Hi all,
It always happens at the same spot. 2.6.37 is fine with almost
identical .config. Dmesg and config attached.
Tvrtko
On Sun, Mar 27, 2011 at 08:18:05PM +0100, Tvrtko Ursulin wrote:
>
> Hi all,
>
> It always happens at the same spot. 2.6.37 is fine with almost
> identical .config. Dmesg and config attached.
Hi,
can you please send dmesg when 2.6.37 is booted.
Also output of "lspci -nnxxxx" would be of interest.
dmidecode (as root) would also help.
Does it help to disable C1e in your BIOS (if there is such an option)?
If there is no such option please try to boot 2.6.38.1 with idle=mwait
on the command line.
Thanks for now,
Andreas
On Wed, 2011-03-30 at 17:53 +0200, Andreas Herrmann wrote:
> On Sun, Mar 27, 2011 at 08:18:05PM +0100, Tvrtko Ursulin wrote:
> >
> > Hi all,
> >
> > It always happens at the same spot. 2.6.37 is fine with almost
> > identical .config. Dmesg and config attached.
>
>
> Hi,
>
> can you please send dmesg when 2.6.37 is booted.
> Also output of "lspci -nnxxxx" would be of interest.
> dmidecode (as root) would also help.
Attached.
> Does it help to disable C1e in your BIOS (if there is such an option)?
> If there is no such option please try to boot 2.6.38.1 with idle=mwait
> on the command line.
Both tests (on their own) booted without hanging.
Hope this helps.
Tvrtko
On Wed, Mar 30, 2011 at 07:18:17PM +0100, Tvrtko Ursulin wrote:
> On Wed, 2011-03-30 at 17:53 +0200, Andreas Herrmann wrote:
> > On Sun, Mar 27, 2011 at 08:18:05PM +0100, Tvrtko Ursulin wrote:
> > >
> > > Hi all,
> > >
> > > It always happens at the same spot. 2.6.37 is fine with almost
> > > identical .config. Dmesg and config attached.
> >
> >
> > Hi,
> >
> > can you please send dmesg when 2.6.37 is booted.
> > Also output of "lspci -nnxxxx" would be of interest.
Sorry, forgot to say that lspci -nnxxxx should also be run as root.
(To get complete PCI config space.)
> > dmidecode (as root) would also help.
>
> Attached.
Thanks.
> > Does it help to disable C1e in your BIOS (if there is such an option)?
> > If there is no such option please try to boot 2.6.38.1 with idle=mwait
> > on the command line.
>
> Both tests (on their own) booted without hanging.
Ok.
This reminds me on some Gigabyte mainboards (for which we couldn't
root cause the problem). The workaround for those boards was to boot
with acpi_skip_timer_override.
The Gigabyte mainboards had some odd lapic_id settings and yours too:
> [ 0.000000] ACPI: LAPIC (acpi_id[0x01] lapic_id[0x00] enabled)
> [ 0.000000] ACPI: LAPIC (acpi_id[0x02] lapic_id[0x01] enabled)
> [ 0.000000] ACPI: LAPIC (acpi_id[0x03] lapic_id[0x02] enabled)
> [ 0.000000] ACPI: LAPIC (acpi_id[0x04] lapic_id[0x03] enabled)
> [ 0.000000] ACPI: LAPIC (acpi_id[0x05] lapic_id[0x84] disabled) <====
> [ 0.000000] ACPI: LAPIC (acpi_id[0x06] lapic_id[0x85] disabled) <====
> [ 0.000000] ACPI: IOAPIC (id[0x04] address[0xfec00000] gsi_base[0])
> [ 0.000000] IOAPIC[0]: apic_id 4, version 33, address 0xfec00000, GSI 0-23
This should rather look like
[ 0.000000] ACPI: LAPIC (acpi_id[0x01] lapic_id[0x00] enabled)
[ 0.000000] ACPI: LAPIC (acpi_id[0x02] lapic_id[0x01] enabled)
[ 0.000000] ACPI: LAPIC (acpi_id[0x03] lapic_id[0x02] enabled)
[ 0.000000] ACPI: LAPIC (acpi_id[0x04] lapic_id[0x03] enabled)
[ 0.000000] ACPI: LAPIC (acpi_id[0x05] lapic_id[0x04] enabled)
[ 0.000000] ACPI: LAPIC (acpi_id[0x06] lapic_id[0x05] enabled)
[ 0.000000] ACPI: IOAPIC (id[0x06] address[0xfec00000] gsi_base[0])
[ 0.000000] IOAPIC[0]: apic_id 6, version 33, address 0xfec00000, GSI 0-23
(That's from my system -- also ASUSTek, but different model.)
So can you please provide "lspci -xxxx" (as root) and try another test
-- boot .38 with acpi_skip_timer_override kernel parameter.
And please add also "apic=debug show_lapic=all" and provide dmesg
of that test run.
Thanks,
Andreas
On Thu, Mar 31, 2011 at 02:43:52PM +0200, Andreas Herrmann wrote:
> On Wed, Mar 30, 2011 at 07:18:17PM +0100, Tvrtko Ursulin wrote:
> > On Wed, 2011-03-30 at 17:53 +0200, Andreas Herrmann wrote:
> > > On Sun, Mar 27, 2011 at 08:18:05PM +0100, Tvrtko Ursulin wrote:
> > > >
> > > > Hi all,
> > > >
> > > > It always happens at the same spot. 2.6.37 is fine with almost
> > > > identical .config. Dmesg and config attached.
> > >
> > >
> > > Hi,
> > >
> > > can you please send dmesg when 2.6.37 is booted.
> > > Also output of "lspci -nnxxxx" would be of interest.
>
> Sorry, forgot to say that lspci -nnxxxx should also be run as root.
> (To get complete PCI config space.)
>
> > > dmidecode (as root) would also help.
> >
> > Attached.
>
> Thanks.
>
> > > Does it help to disable C1e in your BIOS (if there is such an option)?
> > > If there is no such option please try to boot 2.6.38.1 with idle=mwait
> > > on the command line.
> >
> > Both tests (on their own) booted without hanging.
>
> Ok.
>
> This reminds me on some Gigabyte mainboards (for which we couldn't
> root cause the problem). The workaround for those boards was to boot
> with acpi_skip_timer_override.
>
> The Gigabyte mainboards had some odd lapic_id settings and yours too:
>
> > [ 0.000000] ACPI: LAPIC (acpi_id[0x01] lapic_id[0x00] enabled)
> > [ 0.000000] ACPI: LAPIC (acpi_id[0x02] lapic_id[0x01] enabled)
> > [ 0.000000] ACPI: LAPIC (acpi_id[0x03] lapic_id[0x02] enabled)
> > [ 0.000000] ACPI: LAPIC (acpi_id[0x04] lapic_id[0x03] enabled)
> > [ 0.000000] ACPI: LAPIC (acpi_id[0x05] lapic_id[0x84] disabled) <====
> > [ 0.000000] ACPI: LAPIC (acpi_id[0x06] lapic_id[0x85] disabled) <====
> > [ 0.000000] ACPI: IOAPIC (id[0x04] address[0xfec00000] gsi_base[0])
> > [ 0.000000] IOAPIC[0]: apic_id 4, version 33, address 0xfec00000, GSI 0-23
>
> This should rather look like
Sorry, forget this statement. You have a quad-core and not a six-core
CPU so this setting seems to be ok. (But it's still worth to try the
acpi_skip_timer_override option.)
Andreas
On Thu, 2011-03-31 at 14:43 +0200, Andreas Herrmann wrote:
> So can you please provide "lspci -xxxx" (as root) and try another test
> -- boot .38 with acpi_skip_timer_override kernel parameter.
>
> And please add also "apic=debug show_lapic=all" and provide dmesg
> of that test run.
Here they are. And doot did not hang with this option.
Tvrtko