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: <20240723215040.GA760037@bhelgaas>
Date: Tue, 23 Jul 2024 16:50:40 -0500
From: Bjorn Helgaas <helgaas@...nel.org>
To: Wei Huang <wei.huang2@....com>
Cc: linux-pci@...r.kernel.org, linux-kernel@...r.kernel.org,
	linux-doc@...r.kernel.org, netdev@...r.kernel.org,
	Jonathan.Cameron@...wei.com, corbet@....net, davem@...emloft.net,
	edumazet@...gle.com, kuba@...nel.org, pabeni@...hat.com,
	alex.williamson@...hat.com, gospo@...adcom.com,
	michael.chan@...adcom.com, ajit.khaparde@...adcom.com,
	somnath.kotur@...adcom.com, andrew.gospodarek@...adcom.com,
	manoj.panicker2@....com, Eric.VanTassell@....com,
	vadim.fedorenko@...ux.dev, horms@...nel.org, bagasdotme@...il.com,
	bhelgaas@...gle.com
Subject: Re: [PATCH V3 08/10] PCI/TPH: Add TPH documentation

On Wed, Jul 17, 2024 at 03:55:09PM -0500, Wei Huang wrote:
> Provide a document for TPH feature, including the description of
> kernel options and driver API interface.
> 
> Co-developed-by: Eric Van Tassell <Eric.VanTassell@....com>
> Signed-off-by: Eric Van Tassell <Eric.VanTassell@....com>
> Signed-off-by: Wei Huang <wei.huang2@....com>
> Reviewed-by: Ajit Khaparde <ajit.khaparde@...adcom.com>
> Reviewed-by: Somnath Kotur <somnath.kotur@...adcom.com>
> Reviewed-by: Andy Gospodarek <andrew.gospodarek@...adcom.com>
> Reviewed-by: Jonathan Cameron <Jonathan.Cameron@...wei.com>
> ---
>  Documentation/PCI/index.rst          |  1 +
>  Documentation/PCI/tph.rst            | 57 ++++++++++++++++++++++++++++

Wrap file to fit in 80 columns.  Looks like about 86 now, which is
kind of a random size.

>  Documentation/driver-api/pci/pci.rst |  3 ++
>  3 files changed, 61 insertions(+)
>  create mode 100644 Documentation/PCI/tph.rst
> 
> diff --git a/Documentation/PCI/index.rst b/Documentation/PCI/index.rst
> index e73f84aebde3..5e7c4e6e726b 100644
> --- a/Documentation/PCI/index.rst
> +++ b/Documentation/PCI/index.rst
> @@ -18,3 +18,4 @@ PCI Bus Subsystem
>     pcieaer-howto
>     endpoint/index
>     boot-interrupts
> +   tph
> diff --git a/Documentation/PCI/tph.rst b/Documentation/PCI/tph.rst
> new file mode 100644
> index 000000000000..103f4c3251e2
> --- /dev/null
> +++ b/Documentation/PCI/tph.rst
> @@ -0,0 +1,57 @@
> +.. SPDX-License-Identifier: GPL-2.0
> +
> +===========
> +TPH Support
> +===========
> +
> +
> +:Copyright: 2024 Advanced Micro Devices, Inc.
> +:Authors: - Eric van Tassell <eric.vantassell@....com>
> +          - Wei Huang <wei.huang2@....com>
> +
> +Overview
> +========
> +TPH (TLP Processing Hints) is a PCIe feature that allows endpoint devices
> +to provide optimization hints, such as desired caching behavior, for
> +requests that target memory space. These hints, in a format called steering
> +tags, are provided in the requester's TLP headers and can empower the system
> +hardware, including the Root Complex, to optimize the utilization of platform
> +resources for the requests.

I see that this is basically cribbed from sec 6.17, but it sort of
conflates Processing Hints (Bi-directional data structure, Requester,
Target, Target with Priority) with Steering Tags (hints about where to
target a TLP), and "optimize the utilization of platform resources for
the requests" is pretty vague marketing-speak.

I think it might be useful to at least mention the Processing Hints
because the distinction between PH and ST helps motivate the need for
both "pci=notph" and "pci=nostmode".

IIUC, we can enable TPH with the TPH Requester Enable bit, but I don't
see an architected mechanism to control the PH bits in the TLP header.
I assume there might be device-specific ways or it might be built into
the hardware.

Use "Steering Tags" to indicate that this is a term defined by the
spec.

> +User Guide
> +==========
> +
> +Kernel Options
> +--------------
> +There are two kernel command line options available to control TPH feature
> +
> +   * "notph": TPH will be disabled for all endpoint devices.
> +   * "nostmode": TPH will be enabled but the ST Mode will be forced to "No ST Mode".
> +
> +Device Driver API
> +-----------------
> +In brief, an endpoint device driver using the TPH interface to configure
> +Interrupt Vector Mode will call pcie_tph_set_st() when setting up MSI-X
> +interrupts as shown below:

I think we should include a spec reference for more details here,
e.g., PCIe r6.0, sec 6.17, 6.17.3.

> +.. code-block:: c
> +
> +    for (i = 0, j = 0; i < nr_rings; i++) {

"j" is not relevant here and makes the example unnecessarily
complicated.

> +        ...
> +        rc = request_irq(irq->vector, irq->handler, flags, irq->name, NULL);
> +        ...
> +        if (!pcie_tph_set_st(pdev, i, cpumask_first(irq->cpu_mask),
> +                             TPH_MEM_TYPE_VM, PCI_TPH_REQ_TPH_ONLY))
> +               pr_err("Error in configuring steering tag\n");

pci_err().  I don't want to encourage drivers to print messages that
have no connection to a specific device.

> +        ...
> +    }
> +
> +The caller is suggested to check if interrupt vector mode is supported using
> +pcie_tph_intr_vec_supported() before updating the steering tags.

I guess this refers to "Interrupt Vector Mode", an ST mode of
operation (PCIe r6.0, sec 6.17.3).  Helpful to style this as
"Interrupt Vector Mode" to indicate that it is a technical term
defined in the spec.

> If a device only
> +supports TPH vendor specific mode, its driver can call pcie_tph_get_st_from_acpi()
> +to retrieve the steering tag for a specific CPU and uses the tag to control TPH
> +behavior.

I'm not sure what "vendor specific mode" refers to here.  If it's the
"Device Specific" ST Mode, use the exact language from the spec to
help people find the details.

> +.. kernel-doc:: drivers/pci/pcie/tph.c
> +   :export:
> +   :identifiers: pcie_tph_intr_vec_supported pcie_tph_get_st_from_acpi pcie_tph_set_st
> diff --git a/Documentation/driver-api/pci/pci.rst b/Documentation/driver-api/pci/pci.rst
> index aa40b1cc243b..3d896b2cf16e 100644
> --- a/Documentation/driver-api/pci/pci.rst
> +++ b/Documentation/driver-api/pci/pci.rst
> @@ -46,6 +46,9 @@ PCI Support Library
>  .. kernel-doc:: drivers/pci/pci-sysfs.c
>     :internal:
>  
> +.. kernel-doc:: drivers/pci/pcie/tph.c
> +   :export:
> +
>  PCI Hotplug Support Library
>  ---------------------------
>  
> -- 
> 2.45.1
> 

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ