linux-kernel - Re: [PATCH v4 06/10] PCI: Add PCIe Device 3 Extended Capability enumeration

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]

Message-ID: <689690d2a07b4_cff991006c@dwillia2-xfh.jf.intel.com.notmuch>
Date: Fri, 8 Aug 2025 17:05:38 -0700
From: <dan.j.williams@...el.com>
To: Bjorn Helgaas <helgaas@...nel.org>, Dan Williams
	<dan.j.williams@...el.com>
CC: <linux-coco@...ts.linux.dev>, <linux-pci@...r.kernel.org>,
	<linux-kernel@...r.kernel.org>, <bhelgaas@...gle.com>, <aik@....com>,
	<lukas@...ner.de>, Ilpo Järvinen
	<ilpo.jarvinen@...ux.intel.com>, Samuel Ortiz <sameo@...osinc.com>, Xu Yilun
	<yilun.xu@...ux.intel.com>
Subject: Re: [PATCH v4 06/10] PCI: Add PCIe Device 3 Extended Capability
 enumeration

Bjorn Helgaas wrote:
> On Thu, Jul 17, 2025 at 11:33:54AM -0700, Dan Williams wrote:
> > PCIe 6.2 Section 7.7.9 Device 3 Extended Capability Structure,
> > enumerates new link capabilities and status added for Gen 6 devices. One
> 
> s/ added for Gen 6 devices//
> 
> I know the "Gen 6 device" terminology is pervasive, but the spec
> suggests avoiding it because it's so ambiguous.

Ok.

> 
> > of the link details enumerated in that register block is the "Segment
> > Captured" status in the Device Status 3 register. That status is
> > relevant for enabling IDE (Integrity & Data Encryption) whereby
> > Selective IDE streams can be limited to a given Requester ID range
> > within a given segment.
> > 
> > If a device has captured its Segment value then it knows that PCIe Flit
> > Mode is enabled via all links in the path that a configuration write
> > traversed. IDE establishment requires that "Segment Base" in
> > IDE RID Association Register 2 (PCIe 6.2 Section 7.9.26.5.4.2) be
> > programmed if the RID association mechanism is in effect.
> > 
> > When / if IDE + Flit Mode capable devices arrive, the PCI core needs to
> > setup the segment base when using the RID association facility, but no
> > known deployments today depend on this.
> 
> So far this mentions a lot of facts, but only the subject hints at
> what it does.  I guess it just captures the Flit Mode status, inferred
> by Segment Captured?
> 
> I'm OK with basically just saying *that*, and moving some of the
> implications to places where we depend on them.

Agree, too wordy. Trimmed to:

PCIe r7.0 Section 7.7.9 Device 3 Extended Capability Structure, defines the
canonical location for determining the Flit Mode of a device. This status
is a dependency for PCIe IDE enabling. Add a new fm_enabled flag to 'struct
pci_dev'.