[Bug 248860] New: i386 installation locks up on driver load
https://bugzilla.novell.com/show_bug.cgi?id=248860 Summary: i386 installation locks up on driver load Product: openSUSE 10.2 Version: Final Platform: x86-64 OS/Version: Other Status: NEW Severity: Normal Priority: P5 - None Component: Installation AssignedTo: bnc-team-screening@forge.provo.novell.com ReportedBy: imotgm@yahoo.com QAContact: jsrain@novell.com When trying to install, using the SUSE boot disk, the installation proceeds to this point;
openSUSE installation program v2.1.17 (c) 1996-2006 SUSU Linux Products GmbH <<<
Starting udev... ok Loading basic drivers... ok Starting hardware detection... ok (If a driver is not working for you try booting with brokenmodules=driver_name.) Giga-byte CK8S Parallel ATA Controller (v2.5) drivers: amd74xx, pata_amd, sata_nv, generic loading amd74xx Giga-byte CK8S Serial ATA Controller (v2.5) drivers: sata_nv, generic loading sata_nv Integrated Technology Express IT/ITE8212 Dual channel ATA RAID controller drivers: it821x, pata_it821x loading it821x at which point it is a locked, dead duck. I've used brokenmodules=it821x, and brokenmodules=pata_it821x, and both at the same time. Absolutely no go from there. I've waited more than an hour, to see if something was still processing, but in the end had to hit the reset button. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 chrubis@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- AssignedTo|bnc-team- |kernel-maintainers@forge.provo.novell.com |screening@forge.provo.novell| |.com | -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 lmb@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |NEEDINFO Info Provider| |imotgm@yahoo.com ------- Comment #1 from lmb@novell.com 2007-02-26 07:23 MST ------- Can you boot with failsafe settings and attach the hwinfo output? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 ------- Comment #2 from imotgm@yahoo.com 2007-02-26 16:59 MST ------- Created an attachment (id=121170) --> (https://bugzilla.novell.com/attachment.cgi?id=121170&action=view) requested attachment I hope this works. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 ------- Comment #3 from imotgm@yahoo.com 2007-02-26 17:04 MST ------- If you're referring to openSUSE 10.2, the answer is an obvious no. We're talking about an inability to install, because the installation program can't load the drivers. I am currently running SUSE 10.1, and don't need to boot with failsafe, as it's running just fine. I've run hwinfo, and created a file "hwinfo.txt". It's been sent. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 imotgm@yahoo.com changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |NEW Info Provider|imotgm@yahoo.com | ------- Comment #4 from imotgm@yahoo.com 2007-03-23 04:31 MST ------- It's three days short of a month since I sent the hwinfo you asked for. Is anyone looking at this? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 ------- Comment #5 from gregkh@novell.com 2007-03-23 10:28 MST ------- No one was looking at it as the state was still stuck at NEEDINFO. Please, next time when you provide the needed information, check the box that states you did so so the state will change to something that pops back up on our queries. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 gregkh@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- AssignedTo|kernel- |teheo@novell.com |maintainers@forge.provo.nove| |ll.com | -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 teheo@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |NEEDINFO Info Provider| |imotgm@yahoo.com ------- Comment #6 from teheo@novell.com 2007-03-26 01:13 MST ------- Hello, hde is showing some errors during detection in opensuse 10.1. IDE's error handling is fragile and may lead to complete lock up depending on circumstances and of course luck. libata's driver (pata_it821x) is supposed to behave much better and should recover gracefully from most situations. * Does hde work correctly in 10.1? If so, can you post the result of 'hdparm -I /dev/hde' and 'smartctl -a /dev/hde'? * If you give brokenmodules=it821x, the installer hangs after printing "pata_it821x", right? * During boot (before hang), you can switch to kernel message console by pressing 'alt-f4'. What's the kernel's last words? Thanks. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 ------- Comment #7 from imotgm@yahoo.com 2007-03-26 09:14 MST ------- 1. Hde is working on 10.1, I write to it every day. The it821x module, or the library it calls has had problems on SUSE since the MB was new. I need to massage it a bit after each kernel update, but it works fine once I get it soothed. fatman:~ # hdparm -I /dev/hde /dev/hde: ATA device, with non-removable media Model Number: ST3160023A Serial Number: 5LJ1SXJM Firmware Revision: 3.01 Standards: Used: ATA/ATAPI-6 T13 1410D revision 2 Supported: 6 5 4 3 Configuration: Logical max current cylinders 16383 65535 heads 16 1 sectors/track 63 63 -- CHS current addressable sectors: 4128705 LBA user addressable sectors: 268435455 LBA48 user addressable sectors: 312581808 device size with M = 1024*1024: 152627 MBytes device size with M = 1000*1000: 160041 MBytes (160 GB) Capabilities: LBA, IORDY(can be disabled) bytes avail on r/w long: 4 Queue depth: 1 Standby timer values: spec'd by Standard, no device specific minimum R/W multiple sector transfer: Max = 16 Current = 16 Recommended acoustic management value: 128, current value: 0 DMA: mdma0 mdma1 mdma2 udma0 udma1 udma2 udma3 udma4 *udma5 Cycle time: min=120ns recommended=120ns PIO: pio0 pio1 pio2 pio3 pio4 Cycle time: no flow control=240ns IORDY flow control=120ns Commands/features: Enabled Supported: * READ BUFFER cmd * WRITE BUFFER cmd * Host Protected Area feature set * Look-ahead * Write cache * Power Management feature set Security Mode feature set * FLUSH CACHE EXT command * Mandatory FLUSH CACHE command * 48-bit Address feature set SET MAX security extension Security: Master password revision code = 65534 supported not enabled not locked not frozen not expired: security count not supported: enhanced erase HW reset results: CBLID- above Vih Device num = 0 determined by the jumper Checksum: correct -------------------------------------------------------------------------------- fatman:~ # smartctl -a /dev/hde -bash: smartctl: command not found -------------------------------------------------------------------------------- 2. Using brokenmodules=it821x the installer hangs at; loading pata_it821xudevd-event[1504]: wait_for_sysfs: waiting for '/sys/devices/pci0000:00/0000:00:00:0e.0/0000:02:0c.0/host2/target2:0:0/2:0:0:0/ioerr_cnt' failed 3. Alt+F4 reveals a still running system, with the following abbreviated output; ata3: port is slow to respond, please be patient (Status 0x0) ata3: port failed to respond (30 secs, Status 0xd0) ata3: soft resetting port ata3: EH ata3.00: speed down requested but no transfer node left ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x2 frozen ata3.00: cmd 29/00:08:00:00:00/00:00:00:00:00/e0 tag 0cdb 0x0 data 4096 in res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout) This repeats a number of times, then does the same with ata4. There is a jump, which moves too fast to copy, wherein there are references to scsi errors on sda, and sdb, which are, in fact, hde and hdg. I can tell because of the partitioning structure. There is a reset, of sorts with; ata3: limiting speed to DMA25 which works its way down the speed ladder to the above, and loops again. It does this for about 20 minutes, then finally gives up on the it8212, and proceeds until it reaches a point where it can't find the install program and asks if it should continue manually. A yes gets a language choice, and looks promising, until it says it can't find the installation media. It asks for a path, which I give, then reports "can't find the image" and goes into an endless loop to enter the /path/to/image, until I pull the plug. I'm currently running SuSE 8.2, SUSE 10.0, and SUSE 10.1, on this machine, as well as Mandrake 10.1, Mandriva 2007, PCLOS, and Slackware 9.1. Only the SUSE clan has problems with the it8212. It's either an odd module in the SUSE kernel, or the library it calls. I first came across a form of this with SUSE 9.3, but it's only 10.2 that loses it completely, and can't seem to tell the difference between PATA, and SATA, to boot. To clarify, I've had over twenty distros on this machine, at one time or another, over the last two years, and only SUSE shows this continuing problem with the it8212, A couple had no module for it at all, but the ones that did all worked from the get go. SUSE has been my main working platform since 6.4, so I'm not shopping for a new one. We need to get this one fixed, so it doesn't end up still broke in 10.3. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 ------- Comment #8 from teheo@novell.com 2007-04-03 00:38 MST ------- Hmmm... the kernel isn't dead. It's just the driver which is stalling the installation. Well, that's a good news. :-) Devices appearing as /dev/sdX is the expected behavior if libata drivers are used and all ATA disks will appear as /dev/sdX in the future. Your interrupt delivery is working judging from where the driver is stuck but there seems to be transmission problems when transferring data over the wire thus resulting in command timeouts (READ DMA commands are timing out). * How are the two harddisks connected? I think there is hardware problem involved here. From the 10.1 hwinfo you posted, it seems you're using 80c cable and connecting one device per channel. On a PATA cable, there are two connectors, which one did you use to connect the disks (the one at the end or the other in the middle)? * When testing opensuse 10.2, I suppose you put the media into the pioneer dvd-rw, right? * What kind of massaging do you need to get 10.1 working? * smartctl is in the package smartmontools. Please install it and report the result of 'smartctl -a /dev/hde' Thanks and please don't forget to clear NEEDINFO when answering. I have a bunch of errors to track and the bug status helps me a lot. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 imotgm@yahoo.com changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |ASSIGNED Info Provider|imotgm@yahoo.com | ------- Comment #9 from imotgm@yahoo.com 2007-04-03 04:52 MST ------- "all ATA disks will appear as /dev/sdX in the future." Who's stupid idea is that? Is hdparm going to be modified so that it is still usable with ATA drives that identify themselves as /dev/sdx? "How are the two harddisks connected?" Like normal, jumpered master, using the end connector. "When testing opensuse 10.2, I suppose you put the media into the pioneer dvd-rw, right? It's the only optical device. "What kind of massaging do you need to get 10.1 working?" I enter a driver parameter that makes no sense, so the kernel does not attempt to load the driver during boot. Once the system is running, I remove the parameter from modprobe.conf, then "modprobe it821x" which loads, and recognizes the two hard drives. Subsequent reboots are done with no issues. If I install an updated kernel, and don't add the nonsense parameter to modprobe.conf, the next reboot will lock the system when it tries to load the driver. "report the result of 'smartctl -a /dev/hde' fatman:~ # smartctl -a /dev/hde smartctl version 5.33 [i686-pc-linux-gnu] Copyright (C) 2002-4 Bruce Allen Home page is http://smartmontools.sourceforge.net/ Warning! Drive Identity Structure error: invalid SMART checksum. === START OF INFORMATION SECTION === Device Model: ST3160023A Serial Number: 5LJ1SXJM Firmware Version: 3.01 User Capacity: 160,041,885,696 bytes Device is: In smartctl database [for details use: -P show] ATA Version is: 6 ATA Standard is: ATA/ATAPI-6 T13 1410D revision 2 Local Time is: Tue Apr 3 05:11:23 2007 CDT SMART support is: Unavailable - device lacks SMART capability. A mandatory SMART command failed: exiting. To continue, add one or more '-T permissive' options. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 teheo@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |NEEDINFO Info Provider| |imotgm@yahoo.com ------- Comment #10 from teheo@novell.com 2007-04-03 05:42 MST ------- * Oh well, it's because libata uses SCSI mid and high layer drivers instead of IDE ones. The general direction is that users don't need to care about what device node a device appears as. Filesystems are identified by unique device IDs or LABELs so what driver drives the device should make no difference from user's POV. This actually works quite well and actually helpful with hotplugging and all. All the standard tools including hdparm and smartmontools already work with /dev/sdX. * Yeap, you got cabling alright. * Something is very off. Your drive returned corrupt SMART data. Module loading order shouldn't matter. The only possibility I can think of is interaction with another device sharing some resource. it821x and geforce 6600 are sharing IRQ 193 which looks suspicious. Can you please try the followings? 1. Give 'irqpoll' kernel parameter and see whether things work. Please also give a shot at 'acpi=noirq'. 2. Booting with 'brokenmodules=it821x,pata_it821x' should make the installer ignore the controller and let the installer finish to load. After the installer is loaded, press 'ctrl-alt-f2', this should give you a command console. In the command console, you can mount a USB stick or existing partition to /mnt and record info there. Please report 'var/log/boot.msg', 'hwinfo --all', 'cat /proc/interrupts'. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 ------- Comment #11 from teheo@novell.com 2007-04-03 05:43 MST ------- Ah... one more thing to try. Please give 'brokenmodules=nvidia' and see what happens. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 ------- Comment #12 from imotgm@yahoo.com 2007-04-06 02:00 MST ------- Created an attachment (id=129540) --> (https://bugzilla.novell.com/attachment.cgi?id=129540&action=view) hwinfo from Ctl+lt+f2 during installation -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 ------- Comment #13 from imotgm@yahoo.com 2007-04-06 02:01 MST ------- Created an attachment (id=129541) --> (https://bugzilla.novell.com/attachment.cgi?id=129541&action=view) from Ctl+Alt+F2 during installation -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 ------- Comment #14 from imotgm@yahoo.com 2007-04-06 02:06 MST ------- Created an attachment (id=129542) --> (https://bugzilla.novell.com/attachment.cgi?id=129542&action=view) boot.msg using it821x module -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 ------- Comment #15 from imotgm@yahoo.com 2007-04-06 02:08 MST ------- Created an attachment (id=129543) --> (https://bugzilla.novell.com/attachment.cgi?id=129543&action=view) boot.msg using pata_it821x module -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 ------- Comment #16 from imotgm@yahoo.com 2007-04-06 02:29 MST ------- Created an attachment (id=129549) --> (https://bugzilla.novell.com/attachment.cgi?id=129549&action=view) 10.2 /var/log/messages for Apr 5. I added spaces for each time syslog started, to seperate into easier to read blocks. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 imotgm@yahoo.com changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |ASSIGNED Info Provider|imotgm@yahoo.com | ------- Comment #17 from imotgm@yahoo.com 2007-04-06 02:34 MST ------- "All the standard tools including hdparm and smartmontools already work with /dev/sdX Oh, really? [From SUSE 10.1] fatman:~ # hdparm -c1 -d1 -u1 /dev/hde /dev/hde: setting 32-bit IO_support flag to 1 setting unmaskirq to 1 (on) setting using_dma to 1 (on) IO_support = 1 (32-bit) unmaskirq = 1 (on) using_dma = 1 (on) fatman:~ # fatman:~ # hdparm -c1 -d1 -u1 /dev/hdg /dev/hdg: setting 32-bit IO_support flag to 1 setting unmaskirq to 1 (on) setting using_dma to 1 (on) IO_support = 1 (32-bit) unmaskirq = 1 (on) using_dma = 1 (on) fatman:~ # [From SUSE 10.2] fatman:~ # hdparm -c1 -d1 -u1 /dev/sda /dev/sda: setting 32-bit IO_support flag to 1 HDIO_SET_32BIT failed: Invalid argument setting unmaskirq to 1 (on) HDIO_SET_UNMASKINTR failed: Inappropriate ioctl for device setting using_dma to 1 (on) HDIO_SET_DMA failed: Inappropriate ioctl for device IO_support = 0 (default 16-bit) HDIO_GET_UNMASKINTR failed: Inappropriate ioctl for device HDIO_GET_DMA failed: Inappropriate ioctl for device fatman:~ # fatman:~ # hdparm -c1 -d1 -u1 /dev/sdb /dev/sdb: setting 32-bit IO_support flag to 1 HDIO_SET_32BIT failed: Invalid argument setting unmaskirq to 1 (on) HDIO_SET_UNMASKINTR failed: Inappropriate ioctl for device setting using_dma to 1 (on) HDIO_SET_DMA failed: Inappropriate ioctl for device IO_support = 0 (default 16-bit) HDIO_GET_UNMASKINTR failed: Inappropriate ioctl for device HDIO_GET_DMA failed: Inappropriate ioctl for device fatman:~ # Doesn't look that way from here. As you can see, "brokenmodules=it821x irqpoll" actually got this puppy to install. That's the last of the good news. The install froze twice, and had to be completed with an upgrade repair. When it finally rebooted, using the install CD to boot, (grub never got installed) it rebooted into text mode, with a garbled Yast screen that was all but unreadable. I finished the install from that, but only my long history with SUSE got me through. If I was a total newbie, it would have been too much. This is the worst installation experience I've ever had, with any Linux OS. (7 years with SUSE, 20+ distros tried, unknown total number of installations, but it's a lot) While running from the installation, and after, that first reboot, using, I would guess, the kernel from the install CD, I got some of the things you asked for, which I'm sending as attachments. Once I was able to boot from grub, without the install CD, the brokenmodules trick didn't work, and SUSE 10.2 loads the it821x module, goes into a near panic, sees /dev/hdg, but not /dev/hde. Using Yast, I was able to switch to the pata_it821x module, but it does a bit of a panic also, and ends up seeing neither of the drives. I'm including, as attachments, "boot.msg" for both the it821x, and the pata_it821x bootups, as well as the "messages" from the only boot so far, with the pata_it821x. The 10.2 system locked up solid, when I tried to read the messages, to the point that I had to do a hard reset, and booted back to 10.1, to retreive it. FYI, I have seperate /home partitions for 10.2, and 10.1, so there is no cross contamination betweem messages. As you seemed to think there might be a physical problem with the drive on hde, I switched the drive with a new 300GB Seagate. Smartctl has the same problem with it when connected to the it8212, as seen by SUSE. SUSE 10.2's version of it821x, and pata_it821x, doesn't like it any better than the 160GB Seagate it replaced. Booting to Mandriva 2007 gets no panic type messages during the boot process, and once booted recognizes the drive just fine, as /dev/hde. Reads, and writes, are normal, without issue, as the are on SUSE 10.1. As it made no difference whatever, in solving this particular problem, I've re-installed the 160GB Seagate, as it has links to it from the other existing OSs, including SUSE 10.1. Now, my question is, being as I've actually gotten 10.2 to install, however crappy that installation process went, do we continue with the it8212 problem here, or do I have to file another bug report? Jus because a work-around is found, it doesn't change the fact that there's something wrong with SUSE's modules for this controller, that does not exist in the modules used by other distros. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 ------- Comment #18 from imotgm@yahoo.com 2007-04-06 02:48 MST ------- Forgot to mention, when doing Ctl+Alt+F2 during installation, "cat /var/log/boot.msg" gave error "no such file". -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 imotgm@yahoo.com changed: What |Removed |Added ---------------------------------------------------------------------------- Attachment #129541|from Ctl+Alt+F2 during |interrupts from Ctl+Alt+F2 during installation description|installation | -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 teheo@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |NEEDINFO Info Provider| |imotgm@yahoo.com ------- Comment #19 from teheo@novell.com 2007-04-06 05:41 MST ------- Fully agreed that was one crappy installation experience. Kudos to you for getting through. * regarding hdparm: Many control knobs including manual transfer mode configuration is not supported yet. It's most likely that we're gonna drop some of them. But the command issue interface is compatible and features relying on direct command execution works - e.g. -I (-i in devel tree), -Y, -y. smartmontools should work fully. These things are in progress and that's why the IDE drivers are still default for PATA drives. * Yeap, brokenmodules is only effective during installation. * No need to file another bug report and I don't think your drives are broken. Both devices can't be broken the same way and as you said they work fine on other distributions. Yet, I still don't think it821x drivers are to blame. it821x controllers are fairly standard IDE devices and the driver hasn't changed much. I'm sure Mandravia or any other distributions are shipping almost identical drivers, so something else is involved and that's why I'm scratching head here. By doing the following, you can prevent it821x drivers from loading during boot. # echo "STARTMODE='off'" > "/etc/sysconfig/hardware/hwcfg-bus-pci-0000:02:0c.0" You should have clean boot and can load it821x or pata_it821x afterward using modprobe. 1. After doing the above such that no driver is loaded for the controller, load pata_it821x and report /var/log/messages right after kernel starts complaining. 2. Please post kernel message from mandravia 2007. It would be great if you can give a shot to pata_it821x there too as they produce much better error messages. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 ------- Comment #20 from imotgm@yahoo.com 2007-04-06 09:56 MST ------- # echo "STARTMODE='off'" > "/etc/sysconfig/hardware/hwcfg-bus-pci-0000:02:0c.0" If you looked at the "messages" you'll note that the last boot with the pata_it821x took hours. It's still set that way, as it locked up before I had a chance to change it. I don't look forward to doing that again. I mounted the 10.2 / partition on 10.1 and made the file in "/mnt/suse/etc/sysconfig/hardware/" using "konqueror -superuser mode". I suspect I could have done a "chroot /mnt/suse" and used the command line, as above, or just added "/mnt/suse" to the above path, without the "chroot" but this way I could see what I was doing in reference to the other files already there. I like poking around in places I don't normally spend much time in, once I've got some idea what goes on in there. A little off topic, but for future changes, can one do a "chroot" as stated above, from SUSE 10.1 run level 3, and then run 10.2 Yast, either in text mode, or after an "init 5" from 10.2's GUI? I haven't done enough with "chroot" to know if this is possible, without risking some damage to 10.1, but it sounds like fun. :-) I'm right in the middle of setting up an Epson all-in-one scanner printer in a Windows VM, on my SUSE 10.1 host, as Epson has not yet issued Linux drivers for this model yet, so I'll try booting 10.2 when I'm done, and get back to you, with the results, then. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 ------- Comment #21 from teheo@novell.com 2007-04-06 10:14 MST ------- You can probably disable the it821x controller from BIOS menu. But mounting it from 10.1 and writing the file should be fine too. I dunno much about konqeror (have been a gnome user for around eight years now) but as long as the file containing "STARTMODE='off'" (without the double quotes) is created, it should work. Regarding chroot, it should mostly work but kernel API changes so you might meet some weird problems (udev not working properly, X not starting up because certain kernel module cannot be loaded, etc...). Even if things go wrong, your system should be fine after clean reboot. But I can't guarantee anything. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 ------- Comment #22 from imotgm@yahoo.com 2007-04-07 01:15 MST ------- OK. Rebooted 10.2. The STARTMODE='off' didn't do anything; pata_it821x loaded, and boot took forever, again. Looking at other hwcfg-bus-pci-* files present, I'm wondering if the the driver names shouldn't be included also, before the STARTMODE='off' entry? I'm sending "messages" from 10.2, and renamed "Mdv.messages" from Mandriva. I tried to load pata_it821x in Mandriva with the following results; [root@fatman imotgm]# modprobe pata_it821x FATAL: Module pata_it821x not found. I tried the "chroot" trick and started Yast in text mode, but it lacked modules, or something, and wouldn't actually run. No devices in /dev so no mounts possible. All in all, a "no harm, no foul" experience. :) -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 ------- Comment #23 from imotgm@yahoo.com 2007-04-07 01:20 MST ------- Created an attachment (id=129751) --> (https://bugzilla.novell.com/attachment.cgi?id=129751&action=view) 10.2 "messages" after STARTMODE='off' -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 ------- Comment #24 from imotgm@yahoo.com 2007-04-07 01:21 MST ------- Created an attachment (id=129752) --> (https://bugzilla.novell.com/attachment.cgi?id=129752&action=view) "messages" file from Mandriva 2007 -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 ------- Comment #25 from imotgm@yahoo.com 2007-04-07 01:33 MST ------- # fdisk -l in 10.2 shows /dev/hda, /dev/hdc, and /dev/sdd (my 160GB usb hard drive) only. Both it821x, and pata_it821x are loaded, as shown by lsmod, but 10.2 no longer sees either drive on the it8212 controller. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 imotgm@yahoo.com changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |ASSIGNED Info Provider|imotgm@yahoo.com | ------- Comment #26 from imotgm@yahoo.com 2007-04-07 01:35 MST ------- forgot to reset the needinfo flag. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 teheo@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |NEEDINFO Info Provider| |imotgm@yahoo.com ------- Comment #27 from teheo@novell.com 2007-04-07 04:03 MST ------- Weird, as long as the file name matches the device ID and it contains the line, it should work, hmmm, unless you have it loaded in initrd. This is "Load Module in initrd" option in yast->hardware->disk controllers->it8212 controller. If it's checked, uncheck it and press apply, initrd will be generated without it821x drivers. If it's already unchecked, please double check the filename and the content is correct. You can find out the PCI device ID by running 'lspci -nn'. If booting is too painful, I suggesting disabling the controller in the BIOS menu if possible or disconnect the disks physically. In the Mandrivia boot log, the kernel is 2.6.17 so pata_it821x isn't there yet. IRQ routing is different but I don't think that's significant as IRQ seems to be working in the suse kernel too. Hmmm... Okay, this is interesting. The mdv kernel is choking on harddrives too. hdg: 390721968 sectors (200049 MB) w/8192KiB Cache, CHS=24321/255/63, BUG hdg:hdg: recal_intr: status=0x51 { DriveReady SeekComplete Error } hdg: recal_intr: error=0x04 { DriveStatusError } hdc: dma_intr: status=0x51 { DriveReady SeekComplete Error } hdc: dma_intr: error=0x84 { DriveStatusError BadCRC } ide: failed opcode was: unknown hdc: dma_timer_expiry: dma status == 0x60 hdc: DMA timeout retry hdc: timeout waiting for DMA hdg is attached to it821x and hdc to sata_nv. BadCRC and DMA timeout expiry indicate transmission failure on the IDE cables. It seems like something is wrong system-wide. IDE layer sometimes reacts okay to such error conditions, at other times, it can lock the machine up hard. That's one of the reasons why new libata drivers are being implemented. libata error handling is reliable but tends to take a long time under certain conditions - being worked in the devel tree. Well, sometimes, drivers don't even matter because some IDE controllers just hold PCI bus and never releases on certain error conditions. No software can save the machine from that. 10.2 reacting especially badly might be caused not by kernel changes but by userland changes. Automounting, indexing kind of stuff. As you can see from the 10.2 boot.msg w/ pata_it821x, the drive doesn't spew error messages during detection phase and reading the partition proceeds without any problem. So, the driver and controller is working. It's only later that things go bad. In older distributions, after detection nothing much really happens, but we're doing a lot of automatic things lately. Anyways, you need to fix the root cause and my current best bet is power supply. I think it's either overloaded or dying out (they do, really). Please try to lessen power consumption by disconnecting some devices or try to connect them differently. It would be best if you've got an extra powersupply to try out. They're really cheap these days. If it's not PSU, please scream at me. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 imotgm@yahoo.com changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |ASSIGNED Info Provider|imotgm@yahoo.com | ------- Comment #28 from imotgm@yahoo.com 2007-04-08 02:06 MST ------- 7 00:49:06 fatman kernel: ide1: reset: success 7 00:49:07 fatman kernel: hdc: set_drive_speed_status: status=0x58 { DriveReady SeekComplete DataRequest } 7 00:49:07 fatman kernel: ide: failed opcode was: unknown Does not the above indicate that the errors on hdc were because of timing issues in the order of loading drivers? FYI hdc is attached to amd74xx, not sata_nv. I have four IDE controllers, for eight possible drives total; four on amd74xx, and four on the it821x. There are two SATA controllers for four possible drives attached as two to sata_nv, and two to sata_sil. The power supply is new. I contacted Antec and gave them all the specs for the motherboard, and the video card. I told them I would most likely end up with all of the possible drive connections filled, and wanted their recommendation for the best unit they made, to do the job. They recommended a 400 watt top of the line unit, I purchased the 500 watt version of the same. As I only have five of the possible twelve connections filled, I highly doubt the the PSU is overloaded, and you'd have to beat me with a club to convince me that Antec makes anything other than quality products. I've had cheap OEM puwer supplies fail, rather early in most cases, but I've never had a problem with any of the Antec replacement units, some of which are eight to ten years old, have been in multiple computers, and are still in use. I'm also a bit of a fanatic about doing complete tear downs of my computers at least every four to six months, to clean and lubricate all fans, clean all connectors, and replace all internal cabling. I try to schedule the tear downs for every four months, but the longest interval has been six months, give or take a week. I have tested over twenty distrbutions on this machine, with as many as twelve being present at one time. Only SUSE has noticable problems with this it8212 board, and only SUSE 10.2 has major problems with it. If you really think there's any possibility that some conflict with the automated stuff is the cause, rather than the driver itself, tell me what you'd like disabled, and how to do that, or how to remove it entirely, and I'd be more than happy to get that crap off of my machine. I hated it on Windows, and always turned off as much as I could. I loved Linux because there was none of that bullshit present, and I was in control of my machine, not some programmers ego boosting wet dream. Adding all that crap as default, to try to win over the Windows wienees, is the worst thing anyone could do to the Linux users who love Linux for what it is, in it's native form. Gee, I'm glad I got that last part off my chest!! ;-) Back to business, I'll check to see if the pata_it821x is in the initrd, and remove it if so. I've already checked the file name for typos, and all seems correct. Pulling the hard drives to get a quick boot is not a problem, but sort of defeats the purpose for booting. Without the drives in place, one can learn nothing about the problem. Another FYI, in reguards to the above. Until I added the fourth drive, none of the previous versions of SUSE had any problems with the it8212 card. They boot fine with only hde in place. The problem is that SUSE tries to make a RAID when hdg gets added. Mandriva has a parameter "it821x_noraid" in its invocation of the driver. I tried that in SUSE, and It coughs up an error "unknown parameter" and doesn't load the module at all. That's the parameter I refered to earlier as "makes no sense" (to SUSE) to prevent the loading of the module. I've added that to the 10.2 modprobe.conf, from 10.1, so hopefully it will boot quicker, with both drives in place, to change the "load in initrd". Before you ask, or recommend, BIOS already sets the card for JBOD. I'd still prefer the it821x to the pata_it821x. It's bad evough trying to know which USB drive gets which sdx as well as which SATA drive gets which sdx. Adding the normally hdx to the pot of sdx seems like a backward step to me. If my hdx drives are seen as sdx, will the order shift when a real SATA drive is added? I'll bet it does, and I'll bet grub will puke on it when it does. So to the kernel when root=/dev/sda3 makes no sense, because sda suddenly is seen as sdd, or sde. What happens to my hdx16, hdx17 and hdx18 partitions when they're suddenly sdx??, because scsi only allows 15 partitions? It's like going back to ide=scsi for the CD burners. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 teheo@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |NEEDINFO Info Provider| |imotgm@yahoo.com ------- Comment #29 from teheo@novell.com 2007-04-08 05:03 MST ------- * Sorry about the sata_nv confusion. The point I was trying to make was problems occur across two unrelated controllers. They're not sharing any resources and even if they do the CRC error bit's meaning is very specific, it's CRC error on the IDE bus. The most common causes for that are faulty cabling, power problem and broken hardware. Also, it's quite common that IDE controllers using the TF/BMDMA interface indicate transmission failure detected on the host side as timeout. So, those symptoms across multiple independent controllers make me think it's most likely a power problem. I've seen quite a few weird transmission failure cases caused by power issues and because when power is problematic the symptoms vary greatly, it's a good idea to rule it out in the early stage during debugging. Also, another problem with high power multi-rail PSU is that each rail is actually given less wattage compared to e.g. 300w single-rail PSU. So, overloading a rail is actually easier on multi-rail PSUs. I have some first hand experience with that and I always use 300~350w single-rail PSU to power my disk arrays. So, if it isn't too much trouble, please try to balance disks more evenly among power connectors and try to use another power supply to power the system. Mandravia is suffering the same problem. Its symptom is much less drastic but the root cause is still there. Your system isn't working too well currently. I'm suspecting PSU problem but it might as well be a it821x driver problem as the smart mode support on the driver seems a bit flaky. * About the automatic stuff, oh well, I guess we'll just have to go that way to better make it easier for less tech-savvy users. I can probably disable automounting by editing fstab and messing with udev rules. Auto indexing is done by beagle but I can't tell from here whether that's causing your problem. Just de-install beagle if you don't want it. Automatic indexing has been there for long time with unix tho, think about locatedb and how long it has been around unix. I'm an old dog regarding those fancy desktop stuff, so I'm afraid I won't be too helpful. * The parameter is named it8212_noraid not it821x_noraid. git tells me it hasn't changed since its first inclusion in 2005. * Due to the large number of disk ports including USB or what not, it's becoming more an more difficult to maintain static device naming. I also liked /dev/hdX static naming but it's just a fact that they're going away (they're not so static with native/legacy mode switching controllers anyway). So, it seems we'll have to learn the LABEL= and UUID tricks. They're really neat once you get used to them too. And yeap, you're right, adding discontinuous minors to scsi disks in on my todo list. * To sum up, I think it might be a hardware, more specifically, PSU related problem. Please try to rule that out. Suse, redhat, mandravia, what not, most distributions try to keep their kernels as close to the mainline as possible these days and to some extent userland too. According to your report, mandravia kernel is seeing the problems too and I think you're likely to encounter similar severity as suse in later versions. So, please apply some standard hardware debugging methods - swap drives, connectors, disconnect some of them, redistribute power connectors. It takes some time but you know a lot and I think you'll be able to figure out what's going on (or not). Thanks. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 imotgm@yahoo.com changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |ASSIGNED Info Provider|imotgm@yahoo.com | ------- Comment #30 from imotgm@yahoo.com 2007-05-14 07:37 MST ------- Did the PSU switch, cable switch etc. No change. Recompiled kernel without support for the pata_* modules, all is well. No boot hang ups, no problems with drive recognition. System works as expected, in all respects. Don't know what else to say, except the problems vanished along with pata_* modules. Without them present, the older it821x works beautifully; better than in the last three versions, I didn't have to do any massaging, tweaking, it just works. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 teheo@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |NEEDINFO Info Provider| |imotgm@yahoo.com ------- Comment #31 from teheo@novell.com 2007-06-07 02:26 MST ------- Sorry about the delay. Can you post dmesg and 'hdparm -I /dev/hdX' results of the working setup (with it821x)? Thanks. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 imotgm@yahoo.com changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |ASSIGNED Info Provider|imotgm@yahoo.com | ------- Comment #32 from imotgm@yahoo.com 2007-06-08 17:49 MST ------- Attachments sent. Regarding the concern about the dmesg errors, and crc errors from hdparm, it has nothing to do with my particular hardware. It effects all IDE raid controllers, and apparently some SATA controllers as well, and is caused by a bug in the kernel, as yet uncorrected. It is common to later 2.6 kernels, and is directly related to the effort to bring IDE controllers under the control of libata. I've been reading the kernal mailing list entries regarding this and their conclusion is the same as I expressed earlier; it's annoying, but not harmful, as it is self correcting upon reset. This brings us back to my original premiss; this is still highly experimental technology, is far from perfected, and has no place in a default installation kernel. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 ------- Comment #33 from imotgm@yahoo.com 2007-06-08 17:51 MST ------- Created an attachment (id=145135) --> (https://bugzilla.novell.com/attachment.cgi?id=145135&action=view) Requested dmesg output from recompiled kernel -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 ------- Comment #34 from imotgm@yahoo.com 2007-06-08 17:53 MST ------- Created an attachment (id=145136) --> (https://bugzilla.novell.com/attachment.cgi?id=145136&action=view) Requested output from "hdparm -I" with recompiled kernel -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 teheo@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |NEEDINFO Info Provider| |imotgm@yahoo.com ------- Comment #35 from teheo@novell.com 2007-06-08 23:09 MST ------- Can you post /var/log/boot.msg (dmesg result is missing the version string at the top)? IRQ routing in the working and non-working kernels are very different. With ide driver, the failure messages during probing can be due to the controller being in smart mode which isn't supported too well and might be irrelevant to the problem at hand. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 ------- Comment #36 from imotgm@yahoo.com 2007-06-11 17:55 MST ------- Created an attachment (id=145440) --> (https://bugzilla.novell.com/attachment.cgi?id=145440&action=view) requested /var/log/boot/msg -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 ------- Comment #37 from imotgm@yahoo.com 2007-06-11 18:01 MST ------- Created an attachment (id=145441) --> (https://bugzilla.novell.com/attachment.cgi?id=145441&action=view) dmesg output (new) -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 imotgm@yahoo.com changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |ASSIGNED Info Provider|imotgm@yahoo.com | ------- Comment #38 from imotgm@yahoo.com 2007-06-11 18:05 MST ------- Here's requested /var/log/boot.msg, and a new version of dmesg, with the version string at the top. Don't know why it didn't show on the first one. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 teheo@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |NEEDINFO Info Provider| |imotgm@yahoo.com ------- Comment #39 from teheo@novell.com 2007-06-11 21:31 MST ------- We've been barking at the wrong tree. The root cause seems to be IRQ routing not drivers themselves. ftp://ftp.suse.com/pub/projects/kernel/kotd/SL102_BRANCH/i386/kernel-default-2.6.18.8-SL102_BRANCH_20070530164615.i586.rpm Does the above kernel work out of the box? Thanks. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 imotgm@yahoo.com changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |ASSIGNED Info Provider|imotgm@yahoo.com | ------- Comment #40 from imotgm@yahoo.com 2007-06-12 01:48 MST ------- It doesn't work at all! Locks up solid as soon as it sees /dev/hdg, using IRQ 193, for the it8212. My kernel uses IRQ 18, and boots right up. I also let the updater install the latest kernel, vmlinuz-2.6.18.8-0.3-default, which also locks up at the same place as the BRANCH kernel. Can't give you any dmesg, or boot.msg, as it doesn't boot, so there is none. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 teheo@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |NEEDINFO Info Provider| |imotgm@yahoo.com ------- Comment #41 from teheo@novell.com 2007-06-12 01:56 MST ------- OIC, thanks for testing. Please post your .config file. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 ------- Comment #42 from imotgm@yahoo.com 2007-06-12 05:42 MST ------- Created an attachment (id=145578) --> (https://bugzilla.novell.com/attachment.cgi?id=145578&action=view) requested config file -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860 imotgm@yahoo.com changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |ASSIGNED Info Provider|imotgm@yahoo.com | ------- Comment #43 from imotgm@yahoo.com 2007-06-12 05:45 MST ------- I thought you might want that at some time. :) -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c44 Tejun Heo <teheo@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |teheo@novell.com AssignedTo|teheo@novell.com |ak@novell.com Status|ASSIGNED |NEW --- Comment #44 from Tejun Heo <teheo@novell.com> 2007-06-18 07:45:30 MST --- The winner seems to be CONFIG_PCI_MSI. Can you verify this by turning on CONFIG_PCI_MSI in your custom kernel? I don't really know about this part of the kernel. Reassigning to Andi. Andi, IRQ delivery fails if IRQ is routed like the following. ACPI: PCI Interrupt 0000:02:0c.0[A] -> Link [APC1] -> GSI 16 (level, low) -> IRQ 185 But, it works if IRQ is routed this way. ACPI: PCI Interrupt 0000:02:0c.0[A] -> Link [APC1] -> GSI 16 (level, low) -> IRQ 18 The failing dmesg is https://bugzilla.novell.com/attachment.cgi?id=129542 The working dmesg is https://bugzilla.novell.com/attachment.cgi?id=145441 Thanks. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c45 Frank Podgorski <imotgm@yahoo.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |ASSIGNED --- Comment #45 from Frank Podgorski <imotgm@yahoo.com> 2007-06-18 23:11:44 MST --- Are you asking me to recompile another kernel, or is there a way to do this with the existing kernel? The new page looks real nice. I like it a lot. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c46 --- Comment #46 from Frank Podgorski <imotgm@yahoo.com> 2007-06-18 23:18:28 MST --- Another thought, If asking for a recompile, would that be with the old sources, or the sources from the BRANCH kernel you asked me to try, or the current default kernel? If it makes a difference, please specify. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c47 --- Comment #47 from Tejun Heo <teheo@novell.com> 2007-06-18 23:26:35 MST --- Yeap, one more recompile please. The same version/configuration as your custom working kernel but with CONFIG_PCI_MSI turned on and I like the new look too. Easier on the eyes. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c48 Frank Podgorski <imotgm@yahoo.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |NEEDINFO Info Provider| |imotgm@yahoo.com --- Comment #48 from Frank Podgorski <imotgm@yahoo.com> 2007-06-19 00:01:07 MST --- I'm posting this from a PCLinuxOS at the moment, and have a couple of things cooking right now, so it will be a little while, but I'll get back as soon as I can reboot and do the dirty deed. ;-) My old eyes find it much easier. I do have a suggestion though. As the new page opens to the add comments box, and I do realize there is a login button at the bottom of the page, would it be that much work to add a login button right alongside the comments box? It would save a bit of scrolling each time one checks in. Like have it positioned like the commit button, only at the top of the box so as to not add confusion, or accidental hits on the wrong button. Maybe a bright color to make it more obvious which is which. Maybe even a sign; -------------------------- XXXXX <------Login here | | | | | | [negotiation] I'll trade you a kernel for a button ;) [/negotiation] -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c49 --- Comment #49 from Tejun Heo <teheo@novell.com> 2007-06-19 00:18:04 MST --- I'm as much a user as you are as far as the bugzilla is concrned, so I'm not in a position to negotiate. Also, I'm okay with the current layout. I think the best thing to do would be filing a bug report against Product: bugzilla and let the bugzilla guys decide. Just opening a new bug against bugzilla and copying the last paragraph should suffice. Thanks. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c50 Frank Podgorski <imotgm@yahoo.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |ASSIGNED Info Provider|imotgm@yahoo.com | --- Comment #50 from Frank Podgorski <imotgm@yahoo.com> 2007-06-19 04:42:53 MST --- OK, the new kernel works like the other working kernel, with the exception that it uses IRQ 193, instead of 18 for the it8212 controller. Is this what you expected? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c51 --- Comment #51 from Tejun Heo <teheo@novell.com> 2007-06-19 23:54:44 MST --- No, not really. I was suspecting the different IRQ routing affected IRQ delivery. Andi, do you have any idea what's going on? IRQ delivery doesn't work well with the packaged kernel but it does with the custom kernel compiled with config file in comment #42. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c52 Andi Kleen <ak@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |NEEDINFO Info Provider| |imotgm@yahoo.com --- Comment #52 from Andi Kleen <ak@novell.com> 2007-06-20 05:06:15 MST --- MSI is probably broken on the chipset Mainline has a Quirk for CK804. This is a NF3, it might have the same problem. I assume it works with pci=nomsi ? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c53 Frank Podgorski <imotgm@yahoo.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |ASSIGNED Info Provider|imotgm@yahoo.com | --- Comment #53 from Frank Podgorski <imotgm@yahoo.com> 2007-07-13 21:37:46 MST --- It's been a while so I thought I'd stop by. I see the needinfo flag. Are you needing something from me? It's not clear from the last two comments. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c54 --- Comment #54 from Tejun Heo <teheo@novell.com> 2007-07-13 22:00:32 MST --- I think Andi is asking whether giving "pci=nomsi" makes any difference. Thanks. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c55 --- Comment #55 from Frank Podgorski <imotgm@yahoo.com> 2007-07-14 14:36:04 MST --- With which kernel? The last one I compiled, the stock one, or the first one I compiled? You are speaking as a boot parameter, are you not? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c56 Andi Kleen <ak@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |NEEDINFO Info Provider| |imotgm@yahoo.com --- Comment #56 from Andi Kleen <ak@novell.com> 2007-07-17 11:40:46 MST --- Boot 10.2 DVD. Enter pci=nomsi on the boot screen. Report success/failure. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c57 Frank Podgorski <imotgm@yahoo.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |ASSIGNED Info Provider|imotgm@yahoo.com | --- Comment #57 from Frank Podgorski <imotgm@yahoo.com> 2007-07-18 04:45:30 MST --- Locks up at loading it821x -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c58 Stephan Lauffer <lauffer@ph-freiburg.de> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |lauffer@ph-freiburg.de --- Comment #58 from Stephan Lauffer <lauffer@ph-freiburg.de> 2007-10-08 02:27:38 MST --- opensuse-10.3 x86_64 hangs during the boot process here, too. Disabling (brokenmodules...) pata_it821x helps out but now I get endless warnings about /dev/hdc in /var/log/warn: Oct 8 10:20:27 lisa kernel: hdc: status error: status=0x58 { DriveReady SeekComplete DataRequest } Oct 8 10:20:27 lisa kernel: ide: failed opcode was: unknown On /dev/hdc there's a ide dvd connected. USB (a SD-Card in the cardreader) will not be recognized. I'll attach a hwinfo log (ASUS P5LD2 Mainboard with Intel 945P). With 10.2 I've had no problems. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c59 --- Comment #59 from Stephan Lauffer <lauffer@ph-freiburg.de> 2007-10-08 02:31:23 MST --- Created an attachment (id=176727) --> (https://bugzilla.novell.com/attachment.cgi?id=176727) hwinfo --storage-ctrl hwinfo --storage-ctrl from a machine which hangs during boot after loading pata_it821x -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c60 Tejun Heo <teheo@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |NEEDINFO Info Provider| |imotgm@yahoo.com --- Comment #60 from Tejun Heo <teheo@novell.com> 2007-10-08 02:39:09 MST --- Stephan, can you setup serial console and grab kernel log of failing boot with pata_it821x? If that's not possible, taking a photo of locked up screen with kernel messages (pass splash=0 as kernel boot parameter) is good too. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c61 Stephan Lauffer <lauffer@ph-freiburg.de> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |ASSIGNED Info Provider|imotgm@yahoo.com | --- Comment #61 from Stephan Lauffer <lauffer@ph-freiburg.de> 2007-10-08 09:02:49 MST --- Created an attachment (id=176826) --> (https://bugzilla.novell.com/attachment.cgi?id=176826) boot log with freeze after loading pata_it821x For this boot log I've taken the openSuSE-10.3 x86_64 GM DVD1 ("Installation"...) -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c62 --- Comment #62 from Andi Kleen <ak@novell.com> 2007-10-08 09:11:59 MST --- Does it work with pci=nomsi? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860 Andi Kleen <ak@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |NEEDINFO Info Provider| |lauffer@ph-freiburg.de -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c63 Stephan Lauffer <lauffer@ph-freiburg.de> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |ASSIGNED Info Provider|lauffer@ph-freiburg.de | --- Comment #63 from Stephan Lauffer <lauffer@ph-freiburg.de> 2007-10-08 09:23:03 MST --- no, doesn't help - I see there no changes. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c64 Andi Kleen <ak@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- AssignedTo|ak@novell.com |teheo@novell.com Status|ASSIGNED |NEW --- Comment #64 from Andi Kleen <ak@novell.com> 2007-10-08 09:29:00 MST --- Back to Tejun then. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c65 Tejun Heo <teheo@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |NEEDINFO Info Provider| |lauffer@ph-freiburg.de --- Comment #65 from Tejun Heo <teheo@novell.com> 2007-10-08 19:29:32 MST --- Hmmm... Stephan, thanks for the log. While installation system is trying to load modules, switch to console 4 (alt-f4), kernel should be logging messages there. Please report what the kernel says while pata_it821x is loaded and after it locks up. Photos taken with a digital camera should work. Thanks. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c66 Stephan Lauffer <lauffer@ph-freiburg.de> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |NEW Info Provider|lauffer@ph-freiburg.de | --- Comment #66 from Stephan Lauffer <lauffer@ph-freiburg.de> 2007-10-10 01:39:12 MST --- Created an attachment (id=177274) --> (https://bugzilla.novell.com/attachment.cgi?id=177274) log messages during load of pata_it821x Maybe we need to add some debug options to the kernel? Well, after loading the pata_it821x I need to reset the machine. Btw: Without loading it the dvdram ist not useable (and also no usb (only storage?) devices). If it makes sense I change the ide dvdram with another one... -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c67 Tejun Heo <teheo@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |NEEDINFO Info Provider| |imotgm@yahoo.com --- Comment #67 from Tejun Heo <teheo@novell.com> 2007-10-10 01:50:07 MST --- So, after printing those messages, what's the system status? Can you turn switch between virtual consoles (alt-fN) or turn on/off numlock? If the system still responds to inputs, please leave the system alone for a while (several minutes) and see whether anything happens. Thanks. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c68 Stephan Lauffer <lauffer@ph-freiburg.de> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |NEW Info Provider|imotgm@yahoo.com | --- Comment #68 from Stephan Lauffer <lauffer@ph-freiburg.de> 2007-10-10 02:59:23 MST --- It's dead - like I said before "I need to reset the machine" (which means power off or reset button from the mainboard), no keyboard reaction. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c69 Tejun Heo <teheo@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |NEEDINFO Info Provider| |imotgm@yahoo.com --- Comment #69 from Tejun Heo <teheo@novell.com> 2007-10-10 04:34:10 MST --- Does adding the following to kernel parameter make any difference? Options="libata.pata_dma=1" -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c70 Stephan Lauffer <lauffer@ph-freiburg.de> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |NEW Info Provider|imotgm@yahoo.com | --- Comment #70 from Stephan Lauffer <lauffer@ph-freiburg.de> 2007-10-10 07:27:41 MST --- Nothing has changed with that. :-/ But sorry, I need to correct myself (comment #68) in one (important) point: For about 30..40 seconds after I see the loading of pata_ti821x I can turn on/off numlock and change between the consoles. Sorry, I didn't recognized that since I alway was doing some other work while rebooting the machine... (: I'll keep on testing - let me know what to do next! -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c71 Tejun Heo <teheo@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |NEEDINFO Info Provider| |lauffer@ph-freiburg.de --- Comment #71 from Tejun Heo <teheo@novell.com> 2007-10-10 19:40:52 MST --- Stephan, can you connect the ODD to different controller and get the system installed? I think we'll need to test debug patches and it's difficult to do without installed system. Thanks. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c72 --- Comment #72 from Stephan Lauffer <lauffer@ph-freiburg.de> 2007-10-10 23:47:00 MST --- Sorry, what's ODD? Do you mean HDD? Well, I got the system installed with the "brokenmodules=pata_ti821x" kernel append (see #58: "Disabling (brokenmodules...) pata_it821x helps out..."). Now I've just no working usb support (maybe some usb device will work but my cardreader doesn't) and tons of error logs like: Oct 11 06:20:35 lisa kernel: hdc: cdrom_pc_intr: The drive appears confused (ireason = 0x01). Trying to recover by ending request. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c73 --- Comment #73 from Tejun Heo <teheo@novell.com> 2007-10-11 00:44:45 MST --- ODD was optical disc drive. Anyways, so you have a working system, great. 1. Please post the result of "hdparm -I /dev/hdc". 2. Can you unload the IDE driver and load pata_it821x? ie. Does your root partition live on a disk connected to a different controller? If so, please install KOTD and try to load pata_it821x. ftp://ftp.suse.com/pub/projects/kernel/kotd/HEAD/ 3. If you can't do #2, you'll need to create separate grub entry with alternate initrd. Module listto be included in initrd is determined by INITRD_MODULES in /etc/sysconfig/kernel. After replacing ide driver with pata_it821x, run "mkinitrd -k vmlinuz -i initrd-pata_it821x". During boot, exit to grub text mode and edit the boot entry such that initrd= points to the debug initrd. This way, you can test it while keeping the original working configuration. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c74 --- Comment #74 from Stephan Lauffer <lauffer@ph-freiburg.de> 2007-10-11 02:15:47 MST --- ok... (I need help *g) These results here are still taken from the kernel-default-2.6.22.5-31 kernel (not kotd). Info for 1.: ------------ The 1st ODD I was using is: <<snip>> ATAPI CD-ROM, with removable media Model Number: HL-DT-ST DVDRAM GSA-4160B Serial Number: K7948M33234 Firmware Revision: A300 Standards: Likely used CD-ROM ATAPI-1 Configuration: DRQ response: 50us. Packet size: 12 bytes Capabilities: LBA, IORDY(can be disabled) DMA: mdma0 mdma1 mdma2 udma0 udma1 udma2 udma3 *udma4 Cycle time: min=120ns recommended=120ns PIO: pio0 pio1 pio2 pio3 pio4 Cycle time: no flow control=120ns IORDY flow control=120ns HW reset results: CBLID- above Vih Device num = 0 <<snap>> But because of the errors in the /var/log/warn I pluggen on another ODD: <<snip>> ATAPI CD-ROM, with removable media Model Number: _NEC DVD_RW ND-4570A Serial Number: Firmware Revision: 1.02 Standards: Likely used CD-ROM ATAPI-1 Configuration: DRQ response: 3ms. Packet size: 12 bytes Capabilities: LBA, IORDY(cannot be disabled) DMA: mdma0 mdma1 mdma2 udma0 udma1 *udma2 Cycle time: min=120ns recommended=120ns PIO: pio0 pio1 pio2 pio3 pio4 Cycle time: no flow control=120ns IORDY flow control=120ns <<snap>> Actually I'm using the 1st one, the HL-DT-ST. There I've a lot of kernel: hdc: status error: status=0x58 { DriveReady SeekComplete DataRequest } kernel: ide: failed opcode was: unknown kernel: hdc: drive not ready for command Info for 2.: ------------ Well, on one IDE port I've a master and a slave parallel IDE disk. The ODD is plugged at another ide port as master. I don't know how to unload all the needed kernel modules (since some of them are in use and I guess I can't remove the one for my ide/root(?)). So I checked out what I can do in /etc/sysconfig/kernel and there I have now this: INITRD_MODULES="processor thermal it821x pata_it821x ata_piix fan jbd ext3 edd piix" And yes, now pata_it821x is loaded! See here: lisa:/ # lsmod | grep -E 'ide|ata' ide_cd 57232 0 cdrom 52392 1 ide_cd ide_disk 34304 10 ata_piix 37508 0 pata_it821x 29060 0 libata 164096 2 ata_piix,pata_it821x scsi_mod 176536 1 libata ide_core 165648 4 ide_cd,ide_disk,piix,it821x I throw away the ide-generic there. Is this the right ide module? Should I try to unload/substitute some more? I've downloaded the kotd with debuging... I'll install it after this bugzilla entry... Info to 3.: ----------- Ah - well that's what I doing all the time here. With some minor change: I've some different initrd-entries in the menu.conf so I don't need to swith into the edit mode of grub. But there's just one point which makes me unsure: mkinitrd uses a lot more modules than I've in INITRD_MODULES. F.e. rigth now it takes: Kernel Modules: processor thermal ide-core it821x scsi_mod libata pata_it821x ata_piix fan jbd mbcache ext3 edd piix ide-disk usbcore ohci-hcd uhci-hcd ehci-hcd ff-memless hid usbhid I guess it's ok so, but not sure... Well, ok, now I'll give it a try with the kotd-debugin. But I'm wondering why I now can use pata_it821x since it hangs during the boot of the opensuse installatione... (I need to hurry up with this bug/problem... I've some smbfs-problem left) ;) -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c75 Tejun Heo <teheo@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |albertcc@tw.ibm.com --- Comment #75 from Tejun Heo <teheo@novell.com> 2007-10-11 03:34:13 MST --- [cc'ing Albert.] * What happens if you connect the NEC drive to pata_it821x? Does that make the machine die too? * With the above INITRD_MODULES line, it821x (IDE 821x driver) is loaded before pata_it821x and attaches to the controller. pata_it821x is loaded but doesn't attach to the controller. Swapping them or passing "brokenmodules=it821x" will make pata_it821x attach to the controller. * During system boot up, lots of modules are loaded by udev and other services. Nothing wrong there. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c76 --- Comment #76 from Stephan Lauffer <lauffer@ph-freiburg.de> 2007-10-11 06:58:49 MST --- With the NEC drive the machine dies too, yes. Heya, but I got the pata_821x loaded and the devices where recognized as scsi ones. :) The boot process breakes later on than while there will no /dev/hdb6 (my root) appear. Ok, thats correct I guess. I should now rewrite the fstab (and maybe the menu.conf of grub)(?). I'll send two photos... 1st with the top of the successfully boot screen and then the loaded kernel modules which I read out after the stopped boot process (in this case I've a simple bash prompt). -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c77 --- Comment #77 from Stephan Lauffer <lauffer@ph-freiburg.de> 2007-10-11 07:02:46 MST --- Created an attachment (id=177685) --> (https://bugzilla.novell.com/attachment.cgi?id=177685) successfully load of pata_i821x - picture of boot messages successfully load of pata_i821x - picture of boot messages -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c78 --- Comment #78 from Stephan Lauffer <lauffer@ph-freiburg.de> 2007-10-11 07:06:03 MST --- Created an attachment (id=177688) --> (https://bugzilla.novell.com/attachment.cgi?id=177688) cat /proc/modules after the break of the boot process Since there will none expected /dev/hd? device found the boot process failes now. These are the loaded modules at this point (look good to me) -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c79 --- Comment #79 from Tejun Heo <teheo@novell.com> 2007-10-11 17:22:35 MST --- Stephan, so pata_it821x works fine on KOTD, right? You can fix the boot problem by mounting root partition by volume label or UUID. You can switch between hdX and sdX and the system will work just fine. It can be configured in yast but you probably need to update two places - system->partitioner and system->boot loader. In partitioner, it's edit->fstab options. In boot loader, I'm not sure how to actually massage it but the end result should look like the following. kernel /boot/vmlinuz-blahblah root=/dev/disk/by-label/{ROOT_LABEL} other_opts.. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c80 Stephan Lauffer <lauffer@ph-freiburg.de> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |NEW Info Provider|lauffer@ph-freiburg.de | --- Comment #80 from Stephan Lauffer <lauffer@ph-freiburg.de> 2007-10-12 00:42:08 MST --- Tejun, that's interesting: After the successfully load of pata_it821x from kotd I switched back to the 2.6.22.5-31-default (x86_64) from the opensuse-10.3 GM dvd. And with the same module list (that means: disabling it821x and using pata_it821x) it works, too! So what's up with the boot process of the 10.3 installation? Is this a result of a acip-i386-chipset-bios-whatever-problem? The way to my installation now seems to be this one: At the first time booting && installing from DVD: brokenmodules=pata_it821x. After that, reorganize s/it821x/pata_it821x/ and prepare bootloader && fstab. Strange, isn't it? (: Sorry for the long dialog and testing just to figure this out... -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c81 --- Comment #81 from Stephan Lauffer <lauffer@ph-freiburg.de> 2007-10-12 01:48:40 MST --- *grrr* after loading pata why doesn't it found the /dev/sdb6 here? (btw - why do we "need" the pata way?) "[...] Trying manual resume from /dev/sda5 resume device /dev/sda5 not found (ignoring) .. Want me to fall back to /dev/hdb6? (Y/n)" -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c82 Tejun Heo <teheo@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |NEEDINFO Info Provider| |imotgm@yahoo.com --- Comment #82 from Tejun Heo <teheo@novell.com> 2007-10-12 05:06:14 MST --- If you configure online update early during installation, the installation system can download the updated kernel during installation and use it. I'm not sure whether that was the case here tho. Transition to pata_* drivers bring benefits to many cases. It definitely incurs transition overhead and probably should have been done better. But, anyways, in the long term, I think it will be better this way. Keeping two separate ATA stacks will be pretty expensive in development and testing cost and the old IDE layer just couldn't cope with SATA and other new developments. I think you can specify LABEL for resume= too but am not too sure. It's for hibernation. Can you please test booting from 10.3 installation media again? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c83 --- Comment #83 from Stephan Lauffer <lauffer@ph-freiburg.de> 2007-10-12 06:15:40 MST --- Yes, I understand... for sure it makes sens to support only one stack, full ack. Related to your question about another boot of 10.3 install... what should I check there? Loading up pata will end up with the frozen system, setting pata as brokenmodule will work, it821x will be loaded and I'm be able to install the distribution. Should I now try to install twice, without online updates? Wouldn't it be enough to just only "downgrade" the kernel? That's what I did in #80... If you like I can do a 2nd install... oO(The interesting question to me is: What exactly must be changed to switch from the workin it821x environment to the pata_it821x world?) What I did is s/hd/sd/ for fstab and menu.conf, put the pata_it821x in initrd and add a brokenmudules=it821x to the kernel append. And during the boot I can see the successfully load (it looks good for me, see #77) of the pata system but shortly after that /dev/sdb5 (/dev/sdb5 is /boot and /dev/sdb6 is / and so on...) will not be found (see #81) -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c84 Tejun Heo <teheo@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |ASSIGNED Info Provider|imotgm@yahoo.com | --- Comment #84 from Tejun Heo <teheo@novell.com> 2007-10-13 07:51:14 MST --- I just wanted to make sure that the installation media always fails the same way. It would be good to know that the installed system fails to if installation is done w/o online updating but I don't think it's necessary. As for switching from it821x to pata_it821x, the best way would be to start out with LABEL/UUID mounting from the beginning. If done that way, the system wouldn't care which driver is attached to the device. It just looks for the filesystem with the given ID and use it no matter where it is. You probably need to edit /boot/grub/menu.lst to fix the current problem. Thanks. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=248860#c85 Tejun Heo <teheo@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |coolo@novell.com Status|ASSIGNED |RESOLVED Resolution| |FIXED --- Comment #85 from Tejun Heo <teheo@novell.com> 2007-10-24 04:30:48 MST --- Closing this one as FIXED but we need driver update disk or some other means to update pata_it821x during installation. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
participants (1)
-
bugzilla_noreply＠novell.com