linux-kernel - Re: [PATCH] x86/PCI: never allocate PCI MMIO resources below BIOS

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>] [day] [month] [year] [list]

Date:	Mon, 26 Apr 2010 19:02:49 -0700
From:	"H. Peter Anvin" <hpa@...or.com>
To:	Jesse Barnes <jbarnes@...tuousgeek.org>,
	Linus Torvalds <torvalds@...ux-foundation.org>
CC:	Bjorn Helgaas <bjorn.helgaas@...com>,
	Andy Isaacson <adi@...apodia.org>,
	"R. Andrew Bailey" <bailey@...mai.com>,
	Yinghai <yinghai.lu@...cle.com>,
	Thomas Gleixner <tglx@...utronix.de>,
	Ingo Molnar <mingo@...hat.com>, guenter.roeck@...csson.com,
	"linux-pci@...r.kernel.org" <linux-pci@...r.kernel.org>,
	"x86@...nel.org" <x86@...nel.org>,
	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
	Thomas Renninger <trenn@...e.de>, yaneti@...lera.com
Subject: Re: [PATCH] x86/PCI: never allocate PCI MMIO resources below    BIOS_END

The 1 MB range is only one case of a prereserved space.  It's special only in the sense that it is *always* reserved, even if the map doesn't mark it as such.  Fixed (or SMM-used) resources in reserved space above 1 MB are exactly the same issue.

"Jesse Barnes" <jbarnes@...tuousgeek.org> wrote:

>On Mon, 26 Apr 2010 18:27:27 -0700 (PDT)
>Linus Torvalds <torvalds@...ux-foundation.org> wrote:
>
>> 
>> 
>> On Mon, 26 Apr 2010, Jesse Barnes wrote:
>> > 
>> > Glad we agree.  As I said (and echoing Bjorn), I think it would be best
>> > to reserve this space in a way that doesn't just use IORESOURCE_BUSY.
>> > We want and need to do allocations from the special region, so we
>> > should mark it as such.
>> 
>> I think Bjorn's patch to pcibios_align_resource() is really good and 
>> clever, and I think it should take care of the need for IORESOURCE_BUSY, 
>> no? We do want to let devices that are _already_ allocated there insert 
>> their resources, it's just that we never want to allocate new ones in the 
>> low 1M region.
>> 
>> Do we actually have a regression left with Bjorn's patch?
>
>No, I think we're covered.  But it sounded like Peter was also
>concerned about making new allocations from the 1M space, which would
>mean we'd need something other than the IORESOURCE_BUSY bit.  But maybe
>Bjorn's patch plus simply removing the IORESOURCE_BUSY line is
>sufficient for that.  The downside there is that it doesn't clearly
>communicate the special nature of the 1M region.
>
>-- 
>Jesse Barnes, Intel Open Source Technology Center

-- 
Sent from my Android phone with K-9 Mail. Please excuse my brevity.