Bug 9795 - Confusion caused by pci=routeirq console messages
Summary: Confusion caused by pci=routeirq console messages
Status: REJECTED INVALID
Alias: None
Product: ACPI
Classification: Unclassified
Component: Config-Interrupts (show other bugs)
Hardware: All Linux
: P1 low
Assignee: ykzhao
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2008-01-22 10:24 UTC by Artem Goncharov
Modified: 2008-03-12 15:39 UTC (History)
1 user (show)

See Also:
Kernel Version: 2.6.24-rc8
Subsystem:
Regression: ---
Bisected commit-id:


Attachments
dmesg (22.73 KB, application/octet-stream)
2008-01-23 10:40 UTC, Artem Goncharov
Details

Description Artem Goncharov 2008-01-22 10:24:37 UTC
Earliest failing kernel version:
2.6.24-rc8
Distribution:
Fedora8
Hardware Environment:
Acer Aspire 5052
Software Environment:
Problem Description:

After adding pci=routeacpi to kernel following output appeared in /var/log/messages
Jan 22 20:01:10 blending kernel: PCI: Using ACPI for IRQ routing
Jan 22 20:01:10 blending kernel: PCI: Routing PCI interrupts for all devices because "pci=routeirq" specified
Jan 22 20:01:10 blending kernel: ACPI: PCI Interrupt 0000:00:12.0[A] -> GSI 22 (level, low) -> IRQ 22
Jan 22 20:01:10 blending kernel: ACPI: PCI Interrupt 0000:00:13.0[A] -> GSI 19 (level, low) -> IRQ 19
Jan 22 20:01:10 blending kernel: ACPI: PCI Interrupt 0000:00:13.1[A] -> GSI 19 (level, low) -> IRQ 19
Jan 22 20:01:10 blending kernel: ACPI: PCI Interrupt 0000:00:13.2[A] -> GSI 19 (level, low) -> IRQ 19
Jan 22 20:01:10 blending kernel: ACPI: PCI Interrupt 0000:00:14.1[A] -> GSI 16 (level, low) -> IRQ 16
Jan 22 20:01:10 blending kernel: ACPI: PCI Interrupt 0000:00:14.2[A] -> GSI 16 (level, low) -> IRQ 16
Jan 22 20:01:10 blending kernel: ACPI: PCI Interrupt 0000:01:05.0[A] -> GSI 17 (level, low) -> IRQ 17
Jan 22 20:01:10 blending kernel: ACPI: PCI Interrupt 0000:02:00.0[A] -> GSI 16 (level, low) -> IRQ 16
Jan 22 20:01:10 blending kernel: ACPI: PCI Interrupt 0000:08:01.0[A] -> GSI 23 (level, low) -> IRQ 23
Jan 22 20:01:10 blending kernel: ACPI: PCI Interrupt 0000:08:01.1[B] -> GSI 22 (level, low) -> IRQ 22
Jan 22 20:01:10 blending kernel: ACPI: PCI Interrupt 0000:08:01.2[B] -> GSI 22 (level, low) -> IRQ 22
Jan 22 20:01:10 blending kernel: ACPI: PCI Interrupt 0000:08:01.3[B] -> GSI 22 (level, low) -> IRQ 22
Jan 22 20:01:10 blending kernel: ACPI: PCI Interrupt 0000:08:01.4[B] -> GSI 22 (level, low) -> IRQ 22
Jan 22 20:01:10 blending kernel: ACPI: PCI Interrupt 0000:08:02.0[A] -> GSI 20 (level, low) -> IRQ 20
Comment 1 ykzhao 2008-01-22 17:24:17 UTC
Hi, Artem
   There is no boot parameter about pci=routeacpi. Will you please confirm whether the boot option of "pci=routeirq" is added ? 
   If the boot option of "pci=routeirq" is added, the above message should be reported. The option is to enable the interrupt of all PCI devices.
   It will be great if you can attach the output of dmesg.
   Thanks.
Comment 2 Artem Goncharov 2008-01-23 10:40:33 UTC
Created attachment 14547 [details]
dmesg

dmesg
Comment 3 Artem Goncharov 2008-01-23 10:41:51 UTC
Sorry, pci=routeirq option was added (mistyping ;-)
Comment 4 ykzhao 2008-01-23 17:49:19 UTC
Hi, Artem
   Thanks for the confirm. 
   From the dmesg in comment #2 it seems that there is no error about the laptop. The only confusing is that OSPM reports ACPI PCI interrupt message when boot option of pci=routeirq is added. 
   The boot option of "pci=routeirq" is used to enable the interrupt of all PCI devices before loading PCI driver for PCI device and OSPM will report the above message. 
   
   So the bug will be rejected.
Comment 5 Bjorn Helgaas 2008-03-12 15:39:04 UTC
If there's confusion here, I'm afraid I'm responsible for it, since I added the "pci=routeirq" option.

Without "pci=routeirq", PCI interrupts are enabled only when a driver enables a device.  When you specify "pci=routeirq", the only thing that changes is that we enable the interrupts for all devices right away at boot, before any drivers claim devices.

What would make this less confusing to you?  Would you just rather not see the messages about how the interrupts are routed?  That's useful debugging information, so I'd like to keep that in some form.

Note You need to log in before you can comment on or make changes to this bug.