linux-kernel - Re: [bug] mapping multiple BARs

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date:	Tue, 5 Apr 2016 08:09:28 +0200
From:	Borislav Petkov <bp@...e.de>
To:	Jiri Olsa <jolsa@...hat.com>, Bjorn Helgaas <bhelgaas@...gle.com>
Cc:	Stephane Eranian <eranian@...gle.com>,
	Peter Zijlstra <peterz@...radead.org>,
	Andi Kleen <ak@...ux.intel.com>,
	LKML <linux-kernel@...r.kernel.org>,
	Arnaldo Carvalho de Melo <acme@...hat.com>,
	"mingo@...e.hu" <mingo@...e.hu>, Zhang Rui <rui.zhang@...el.com>,
	"Rafael J. Wysocki" <rjw@...ysocki.net>
Subject: Re: [bug] mapping multiple BARs

On Thu, Mar 24, 2016 at 02:36:56PM +0100, Jiri Olsa wrote:
> On Thu, Mar 24, 2016 at 11:15:06AM +0100, Borislav Petkov wrote:
> > On Thu, Mar 24, 2016 at 10:44:41AM +0100, Jiri Olsa wrote:
> > > hi,
> > > I'm hitting following lines in dmesg:
> > >   resource sanity check: requesting [mem 0xfed10000-0xfed15fff], which spans more than reserved [mem 0xfed10000-0xfed13fff]
> > >   caller snb_uncore_imc_init_box+0x7c/0xa0 mapping multiple BARs
> > > 
> > > I'm on IvyBridge model 58
> > >   vendor_id       : GenuineIntel
> > >   cpu family      : 6
> > >   model           : 58
> > >   model name      : Intel(R) Core(TM) i7-3720QM CPU @ 2.60GHz
> > >   stepping        : 9
> > >   microcode       : 0x1c
> > > 
> > > I've read following threads:
> > >   http://marc.info/?w=2&r=1&s=mapping+multiple+BARs&q=t
> > > 
> > > and it seems this is supposed to be already fixed by:
> > >   cb171f7abb9a PNP: Work around BIOS defects in Intel MCH area reporting
> > 
> > And as we talked on IRC, it looked like the quirk gets applied on my
> > machine and not on yours. I.e., I see:
> > 
> > [    0.669351] pnp 00:01: [Firmware Bug]: PNP resource [mem 0xfed10000-0xfed13fff] covers only part of 0000:00:00.0 Intel MCH; extending to [mem 0xfed10000-0xfed17fff]
> > 
> > and you don't.
> > 
> > Can you add some debug printks to quirk_intel_mch() to find out why?
> 
> seems like this function did not find any quirk
> (patch for the output attached)
> 
> [root@...el-chiefriver-04 ~]# dmesg | grep quirk_intel_mch
> quirk_intel_mch dev WDT0
> quirk_intel_mch   region start ffff880244617778, end fed10000
> quirk_intel_mch   mch    start fed10000, end fed17fff
> quirk_intel_mch   res    start 6a4, end 6a4
> quirk_intel_mch   res    start 6a0, end 6a0
> quirk_intel_mch dev LDRC
> quirk_intel_mch   region start ffff880244617378, end fed10000
> quirk_intel_mch   mch    start fed10000, end fed17fff
> quirk_intel_mch   res    start 2e, end 2f
> quirk_intel_mch   res    start 4e, end 4f
> quirk_intel_mch   res    start 61, end 61
> quirk_intel_mch   res    start 63, end 63
> quirk_intel_mch   res    start 65, end 65
> quirk_intel_mch   res    start 67, end 67
> quirk_intel_mch   res    start 70, end 70
> quirk_intel_mch   res    start 80, end 80
> quirk_intel_mch   res    start 92, end 92
> quirk_intel_mch   res    start b2, end b3
> quirk_intel_mch   res    start 680, end 69f
> quirk_intel_mch   res    start 1004, end 1013
> quirk_intel_mch   res    start ffff, end ffff
> quirk_intel_mch   res    start ffff, end ffff
> quirk_intel_mch   res    start 400, end 453
> quirk_intel_mch   res    start 458, end 47f
> quirk_intel_mch   res    start 500, end 57f
> quirk_intel_mch   res    start 164e, end 164f
> quirk_intel_mch dev CWDT
> quirk_intel_mch   region start ffff8802446b8b78, end fed10000
> quirk_intel_mch   mch    start fed10000, end fed17fff
> quirk_intel_mch   res    start 454, end 457
> quirk_intel_mch dev PDRC
> quirk_intel_mch   region start ffff8802446bcb78, end fed10000
> quirk_intel_mch   mch    start fed10000, end fed17fff
> quirk_intel_mch   res    start fed1c000, end fed1ffff
> quirk_intel_mch   res    start fed10000, end fed17fff

Here's an exact match.

> quirk_intel_mch   res    start fed18000, end fed18fff
> quirk_intel_mch   res    start fed19000, end fed19fff
> quirk_intel_mch   res    start f8000000, end fbffffff
> quirk_intel_mch   res    start fed20000, end fed3ffff
> quirk_intel_mch   res    start fed90000, end fed93fff
> quirk_intel_mch   res    start fed45000, end fed8ffff
> quirk_intel_mch   res    start ff000000, end ffffffff
> quirk_intel_mch   res    start fee00000, end feefffff
> quirk_intel_mch   res    start b15b0000, end b15b0fff
> 
> 
> but AFAICS the following dmesg output:
> 
>   resource sanity check: requesting [mem 0xfed10000-0xfed15fff], which spans more than reserved [mem 0xfed10000-0xfed13fff]
>   caller snb_uncore_imc_init_box+0x7c/0xa0 mapping multiple BARs
> 
> is related to checking of iomem_resource, which I haven't
> found connected to the pnp_dev resources from the quirk code

Hmm, so mine says:

[    0.643771] pnp 00:01: [Firmware Bug]: PNP resource [mem 0xfed10000-0xfed13fff] covers only part of 0000:00:00.0 Intel MCH; extending to [mem 0xfed10000-0xfed17fff]

which, AFAIR, means, the PNP resource needs to be extended. And which
the quirk does.

In your case, snb_uncore_imc_init_box() requests bigger resource than
what is reserved, AFAICT.

And your e820 map has those ranges too:

BIOS-e820: [mem 0x00000000fed10000-0x00000000fed13fff] reserved
BIOS-e820: [mem 0x00000000fed18000-0x00000000fed19fff] reserved

There's also

system 00:07: [mem 0xfed10000-0xfed17fff] could not be reserved

in your dmesg.

This all looks like a funny BIOS to me. And pretty old at that:

DMI: Intel Corporation 2012 Client Platform/Emerald Lake 2, BIOS ACRVMBY1.86C.0078.P00.1201161002 01/16/2012

Can you get an update or so?

Bjorn, can you make any sense of this?

Thanks.

-- 
Regards/Gruss,
    Boris.

SUSE Linux GmbH, GF: Felix Imendörffer, Jane Smithard, Graham Norton, HRB 21284 (AG Nürnberg)
--