent the drivers from being in kernel? Petr, The kernel 2.6.26-12, uploaded today, should have the XVR drivers enabled, as I understand it's going to be used to build the d-i RC2, so it should be all good. Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Reopening, bug not fixed in 2.6.26-12
found 2.6.26-12 thanks Hi, Contrary to what 2.6.26-12 changelog says, the Sparc XVR framebuffer drivers (CONFIG_FB_XVR500, CONFIG_FB_XVR2500) were NOT enabled in it: ju...@droopy:~/tmp$ dpkg -x linux-image-2.6.26-1-sparc64-smp_2.6.26-12_sparc.deb tmp ju...@droopy:~/tmp$ grep -i xvr tmp/boot/config-2.6.26-1-sparc64-smp # CONFIG_FB_XVR500 is not set # CONFIG_FB_XVR2500 is not set Please enable them for the next upload (which will hopefully make it into lenny). Thanks. -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: Call for testing of RC2 images
On Thu, Jan 29, 2009 at 11:19:17AM -0200, Otavio Salvador wrote: > -BEGIN PGP SIGNED MESSAGE- > Hash: SHA1 > > Hello, > > The Debian-CD Team has built the images based on RC2 installer and > I'd like to ask for some tests before we finally announce the release. > > The images are available at http://cdimage.debian.org/cdimage/.lenny_bibble/ > > Please send feedback to the mailing list (even successful ones). Sparc (SunBlade 1000) 1. Serial console/Guided partitioning/All files in one partition - ok. 2. Terminal/Guided partitioning with LVM - fails to boot. It looks like the LVM volumes are set up correctly, however the initramfs script scripts/local-top/lvm2 which is supposed to start the volumes up is not getting invoked. If I run it by hand, the device nodes in /dev/mapper get created, and I can exit initramfs shell to successfully complete the boot. This machine needs firmware loaded from the USB stick, and this causes (known?) issues: if I insert the stick to load the firmware, and then remove it immediately after the loading step, the real disk is still detected as /dev/sdb, and that's what gets written to the configuration. So if I later boot without the stick, the disk is detected as /dev/sda, causing problems. Amusingly, I can boot when using LVM, because /boot is a separate (non-LVM) partition and is only used to only read kernel and initrd, however filesystem check for /dev/sdb1 (which is /boot) fails, as device does not exist. Cheers. -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#513976: Problem with LVM (RC2 on Sparc)
Package: installation-reports > Hello, > > The Debian-CD Team has built the images based on RC2 installer and > I'd like to ask for some tests before we finally announce the release. > > The images are available at http://cdimage.debian.org/cdimage/.lenny_bibble/ > > Please send feedback to the mailing list (even successful ones). I had some problems installing RC2 on my Sparc (SunBlade 1000), when I chose the Guided partitioning with LVM option. The machine failed to boot after reboot at the end of the installation. It looks like the LVM volumes are set up correctly, however the initramfs script scripts/local-top/lvm2 which is supposed to start the volumes up is not getting invoked. If I run it by hand, the device nodes in /dev/mapper get created, and I can exit initramfs shell to successfully complete the boot. This machine needs firmware loaded from the USB stick, and this causes (known?) issues: if I insert the stick to load the firmware, and then remove it immediately after the loading step, the real disk is still detected as /dev/sdb, and that's what gets written to the configuration. So if I later boot without the stick, the disk is detected as /dev/sda, causing problems. Amusingly, I can boot when using LVM, because /boot is a separate (non-LVM) partition and is only used to only read kernel and initrd, however filesystem check for /dev/sdb1 (which is /boot) fails, as device does not exist. Cheers. -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: Breaking X.Org on sparc (was: Bug#514418: [FIX]: ultra45 boot failing...)
On Sun, Feb 08, 2009 at 03:49:22PM +0100, W. Martin Borgert wrote: > On 2009-02-08 00:07, Luk Claes wrote: > > I think it's best to delay that to r1. Can someone please provide a text > > for the release notes to describe the problem, TIA? > > Hi all: Please send me a paragraph of text ASAP. > (Or file a bug with the proposed text against "release-notes".) Something like this: Lenny kernel fails to boot on Sparc workstations with PCI-Express Due to an unfortunate interaction of a kernel fix with PCI-Express subsystem, Lenny default kernel will fail to boot on Sparc workstations with PCI-Express slots, like Ultra 25 and Ultra 45. As this problem has been discovered very late in the release cycle, we were unable to include a fix in the original Lenny release, but we'll do our best to eliminate the problem for the first point release. Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#464445: Installation report - SunBlade 1000 (sparc)
7fe0012a000 (32-bit, non-prefetchable) [size=8K] Capabilities: [40] Power Management version 1 Kernel driver in use: sym53c8xx Kernel modules: sym53c8xx 0001:00:00.0 Host bridge [0600]: Sun Microsystems Computer Corp. Schizo PCI Bus Module [108e:8001] Flags: bus master, 66MHz, medium devsel, latency 64, IRQ 2147483647 0001:00:04.0 SCSI storage controller [0100]: QLogic Corp. QLA2200 64-bit Fibre Channel Adapter [1077:2200] (rev 05) Flags: bus master, 66MHz, medium devsel, latency 64, IRQ 21 I/O ports at 7ffed000300 [size=256] Memory at 7fd0010 (32-bit, non-prefetchable) [size=8K] Expansion ROM at 7fd0012 [disabled] [size=128K] Capabilities: [44] Power Management version 1 Kernel driver in use: qla2xxx Kernel modules: qla2xxx Base System Installation Checklist: [O] = OK, [E] = Error (please elaborate below), [ ] = didn't try it Initial boot: [O] Detect network card:[O] Configure network: [O] Detect CD: [O] Load installer modules: [O] Detect hard drives: [E] Partition hard drives: [O] Install base system:[O] Clock/timezone setup: [O] User/password setup:[O] Install tasks: [O] Install boot loader:[O] Overall install:[O-] Comments/Problems: Installed from serial console using netinst CD. Everything went smoothly except for hard drive detection: the drive is hooked up to a Qlogic controller, which requires non-free firmware to be initialized. As far as I can tell, d-i still does not have support for it, so before running the hard drive detection I had to drop into a shell and manually run the following commands: wget http://ftp.ie.debian.org/debian/pool/non-free/f/firmware-nonfree/firmware-qlogic-di_0.9_all.udeb udpkg -i firmware-qlogic-di_0.9_all.udeb modprobe -r qla2xxx modprobe qla2xxx After that I was able to proceed with the installation. Same thing needed to be done before reboot: chroot into the installed system, mount /proc, install firmware-qlogic package (had to add non-free to sources.list), regenerate initrd using update-initramfs -k all -u. Cheers. -- Jurij Smakov [EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Latest sarge sparc32 netboot image problems
Hi, JFYI, the current sarge sparc32 netboot 2.4 image downloaded from http://ftp.debian.org/debian/dists/sarge/main/installer-sparc/current/images/sparc32/netboot/boot.img fails to load the network modules due to missing crc32_* symbols. The module providing these symbols appears to be missing from initrd. The 2.6 netboot image downloaded from the same location appears to have screwed-up display (no pseudographics) and hangs during the device detection while loading sd_mod module on my SS20. Best regards, Jurij Smakov[EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Preparing linux-2.6 2.6.18-1
On Wed, Sep 20, 2006 at 06:16:32PM -0700, Steve Langasek wrote: > On Wed, Sep 20, 2006 at 03:38:50PM +0200, Frederik Schueler wrote: > > Second: this release contains ALL binary firmware blobs shipped > > upstream, even those we kept pruning since the day Herbert Xu removed > > them the first time in 2004. > > What in the world? Why would you do that anyway? It does not make sense to me either. I would expect the status quo to be maintained until the issue is resolved through the GR. Best regards, Jurij Smakov[EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: how to control/override dev assigment 3ware and USB
On Thu, Sep 21, 2006 at 05:25:13PM -0700, RParr wrote: > > I am still unable to get Debian installed on this machine. > > I AM able to boot Knoppix LiveCD and NOT have any device conflicts. > Knoppix (5.0.1) assigns the 3ware /dev/sde and the USB card reader as > /dev/sda, sdb, sdc, and sdd and thus has no conflict. > > I used the Debian netinst to boot into rescue mode, change the apt > sources.list from etch to sid and dist-upgrade. That went well and the > system is now booting the 2.6.17 kernel but still freezing because > /dev/sde is reassigned. > > I have already edited /etc/default/bootlogd to enable boot logging but I > get no logs. > > 1) Does anyone have any ideas as to how I can log the initial boot up > stuff and figure out if something is getting assigned to /dev/sda (and > thus starting the USB reading at /dev/sdb) or if the USB assignment is > just wrong. > > 2) Does anyone know how best to prevent the USB drive assignments from > happening. Hi, You can "blacklist" a module, so that it's not going to be automatically loaded by udev, even if a device is present. To do that, assuming the name of the module is foo, you should create an arbitrarily named file in /etc/modprobe.d directory, with contents blacklist foo You can (hopefully) try to blacklist modules usbcore, ehci_hcd and uhci_hcd to disable USB devices. To be honest, I don't think that device naming conflict is responsible for the hang, but so far I have no clue what might cause it. What CD image are you using? What are the last messages on the screen that you see? Best regards, -- Jurij Smakov[EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#389516: Supported hardware chapter for installation manual
Package: installation-guide-sparc Severity: wishlist Hi, The attached file (sparc.xml) contains the description of supported sparc hardware. It is intended to replace en/hardware/supported.sparc.xml. Please review and apply. Thanks, -- Jurij Smakov[EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC sparc.xml Description: application/xml
Bug#384342: installation-report: doesn't boot kernel: Illegal Instruction (sparc)
Hi Christopher, Can you please retry booting it with some recent image (like a recent daily)? In order to do any useful debugging I need to know exactly what image it was and what kernel version it was running, along with output of ctrace. Thanks, -- Jurij Smakov[EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
sparc64 installation reports - Ultra5
Hi, Just tried a few installs using netinst 20061019 on my Ultra5 (Ultrasparc CPU, 128M of RAM, 8GB IDE hard drive, atyfb framebuffer hooked up to a monitor) * English install, manual partitioning, do not use network mirror, standard system. * Russian install, automatic partitioning, everything in one partition, use network mirror, file server. Some minor translation comments/suggestions, sent to russian l10n list. * Lithuanian install: automatic partitioning, use entire disk and setup lvm, separate /home partition, use network mirror, web server. Initial partitioning screen, and partitioning erase warning (and some other minor stuff) are not translated. Very impressive! Everything was very smooth, hardly any comments at all. I think that's the best quality sparc installer I've ever seen. Framebuffer on this box is working, making pseudographics and fonts look very nice. More reports to come tomorrow and over the weekend. Best regards, -- Jurij Smakov [EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#395230: installation-reports: [sparc64] ultra10 creator 3d install failed
On Wed, Oct 25, 2006 at 11:21:59AM -0700, Blars Blarson wrote: > Package: installation-reports > Severity: important > > Due to bug 395147, the sparc buisinesscard cd 2006-10-23 version > cannot be used. The system displays nothing past the "Booting > Linux..." message. Please try booting with 'video=atyfb:off' boot parameter. Thanks. -- Jurij Smakov [EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#400408: Modules in different udebs depending on the same module == problem
Package: kernel-wedge Version: 2.29 Severity: normal Hi, I've recently encountered a situation in which both SCSI module and a NIC module depend on firmware_class.ko (on sparc). Since they go into different udebs, kernel-wedge complains about duplicate modules in udebs. Looks like we need something like common-modules udeb for situation like this. Or is there some existing workaround? Best regards, -- Jurij Smakov [EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#400837: Sun Netra-X1 install configures dmfe driver rather than Tulip driver
Hi Richard, Thanks for your report. Unfortunately, this issue is pretty well known (check out the bug 334104, for example), and so far we have failed to come up with a decent solution to it. The problem is that both dmfe and tulip cards advertise support for this particular PCI id. The sad part is that there are some cards out there, having the same PCI id, but requiring different drivers to work! So far I could not come up with anything better than patching both modules to introduce a parameter which can be used to prevent them from initializing any devices. Unfortunately, it will still be up to user to recognize that the problem comes from tulip/dmfe driver conflict, and disable the "faulty" one. Best regards, -- Jurij Smakov [EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Retiring the sparc32 port
Hi, First of all, I would like to apologize for falling out of the loop for almost 4 months. My move and settling-in period took quite a bit longer than I expected. In the meantime there was no further progress on the decision about continued sparc32 support, so I would like to address it as soon as possible. The discussion which took place in April and May did not in any way resolve the fundamental problem of sparc32: lack of people actively interested in maintaining it (please correct me if I'm wrong). Most developers involved with the sparc port didn't object to dropping it. There were a few objections, but I haven't seen any realistic plan for keeping the port afloat. Even though it appears that there were a few bugfixes for it since my departure, there is still no active upstream maintenance, so I don't think we have sufficient reasons to re-enable the sparc32 kernel builds. Without the updated kernels the debian-boot team cannot really keep the installer in sync with other architectures, which means that we have to make a decision. I don't see other option than dropping it at that point, and I would like to make it official. If there are no last-minute objections, I would like to make an official announcement on d-d-a that Debian is dropping support for sparc32 for lenny within the next couple of days. Best regards, -- Jurij Smakov [EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC signature.asc Description: Digital signature
Bug#433667: Please OK licence change for silo-installer udeb
On Wed, Jul 18, 2007 at 07:38:05PM +0200, Frans Pop wrote: > (Contributors to silo-installer BCC'ed.) > > We were recently alerted to the fact that the source for silo-installer > udeb did not include a copyright file. After checking with Ben Collins, > the original author, I have added a copyright statement as in [1] in the > D-I SVN repository (gpl version 2 or later; text consistent with other > copyright files in the installer). > > As this can be considered a licence change and as you have contributed to > silo-installer in the past, please OK this licence change by replying to > [EMAIL PROTECTED] with a short statement of your agreement. I approve this change. Best regards, -- Jurij Smakov [EMAIL PROTECTED] Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC signature.asc Description: Digital signature
Bug#602740: Squeeze beta1 installer on SunBlade 1000 (sparc) - keyboard not working
Package: installation-reports Hello, I've tried a couple of installs on my SunBlade 1000 box (sparc) using the Squeeze beta1 netinst image. Installation using serial console with guided partitioning worked like a charm. However, when I tried to install from regular console, keyboard (Sun USB Type 6) did not work in the installer. I've tried the latest Lenny installer (5.06 netinst) and the same keyboard works there, so it looks like a regression. Judging by the entries in dmesg, keyboard is detected properly in both cases and expected drivers are loaded - it just does not work with Squeeze installer. Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20101107175057.ga4...@droopy.oc.cox.net
Bug#602740: Correction
Correction: keyboard actually works in the system installed using Squeeze beta1, but not in installer itself. There is a kernel version difference, installer has 2.6.32-25 while installed system ends up with 2.6.32-27. Trying to figure out whether this makes a difference now. Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20101108183246.ga4...@droopy.oc.cox.net
Bug#602740: Caused by missing usbhid/hid modules
reassign linux-kernel-di-sparc-2.6 severity serious thanks This bug is caused by the lack of usbhid/hid modules in the installer. It worked for lenny because in lenny these modules were compiled into the kernel: ju...@droopy:~$ grep CONFIG_HID config-2.6.26-2-sparc64-smp CONFIG_HID_SUPPORT=y CONFIG_HID=y # CONFIG_HID_DEBUG is not set CONFIG_HIDRAW=y CONFIG_HID_FF=y CONFIG_HID_PID=y ju...@droopy:~$ grep CONFIG_USB_HID config-2.6.26-2-sparc64-smp CONFIG_USB_HID=y # CONFIG_USB_HIDINPUT_POWERBOOK is not set # CONFIG_USB_HIDDEV is not set However the current image used by the installer has: ju...@droopy:~$ grep CONFIG_HID config-2.6.32-5-sparc64-smp CONFIG_HID_SUPPORT=y CONFIG_HID=m CONFIG_HIDRAW=y CONFIG_HID_PID=y CONFIG_HID_A4TECH=m [...] ju...@droopy:~$ grep CONFIG_USB_HID config-2.6.32-5-sparc64-smp CONFIG_USB_HID=m CONFIG_USB_HIDDEV=y We need to add input-modules on sparc, which should pull in usbhid (and, hopefully, the modules it depends on). Otherwise any install using USB keyboard will not work. Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20101108205645.ga6...@droopy.oc.cox.net
Bug#602740: Caused by missing usbhid/hid modules
reassign 602740 linux-kernel-di-sparc-2.6 severity 602740 serious thanks -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20101108210507.ga6...@droopy.oc.cox.net
Old kernels used for sparc daily builds
Hello, It looks like the sparc daily builds [0] are happening in a broken environment, because the kernel image which gets installed into the /boot directory of the ISO image dates back to June: ju...@droopy:~$ grep 'built on' /mnt/iso/boot/debian.txt This is a Debian installation CDROM, built on 20101110-15:27. ju...@droopy:~$ strings /mnt/iso/boot/sparc64 | grep 'Linux version' Linux version 2.6.32-5-sparc64 (Debian 2.6.32-15) (b...@decadent.org.uk) (gcc version 4.3.5 (Debian 4.3.5-1) ) #1 Tue Jun 1 06:56:43 UTC 2010 ju...@droopy:~$ It would be great to fix it as testing the images with such an old kernel is meaningless (for example, it hangs on my workstation early during boot). [0] http://cdimage.debian.org/cdimage/daily-builds/daily/arch-latest/sparc/iso-cd/ Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20101110182436.ga6...@droopy.oc.cox.net
Re: Fwd: Old kernels used for sparc daily builds
On Wed, Nov 10, 2010 at 10:43:37PM +0100, Gaudenz Steinlin wrote: > Hi Geert > > You used to build the sparc daily images for d-i. But these images are > no longer built since June. Are you still intending to provide daily > d-i images for sparc or should someone else take over this job? I'm pretty sure that the images are getting built, as, for example, today's daily contains the binary packages build from version 1.60 of linux-kernel-di-sparc-2.6, which was uploaded to the archive just yesterday. It's just that the kernel which gets installed into the image is old. I don't know how the build works, but it might be that the kernel which makes it onto the iso is the same one the build machine is running (or something :-). Best regards. > Gaudenz > > --- Begin forwarded message from Jurij Smakov --- > From: Jurij Smakov > To: debian-boot > Date: Wed, 10 Nov 2010 19:24:36 +0100 > Subject: Old kernels used for sparc daily builds > > Hello, > > It looks like the sparc daily builds [0] are happening in a broken > environment, because the kernel image which gets installed into the > /boot directory of the ISO image dates back to June: > > ju...@droopy:~$ grep 'built on' /mnt/iso/boot/debian.txt > This is a Debian installation CDROM, built on 20101110-15:27. > ju...@droopy:~$ strings /mnt/iso/boot/sparc64 | grep 'Linux version' > Linux version 2.6.32-5-sparc64 (Debian 2.6.32-15) (b...@decadent.org.uk) (gcc > version 4.3.5 (Debian 4.3.5-1) ) #1 Tue Jun 1 06:56:43 UTC 2010 > ju...@droopy:~$ > > It would be great to fix it as testing the images with such an old > kernel is meaningless (for example, it hangs on my workstation early > during boot). > > [0] > http://cdimage.debian.org/cdimage/daily-builds/daily/arch-latest/sparc/iso-cd/ > > Best regards, > --- End forwarded message --- > -- > Ever tried. Ever failed. No matter. > Try again. Fail again. Fail better. > ~ Samuel Beckett ~ -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20101110220248.ga7...@droopy.oc.cox.net
Re: [ANNOUNCE] Call for testing Debian Installer 6.0 Beta2 images - soon to release
On Fri, Dec 03, 2010 at 09:00:21PM -0200, Otavio Salvador wrote: > Hello, > > We've managed to get the images available at > http://cdimage.debian.org/cdimage/.squeeze_beta2/ so please take some > time during this weekend and test it. > > I plan to release it all on Monday so RUN! :-) Looks great on sparc. Two installs using netinst image (serial console + guided partitioning and regular console + guided partitioning + LVM) went without a hitch. Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20101204183026.ga3...@droopy.oc.cox.net
Bug#608516: T2+ sparc machines need niu network driver
Package: linux-kernel-di-sparc-2.6 Version: 1.64 Severity: wishlist Hello, As recently reported on the debian-sparc mailing list [0], network is not automatically configured during installation on the latest generation of Sparc boxes with T2+ CPU. Apparently, they use niu.ko network driver, which is available in kernel packages, but not in d-i. Please add it to the list of network drivers included in udebs on sparc. Thanks. [0] http://lists.debian.org/debian-sparc/2010/12/msg00011.html -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20101231181622.ga2...@droopy.oc.cox.net
Bug#608516: T2+ sparc machines need niu network driver
On Fri, Dec 31, 2010 at 06:42:49PM +, Miguel Figueiredo wrote: > Hi, > > added niu module to sparc64. > This needs to be reviewed. I believe you need to add it to linux-kernel-di-sparc-2.6 as well (for which this bug was filed). linux-kernel-di-sparc64-2.6 which your patch affects is for sparc64 port (64-bit userspace and kernel) which is not going to be released with squeeze (yes, it's confusing :-). Best regards, > > Sex, 2010-12-31 às 18:16 +, Jurij Smakov escreveu: > > Package: linux-kernel-di-sparc-2.6 > > Version: 1.64 > > Severity: wishlist > > > > Hello, > > > > As recently reported on the debian-sparc mailing list [0], network is > > not automatically configured during installation on the latest > > generation of Sparc boxes with T2+ CPU. Apparently, they use niu.ko > > network driver, which is available in kernel packages, but not in d-i. > > Please add it to the list of network drivers included in udebs on > > sparc. > > > > Thanks. > > > > [0] http://lists.debian.org/debian-sparc/2010/12/msg00011.html > > -- > > Jurij Smakov ju...@wooyd.org > > Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC > > > > > > > > -- > Melhores cumprimentos/Best regards, > > Miguel Figueiredo > http://www.DebianPT.org > Index: packages/kernel/linux-kernel-di-sparc64-2.6/modules/sparc64/nic-modules > === > --- packages/kernel/linux-kernel-di-sparc64-2.6/modules/sparc64/nic-modules > (revision 66169) > +++ packages/kernel/linux-kernel-di-sparc64-2.6/modules/sparc64/nic-modules > (working copy) > @@ -39,6 +39,7 @@ > uli526x > qla3xxx > atl1 > +niu > > # Support for virtual network devices under Sun Logical Domains. > sunvnet ? -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110101152556.ga2...@droopy.oc.cox.net
Re: SPARC daily d-i builds
On Wed, Jan 12, 2011 at 02:46:24AM +0100, Philipp Kern wrote: > Hi, > > am Wed, Jan 12, 2011 at 01:53:59AM +0100 hast du folgendes geschrieben: > > Gaudenz Steinlin (11/01/2011): > > > As all my attempts to find someone willing to setup d-i daily builds > > > on a sparc buildd [1] failed. I'm going to give up now and will not > > > bother you again until some steps up to do the necessary work or > > > provide me with buildd access. I added the following note to the d-i > > > daily builds overview page: Due to the lack of porter and buildd > > > admin interest there are currently no daily builds for the SPARC > > > architecture. These builds will be reenable as soon as someone finds > > > the time to do the necessary buildd setup. > > > > > > I would also be willing to do the neccessary setup myself if someone > > > provides me with the necessary access to a sparc buildd. > > > > So, again, it's not about lack of interest, it's about lack of time. > > > > Also, you could have pinged debian-wb-t...@. I believe aba did most of > > the d-i autobuilding stuff, so he might have some clue to share. I've > > added this list to the loop. > > and I did tell on #d-boot the following on Nov 13 2010: > > [...] > 15:27 < trave11er> ok, we didn't get any reply from stappers, so sparc > dailies still have the old kernels, afaict > [...] > 15:31 < otavio> trave11er: buildd people were going to put sparc into it > 15:31 < otavio> trave11er: dunno if it has been done > 15:31 < otavio> adsb: ^? > [...] > 15:33 < phil> otavio: "were going to" is quite untrue. There still isn't a > bug > report, and I did report the result of my investigation, i.e. there's no space > on the only LVMed sparc buildd we have. > 15:37 < otavio> phil: I didn't recall about the space issue > 15:37 < otavio> phil: indeed > > So meh, whatever. To my knowledge that's still true. We currently don't have > the capacities and we admitted that. Furthermore not asking -wb-team might > not give you answers, indeed. Sorry if I missed something along the way. That got me thinking... We had this T2K box (zee) donated a while ago (http://lists.debian.org/debian-sparc/2009/08/msg9.html, for example), and it shows up as a buildd on a machine page, however I don't really remember ever seeing a package built on it. Can you please check what's its state and whether it can be used for these builds? Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110112072328.ga3...@droopy.oc.cox.net
Bug#609466: Patch may be available for this issue
Hello, Recently David Miller has posted a patch for this issue, see this thread: http://marc.info/?t=12942617841&r=1&w=2 I'm CC'ing Frans van Berckel who has confirmed that the patch fixed the issue for him. If you can confirm that it fixes it for you as well, feel free to file a bug report agains linux-2.6 with this information. There is probably no chance at this point to get it into squeeze r0, but we can target it for the first update. Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110114220429.ga4...@droopy.oc.cox.net
Bug#562594: SunBlade 1000 installation report - no option to load firmware
Package: installation-reports Boot method: Dec 26 netinst daily image for sparc Image version: http://cdimage.debian.org/cdimage/daily-builds/daily/arch-latest/sparc/iso-cd/debian-testing-sparc-netinst.iso Date: Sat Dec 26 12:20:39 GMT 2009 Machine: SunBlade 1000 Processor: Ultrasparc III x 2 Memory: 2GB Partitions: did not get that far Output of lspci -knn (or lspci -nn): :00:01.0 VGA compatible controller [0300]: Intergraph Corporation Sun Expert3D-Lite Graphics Accelerator [1091:07a0] Subsystem: Intergraph Corporation Device [1091:0140] :00:05.0 Bridge [0680]: Sun Microsystems Computer Corp. RIO EBUS [108e:1100] (rev 01) :00:05.1 Ethernet controller [0200]: Sun Microsystems Computer Corp. RIO 10/100 Ethernet [eri] [108e:1101] (rev 01) Kernel driver in use: gem :00:05.2 FireWire (IEEE 1394) [0c00]: Sun Microsystems Computer Corp. RIO 1394 [108e:1102] (rev 01) :00:05.3 USB Controller [0c03]: Sun Microsystems Computer Corp. RIO USB [108e:1103] (rev 01) Kernel driver in use: ohci_hcd :00:06.0 SCSI storage controller [0100]: LSI Logic / Symbios Logic 53c875 [1000:000f] (rev 37) Kernel driver in use: sym53c8xx :00:06.1 SCSI storage controller [0100]: LSI Logic / Symbios Logic 53c875 [1000:000f] (rev 37) Kernel driver in use: sym53c8xx 0001:00:04.0 SCSI storage controller [0100]: QLogic Corp. QLA2200 64-bit Fibre Channel Adapter [1077:2200] (rev 05) Base System Installation Checklist: [O] = OK, [E] = Error (please elaborate below), [ ] = didn't try it Initial boot: [O] Detect network card:[O] Configure network: [O] Detect CD: [O] Load installer modules: [O] Detect hard drives: [E] Partition hard drives: [ ] Install base system:[ ] Clock/timezone setup: [ ] User/password setup:[ ] Install tasks: [ ] Install boot loader:[ ] Overall install:[ ] Comments/Problems: The QLA2200 SCSI controller in this box is supported by qla2xxx kernel driver, however it requires (non-free) firmware to operate. Earlier versions of installer have offered an option to load firmware from a removable drive during the installation, but this version does not offer this option, fails to detect the disks and jumps directly to the manual driver choice screen, so this is a regression. -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#562594: More information
It turns out that the installer does offer an option to load firmware from external drive, but only *after* I have loaded the firmware by hand and re-ran 'Detect disks' option from installer menu. -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#562594: More information
On Sat, Dec 26, 2009 at 05:03:03PM +0100, Frans Pop wrote: > On Saturday 26 December 2009, Jurij Smakov wrote: > > It turns out that the installer does offer an option to load firmware > > from external drive, but only *after* I have loaded the firmware by > > hand and re-ran 'Detect disks' option from installer menu. > > There are calls to a script 'check missing firmware.sh' during all three > stages of hardware detection. > > Please do an install in expert mode, before any hardware detection is > performed, add a 'set -x' in: > - /bin/hw-detect > - /bin/disk-detect > - /bin/check-missing-firmware > It appears that logic in check-missing-firmware is flawed. It checks the /dev/.udev/firmware-missing directory first to see if any missing firmwares are listed there, and on my system the needed firmware shows up there: ~ # ls -al /dev/.udev/firmware-missing/ drwxr-xr-x2 root root 60 Dec 26 19:05 . drwxr-xr-x8 root root 180 Dec 26 19:08 .. lrwxrwxrwx1 root root 68 Dec 26 19:05 ql2200_fw.bin -> /devices/root/f006a214/pci0001:00/0001:00:04.0/firmware/0001:00:04.0 ~ # It then munges the link destination to come up with the device path of /sys/devices/root/f006a214/pci0001:00/0001:00:04.0. That's the correct device path for SCSI controller, confirmed by matching to the lspci output. It then expects to find a symlink driver/module in this directory, pointing to the driver module. It's not present in my case: ~ # ls -la /sys/devices/root/f006a214/pci0001:00/0001:00:04.0 drwxr-xr-x2 root root0 Dec 26 19:05 . drwxr-xr-x4 root root0 Dec 26 19:02 .. -rw-r--r--1 root root 8192 Dec 26 19:11 broken_parity_status -r--r--r--1 root root 8192 Dec 26 19:04 class -rw-r--r--1 root root 256 Dec 26 19:11 config -r--r--r--1 root root 8192 Dec 26 19:04 device -rw---1 root root 8192 Dec 26 19:11 enable -r--r--r--1 root root 8192 Dec 26 19:04 irq -r--r--r--1 root root 8192 Dec 26 19:11 local_cpulist -r--r--r--1 root root 8192 Dec 26 19:11 local_cpus -r--r--r--1 root root 8192 Dec 26 19:11 modalias -rw-r--r--1 root root 8192 Dec 26 19:11 msi_bus -r--r--r--1 root root 8192 Dec 26 19:11 obppath --w--w1 root root 8192 Dec 26 19:11 remove --w--w1 root root 8192 Dec 26 19:11 rescan -r--r--r--1 root root 8192 Dec 26 19:04 resource -rw---1 root root 256 Dec 26 19:11 resource0 -rw---1 root root 8192 Dec 26 19:11 resource1 -r1 root root 131072 Dec 26 19:11 rom lrwxrwxrwx1 root root0 Dec 26 19:11 subsystem -> ../../../../../bus/pci -r--r--r--1 root root 8192 Dec 26 19:11 subsystem_device -r--r--r--1 root root 8192 Dec 26 19:11 subsystem_vendor -rw-r--r--1 root root 8192 Dec 26 19:02 uevent -r--r--r--1 root root 8192 Dec 26 19:04 vendor ~ # The logic in check_missing() function appends the missing firmware file name to the $files list and the module to $modules list. However, if such mapping fails (as in this case), nothing is appended to $files or $modules, and the file is simply ignored as a result: if [ -n "$modules" ]; then log "missing firmware files ($files) for $modules" return 0 else log "no missing firmware in $MISSING" return 1 fi The problem here is that empty $modules does not mean that there are no files in $MISSING which is set to /dev/.udev/firmware-missing (the log message is wrong as well). Not sure whether this problem is caused by changes to check-missing-firmware or kernel, I can try to boot a lenny installer tomorrow to see how it worked there. Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#562594: Use modalias to detect modules when driver does not register
On Mon, Dec 28, 2009 at 03:17:49PM +0100, Frans Pop wrote: > On Monday 28 December 2009, Jérémy Bobbio wrote: > > This problem looks similar to the one I had while trying to use a daily > > build image on a PowerEdge R410. > > > > The network interfaces on these machines are supported by the bnx2 > > module, but when no firmware is available, the driver does not register > > itself for the device, and there is no `$devpath/driver` directory > > available. > > > > The attached patch adds to `check-missing-firmware` another way to > > determine the module by using `modprobe` and `$devpath/modalias`. > > Thanks Jérémy. Looks good. > > I think the patch needs an added test that modalias exists, something like: > elif [ -e $devpath/modalias ]; then > > And that long line could do with a split to the next line after the pipe. > The "|| true" is probably not needed as the sed statement will absorb any > errors from modprobe, but maybe there should be a redirect of errors: > module=$(modprobe --show-depends "$modalias" 2>/dev/null | \ >sed -n -e '$s#^.*/\([^.]*\)\.ko.*$#\1#p') I see that modprobe has --resolve-alias flag which prints out the module(s) matching an alias, which renders the whole sed concoction unnecessary. > A comment that the module supporting the device is the last line returned > by modprobe would be useful. It's not quite self-evident. > > > Once this changes applied, it sucessfully prompted for the relevant > > firmware files. > > Jurij: does this patch also work for you? Looks reasonable, thanks. Unfortunately, I'm on vacation for the next two weeks without access to my sparc, so I'll not be able to test it until the third week of January. > Patch available at: http://bugs.debian.org/562594#39 Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#565639: SunBlade 1000 installation report - boot fails after installation
Package: installation-reports
Boot method: Jan 14th netinst daily image for sparc
Image version:
http://cdimage.debian.org/cdimage/daily-builds/daily/arch-latest/sparc/iso-cd/debian-testing-sparc-netinst.iso
Date: Sun Jan 17 15:37:37 GMT 2010
Machine: SunBlade 1000, same as in http://bugs.debian.org/562594
Base System Installation Checklist:
[O] = OK, [E] = Error (please elaborate below), [ ] = didn't try it
Initial boot: [O]
Detect network card:[O]
Configure network: [O]
Detect CD: [O]
Load installer modules: [O]
Detect hard drives: [O]
Partition hard drives: [O]
Install base system:[O]
Clock/timezone setup: [O]
User/password setup:[O]
Install tasks: [O]
Install boot loader:[O]
Overall install:[E]
Comments/Problems:
Now that http://bugs.debian.org/562594 is fixed, I was able to
continue the installation and it all went fine until the end.
However, after finishing the installation machine fails to boot with
messages:
SUNW,Sun-Blade-1000 (2 X UltraSPARC-III) , No Keyboard
Copyright 1998-2004 Sun Microsystems, Inc. All rights reserved.
OpenBoot 4.16.4, 2048 MB memory installed, Serial #51831895.
Ethernet address 0:3:ba:16:e4:57, Host ID: 8316e457.
Rebooting with command: boot
Boot device: disk File and args:
ERROR: Last Trap: Fast Data Access MMU Miss
Error -256
{0} ok
I thought it was some kind of random hardware quirk, so I tried
reinstalling a few times, but always got the same result, even if I
was zero-filling the first 200M of the hard drive before partitioning.
I then tried Lenny netinst CD (debian-503-sparc-netinst.iso), and it
worked without problems, so it must be some regression in the recent
CD images. At this point I have no clue what might be causing it, as
both Lenny and the current CD builds install the same SILO version and
partitions are set up identically (~100MB ext2 /boot partition is set
up to hold the boot files in both cases). I'm going to ask on
debian-sparc whether anyone can reproduce it.
Another minor detail: serial console output in daily builds is
black-and-white, while Lenny installer was showing up in color (at
least, red/blue where appropriate).
--
Jurij Smakov ju...@wooyd.org
Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC
--
To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#565639: SunBlade 1000 installation report - boot fails after installation
On Sun, Jan 17, 2010 at 05:08:55PM +0100, Frans Pop wrote: > On Sunday 17 January 2010, Jurij Smakov wrote: > > Comments/Problems: > > Now that http://bugs.debian.org/562594 is fixed, I was able to > > continue the installation and it all went fine until the end. > > However, after finishing the installation machine fails to boot [...] > > As the installation itself works fine I would guess this is not an > installer problem, but a kernel issue. Maybe it leaves the hardware in an > incorrect state for the reboot. > > Have you tried a cold boot of the installed system? Yeah, neither cold-boot nor reinstalling/rerunning silo in the /target after installation is done help. I'll try to investigate next week. -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#565639: SunBlade 1000 installation report - boot fails after installation
On Sun, Jan 17, 2010 at 05:08:55PM +0100, Frans Pop wrote: > On Sunday 17 January 2010, Jurij Smakov wrote: > > Comments/Problems: > > Now that http://bugs.debian.org/562594 is fixed, I was able to > > continue the installation and it all went fine until the end. > > However, after finishing the installation machine fails to boot [...] > > As the installation itself works fine I would guess this is not an > installer problem, but a kernel issue. Maybe it leaves the hardware in an > incorrect state for the reboot. > > Have you tried a cold boot of the installed system? Adding kernel team for comment: it looks like the value returned by uname -m has changed from 'sparc64' in kernel 2.6.26 to 'sparc' in 2.6.30 (and, probably, later ones). This breaks SILO, which uses this value to determine what first-stage bootloader to install, as a result the sparc machines installed with current daily installer builds turn unbootable by the end of installation. Please confirm that this change is not a result of some Debian-specific patch or build procedure, as it should be reported upstream if that's not the case. For reference: ju...@debian:~$ uname -a Linux debian 2.6.26-2-sparc64-smp #1 SMP Thu Nov 5 03:34:29 UTC 2009 sparc64 GNU/Linux ju...@debian:~$ uname -m sparc64 ~ # uname -a Linux debian 2.6.30-2-sparc64 #1 Sat Sep 26 07:26:31 UTC 2009 sparc GNU/Linux ~ # uname -m sparc I've confirmed that forcing the machine type by running 'silo -u -f' makes the machine bootable again. Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: Bug#512740: Sparc disk labels broken on LDOM and Parallel installs
On Tue, Feb 09, 2010 at 06:54:39PM +0100, Frans Pop wrote: > Below are instructions for building a custom D-I image that includes a > patched libparted and partman. It's not that hard. I've built a mini iso image according to these instructions. It is available from http://www.wooyd.org/debian/512740, along with intermediate packages. Minor issues with the build and the resulting image: 1. dh-di needed for build of partman-base is not pulled in as a result of 'apt-get build-dep debian-installer', probably should be? 2. Firmware loading is broken in the mini iso image, error is 'mountmedia not found'. 3. The installation is still affected by the kernel issue where it misreports the architecture and SILO installs the incorrect second-stage bootloader as a result. I've ran a couple of installs with this image, zero-filling beginning of the disk with zeros during one of them, and did not see any regressions. Mike, can you try downloading the mini.iso and testing that it fixes the issue for you? If your machine needs firmware loaded for the disks to be detected, you'll need to do it manually. Also, you will need to drop into installer shell at the end of the installation, chroot into /target, mount /proc and run 'silo -u -f' to unbreak the booloader. Let me know if you would like more specific instructions. Thanks. -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20100214155149.ga5...@droopy.oc.cox.net
Re: [RFH] Update of Debian Installer for 2.6.29
On Tue, Apr 21, 2009 at 09:51:56PM -0300, Otavio Salvador wrote: > Hello porters, > > Most arches has 2.6.29 packages ready (AFAIK only hppa lacks them > right now) and I also did most changes for all arches on SVN to get > 2.6.29 updating as easy as possible. Now, we need porters' help. > > Please check your pet arch and update the kernel and modules for > 2.6.29. This is quite important to us to start testing installer with > this kernel. > > Currently those are the arches that have already switch: > > - i386 > - amd64 > - mips* > - mipsel* > - armel* > > I taged the ones that has pending changes in SVN with "*". > > So if your arch is not listed above, you should take a look and try to > update it. If you can't do that in near future please speakup and > we'll try to find someone to handle it. > > Thanks a lot by looking at it. Otavio, what exactly needs to be done? Are there instructions available somewhere? I thought that you would just push the new kernels into d-i, do we just need to check that kernel-wedge config is still valid and includes all important modules? I could look at it over the weekend, but I'll need some more guidance :-). Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: [RFH] Update of Debian Installer for 2.6.29
On Tue, Apr 21, 2009 at 09:51:56PM -0300, Otavio Salvador wrote:
> Hello porters,
>
> Most arches has 2.6.29 packages ready (AFAIK only hppa lacks them
> right now) and I also did most changes for all arches on SVN to get
> 2.6.29 updating as easy as possible. Now, we need porters' help.
>
> Please check your pet arch and update the kernel and modules for
> 2.6.29. This is quite important to us to start testing installer with
> this kernel.
Well, 2.6.29 does not even boot on my SunBlade 1000, failing early in
the boot process:
Rebooting with command: boot
Boot device: disk File and args:
SILO Version 1.4.13
boot:
Allocated 8 Megs of memory at 0x4000 for kernel
Kernel doesn't support loading to high memory, relocating...done.
Loaded kernel version 2.6.29
Loading initial ramdisk (6348059 bytes at 0x40 phys, 0x40C0 virt)...
ERROR: Last Trap: Fast Data Access MMU Miss
Error -256
{0} ok
2.6.26 boots fine under the same conditions.
--
Jurij Smakov ju...@wooyd.org
Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC
--
To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: [RFH] Update of Debian Installer for 2.6.29
On Wed, Apr 22, 2009 at 09:02:23PM -0300, Otavio Salvador wrote: > On Wed, Apr 22, 2009 at 7:28 PM, Jurij Smakov wrote: > >> So if your arch is not listed above, you should take a look and try to > >> update it. If you can't do that in near future please speakup and > >> we'll try to find someone to handle it. > >> > >> Thanks a lot by looking at it. > > > > Otavio, what exactly needs to be done? Are there instructions available > > somewhere? I thought that you would just push the new kernels into d-i, > > do we just need to check that kernel-wedge config is still valid and > > includes all important modules? > > Yes; you should check if we didn't forgot anything important. We've done > our best to try to properly handle it and we also have updated kernel-wedge > but we surely can forget a specific module or change. > > Besides that, you need to properly build the modules, test and upload. It turns out that 2.6.29-3 is not bootable on sparc due to http://bugs.debian.org/525926. Kernel team plans another upload in the next couple of days, and I'll try to proceed with getting it into d-i, after confirming that it works as expected. Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Switching to by-label mounting of partitions
Hi, When using d-i on my sparc box I have to use the firmware loading feature in order to make the firmware for the SCSI controller available to the installer, otherwise the disks are not detected. A minor annoyance is that once you insert the USB flash drive with firmware, it's assigned 'sda' as as a device name, and then the real disk gets 'sdb'. This causes problems later, as /dev/sdb2 is listed as root device in /etc/fstab and /etc/silo.conf. If one tries to boot without the flash drive inserted later, the real disk is recognized as 'sda' instead of 'sdb', causing a boot failure, as the filesystem cannot be found. Today I tried to set the labels in the installer, and then manually edit /etc/fstab and /etc/silo.conf at the end of installation, specifying the devices by LABEL instead of device name. It appears to work fine, you only need to quote the value for root put in silo.conf like this: root="LABEL=root" As the amount of users who are using the firmware loading feature is probably non-negligible, it would be cool to get the installer to assign labels by default to all partitions, and generate the /etc/fstab and bootloader configs using labels. What do you think? Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: debian 5.02 install
On Mon, Jul 06, 2009 at 03:55:22PM +0100, Allan Fearon wrote: > after doing clean install of Debian 5.02 desktop found that > network-manager and update-manager no longer included by default. is > there a reason for this. Rerouting to debian-boot, where this discussion is more appropriate, debian-release -> bcc. -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: unblock fontconfig?
On Mon, Jul 06, 2009 at 06:40:19PM +0800, Paul Wise wrote: > Hi all, > > fontconfig has been waiting 21 days for migration, perhaps it should be > unblocked? No RC bugs were filed during that time. debian-boot, please approve, this is blocked due to a udeb. Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#645657: Debian daily image isn't bootable on sparc
retitle 645657 Daily sparc netboot images are too big, fail to boot thanks The netboot images are too big and are larger than OBP can handle (it seems the limit is around 10MB). Here's the list of files in the initrd, sorted by size: http://www.wooyd.org/debian/initrd_contents.txt Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20111020220806.ga5...@wooyd.org
Dropping wpasupplicant from sparc netboot image
Hello, As we discovered recently, the daily sparc netboot image exceeds 10MB limit imposed by OpenBoot, so it fails to boot. I've poked at it a bit, and it seems that one of the most straightforward ways to make room is to drop wpasupplicant and all its dependencies (libcrypto, crypto kernel modules) which account for significant amount of space on the initrd. AFAIK, sparc does not have any hardware which does wifi natively, and while it's in theory possible to use some usb-wifi dongle to do the installation, I don't remember any reports of anyone doing it. The straightforward way to achieve this goal is to make netcfg not depend on wpasupplicant for sparc (same way as s390 currently does). This will drop wpasupplicant (and, subsequently, installation over wireless support) not only from netboot, but also from all other installation media (like businesscard iso). Let me know if you have any objections, otherwise I'll commit necessary fixes in a couple of days. Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20111023123706.ga5...@wooyd.org
Re: Dropping wpasupplicant from sparc netboot image
[correct the debian-sparc address] On Sun, Oct 23, 2011 at 01:37:06PM +0100, Jurij Smakov wrote: > Hello, > > As we discovered recently, the daily sparc netboot image > exceeds 10MB limit imposed by OpenBoot, so it fails to boot. I've > poked at it a bit, and it seems that one of the most straightforward > ways to make room is to drop wpasupplicant and all its dependencies > (libcrypto, crypto kernel modules) which account for significant > amount of space on the initrd. AFAIK, sparc does not have any hardware > which does wifi natively, and while it's in theory possible to use > some usb-wifi dongle to do the installation, I don't remember any > reports of anyone doing it. > > The straightforward way to achieve this goal is to make netcfg not > depend on wpasupplicant for sparc (same way as s390 currently does). > This will drop wpasupplicant (and, subsequently, installation over > wireless support) not only from netboot, but also from all other > installation media (like businesscard iso). > > Let me know if you have any objections, otherwise I'll commit > necessary fixes in a couple of days. > > Best regards, > -- > Jurij Smakov ju...@wooyd.org > Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20111023124130.ga5...@wooyd.org
Re: Dropping wpasupplicant from sparc netboot image
On Sun, Oct 23, 2011 at 01:41:30PM +0100, Jurij Smakov wrote: > [correct the debian-sparc address] > > On Sun, Oct 23, 2011 at 01:37:06PM +0100, Jurij Smakov wrote: > > Hello, > > > > As we discovered recently, the daily sparc netboot image > > exceeds 10MB limit imposed by OpenBoot, so it fails to boot. I've > > poked at it a bit, and it seems that one of the most straightforward > > ways to make room is to drop wpasupplicant and all its dependencies > > (libcrypto, crypto kernel modules) which account for significant > > amount of space on the initrd. AFAIK, sparc does not have any hardware > > which does wifi natively, and while it's in theory possible to use > > some usb-wifi dongle to do the installation, I don't remember any > > reports of anyone doing it. > > > > The straightforward way to achieve this goal is to make netcfg not > > depend on wpasupplicant for sparc (same way as s390 currently does). > > This will drop wpasupplicant (and, subsequently, installation over > > wireless support) not only from netboot, but also from all other > > installation media (like businesscard iso). > > > > Let me know if you have any objections, otherwise I'll commit > > necessary fixes in a couple of days. Nobody screamed loudly, so I've committed the necessary change to netcfg [0] and will release a new version later today - this is your last chance to object. [0] http://anonscm.debian.org/gitweb/?p=d-i/netcfg.git;a=commitdiff;h=630d929f282ac8b45243acf6ad6d4425779828db Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20111029110450.ga16...@wooyd.org
Re: Debian installer build: failed or old builds
On Mon, Jan 23, 2012 at 07:16:37AM +0100, Christian PERRIER wrote: > Quoting Daily build aggregator (debian-boot@lists.debian.org): > > Debian installer build overview > > --- > > > > Failed or old builds: > > > Houston, we have a problem. Hopefully, this will be fixed by http://anonscm.debian.org/gitweb/?p=d-i/debian-installer.git;a=commitdiff;h=3d1a7e02fc08a65a21d8181c30db3d20035698d2 Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120123213553.ga9...@wooyd.org
Bug#657389: Tasksel is to blame
reassign 657389 tasksel found 657389 3.07 thanks I noticed that as well with today's dailies. This is displayed by tasksel, installer just invokes it in /target. I can also reproduce it on an installed system by running 'tasksel -t'. Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120125224126.ga6...@wooyd.org
Bug#657389: Tasksel is to blame
On Sun, Jan 29, 2012 at 02:26:54AM -0400, Joey Hess wrote:
> Jurij Smakov wrote:
> > I noticed that as well with today's dailies. This is displayed by
> > tasksel, installer just invokes it in /target. I can also reproduce it
> > on an installed system by running 'tasksel -t'.
>
> Since I can't reproduce this, I can only guess.
>
> tasksel contains a sub list_installed() that parses
> /var/lib/dpkg/status. Perhaps something about the status file format has
> changed?
>
> Parsing the file is a bit gratuitous, so I've attached a patch that
> switches it to dpkg-query. Does it fix the issue?
No, that does not help. I investigated a bit and I believe that the
problem is in getdescriptions() function. In particular, it does the
following while processing the output of "apt-cache show task-${task}":
my ($description)=/^Description-.*: (.*)$/m;
($description)=/^Description: (.*)$/m
unless defined $description;
This inadvertently catches the Description-md5 field (which is new, I
guess?):
# apt-cache show task-desktop | grep Description
Description: Debian desktop environment
Description-md5: 17cb4a1ed6025b48045cc576bf52317c
#
Best regards,
--
Jurij Smakov ju...@wooyd.org
Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC
--
To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20120129195909.ga7...@wooyd.org
Sparc netboot image is too large to boot (again)
Hello, I've noticed that, yet again, sparc daily netboot image is too large to boot. Last time we mitigated the problem by removing the support for wireless networking. I'm going to poke around again to see what else can be get rid of, please let me know if you have any ideas. Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120204121625.ga6...@wooyd.org
Re: Sparc netboot image is too large to boot (again)
[Corrected CC address] On Sat, Feb 04, 2012 at 12:16:25PM +, Jurij Smakov wrote: > Hello, > > I've noticed that, yet again, sparc daily netboot image is too large > to boot. Last time we mitigated the problem by removing the support > for wireless networking. I'm going to poke around again to see what > else can be get rid of, please let me know if you have any ideas. > > Best regards, > -- > Jurij Smakov ju...@wooyd.org > Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC > > > -- > To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org > with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org > Archive: http://lists.debian.org/20120204121625.ga6...@wooyd.org -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120204150914.ga3...@wooyd.org
Re: Sparc netboot image is too large to boot (again)
On Sun, Feb 05, 2012 at 03:11:48PM +0100, Samuel Thibault wrote: > Bastian Blank, le Sun 05 Feb 2012 14:55:49 +0100, a écrit : > > On Sun, Feb 05, 2012 at 02:09:11PM +0100, Samuel Thibault wrote: > > > AIUI, linux can now have support for uncompressing bzip2, lzma and lzo. > > > > CONFIG_RD_GZIP=y > > CONFIG_RD_BZIP2=y > > CONFIG_RD_LZMA=y > > CONFIG_RD_XZ=y > > CONFIG_RD_LZO=y > > > > The amd64 image have support for gzip, bzip2, lzma and the xz variant. > > Ah, right, it's not explicitly set in the linux-2.6 configs, but > automatically enabled by Kconfig. > > So, anything against using xz compression instead? (except that there > might be quite a few scripts to update from initrd.gz to initrd.xz...) Samuel, thanks a lot for all ideas. Using xz on the sparc's netboot initrd image reduces its size dramatically: -rw-r--r-- 1 jurij jurij 4341553 Feb 17 18:50 initrd.gz -rw-r--r-- 1 jurij jurij 3083408 Feb 17 18:51 initrd.xz Attached patch implements the switch. It's not particularly elegant, but minimally intrusive, only affecting the building ot sparc's netboot image. I would like to commit it some time next week, unless anyone objects. Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC >From 821362eaaba0982132ad2ed54c9aff4be01eee38 Mon Sep 17 00:00:00 2001 From: Jurij Smakov Date: Fri, 17 Feb 2012 19:07:45 + Subject: [PATCH] Switch sparc netboot image to xz-compressed initrd to get its size under the magic limit of 10MB (otherwise it does not boot). --- build/config/sparc/netboot.cfg | 10 -- debian/control |4 +++- 2 files changed, 11 insertions(+), 3 deletions(-) diff --git a/build/config/sparc/netboot.cfg b/build/config/sparc/netboot.cfg index 23b3b46..828ce73 100644 --- a/build/config/sparc/netboot.cfg +++ b/build/config/sparc/netboot.cfg @@ -7,7 +7,13 @@ EXTRANAME = $(MEDIUM)/ MANIFEST-BOOT = "tftp boot image for sparc64" +TEMP_INITRD_XZ = $(TEMP)/initrd.xz + +$(TEMP_INITRD_XZ): $(TEMP_INITRD) + # xz options ensure that kernel's decoder can handle the image. + zcat $(TEMP_INITRD) | xz --check=crc32 --lzma2=dict=512KiB > $(TEMP_INITRD_XZ) + .PHONY: arch_boot -arch_boot: +arch_boot: $(TEMP_INITRD_XZ) gzip -c $(TEMP)/System.map-$(KERNELIMAGEVERSION) >$(TEMP)/System.map.gz - tftpboot.sh $(TEMP_KERNEL) $(TEMP)/System.map.gz $(TEMP_INITRD) $(TEMP_BOOT) + tftpboot.sh $(TEMP_KERNEL) $(TEMP)/System.map.gz $(TEMP_INITRD_XZ) $(TEMP_BOOT) diff --git a/debian/control b/debian/control index 5eef76a..949e14b 100644 --- a/debian/control +++ b/debian/control @@ -1,4 +1,4 @@ -Source: debian-installer +source: debian-installer Section: devel Priority: optional Maintainer: Debian Install System Team @@ -78,6 +78,8 @@ Build-Depends: # m68k uses it for atari floppies cpio, # For creating initramfs images. + xz-utils, +# For compressing initramfs images. devio [armeb armel], # Tool to read and write from block devices, used to byteswap # kernels and add nslu2 boot magic. -- 1.7.9
Re: Sparc netboot image is too large to boot (again)
On Sat, Feb 18, 2012 at 09:58:25PM +0100, Bastian Blank wrote: > On Fri, Feb 17, 2012 at 08:00:24PM +0000, Jurij Smakov wrote: > > Attached patch implements the switch. It's not particularly elegant, > > but minimally intrusive, only affecting the building ot sparc's > > netboot image. I would like to commit it some time next week, unless > > anyone objects. > > I would do that for all architectures. There is no reason to use gzip on > any of them. I wouldn't want to break the netboot images for other arches by making changes which are not really required. Each architecture probably has some considerations I'm not necessarily aware of, so I would suggest to individual architecture porters to test xz compression out, paving the way to making it the default. Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120219115009.ga3...@wooyd.org
Bug#609466: console [tty0] enabled, bootconsole disabled
On Fri, Jan 21, 2011 at 07:01:47PM +0100, Tomten Larsson wrote: > > Hello, > I have tested this image. > This build finished at Fri Jan 21 09:12:54 UTC 2011. > http://cdimage.debian.org/cdimage/daily-builds/daily/arch-latest/sparc/iso-cd/debian-testing-sparc-netinst.iso > > It still has the same problem. > > console [tty0] enabled, bootconsole disabled > > after this nothing more happens Well, this is hardly surprising. As I wrote in my previous message, we need to build a patched kernel to see whether it fixes the problem. However, I did not recognize that you cannot even boot the installer, so you are unable to build/test such kernel. One thing you can try is disable aty framebuffer to see whether it will allow you to boot the installer. After you see the installer welcome screen with the 'boot:' prompt, instead of simply pressing Enter, try entering 'install video=atyfb:off' (without quotes) and then press Enter. That should boot the kernel with framebuffer disabled. Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110122180206.ga2...@droopy.oc.cox.net
sparc errata for squeeze
Hello, Here are a couple of sparc-specific items which are unlikely to be resolved for r0, so please add them to errata: 1. niu network driver, required by newer T2+ sparc systems is not included in d-i, making it impossible to install on them using any method which relies on early network setup. Driver is included in kernel packages, so network is expected to operate normally after installation is completed. This is tracked in bug 608516, I hope to push the fix into r1. 2. Systems using aty graphics cards (for example, Ultra 10) may not boot the installer correctly, with kernel freezing early in the boot stage with last message "console [tty0] enabled, bootconsole disabled". This is a kernel bug, for which a patch is available, however I only seen one report of it fixing the issue, and it's unlikely to be included into the kernels released in r0. The issue may be worked around by adding a kernel boot parameter 'video=atyfb:off' to turn off the framebuffer during boot, which allows the installer (and regular kernel) to be booted on such systems. This was reported in bug 609466, will try to build the patched kernel to verify that it fixes it, and push into r1. Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110123131833.ga4...@droopy.oc.cox.net
Re: SPARC daily d-i builds
On Wed, Jan 12, 2011 at 09:16:18AM +0100, Julien Cristau wrote: > On Wed, Jan 12, 2011 at 07:23:29 +0000, Jurij Smakov wrote: > > > That got me thinking... We had this T2K box (zee) donated a while ago > > (http://lists.debian.org/debian-sparc/2009/08/msg9.html, for > > example), and it shows up as a buildd on a machine page, however I > > don't really remember ever seeing a package built on it. Can you > > please check what's its state and whether it can be used for these > > builds? > > > Last I knew, zee was being used as a buildd for the prospective sparc64 > port. > http://buildd.debian-ports.org/fetch.php?&pkg=scid&ver=1%3A4.2.2.cvs20110111-1&arch=sparc64&stamp=1294809847&file=log&as=raw > suggests that's still true. Fair enough. However, the machine has 32 CPUs and 32G of RAM, so even while used as a buildd, significant amounts of resources are probably just sitting idle. I see that it only has 72G of disk though (2x72G disks in RAID-1 config), however we could probably solicit donation of the disks needed or even convince DPL to throw some money at it. Hence the questions: 1. Is there any virtualization solution on zee which would allow us to run more than a single buildd on it? 2. If disks are the bottleneck, what are the part numbers for them? I guess that if we would get another two, we could simply make another RAID-1 volume out of that, to not interfere too much with existing setup? 3. Who is the right person to talk to about arranging various reconfigurations like that? Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110123220540.ga4...@droopy.oc.cox.net
Bug#608516: Fix committed, tagging as pending
tags 608516 pending thanks I've committed the fixes for both sparc and sparc64 packages: http://git.debian.org/?p=d-i/linux-kernel-di-sparc-2.6.git;a=commit;h=255af4217d51fc32eeb0b844884a7244656b0675 http://git.debian.org/?p=d-i/linux-kernel-di-sparc64-2.6.git;a=commit;h=38858cb83eef0324d6bba803723c39e99bde77d6 They are too late for squeeze's initial release, most likely will be included in the first point release (r1). Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110123233436.ga5...@droopy.oc.cox.net
Bug#609466: console [tty0] enabled, bootconsole disabled
On Sun, Jan 23, 2011 at 01:23:52PM +0100, Tomten Larsson wrote: > > Hello, > Thanks for the quick reply. > entering 'install video=atyfb:off' (without quotes) > did make the trick. > > After choosing language, keyboard, > > It can't find the cdrom. > ^^^ > > lsmod indicates the cdrom driver is loaded. Which driver do you see loaded? I think your system needs cmd64x IDE driver which, unfortunately, is missing from the d-i's initrd. There is now a bug filed for it (http://bugs.debian.org/610906), and we'll try to make sure that it is fixed for the first point release (it's too late for squeeze's original release). Unfortunately, I can't offer much in terms of workaround at this point. If you can set up netbooting for your machine, you should be able to install this way as this will not require accessing the CD-ROM to retrieve udebs, they will be downloaded off the network, and udebs should contain theh pata_cmd64x driver that you need. Best regards, > bug=609733 indicates that the issue with the cdrom is solved, but on my > system it doesn't work. > > I tried all the mentioned images, in bug 609466 > > Best regards > Otto > > > > Date: Sat, 22 Jan 2011 18:02:06 + > > From: ju...@wooyd.org > > To: tomten1...@live.se > > CC: 609...@bugs.debian.org; ju...@wooyd.org; fberc...@xs4all.nl; > > el...@debianpt.org > > Subject: Re: console [tty0] enabled, bootconsole disabled > > > > On Fri, Jan 21, 2011 at 07:01:47PM +0100, Tomten Larsson wrote: > > > > > > Hello, > > > I have tested this image. > > > This build finished at Fri Jan 21 09:12:54 UTC 2011. > > > http://cdimage.debian.org/cdimage/daily-builds/daily/arch-latest/sparc/iso-cd/debian-testing-sparc-netinst.iso > > > > > > It still has the same problem. > > > > > > console [tty0] enabled, bootconsole disabled > > > > > > after this nothing more happens > > > > Well, this is hardly surprising. As I wrote in my previous message, we > > need to build a patched kernel to see whether it fixes the problem. > > However, I did not recognize that you cannot even boot the installer, > > so you are unable to build/test such kernel. > > > > One thing you can try is disable aty framebuffer to see whether it > > will allow you to boot the installer. After you see the installer > > welcome screen with the 'boot:' prompt, instead of simply pressing > > Enter, try entering 'install video=atyfb:off' (without quotes) and > > then press Enter. That should boot the kernel with framebuffer > > disabled. > > > > Best regards, > > -- > > Jurij Smakov ju...@wooyd.org > > Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC > -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110123235615.ga5...@droopy.oc.cox.net
Bug#610906: Suggested patch
reassign 610906 debian-installer
tags 610906 patch
thanks
Attached is a patch which, I believe, should do the trick. The idea is
to add pata-modules to cdrom initrd on sparc, so that pata_cmd64x.ko
gets included. Complete list of new files resulting from pulling in
pata-modules:
-rw-r--r-- 1 jurij jurij 13744 Dec 10 18:44 pata_artop.ko
-rw-r--r-- 1 jurij jurij 13984 Dec 10 18:44 pata_cmd64x.ko
-rw-r--r-- 1 jurij jurij 23880 Dec 10 18:44 pata_it821x.ko
-rw-r--r-- 1 jurij jurij 8016 Dec 10 18:44 pata_jmicron.ko
-rw-r--r-- 1 jurij jurij 9512 Dec 10 18:44 pata_marvell.ko
-rw-r--r-- 1 jurij jurij 9432 Dec 10 18:44 pata_sch.ko
This is targeted for squeeze r1.
Best regards,
--
Jurij Smakov ju...@wooyd.org
Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC
>From be086ccd9915f029a5de5b7fc1829fdab80f0a67 Mon Sep 17 00:00:00 2001
From: Jurij Smakov
Date: Mon, 24 Jan 2011 21:31:29 +
Subject: [PATCH] Add pata-modules to cdrom config for sparc and sparc64.
---
build/pkg-lists/cdrom/sparc.cfg |1 +
build/pkg-lists/cdrom/sparc64.cfg |1 +
debian/changelog |8
3 files changed, 10 insertions(+), 0 deletions(-)
diff --git a/build/pkg-lists/cdrom/sparc.cfg b/build/pkg-lists/cdrom/sparc.cfg
index 116c905..763cf12 100644
--- a/build/pkg-lists/cdrom/sparc.cfg
+++ b/build/pkg-lists/cdrom/sparc.cfg
@@ -6,6 +6,7 @@ kbd-chooser
input-modules-${kernel:Version}
ide-modules-${kernel:Version}
+pata-modules-${kernel:Version}
usb-modules-${kernel:Version}
usb-storage-modules-${kernel:Version}
diff --git a/build/pkg-lists/cdrom/sparc64.cfg b/build/pkg-lists/cdrom/sparc64.cfg
index 116c905..763cf12 100644
--- a/build/pkg-lists/cdrom/sparc64.cfg
+++ b/build/pkg-lists/cdrom/sparc64.cfg
@@ -6,6 +6,7 @@ kbd-chooser
input-modules-${kernel:Version}
ide-modules-${kernel:Version}
+pata-modules-${kernel:Version}
usb-modules-${kernel:Version}
usb-storage-modules-${kernel:Version}
diff --git a/debian/changelog b/debian/changelog
index bc475ac..b0ce0e6 100644
--- a/debian/changelog
+++ b/debian/changelog
@@ -1,3 +1,11 @@
+debian-installer (20110124) UNRELEASED; urgency=low
+
+ [ Jurij Smakov ]
+ * Add pata-modules to cdrom config for sparc and sparc64.
+Closes: #610906.
+
+ -- Jurij Smakov Mon, 24 Jan 2011 21:27:36 +
+
debian-installer (20110106) unstable; urgency=low
* Use SpaceFun theme. Closes #603554.
--
1.7.2.3
Bug#610906: Fix committed, tagging pending
tag 610906 pending thanks Fix committed to git repo: http://git.debian.org/?p=d-i/debian-installer.git;a=commit;h=be086ccd9915f029a5de5b7fc1829fdab80f0a67 Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110130214206.ga5...@droopy.oc.cox.net
Another d-i errata item for sparc
Hello, Please commit the following to d-i errata for sparc: --8< Some sparc systems cannot be installed using CD-ROM Debian installer for Squeeze does not include PATA kernel drivers, which makes it impossible to complete the installation from CD media on systems which require these drivers to access the CD-ROM drive (for example, Ultra 10), as the installer will fail to detect it. The problem may be worked around by netbooting the installer, thus eliminating the need to access CD-ROM drive during installation. A fix for this problem will be included in the first point release. --8< I've committed the fix to git, bug is #610906. Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110202230619.ga3...@droopy.oc.cox.net
Bug#609466: Workaround nonfunctional (SunBlade 150)
On Sun, Feb 06, 2011 at 05:01:45PM +0100, Christoph Egger wrote: > Hi! > > Just a quick note: the workaround `video=atyfb:off` doesn't seem to > help on my SunBlade 150 (installer) If you already didn't, can you try cold-booting the installer, i.e. power-off the machine completely and then start again (not just reset it). -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110206164036.ga21...@droopy.oc.cox.net
Unofficial Squeeze installer image available
[bcc: debian-boot@lists.debian.org and some bug reporters] Hello, Over the last couple of days I've managed to build a functional Squeeze installer mini iso image for sparc, which addresses all known sparc-specific problems described in Squeeze errata. You can find the image, patches used to build it and a README file describing the bugs it's supposed to address (as well as rebuild instructions) at http://www.wooyd.org/debian/squeeze/ If you have encountered problems during Squeeze installation, please test this image and report the results, as a positive confirmation of fixes on a variety of different systems is essential for getting all these fixes into the first Squeeze point release. Best regards, -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110209214321.ga3...@droopy.oc.cox.net
Proposed d-i fixes on sparc for 6.0.1
Hello, I would like to push a couple of fixes into 6.0.1 which prevent people from completing the installation on a wide array of sparc machines: http://bugs.debian.org/608516: niu network driver is not included in the installer udebs. This prevents network detection during the installation on newer T2+ sparc systems. Proposed patch is committed to git: http://git.debian.org/?p=d-i/linux-kernel-di-sparc-2.6.git;a=commit;h=255af4217d51fc32eeb0b844884a7244656b0675 http://bugs.debian.org/610906 pata_cmd6x (and other pata drivers) are missing from installer initrd. This prevents CD-ROM drive detection during installation, making install from any CD media impossible (and, unfortunately, there is no simple workaround - the only way to avoid this is to netboot). Proposed patch is committed to git: http://git.debian.org/?p=d-i/debian-installer.git;a=commit;h=be086ccd9915f029a5de5b7fc1829fdab80f0a67 I would like your blessing to upload these changes to stable. Thanks. -- Jurij Smakov ju...@wooyd.org Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20110212214218.ga4...@droopy.oc.cox.net
