linux-kernel - [PATCH v4 0/4] PCI: Allow D3Hot for PCI bridges in Devicetree based platforms

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-next>] [day] [month] [year] [list]

Message-Id: <20240326-pci-bridge-d3-v4-0-f1dce1d1f648@linaro.org>
Date: Tue, 26 Mar 2024 16:18:16 +0530
From: Manivannan Sadhasivam <manivannan.sadhasivam@...aro.org>
To: Bjorn Helgaas <bhelgaas@...gle.com>, 
 "Rafael J. Wysocki" <rafael@...nel.org>, Len Brown <lenb@...nel.org>
Cc: linux-pci@...r.kernel.org, linux-kernel@...r.kernel.org, 
 linux-acpi@...r.kernel.org, lukas@...ner.de, 
 mika.westerberg@...ux.intel.com, 
 Manivannan Sadhasivam <manivannan.sadhasivam@...aro.org>, 
 Bjorn Helgaas <helgaas@...nel.org>
Subject: [PATCH v4 0/4] PCI: Allow D3Hot for PCI bridges in Devicetree
 based platforms

Hi,

This series allows D3Hot for PCI bridges in Devicetree based platforms.
Even though most of the bridges in Devicetree platforms support D3Hot, PCI
core will allow D3Hot only when one of the following conditions are met:

1. Platform is ACPI based
2. Thunderbolt controller is used
3. pcie_port_pm=force passed in cmdline

While options 1 and 2 do not apply to most of the DT based platforms,
option 3 will make the life harder for distro maintainers.

Initially, I tried to fix this issue by using a Devicetree property [1], but
that was rejected by Bjorn and it was concluded that D3Hot should be allowed by
default for all the Devicetree based platforms.

During the review of v3 series, Bjorn noted several shortcomings of the
pci_bridge_d3_possible() API [2] and I tried to address them in this series as
well.

But please note that the patches 2 and 3 needs closer review from ACPI and x86
folks since I've splitted the D3Hot and D3Cold handling based on my little
understanding of the code.

Testing
=======

This series is tested on SM8450 based development board on top of [3].

- Mani

[1] https://lore.kernel.org/linux-pci/20240214-pcie-qcom-bridge-v3-1-3a713bbc1fd7@linaro.org/
[2] https://lore.kernel.org/linux-pci/20240305175107.GA539676@bhelgaas/
[3] https://lore.kernel.org/linux-arm-msm/20240321-pcie-qcom-bridge-dts-v2-0-1eb790c53e43@linaro.org/

Changes in v4:
- Added pci_bridge_d3_possible() rework based on comments from Bjorn
- Got rid of the DT property and allowed D3Hot by default on all DT platforms

Changes in v3:
- Fixed kdoc, used of_property_present() and dev_of_node() (Lukas)
- Link to v2: https://lore.kernel.org/r/20240214-pcie-qcom-bridge-v2-1-9dd6dbb1b817@linaro.org

Changes in v2:
- Switched to DT based approach as suggested by Lukas.
- Link to v1: https://lore.kernel.org/r/20240202-pcie-qcom-bridge-v1-0-46d7789836c0@linaro.org

Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@...aro.org>
---
Manivannan Sadhasivam (4):
      PCI/portdrv: Make use of pci_dev::bridge_d3 for checking the D3 possibility
      PCI: Rename pci_bridge_d3_possible() to pci_bridge_d3_allowed()
      PCI: Decouple D3Hot and D3Cold handling for bridges
      PCI: Allow PCI bridges to go to D3Hot on all Devicetree based platforms

 drivers/pci/bus.c          |  2 +-
 drivers/pci/pci-acpi.c     |  9 ++---
 drivers/pci/pci-sysfs.c    |  2 +-
 drivers/pci/pci.c          | 90 ++++++++++++++++++++++++++++++++--------------
 drivers/pci/pci.h          | 12 ++++---
 drivers/pci/pcie/portdrv.c | 16 ++++-----
 drivers/pci/remove.c       |  2 +-
 include/linux/pci.h        |  3 +-
 8 files changed, 89 insertions(+), 47 deletions(-)
---
base-commit: 705c1da8fa4816fb0159b5602fef1df5946a3ee2
change-id: 20240320-pci-bridge-d3-092e2beac438

Best regards,
-- 
Manivannan Sadhasivam <manivannan.sadhasivam@...aro.org>