Bug 180773
| Summary: | ahci probe failure | ||
|---|---|---|---|
| Product: | [Fedora] Fedora | Reporter: | Richard Tresidder <rtresidd> |
| Component: | kernel | Assignee: | Jeff Garzik <jgarzik> |
| Status: | CLOSED RAWHIDE | QA Contact: | Brian Brock <bbrock> |
| Severity: | medium | Docs Contact: | |
| Priority: | medium | ||
| Version: | 5 | CC: | davej, peterm, wtogami |
| Target Milestone: | --- | ||
| Target Release: | --- | ||
| Hardware: | i686 | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Fixed In Version: | Doc Type: | Bug Fix | |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2006-03-06 16:41:53 UTC | Type: | --- |
| Regression: | --- | Mount Type: | --- |
| Documentation: | --- | CRM: | |
| Verified Versions: | Category: | --- | |
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
| Cloudforms Team: | --- | Target Upstream Version: | |
| Embargoed: | |||
| Bug Depends On: | |||
| Bug Blocks: | 172490 | ||
|
Description
Richard Tresidder
2006-02-10 08:50:20 UTC
What actual message does it show, is there something on tty3 (alt +f3) like: moving (1) to step accounts moving (1) to step reposetup Warning, could not load sqlite, falling back to pickle also check tty4 (alt + F4) for OOM killer message when it hangs. I'm pretty sure you are experiencing bug #176492 and that will confirm it for me. Hi again.
For a straight "Enter" install
TTY3 gives
.....
inserted /tmp/ieee1394.ko
inserted /tmp/ohci1394.ko
load module set done
probing for firewire scsi devices
modules to insert pcmcia_core yenta_core pcmcia
load module set done
modules to insert mii b44 yenta_socket libata ahci
loaded mii from /modules/modules.cgz
loaded b44 from /modules/modules.cgz
loaded libata from /modules/modules.cgz
loaded ahci from /modules/modules.cgz
inserted /tmp/mii.ko
inserted /tmp/b44.ko
inserted /tmp/libata.ko
inserted /tmp/ahci.ko
load module set done
trying to mount CD device hdc
no more messages..
TTY4
......
eth0: Broadcomm 4400 10/100BaseT Ethernet 00:76:.. (MAC)
libata version 1.20 loaded
ahci 0000:00:1f.2: version 1.2
ACPI: PCI Interrupt 0000:00:1f.2[B] - > Link [LINKB] -> GSI 5 (level, low) -> IRQ5
ACPI: PCI Interrupt for device 0000:00:1f.2 disabled
ahci: probe of 0000:00:1f.2 failed with error -12
hdc: lost interrupt
hdc: lost interrupt
hdc: lost interrupt
ide-cd: cmd 0x1e timed out
hdc: lost interrupt
repeats...
Turning ACPI off ie "linux acpi=off" does not improve the situation. you still
get the ahci probe failure followed by the lost interrupts as above.
When doing a network install "linux askmethod" and selecting ftp.
we get through to the selection of language.
on TTY3 the message saying "trying to mount CD device hdc"
does not appear as before.
After getting to the point of selecting th IP information, I select DHCP and set
it off, however the information is not retrieved.
TTY3
continuing from after "inserted /tmp/ahci.ko"
load module set done
setting language to en_US.UTF-8
53 keympas are available
need to set up networking
going to pick an interface
only have one network device: eth0 (odd the wirless device isn't found)
going to do getNetConfig
eth0 isn't a wireless adaptor
running dhcp for eth0
TTY4
irq9: nobody cared (try booting with "irqpoll" option)
[<c01368a5>] __report_bad_irq+0x2b/0x69 [<c0136969>] note_interrupt+0x6d/0xa0
[<c01363cd>] do_IRQ+0x9a/0xcd [<c0010548d>] do_IRQ+0x69/0x77
[<c0103c9a>] common_interrupt+0x1a/0x20 [<c01362fe>] handle_IRQ_event+0x17/0x4c
[<c01363ad>] do_IRQ+0x7a/0xcd [<c00105480>] do_IRQ+0x5c/0x77
[<c0103c9a>] common_interrupt+0x1a/0x20 [<c01362fe>] gen_new_estimator+0x5d/0x15d
more of the same.. including "do_softirq" messages
inode_has_perm
Disabling IRQ #9 appears on last line....
OK now there is a hint... I tried doing the network install with the irqpoll
kernel option and the network worked :)
HOWEVER the normal install still failed on trying to mount the cdrom..
I have to go out for dinner now I will continue to finish the installation later.
I think this still needs to remain open to solve the normal install issue
Thanks
Richard
Hmm quickly filled out the login info to see how much further it would go... and after filling out login info and saying continue, the installer stopped again.. The same maessage appears on TTY3 trying to mount CD device hdc on TTY4 we get lost interrupt OK no time to try anything else at the moment will get back later Done some more reading and this appears to be a rather widespread issue, and seems to be assosciated with changes that were made to the acpi code However I haven't been able to look much further than that I booted into the older FC3 installation on this laptop to get model info etc from the cdrom. The kernel used for this was 2.6.11-1.14_FC3 and the CDROM works but the network interface won't. previously 2.6.9-1.667 kernel both CDROM and network were fine CDROM is a NEC DVD+/-RW ND-6500A I'm unable to set DMA to on with this drive... comes back with operation not permitted.. hmm I have tried burning the install CD at 1x and the install still fails at the same spot. I've tried lots of different combinations with regard to dma and apic settings such as off, noirq etc and nothing changes the situation, I'd really like to get this going for obvious reasons and it is frustrating that this is obviously a regression of some sort Thanks Richard Hi I've managed to find a possible kernel trap discussion on this http://kerneltrap.org/node/5836 I have tried using the libata.atapi_enabled=1 as a boot parameter but it will not work It would be REALLY nice if TTY2 was up and actually working at this point so you could do some manual fishing around. I've tried using the noprobe boot parameter option to enable manual loading of the drivers, however I can't seem to find any reference to libata in the driver list. Using the noprobe option does bring tty2 up but the instant you enter a command it locks the terminal.... Thanks Richard This problem does not exist in FC5 Test 3 anymore. Everything seems to get probed correctly so I think this can be closed. (well at least for this Dell inspiron 9300 system) Thanks Richard OK maybe not fully solved...
Although a straight CD install will work fine, Trying a network install will
not bring up the network card correctly.
However using the kernel config option "irqpoll" will get around this.
Seems that without this option we are still seeing irq9 being disabled.
So maybe don't close this yet, but I'm not sure how I can help debug this?
Thanks
Richard
please file a seperate bug for the networking problem. multiple problems in one bug causes all kinds of confusion especially when searching for bugs. |