Bug 7292
Summary: | irq 7: nobody cared - 945GM/ICH7/T1300 | ||
---|---|---|---|
Product: | ACPI | Reporter: | Boris B. Zhmurov (bb) |
Component: | Config-Interrupts | Assignee: | Len Brown (lenb) |
Status: | REJECTED INVALID | ||
Severity: | normal | CC: | acpi-bugzilla, akpm |
Priority: | P2 | ||
Hardware: | i386 | ||
OS: | Linux | ||
Kernel Version: | 2.6.18 | Subsystem: | |
Regression: | --- | Bisected commit-id: | |
Attachments: |
dmesg
.config dmesg of 2.6.16 acpidump of 2.6.16 interrupts of 2.6.26 lspci -vv of 2.6.16 |
Description
Boris B. Zhmurov
2006-10-09 10:48:28 UTC
Created attachment 9192 [details]
dmesg
dmesg
Created attachment 9193 [details]
.config
kernel's config
On Mon, 9 Oct 2006 10:59:57 -0700 bugme-daemon@bugzilla.kernel.org wrote: > Most recent kernel where this bug did not occur: 2.6.18 (vanilla) You say it occurred in 2.6.18, and that it didn't occur in 2.6.18. Please update the bug to tell us the most recent kernel version which did _not_ have this bug, thanks. Most recent kernel where this bug did not occur: 2.6._16_ (vanilla) My mistake, sorry. Kernel, which did not have this bug - 2.6.16! I tried 2.6.18 without nvidia - bug still exists. And it didn't occur on 2.6.16 with nvidia driver. Please attach the dmesg and /proc/interrupts from the successful 2.6.16 boot. Please test that the problem goes away with "acpi=off". Please attach the output from lspci -vv and the output from acpidump, available in the latest pmtools here: http://ftp.kernel.org/pub/linux/kernel/people/lenb/acpi/utils/ Academic question: > # CONFIG_X86_UP_APIC is not set Curious why you have the IOAPIC disabled, the machine seems to have one: > ACPI: MADT (v001 A M I OEMAPIC 0x01000627 MSFT 0x00000097) @ 0x3ffd0390 Created attachment 9228 [details]
dmesg of 2.6.16
2.6.16's dmesg
Created attachment 9229 [details]
acpidump of 2.6.16
acpidump of 2.6.16
Created attachment 9230 [details]
interrupts of 2.6.26
interrupts of 2.6.26
It's strange, but on 2.6.16.29 now I see the same problem, but it never happend
like for three or four months. :( And now I see it on 2.6.16. What should I try?
Should I enable some debug options in .config?
>> # CONFIG_X86_UP_APIC is not set
> Curious why you have the IOAPIC disabled, the machine seems to have one:
On one of my HP Proliant DL320 servers, I have completely freeze at boot time
with that option on 2.6.12 kernel. Due to server is located about 1000
kilometers far from me, I'm afraid to turn it on again :)
Created attachment 9231 [details]
lspci -vv of 2.6.16
lspci -vv of 2.6.16
When I boot with "irqpoll" command line, after 2-3 hours, notebook hang completely. Only poweroff helps... Trying with acpi=off same errors in dmesg with acpi=off > on 2.6.16.29 now I see the same problem Good to know this isn't a recent Linux regression. > IOAPIC It is safe to venture that enabling the IOAPIC on your laptop will not have any bad effect on your 2.6.12 proliant server 1000 km away:-) > ERR: 137813 This is not good news, it means your 8259 is getting lots of spurious interrupts. Curious that you don't see the following line in your dmesg to identify which IRQ is taking the spurious interrupts: mask_and_ack_8259A(unsigned int irq) ... if (!(spurious_irq_mask & irqmask)) { printk(KERN_DEBUG "spurious 8259A interrupt: IRQ%d.\n", irq); spurious_irq_mask |= irqmask; } atomic_inc(&irq_err_count); > same errors in dmesg with acpi=off Good to know -- then by definition, this issue does not belong in the ACPI bugzilla category. > nvidia: module license 'NVIDIA' taints kernel Please re-open when you can reproduce the issue and supply debug output without this binary module loaded. |