Re: [Xen-devel] linux-3.9-rc0 regression from 3.8 SATA controller not detected under xen

[Sander Eikelenboom <linux@xxxxxxxxxxxxxx>]

Tuesday, February 26, 2013, 4:20:38 PM, you wrote:

> On Mon, Feb 25, 2013 at 11:18:30PM +0100, Sander Eikelenboom wrote:
>> Hi Konrad,
>> 
>> I can't get linux-3.9 rc0 to boot under xen-unstable.
>> It doesn't detect the s-ata controller, so it ends op with udev timing and 
>> bailing out to busybox.
>> 
>> I don't see a obvious error in the logs.
>> 
>> The same kernel boots fine on baremetal.
>> With linux 3.8 it boots fine under this version of xen-unstable.
>> 
>> Can you spot something from the logs or you have a hunch from where to start 
>> a bisect ?
>> (bisecting could be a problem due to multiple boot issues and entangled 
>> patchsets i presume ..)

> There are at least two pitfalls - the
> 14e568e78f6f80ca1e27256641ddf524c7dbdc51 (which you can revert) and you need
> to make sure you have 0cc9129d75ef8993702d97ab0e49542c15ac6ab9
> if you end up in the merge 2ef14f465b9e096531343f5b734cffc5f759f4a6
> ("Merge branch 'x86-mm-for-linus' of 
> git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip")


>> 
>> 3.9-rc0 last commit is: ab7826595e9ec51a51f622c5fc91e2f59440481a   Merge tag 
>> 'mfd-3.9-1' of git://git.kernel.org/pub/scm/linux/kernel/git/sameo/mfd-2.6

> Interestingly I can reproduce it on my F1A75-M box as well.

> With baremetal I get:
> [   12.701221] ahci 0000:00:11.0: version 3.0
> [   12.701424] initcall radeon_init+0x0/0x1000 [radeon] returned 0 after 
> 528175 usecs
> [   12.712999] ahci 0000:00:11.0: irq 46 for MSI/MSI-X
> [   12.717967] ahci 0000:00:11.0: AHCI 0001.0300 32 slots 6 ports 6 Gbps 0x3f 
> impl SATA mode
> [   12.726132] ahci 0000:00:11.0: flags: 64bit ncq sntf ilck pm led clo pmp 
> pio slum part
> [   12.735405] scsi0 : ahci
> [   12.738087] scsi1 : ahci
> [   12.740734] scsi2 : ahci
> [   12.743511] scsi3 : ahci
> [   12.746278] scsi4 : ahci
> [   12.748930] scsi5 : ahci

> And when booting under Xen 4.1 (note, no IOMMU):

> [   19.178696] ahci 0000:00:11.0: version 3.0
> [   19.178721] xen: registering gsi 19 triggering 0 polarity 1
> [   19.178743] xen: --> pirq=19 -> irq=19 (gsi=19)
> [   19.179040] ahci 0000:00:11.0: AHCI 0001.0300 32 slots 6 ports 6 Gbps 0x3f 
> impl SATA mode
> [   19.179043] ahci 0000:00:11.0: flags: 64bit ncq sntf ilck pm led clo pmp 
> pio slum part
> [   19.241748] ahci: probe of 0000:00:11.0 failed with error -22
> [   19.249124] initcall ahci_pci_driver_init+0x0/0x1000 [ahci] returned 0 
> after 460418 usecs

> Had you tried to checkout right before this mega update from Jeff Garzik:

> ab78265 Merge tag 'mfd-3.9-1' of 
> git://git.kernel.org/pub/scm/linux/kernel/git/sameo/mfd-2.6
> 21fbd58 Merge branch 'v4l_for_linus' of 
> git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media
> d9978ec Merge tag 'upstream-linus' of 
> git://git.kernel.org/pub/scm/linux/kernel/git/jgarzik/libata-dev
==>> 77be36d Merge tag 'stable/for-linus-3.9-rc0-tag' of 
git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen

> And seeing if that works?

Hmm that's worth the try, one of the changes seems to be to try to enable 
multiple msi per device, somehow that sounds like something that could blow up 
under xen.
Compiling now ..

>> 
>> --
>> Sander
>> 
>> Attached:
>> - Serial log with 3.9-rc0 kernel (missing sata)
>> - Serial log with 3.8 kernel (booting fine
>> - .config







_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel