[Intel-gfx] Info: mapping multiple BARs. Your kernel is fine.

[Stephane Eranian]

Hi,

I am trying to understand your test case.
Were you actually measure uncore_imc events at the time you suspended?

I tried on my IvyBridge Lenovo and it works fine with 3.14-rc4+ (tip.git).
I used: echo -n disk >/sys/power/state




On Tue, Feb 25, 2014 at 4:48 PM, H. Peter Anvin <hpa at zytor.com> wrote:
> On 02/24/2014 12:19 PM, Borislav Petkov wrote:
>> Btw,
>>
>> I don't know whether the following observation is related or not, but it
>> so happens that after resume from suspend-to-disk, I see the booting up
>> of the resume kernel on the console but when it is time for the original
>> kernel to take over and switch to graphics, the screen remains black but
>> the machine is responsive over the network.
>>
>> And this doesn't happen on every resume but only sporadically.
>>
>> And yep, -rc3 was fine.
>>
>> On Mon, Feb 24, 2014 at 05:24:00PM +0100, Borislav Petkov wrote:
>>> This started happening this morning after booting -rc4+tip, let's
>>> add *everybody* to CC :-)
>>>
>>> We have intel_uncore_init, snb_uncore_imc_init_box, uncore_pci_probe and
>>> other goodies on the stack.
>>>
>
> snb_uncore_imc_init_box() is introduced new in tip:perf/core, and is a
> relatively recent commit (b9e1ab6d4c0582cad97699285a6b3cf992251b00), so
> I suspect that that wasn't in whatever -rc3 mix you were testing.
>
> I am wondering if backing/disabling out that support (perhaps by
> removing the relevant PCI ID) fixes the problem?
>
>         -hpa
>