[Bug 576540] New: boot process: several issues with(out) ACPI (no scanner, no serial port, no floppy ...)
http://bugzilla.novell.com/show_bug.cgi?id=576540 http://bugzilla.novell.com/show_bug.cgi?id=576540#c0 Summary: boot process: several issues with(out) ACPI (no scanner, no serial port, no floppy ...) Classification: openSUSE Product: openSUSE 11.2 Version: Final Platform: i586 OS/Version: openSUSE 11.2 Status: NEW Severity: Normal Priority: P5 - None Component: Basesystem AssignedTo: bnc-team-screening@forge.provo.novell.com ReportedBy: kunsche@t-online.de QAContact: qa@suse.de Found By: --- Blocker: --- User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; de; rv:1.9.0.1) Gecko/2008070208 Firefox/3.0.1 (.NET CLR 3.5.30729) First of all I filed parts of these issues earlier (see bug #570473) but in the mean time some more issues occurred, so it seems to be a more general issue with the boot process. OK, here's the whole story: When ACPI is enabled on my somewhat outdated 32 bit PIII board (GA-6va7+) everything is fine. But as this board is fully equipped with hardware (which partly doesn't allow to share interrupts) I need the Interrupt that is normally occupied by the ACPI controller. Therefore I have to switch it off (ACPI=off at boot prompt). This now causes several issues, like that no floppy device (/dev/fd0) is generated, the execution of the boot.clock script failed (that’s the already filed bug) and the parallel port is not working (here one can see faulty interrupts) and last (but probably) not least the usb scanner doesn't work (this is an Printer/Scanner combination, where the printing functionality is still available). This behaviour was verified with the following kernels: 2.6.31.5-default, 2.6.31.8-default, 2.6.31.8-pae , 2.6.31.8-vanilla , 2.6.32.41-default (from ...Kernel:/HEAD/openSUSE_11.2/i586) By the way, switching of parallel booting by setting the sysconfig variable PARALLELL=no gave no difference. To say it again, when ACPI is enabled there are none of these issues (except that the GPIB card isn't usable as the ACPI controller occupies the interrupt). Please have a look at the output of cat /proc/interrupts ############## acpi=on #################### CPU0 0: 36891 XT-PIC-XT timer 1: 85 XT-PIC-XT i8042 2: 0 XT-PIC-XT cascade 3: 1 XT-PIC-XT 4: 1 XT-PIC-XT 5: 0 XT-PIC-XT ni_atmio 6: 7 XT-PIC-XT floppy 7: 78 XT-PIC-XT parport0 8: 0 XT-PIC-XT rtc0 9: 0 XT-PIC-XT acpi 10: 430 XT-PIC-XT ohci_hcd:usb2, ohci_hcd:usb3, ohci1394, EMU10K1, eth0, eth1 11: 16372 XT-PIC-XT ehci_hcd:usb1, ivtv0, nvidia 12: 1821 XT-PIC-XT i8042 14: 16133 XT-PIC-XT pata_via 15: 708 XT-PIC-XT pata_via .. MCP: 1 Machine check polls ERR: 0 MIS: 0 ############## acpi=off #################### CPU0 0: 79162 XT-PIC-XT timer 1: 43 XT-PIC-XT i8042 2: 0 XT-PIC-XT cascade 3: 1 XT-PIC-XT 4: 1 XT-PIC-XT 5: 0 XT-PIC-XT ni_atmio 7: 292 XT-PIC-XT 8: 0 XT-PIC-XT rtc0 9: 0 XT-PIC-XT atgpib 10: 430 XT-PIC-XT ohci_hcd:usb2, ohci_hcd:usb3, ohci1394, EMU10K1, eth0, eth1 11: 34501 XT-PIC-XT ehci_hcd:usb1, ivtv0, nvidia 12: 1925 XT-PIC-XT i8042 14: 15252 XT-PIC-XT pata_via 15: 1017 XT-PIC-XT pata_via .. MCP: 2 Machine check polls ERR: 292 MIS: 0 Prior distributions on this system (SuSE 8.2 , openSUSE 10.2 , 10.3 and 11.1) never had these problems. Using openSUSE 11.1 with an 2.6.32.41-default kernel (from ...Kernel:/HEAD/openSUSE_11.1/i586) also indicated no problem, so the described issues seems not to be kernel related, but a weakness of the boot process. Maybe one should also have a look at bug#576095, it's not triggered by ACPI on or off, but it seems also be related to the boot process. Reproducible: Always Steps to Reproduce: 1. switch off ACPI with the kernel parameter acpi=off at the boot prompt 2. 3. Actual Results: parallel port not working, boot.clock script failed, no /dev/fd0 generated, usb scanner doesn't work Expected Results: functionality of the floppy, scanner and parallel port should not depend on the usage of ACPI -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=576540 http://bugzilla.novell.com/show_bug.cgi?id=576540#c yang xiaoyu <xyyang@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |xyyang@novell.com AssignedTo|bnc-team-screening@forge.pr |jsrain@novell.com |ovo.novell.com | -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=576540 http://bugzilla.novell.com/show_bug.cgi?id=576540#c1 Jiri Srain <jsrain@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Component|Basesystem |Kernel AssignedTo|jsrain@novell.com |kernel-maintainers@forge.pr | |ovo.novell.com --- Comment #1 from Jiri Srain <jsrain@novell.com> 2010-02-08 10:25:51 UTC --- This sounds to me more like a kernel problem, not bootloader. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=576540 http://bugzilla.novell.com/show_bug.cgi?id=576540#c Jeff Mahoney <jeffm@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |jeffm@novell.com AssignedTo|kernel-maintainers@forge.pr |astarikovskiy@novell.com |ovo.novell.com | -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=576540 http://bugzilla.novell.com/show_bug.cgi?id=576540#c2 --- Comment #2 from none none <kunsche@t-online.de> 2010-03-17 08:48:18 UTC --- ..OK in the mean time I digged in a bit deeper into the system and dug out that for whatever reason some modules will/can not be loaded when ACPI is off. A workaround is to force the system to load these modules at boot time. This can be done by modifying /etc/sysconfig with the sysconfig editor via YAST. Here one has to edit in the -> system -> kernel submenu the variable MODULES_LOADED_ON_BOOT and fill it with the parameters parport_pc floppy (in my case) This will fix all above mentioned issues. I don't know why the boot.clock script will be executed also without forcing it's module "rtc_cmos" explicitly, but it's reproducable the case. (The boot.clock issue can be fixed separately by forcing "rtc_cmos", but it's also covered by forcing only "parport_pc" or "floppy") I've also mentioned an issue with the scanner but this was probably sporadic and not related to this bug, anyhow it's not an issue anymore. IMHO I would not close this bug, as the behaviour is (at least for me) somewhat non-transparent. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=576540 http://bugzilla.novell.com/show_bug.cgi?id=576540#c3 --- Comment #3 from Petr Uzel <puzel@novell.com> 2010-05-28 13:16:30 UTC --- *** Bug 570473 has been marked as a duplicate of this bug. *** http://bugzilla.novell.com/show_bug.cgi?id=570473 -- Configure bugmail: http://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=576540 https://bugzilla.novell.com/show_bug.cgi?id=576540#c6 none none <kunsche@t-online.de> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |REOPENED Component|Kernel |Basesystem Version|Final |Milestone 6 of 6 Resolution|WONTFIX | AssignedTo|kernel-maintainers@forge.pr |bnc-team-screening@forge.pr |ovo.novell.com |ovo.novell.com Product|openSUSE 11.2 |openSUSE 11.4 Target Milestone|--- |RC 1 --- Comment #6 from none none <kunsche@t-online.de> 2011-02-09 20:06:01 UTC --- The behaviour is still present with openSUSE 11.3 as well as with openSUSE 11.4M6, so I reopened it. I've also adjusted it to product openSUSE 11.4 and switched the relation from kernel to basesystem as a workaround by forcing the related modules at boot time is possible, see comment #2. There's one difference within 11.4M6, the RTC isn't affected anymore. In the mean time I could observe the same behaviour on 2 additional P3 based machines and also with a Gericon FX5600 Notebook (equipped with a P4 2,4GHz processor) -- 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=576540 https://bugzilla.novell.com/show_bug.cgi?id=576540#c zj jia <zjjia@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |zjjia@novell.com AssignedTo|bnc-team-screening@forge.pr |kernel-maintainers@forge.pr |ovo.novell.com |ovo.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.
https://bugzilla.novell.com/show_bug.cgi?id=576540 https://bugzilla.novell.com/show_bug.cgi?id=576540#c7 Jiri Slaby <jslaby@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|REOPENED |NEEDINFO CC| |jslaby@novell.com InfoProvider| |kunsche@t-online.de --- Comment #7 from Jiri Slaby <jslaby@novell.com> 2011-07-12 21:06:58 UTC --- You mostly cannot use acpi=off nowadays. The system will encounter exactly the problems you are seeing. Instead the IRQ assignment for the card should be changed (e.g. in BIOS) or the driver fixed. Why doesn't the driver support sharing interrupts? Anyway it's hard to say without better understanding of what kind of hardware and card is that. -- 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=576540 https://bugzilla.novell.com/show_bug.cgi?id=576540#c8 none none <kunsche@t-online.de> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |ASSIGNED Version|Milestone 6 of 6 |Final InfoProvider|kunsche@t-online.de | Target Milestone|RC 1 |Final --- Comment #8 from none none <kunsche@t-online.de> 2011-07-20 17:14:37 UTC --- Hmm.. , I'm not sure if I got the point of your comment, anyhow here's the attempt of an reply: You mostly cannot use acpi=off nowadays. The system will encounter exactly the problems you are seeing -> This I don't get, ACPI is an additional functionality which should not influence the above mentioned hardware when it's not used. And beside this, what about PC's without (error free) ACPI functionality at all? Instead the IRQ assignment for the card should be changed (e.g. in BIOS) or the driver fixed. -> The problem is not related to a special card. If you reflect to ISA cards than there's anyhow the problem that BIOS settings are ignored by openSUSE (The settings which are generated by the ISAPNP-tools are also ignored, see bug#576095). Why doesn't the driver support sharing interrupts? -> It's not a question of a driver, it's depending on the hardware. So normally ISA-bus hardware doesn't allow to share interrupts as well as some board components, e.g. the parallel port, RS232 ports and the ACPI controller itself. An additional problem is that some PCI-Hardware can't in fact share it's interrupt with every other PCI-Card. E.g you'll find very often the problem that a PC doesn't work properly when the graphics card is sharing it's interrupt with the sound card. That means if you have a fully equipped mainboard you have to take care for enough independent physical IRQ's, wherefore you can for example disable the not needed ACPI controller to set free an additional IRQ (That was done in the above mentioned example to gain an additional IRQ (9) for the GPIB-Card). -- 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=576540 https://bugzilla.novell.com/show_bug.cgi?id=576540#c9 Greg Kroah-Hartman <gregkh@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |RESOLVED Resolution| |INVALID --- Comment #9 from Greg Kroah-Hartman <gregkh@suse.com> 2011-08-31 19:52:46 UTC --- This is not a kernel issue, but a BIOS one, and one that the kernel can do nothing about. Please get an updated BIOS from your vendor, sorry. -- 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