linux-cxl.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: <dan.j.williams@intel.com>
To: "Bowman, Terry" <terry.bowman@amd.com>,
	<dan.j.williams@intel.com>, <dave@stgolabs.net>,
	<jonathan.cameron@huawei.com>, <dave.jiang@intel.com>,
	<alison.schofield@intel.com>, <bhelgaas@google.com>,
	<shiju.jose@huawei.com>, <ming.li@zohomail.com>,
	<Smita.KoralahalliChannabasappa@amd.com>, <rrichter@amd.com>,
	<dan.carpenter@linaro.org>, <PradeepVineshReddy.Kodamati@amd.com>,
	<lukas@wunner.de>, <Benjamin.Cheatham@amd.com>,
	<sathyanarayanan.kuppuswamy@linux.intel.com>,
	<linux-cxl@vger.kernel.org>
Subject: Re: [PATCH v10 02/17] PCI/CXL: Add pcie_is_cxl()
Date: Thu, 24 Jul 2025 11:00:52 -0700	[thread overview]
Message-ID: <688274d46adc8_134cc7100e4@dwillia2-xfh.jf.intel.com.notmuch> (raw)
In-Reply-To: <4bcbde9e-8795-40ad-b948-56721cfcff8e@amd.com>

Bowman, Terry wrote:
> 
> 
> On 7/23/2025 5:30 PM, dan.j.williams@intel.com wrote:
> > Terry Bowman wrote:
> >> CXL and AER drivers need the ability to identify CXL devices.
> >>
> >> Add set_pcie_cxl() with logic checking for CXL Flexbus DVSEC presence. The
> >> CXL Flexbus DVSEC presence is used because it is required for all the CXL
> >> PCIe devices.[1]
> >>
> >> Add boolean 'struct pci_dev::is_cxl' with the purpose to cache the CXL
> >> Flexbus presence.
> >>
> >> Add function pcie_is_cxl() to return 'struct pci_dev::is_cxl'.
> >>
> >> [1] CXL 3.1 Spec, 8.1.1 PCIe Designated Vendor-Specific Extended
> >>     Capability (DVSEC) ID Assignment, Table 8-2
> >>
[..]
> >> +static void set_pcie_cxl(struct pci_dev *dev)
> >> +{
> >> +	u16 dvsec = pci_find_dvsec_capability(dev, PCI_VENDOR_ID_CXL,
> >> +					      PCI_DVSEC_CXL_FLEXBUS);
> >> +	if (dvsec)
> >> +		dev->is_cxl = 1;
> >> +}
> >> +
> >>  static void set_pcie_untrusted(struct pci_dev *dev)
> >>  {
> >>  	struct pci_dev *parent = pci_upstream_bridge(dev);
> >> @@ -2021,6 +2029,8 @@ int pci_setup_device(struct pci_dev *dev)
> >>  	/* Need to have dev->cfg_size ready */
> >>  	set_pcie_thunderbolt(dev);
> >>  
> >> +	set_pcie_cxl(dev);
> > Per the comment in the header below, in the case of upstream ports and
> > endpoints, this should walk to the parent downstream port and make sure
> > the cxl setting matches. I.e. with hotplug the downstream port may
> > transition from not-cxl to is-cxl. Update downstream-port parents at the
> > beginning of life of their CXL child-devices.
> Ok.
> 
> The check you describe for EPs and UPs would be in is_cxl(), right?

Hmm, no I was considering something like this:

static void set_pcie_cxl(struct pci_dev *dev)
{
        struct pci_dev *parent;
        u16 dvsec = pci_find_dvsec_capability(dev, PCI_VENDOR_ID_CXL,
                                              PCI_DVSEC_CXL_FLEXBUS);
        if (dvsec)
                dev->is_cxl = 1;
        
        if (!pci_is_pcie(dev) ||
            !(pci_pcie_type(dev) == PCI_EXP_TYPE_ENDPOINT ||
              pci_pcie_type(dev) == PCI_EXP_TYPE_UPSTREAM))
                return;

        parent = pci_upstream_bridge(dev);
        set_pcie_cxl(parent);
}

I.e. whenever a new device is scanned there is a chance its upstream
bridge's CXL link has changed state. There is still a chance that the
kernel sees them out of sync for a moment if the link is bouncing, but
for the happy case I think this closes some of the mismatch gap.

  reply	other threads:[~2025-07-24 18:00 UTC|newest]

Thread overview: 87+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-06-26 22:42 [PATCH v10 00/17] Enable CXL PCIe Port Protocol Error handling and logging Terry Bowman
2025-06-26 22:42 ` [PATCH v10 01/17] cxl/pci: Remove unnecessary CXL Endpoint handling helper functions Terry Bowman
2025-07-18 17:55   ` Dave Jiang
2025-07-23 21:58   ` dan.j.williams
2025-07-23 22:15     ` Dave Jiang
2025-06-26 22:42 ` [PATCH v10 02/17] PCI/CXL: Add pcie_is_cxl() Terry Bowman
2025-07-23 22:30   ` dan.j.williams
2025-07-23 23:21     ` Bowman, Terry
2025-07-24 18:00       ` dan.j.williams [this message]
2025-08-09 10:56   ` Alejandro Lucero Palau
2025-08-11 19:14     ` Bowman, Terry
2025-08-11 23:14       ` dan.j.williams
2025-06-26 22:42 ` [PATCH v10 03/17] PCI/AER: Report CXL or PCIe bus error type in trace logging Terry Bowman
2025-06-26 23:25   ` Sathyanarayanan Kuppuswamy
2025-06-27 14:14     ` Bowman, Terry
2025-06-27  9:53   ` Jonathan Cameron
2025-07-02 16:00     ` Bowman, Terry
2025-06-27 11:32   ` Shiju Jose
2025-06-27 14:24     ` Bowman, Terry
2025-07-01 21:27   ` Dave Jiang
2025-07-23 22:56   ` dan.j.williams
2025-06-26 22:42 ` [PATCH v10 04/17] CXL/AER: Introduce CXL specific AER driver file Terry Bowman
2025-06-26 23:42   ` Sathyanarayanan Kuppuswamy
2025-06-27 10:12     ` Jonathan Cameron
2025-06-27 14:29     ` Bowman, Terry
2025-07-24  0:01   ` dan.j.williams
2025-07-24 17:06     ` Bowman, Terry
2025-07-24 20:32       ` dan.j.williams
2025-07-24  1:16   ` dan.j.williams
2025-07-24 17:02     ` Bowman, Terry
2025-07-24 20:23       ` dan.j.williams
2025-06-26 22:42 ` [PATCH v10 05/17] CXL/AER: Introduce kfifo for forwarding CXL errors Terry Bowman
2025-06-27 10:24   ` Jonathan Cameron
2025-07-02 16:21     ` Bowman, Terry
2025-07-02 19:54       ` Dan Carpenter
2025-07-02 19:57         ` Bowman, Terry
2025-07-03 10:06       ` Jonathan Cameron
2025-07-01 21:53   ` Dave Jiang
2025-07-02 17:10     ` Bowman, Terry
2025-07-24  2:01   ` dan.j.williams
2025-07-24 17:21     ` Bowman, Terry
2025-07-24 20:55       ` dan.j.williams
2025-06-26 22:42 ` [PATCH v10 06/17] PCI/AER: Dequeue forwarded CXL error Terry Bowman
2025-06-27 11:00   ` Jonathan Cameron
2025-07-02 17:51     ` Bowman, Terry
2025-07-01 23:04   ` Dave Jiang
2025-07-02 17:56     ` Bowman, Terry
2025-07-03 10:11       ` Jonathan Cameron
2025-07-25  0:38   ` dan.j.williams
2025-06-26 22:42 ` [PATCH v10 07/17] CXL/PCI: Introduce CXL uncorrectable protocol error recovery Terry Bowman
2025-06-27 11:05   ` Jonathan Cameron
2025-07-02 21:06     ` Bowman, Terry
2025-06-27 12:27   ` Shiju Jose
2025-07-02 21:34     ` Bowman, Terry
2025-06-26 22:42 ` [PATCH v10 08/17] cxl/pci: Move RAS initialization to cxl_port driver Terry Bowman
2025-06-27 11:12   ` Jonathan Cameron
2025-07-18 18:01   ` Dave Jiang
2025-06-26 22:42 ` [PATCH v10 09/17] cxl/pci: Map CXL Endpoint Port and CXL Switch Port RAS registers Terry Bowman
2025-06-27 11:17   ` Jonathan Cameron
2025-07-02 21:41     ` Bowman, Terry
2025-07-18 21:28   ` Dave Jiang
2025-07-18 21:55     ` Bowman, Terry
2025-07-18 22:01       ` Dave Jiang
2025-07-18 22:40         ` Bowman, Terry
2025-07-18 22:45           ` Dave Jiang
2025-06-26 22:42 ` [PATCH v10 10/17] cxl/pci: Update RAS handler interfaces to also support CXL Ports Terry Bowman
2025-06-26 22:42 ` [PATCH v10 11/17] cxl/pci: Log message if RAS registers are unmapped Terry Bowman
2025-07-21 21:56   ` Dave Jiang
2025-06-26 22:42 ` [PATCH v10 12/17] cxl/pci: Unify CXL trace logging for CXL Endpoints and CXL Ports Terry Bowman
2025-06-27 12:22   ` Shiju Jose
2025-07-02  1:18     ` Alison Schofield
2025-07-02 22:07       ` Bowman, Terry
2025-07-02 21:56     ` Bowman, Terry
2025-06-26 22:42 ` [PATCH v10 13/17] cxl/pci: Update cxl_handle_cor_ras() to return early if no RAS errors Terry Bowman
2025-06-27 11:48   ` Jonathan Cameron
2025-07-21 22:17   ` Dave Jiang
2025-06-26 22:42 ` [PATCH v10 14/17] cxl/pci: Introduce CXL Endpoint protocol error handlers Terry Bowman
2025-06-27 11:52   ` Jonathan Cameron
2025-06-27 12:27   ` Shiju Jose
2025-07-21 22:35   ` Dave Jiang
2025-07-22 18:23     ` Bowman, Terry
2025-06-26 22:42 ` [PATCH v10 15/17] CXL/PCI: Introduce CXL Port " Terry Bowman
2025-06-26 22:42 ` [PATCH v10 16/17] CXL/PCI: Enable CXL protocol errors during CXL Port probe Terry Bowman
2025-06-26 22:42 ` [PATCH v10 17/17] CXL/PCI: Disable CXL protocol error interrupts during CXL Port cleanup Terry Bowman
2025-07-23 21:55 ` [PATCH v10 00/17] Enable CXL PCIe Port Protocol Error handling and logging dan.j.williams
2025-07-24 15:58   ` Bowman, Terry
2025-08-18 15:18 ` Joshua Hahn

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=688274d46adc8_134cc7100e4@dwillia2-xfh.jf.intel.com.notmuch \
    --to=dan.j.williams@intel.com \
    --cc=Benjamin.Cheatham@amd.com \
    --cc=PradeepVineshReddy.Kodamati@amd.com \
    --cc=Smita.KoralahalliChannabasappa@amd.com \
    --cc=alison.schofield@intel.com \
    --cc=bhelgaas@google.com \
    --cc=dan.carpenter@linaro.org \
    --cc=dave.jiang@intel.com \
    --cc=dave@stgolabs.net \
    --cc=jonathan.cameron@huawei.com \
    --cc=linux-cxl@vger.kernel.org \
    --cc=lukas@wunner.de \
    --cc=ming.li@zohomail.com \
    --cc=rrichter@amd.com \
    --cc=sathyanarayanan.kuppuswamy@linux.intel.com \
    --cc=shiju.jose@huawei.com \
    --cc=terry.bowman@amd.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).