Prev: Add targets to use the Coccinelle checker
Next: Fix performance issue with ondemand governor
From: Bjorn Helgaas on 27 Apr 2010 11:20
On Monday 26 April 2010 07:41:55 pm Yinghai wrote:
> On 04/26/2010 06:27 PM, Linus Torvalds wrote:
> > Do we actually have a regression left with Bjorn's patch?

After the pcibios_align_resource() patch, I'm not aware of any regressions.

But let's double-check this:

> also find one AMD system:
> [ 7.056011] pci_root PNP0A03:00: host bridge window [mem 0xfed20000-0xffffffff]
> ...
> pci assign unassign code could use range like [mem 0xfed20000-0xffffffff] wrongly.

I agree, it's very unlikely that it's safe to put PCI devices all the
way up to 0xffffffff. I suspect this might be fixed by d558b483d5a,
which computes the end of the bridge window using _MAX rather than _LEN.

See https://bugzilla.kernel.org/show_bug.cgi?id=15480#c15 for an example
similar to the one above: we originally thought the window was
[mem 0xcff00000-0xffffffff], but d558b483d5a changes that to
[mem 0xcff00000-0xfebfffff], which matches what Windows found.

Yinghai, can you take a look at your AMD system again with a kernel that
includes d558b483d5a, and see whether we still have a problem? If we
*do* still have a problem, please open a bugzilla and attach a dmesg log
with ACPI resource info collected with the debug patch here:
https://bugzilla.kernel.org/show_bug.cgi?id=15533#c5

Bjorn
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
From: Bjorn Helgaas on 28 Apr 2010 12:10
Yinghai, ping, do you have any more information about this?

On Tuesday 27 April 2010 09:11:10 am Bjorn Helgaas wrote:
> On Monday 26 April 2010 07:41:55 pm Yinghai wrote:
> But let's double-check this:
>
> > also find one AMD system:
> > [ 7.056011] pci_root PNP0A03:00: host bridge window [mem 0xfed20000-0xffffffff]
> > ...
> > pci assign unassign code could use range like [mem 0xfed20000-0xffffffff] wrongly.
>
> I agree, it's very unlikely that it's safe to put PCI devices all the
> way up to 0xffffffff. I suspect this might be fixed by d558b483d5a,
> which computes the end of the bridge window using _MAX rather than _LEN.
>
> See https://bugzilla.kernel.org/show_bug.cgi?id=15480#c15 for an example
> similar to the one above: we originally thought the window was
> [mem 0xcff00000-0xffffffff], but d558b483d5a changes that to
> [mem 0xcff00000-0xfebfffff], which matches what Windows found.
>
> Yinghai, can you take a look at your AMD system again with a kernel that
> includes d558b483d5a, and see whether we still have a problem? If we
> *do* still have a problem, please open a bugzilla and attach a dmesg log
> with ACPI resource info collected with the debug patch here:
> https://bugzilla.kernel.org/show_bug.cgi?id=15533#c5
>
> Bjorn
>


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
From: Yinghai Lu on 28 Apr 2010 13:20
Never mind, for 2.6.34 your patch should be good enough.

On 04/28/2010 09:07 AM, Bjorn Helgaas wrote:
> Yinghai, ping, do you have any more information about this?
>
> On Tuesday 27 April 2010 09:11:10 am Bjorn Helgaas wrote:
>
>> On Monday 26 April 2010 07:41:55 pm Yinghai wrote:
>> But let's double-check this:
>>
>>
>>> also find one AMD system:
>>> [ 7.056011] pci_root PNP0A03:00: host bridge window [mem 0xfed20000-0xffffffff]
>>> ...
>>> pci assign unassign code could use range like [mem 0xfed20000-0xffffffff] wrongly.
>>>
>> I agree, it's very unlikely that it's safe to put PCI devices all the
>> way up to 0xffffffff. I suspect this might be fixed by d558b483d5a,
>> which computes the end of the bridge window using _MAX rather than _LEN.
>>
>> See https://bugzilla.kernel.org/show_bug.cgi?id=15480#c15 for an example
>> similar to the one above: we originally thought the window was
>> [mem 0xcff00000-0xffffffff], but d558b483d5a changes that to
>> [mem 0xcff00000-0xfebfffff], which matches what Windows found.
>>
>> Yinghai, can you take a look at your AMD system again with a kernel that
>> includes d558b483d5a, and see whether we still have a problem? If we
>> *do* still have a problem, please open a bugzilla and attach a dmesg log
>> with ACPI resource info collected with the debug patch here:
>> https://bugzilla.kernel.org/show_bug.cgi?id=15533#c5
>>
>> Bjorn
>>
>>
>
>

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
From: Bjorn Helgaas on 28 Apr 2010 15:10
On Wednesday 28 April 2010 11:14:45 am Yinghai Lu wrote:
> Never mind, for 2.6.34 your patch should be good enough.

Uh, OK.

I'm not trying to be a nuisance, but if there's a machine where
we think there's a bridge window like [mem 0xfed20000-0xffffffff],
I want to fix it, even if you think it's "good enough for 2.6.34."

I was hoping you could specifically confirm that "yes, that AMD
machine was fixed by d558b483d5a," or else give me some more
information that would help me figure out what's going on.

Bjorn

> On 04/28/2010 09:07 AM, Bjorn Helgaas wrote:
> > Yinghai, ping, do you have any more information about this?
> >
> > On Tuesday 27 April 2010 09:11:10 am Bjorn Helgaas wrote:
> >
> >> On Monday 26 April 2010 07:41:55 pm Yinghai wrote:
> >> But let's double-check this:
> >>
> >>
> >>> also find one AMD system:
> >>> [ 7.056011] pci_root PNP0A03:00: host bridge window [mem 0xfed20000-0xffffffff]
> >>> ...
> >>> pci assign unassign code could use range like [mem 0xfed20000-0xffffffff] wrongly.
> >>>
> >> I agree, it's very unlikely that it's safe to put PCI devices all the
> >> way up to 0xffffffff. I suspect this might be fixed by d558b483d5a,
> >> which computes the end of the bridge window using _MAX rather than _LEN.
> >>
> >> See https://bugzilla.kernel.org/show_bug.cgi?id=15480#c15 for an example
> >> similar to the one above: we originally thought the window was
> >> [mem 0xcff00000-0xffffffff], but d558b483d5a changes that to
> >> [mem 0xcff00000-0xfebfffff], which matches what Windows found.
> >>
> >> Yinghai, can you take a look at your AMD system again with a kernel that
> >> includes d558b483d5a, and see whether we still have a problem? If we
> >> *do* still have a problem, please open a bugzilla and attach a dmesg log
> >> with ACPI resource info collected with the debug patch here:
> >> https://bugzilla.kernel.org/show_bug.cgi?id=15533#c5
> >>
> >> Bjorn
> >>
> >>
> >
> >
>
>


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
From: Yinghai on 28 Apr 2010 15:20
On 04/28/2010 12:06 PM, Bjorn Helgaas wrote:
> On Wednesday 28 April 2010 11:14:45 am Yinghai Lu wrote:
>> Never mind, for 2.6.34 your patch should be good enough.
>
> Uh, OK.
>
> I'm not trying to be a nuisance, but if there's a machine where
> we think there's a bridge window like [mem 0xfed20000-0xffffffff],
> I want to fix it, even if you think it's "good enough for 2.6.34."
>
> I was hoping you could specifically confirm that "yes, that AMD
> machine was fixed by d558b483d5a," or else give me some more
> information that would help me figure out what's going on.


[ 0.000000] BIOS-e820: 0000000000000100 - 0000000000098c00 (usable)
[ 0.000000] BIOS-e820: 0000000000098c00 - 00000000000a0000 (reserved)
[ 0.000000] BIOS-e820: 00000000000e6000 - 0000000000100000 (reserved)
[ 0.000000] BIOS-e820: 0000000000100000 - 00000000d7fa0000 (usable)
[ 0.000000] BIOS-e820: 00000000d7fae000 - 00000000d7fb0000 (reserved)
[ 0.000000] BIOS-e820: 00000000d7fb0000 - 00000000d7fbe000 (ACPI data)
[ 0.000000] BIOS-e820: 00000000d7fbe000 - 00000000d7ff0000 (ACPI NVS)
[ 0.000000] BIOS-e820: 00000000d7ff0000 - 00000000f0000000 (reserved)
[ 0.000000] BIOS-e820: 00000000fec00000 - 00000000fec01000 (reserved)
[ 0.000000] BIOS-e820: 00000000fee00000 - 00000000fee01000 (reserved)
[ 0.000000] BIOS-e820: 00000000ff700000 - 0000000100000000 (reserved)
[ 0.000000] BIOS-e820: 0000000100000000 - 0000008028000000 (usable)

[ 6.960006] PCI: Using host bridge windows from ACPI; if necessary, use "pci=nocrs" and report a bug
[ 6.984225] ACPI: PCI Root Bridge [PCI0] (domain 0000 [bus 00-06])
[ 7.023528] pci_root PNP0A03:00: host bridge window [io 0x0000-0x03af]
[ 7.024014] pci_root PNP0A03:00: host bridge window [io 0x03b0-0x03bb]
[ 7.028005] pci_root PNP0A03:00: host bridge window [io 0x03bc-0x03bf]
[ 7.032005] pci_root PNP0A03:00: host bridge window [io 0x03c0-0x03df]
[ 7.036005] pci_root PNP0A03:00: host bridge window [io 0x03e0-0xefff]
[ 7.040011] pci_root PNP0A03:00: host bridge window [mem 0xd8000000-0xe7ffffff]
[ 7.044005] pci_root PNP0A03:00: host bridge window [mem 0xf0000000-0xfe9fffff]
[ 7.048005] pci_root PNP0A03:00: host bridge window [mem 0xfec00000-0xfed0ffff]
[ 7.052005] pci_root PNP0A03:00: host bridge window [mem 0x000a0000-0x000bffff]
[ 7.056011] pci_root PNP0A03:00: host bridge window [mem 0xfed20000-0xffffffff]


d7ff0000-efffffff : reserved
d8000000-e7ffffff : PCI Bus 0000:00
d8000000-dfffffff : GART
e8000000-efffffff : PCI Bus 0000:80
fec00000-fed0ffff : PCI Bus 0000:00
fec00000-fec00fff : reserved
fec00000-fec003ff : IOAPIC 0
fed00000-fed003ff : HPET 0
fed10000-fed1ffff : PCI Bus 0000:80
fed20000-ffffffff : PCI Bus 0000:00
fee00000-fee00fff : Local APIC
fee00000-fee00fff : reserved
fefff000-feffffff : pnp 00:0a
ff700000-ffffffff : reserved
ffb80000-fffffffe : pnp 00:06

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
First  |  Prev  |  Next  |  Last
Pages: 1 2 3
Prev: Add targets to use the Coccinelle checker
Next: Fix performance issue with ondemand governor