mbox series

[v7,0/7] Add suspend to ram support for PCIe on J7200

Message ID 20240102-j7200-pcie-s2r-v7-0-a2f9156da6c3@bootlin.com
Headers show
Series Add suspend to ram support for PCIe on J7200 | expand

Message

Thomas Richard June 19, 2024, 10:15 a.m. UTC 
This adds suspend to ram support for the PCIe (RC mode) on J7200 platform.

In this 7th iteration, the i2c and mux patches were moved to dedicated
series ([1] and [2]).
The patch for the gpio-pca953x driver was removed. It will be sent
separately for further testing and discussion.

No merge conflict with 6.10-rc4.

[1]: https://lore.kernel.org/all/20240613-i2c-omap-wakeup-controller-during-suspend-v1-0-aab001eb1ad1@bootlin.com/
[2]: https://lore.kernel.org/all/20240613-mux-mmio-resume-support-v1-0-4525bf56024a@bootlin.com/

Regards,

Thomas

Signed-off-by: Thomas Richard <thomas.richard@bootlin.com>
---
Changes in v7:
- all: series rebased on Linux 6.10-rc4.
- i2c: patches moved to a dedicated series.
- mux: patches moved to a dedicated series.
- gpio-pca953x: patch removed, will be sent separately.
- Link to v6: https://lore.kernel.org/r/20240102-j7200-pcie-s2r-v6-0-4656ef6e6d66@bootlin.com

Changes in v6:
- i2c-omap: add a patch to remove __maybe_unused attribute of
  omap_i2c_runtime_suspend() and omap_i2c_runtime_resume()
- i2c-omap: fix compile issue if CONFIG_PM_SLEEP is not set
- Link to v5: https://lore.kernel.org/r/20240102-j7200-pcie-s2r-v5-0-4b8c46711ded@bootlin.com

Changes in v5:
- all: series rebased on Linux 6.9-rc1
- pinctrl-single: patch removed (already applied to the pinctrl tree)
- phy: patches moved to a dedicated series.
- pci: add T_PERST_CLK_US macro.
- pci-j721e: update the comments about T_PERST_CLK_US.
- Link to v4: https://lore.kernel.org/r/20240102-j7200-pcie-s2r-v4-0-6f1f53390c85@bootlin.com

Changes in v4:
- all: use SoB/Co-developed-by for patches initially developed by Théo
  Lebrun.
- pinctrl-single: squash the two commits.
- i2c-omap: fix line lenghts of the comment in omap_i2c_suspend().
- mux: mux_chip_resume() return 0 or at the first error.
- phy-j721e-wiz: clean code around dev_err_probe().
- phy-j721e-wiz: use REF_CLK_100MHZ macros.
- pci: fix subject line for all PCI patches.
- pci-cadence: use fsleep() instead of usleep_range().
- pci-cadence: remove cdns_torrent_clk_cleanup() call in
  cdns_torrent_phy_resume_noirq().
- pci-j721e: add a patch to use dev_err_probe() instead of dev_err() in the probe().
- pci-j721e: fix unordered header files.
- pci-j721e: remove some log spammers.
- pci-j721e: add a missing clock disable in j721e_pcie_resume_noirq().
- pci-j721e: simplify the patch "Add reset GPIO to struct j721e_pcie"
- Link to v3: https://lore.kernel.org/r/20240102-j7200-pcie-s2r-v3-0-5c2e4a3fac1f@bootlin.com

Changes in v3:
- pinctrl-single: split patch in two parts, a first patch to remove the
  dead code, a second to move suspend()/resume() callbacks to noirq.
- i2c-omap: add a comments above the suspend_noirq() callback.
- mux: now mux_chip_resume() try to restores all muxes, then return 0 if
  all is ok or the first failure.
- mmio: fix commit message.
- phy-j721e-wiz: add a patch to use dev_err_probe() instead of dev_err() in
  the wiz_clock_init() function.
- phy-j721e-wiz: remove probe boolean for the wiz_clock_init(), rename the
  function to wiz_clock_probe(), extract hardware configuration part in a
  new function wiz_clock_init().
- phy-cadence-torrent: use dev_err_probe() and fix commit messages
- pcie-cadence-host: remove probe boolean for the cdns_pcie_host_setup()
  function and extract the link setup part in a new function
  cdns_pcie_host_link_setup().
- pcie-cadence-host: make cdns_pcie_host_init() global.
- pci-j721e: use the cdns_pcie_host_link_setup() cdns_pcie_host_init()
  functions in the resume_noirq() callback.
- Link to v2: https://lore.kernel.org/r/20240102-j7200-pcie-s2r-v2-0-8e4f7d228ec2@bootlin.com

Changes in v2:
- all: fix commits messages.
- all: use DEFINE_NOIRQ_DEV_PM_OPS and pm_sleep_ptr macros.
- all: remove useless #ifdef CONFIG_PM.
- pinctrl-single: drop dead code
- mux: add mux_chip_resume() function in mux core.
- mmio: resume sequence is now a call to mux_chip_resume().
- phy-cadence-torrent: fix typo in resume sequence (reset_control_assert()
  instead of reset_control_put()).
- phy-cadence-torrent: use PHY instead of phy.
- pci-j721e: do not shadow cdns_pcie_host_setup return code in resume
  sequence.
- pci-j721e: drop dead code.
- Link to v1: https://lore.kernel.org/r/20240102-j7200-pcie-s2r-v1-0-84e55da52400@bootlin.com

---
Thomas Richard (5):
      PCI: cadence: Extract link setup sequence from cdns_pcie_host_setup()
      PCI: cadence: Set cdns_pcie_host_init() global
      PCI: j721e: Use dev_err_probe() in the probe() function
      PCI: Add T_PERST_CLK_US macro
      PCI: j721e: Use T_PERST_CLK_US macro

Théo Lebrun (2):
      PCI: j721e: Add reset GPIO to struct j721e_pcie
      PCI: j721e: Add suspend and resume support

 drivers/pci/controller/cadence/pci-j721e.c         | 121 ++++++++++++++++++---
 drivers/pci/controller/cadence/pcie-cadence-host.c |  44 +++++---
 drivers/pci/controller/cadence/pcie-cadence.h      |  12 ++
 drivers/pci/pci.h                                  |   3 +
 4 files changed, 146 insertions(+), 34 deletions(-)
---
base-commit: 7510725e693fb5e4b4cd17086cc5a49a7a065b9c
change-id: 20240102-j7200-pcie-s2r-ecb1a979e357

Best regards,

Comments

Gregory CLEMENT July 16, 2024, 9:44 a.m. UTC | #1 
Hello PCIe maintainers!

> This adds suspend to ram support for the PCIe (RC mode) on J7200 platform.
>
> In this 7th iteration, the i2c and mux patches were moved to dedicated
> series ([1] and [2]).
> The patch for the gpio-pca953x driver was removed. It will be sent
> separately for further testing and discussion.
>
> No merge conflict with 6.10-rc4.
>
> [1]: https://lore.kernel.org/all/20240613-i2c-omap-wakeup-controller-during-suspend-v1-0-aab001eb1ad1@bootlin.com/
> [2]:
> https://lore.kernel.org/all/20240613-mux-mmio-resume-support-v1-0-4525bf56024a@bootlin.com/

I noticed that this series has not been merged yet, even though it
seemed that there were no remaining comments to address.

Is there any reason why it has not been applied yet?

Thanks,

Gregory

>
> Regards,
>
> Thomas
>
> Signed-off-by: Thomas Richard <thomas.richard@bootlin.com>
> ---
> Changes in v7:
> - all: series rebased on Linux 6.10-rc4.
> - i2c: patches moved to a dedicated series.
> - mux: patches moved to a dedicated series.
> - gpio-pca953x: patch removed, will be sent separately.
> - Link to v6: https://lore.kernel.org/r/20240102-j7200-pcie-s2r-v6-0-4656ef6e6d66@bootlin.com
>
> Changes in v6:
> - i2c-omap: add a patch to remove __maybe_unused attribute of
>   omap_i2c_runtime_suspend() and omap_i2c_runtime_resume()
> - i2c-omap: fix compile issue if CONFIG_PM_SLEEP is not set
> - Link to v5: https://lore.kernel.org/r/20240102-j7200-pcie-s2r-v5-0-4b8c46711ded@bootlin.com
>
> Changes in v5:
> - all: series rebased on Linux 6.9-rc1
> - pinctrl-single: patch removed (already applied to the pinctrl tree)
> - phy: patches moved to a dedicated series.
> - pci: add T_PERST_CLK_US macro.
> - pci-j721e: update the comments about T_PERST_CLK_US.
> - Link to v4: https://lore.kernel.org/r/20240102-j7200-pcie-s2r-v4-0-6f1f53390c85@bootlin.com
>
> Changes in v4:
> - all: use SoB/Co-developed-by for patches initially developed by Théo
>   Lebrun.
> - pinctrl-single: squash the two commits.
> - i2c-omap: fix line lenghts of the comment in omap_i2c_suspend().
> - mux: mux_chip_resume() return 0 or at the first error.
> - phy-j721e-wiz: clean code around dev_err_probe().
> - phy-j721e-wiz: use REF_CLK_100MHZ macros.
> - pci: fix subject line for all PCI patches.
> - pci-cadence: use fsleep() instead of usleep_range().
> - pci-cadence: remove cdns_torrent_clk_cleanup() call in
>   cdns_torrent_phy_resume_noirq().
> - pci-j721e: add a patch to use dev_err_probe() instead of dev_err() in the probe().
> - pci-j721e: fix unordered header files.
> - pci-j721e: remove some log spammers.
> - pci-j721e: add a missing clock disable in j721e_pcie_resume_noirq().
> - pci-j721e: simplify the patch "Add reset GPIO to struct j721e_pcie"
> - Link to v3: https://lore.kernel.org/r/20240102-j7200-pcie-s2r-v3-0-5c2e4a3fac1f@bootlin.com
>
> Changes in v3:
> - pinctrl-single: split patch in two parts, a first patch to remove the
>   dead code, a second to move suspend()/resume() callbacks to noirq.
> - i2c-omap: add a comments above the suspend_noirq() callback.
> - mux: now mux_chip_resume() try to restores all muxes, then return 0 if
>   all is ok or the first failure.
> - mmio: fix commit message.
> - phy-j721e-wiz: add a patch to use dev_err_probe() instead of dev_err() in
>   the wiz_clock_init() function.
> - phy-j721e-wiz: remove probe boolean for the wiz_clock_init(), rename the
>   function to wiz_clock_probe(), extract hardware configuration part in a
>   new function wiz_clock_init().
> - phy-cadence-torrent: use dev_err_probe() and fix commit messages
> - pcie-cadence-host: remove probe boolean for the cdns_pcie_host_setup()
>   function and extract the link setup part in a new function
>   cdns_pcie_host_link_setup().
> - pcie-cadence-host: make cdns_pcie_host_init() global.
> - pci-j721e: use the cdns_pcie_host_link_setup() cdns_pcie_host_init()
>   functions in the resume_noirq() callback.
> - Link to v2: https://lore.kernel.org/r/20240102-j7200-pcie-s2r-v2-0-8e4f7d228ec2@bootlin.com
>
> Changes in v2:
> - all: fix commits messages.
> - all: use DEFINE_NOIRQ_DEV_PM_OPS and pm_sleep_ptr macros.
> - all: remove useless #ifdef CONFIG_PM.
> - pinctrl-single: drop dead code
> - mux: add mux_chip_resume() function in mux core.
> - mmio: resume sequence is now a call to mux_chip_resume().
> - phy-cadence-torrent: fix typo in resume sequence (reset_control_assert()
>   instead of reset_control_put()).
> - phy-cadence-torrent: use PHY instead of phy.
> - pci-j721e: do not shadow cdns_pcie_host_setup return code in resume
>   sequence.
> - pci-j721e: drop dead code.
> - Link to v1: https://lore.kernel.org/r/20240102-j7200-pcie-s2r-v1-0-84e55da52400@bootlin.com
>
> ---
> Thomas Richard (5):
>       PCI: cadence: Extract link setup sequence from cdns_pcie_host_setup()
>       PCI: cadence: Set cdns_pcie_host_init() global
>       PCI: j721e: Use dev_err_probe() in the probe() function
>       PCI: Add T_PERST_CLK_US macro
>       PCI: j721e: Use T_PERST_CLK_US macro
>
> Théo Lebrun (2):
>       PCI: j721e: Add reset GPIO to struct j721e_pcie
>       PCI: j721e: Add suspend and resume support
>
>  drivers/pci/controller/cadence/pci-j721e.c         | 121 ++++++++++++++++++---
>  drivers/pci/controller/cadence/pcie-cadence-host.c |  44 +++++---
>  drivers/pci/controller/cadence/pcie-cadence.h      |  12 ++
>  drivers/pci/pci.h                                  |   3 +
>  4 files changed, 146 insertions(+), 34 deletions(-)
> ---
> base-commit: 7510725e693fb5e4b4cd17086cc5a49a7a065b9c
> change-id: 20240102-j7200-pcie-s2r-ecb1a979e357
>
> Best regards,
> -- 
> Thomas Richard <thomas.richard@bootlin.com>
Thomas Richard Sept. 3, 2024, 12:41 p.m. UTC | #2 
On 6/19/24 12:15, Thomas Richard wrote:
> This adds suspend to ram support for the PCIe (RC mode) on J7200 platform.
> 
> In this 7th iteration, the i2c and mux patches were moved to dedicated
> series ([1] and [2]).
> The patch for the gpio-pca953x driver was removed. It will be sent
> separately for further testing and discussion.
> 
> No merge conflict with 6.10-rc4.
> 
> [1]: https://lore.kernel.org/all/20240613-i2c-omap-wakeup-controller-during-suspend-v1-0-aab001eb1ad1@bootlin.com/
> [2]: https://lore.kernel.org/all/20240613-mux-mmio-resume-support-v1-0-4525bf56024a@bootlin.com/

Hello !!

This series has no remaining comment to address.
Is there any chance to get this series merged ?

Best Regards,

Thomas

> 
> Regards,
> 
> Thomas
> 
> Signed-off-by: Thomas Richard <thomas.richard@bootlin.com>
> ---
> Changes in v7:
> - all: series rebased on Linux 6.10-rc4.
> - i2c: patches moved to a dedicated series.
> - mux: patches moved to a dedicated series.
> - gpio-pca953x: patch removed, will be sent separately.
> - Link to v6: https://lore.kernel.org/r/20240102-j7200-pcie-s2r-v6-0-4656ef6e6d66@bootlin.com
> 
> Changes in v6:
> - i2c-omap: add a patch to remove __maybe_unused attribute of
>   omap_i2c_runtime_suspend() and omap_i2c_runtime_resume()
> - i2c-omap: fix compile issue if CONFIG_PM_SLEEP is not set
> - Link to v5: https://lore.kernel.org/r/20240102-j7200-pcie-s2r-v5-0-4b8c46711ded@bootlin.com
> 
> Changes in v5:
> - all: series rebased on Linux 6.9-rc1
> - pinctrl-single: patch removed (already applied to the pinctrl tree)
> - phy: patches moved to a dedicated series.
> - pci: add T_PERST_CLK_US macro.
> - pci-j721e: update the comments about T_PERST_CLK_US.
> - Link to v4: https://lore.kernel.org/r/20240102-j7200-pcie-s2r-v4-0-6f1f53390c85@bootlin.com
> 
> Changes in v4:
> - all: use SoB/Co-developed-by for patches initially developed by Théo
>   Lebrun.
> - pinctrl-single: squash the two commits.
> - i2c-omap: fix line lenghts of the comment in omap_i2c_suspend().
> - mux: mux_chip_resume() return 0 or at the first error.
> - phy-j721e-wiz: clean code around dev_err_probe().
> - phy-j721e-wiz: use REF_CLK_100MHZ macros.
> - pci: fix subject line for all PCI patches.
> - pci-cadence: use fsleep() instead of usleep_range().
> - pci-cadence: remove cdns_torrent_clk_cleanup() call in
>   cdns_torrent_phy_resume_noirq().
> - pci-j721e: add a patch to use dev_err_probe() instead of dev_err() in the probe().
> - pci-j721e: fix unordered header files.
> - pci-j721e: remove some log spammers.
> - pci-j721e: add a missing clock disable in j721e_pcie_resume_noirq().
> - pci-j721e: simplify the patch "Add reset GPIO to struct j721e_pcie"
> - Link to v3: https://lore.kernel.org/r/20240102-j7200-pcie-s2r-v3-0-5c2e4a3fac1f@bootlin.com
> 
> Changes in v3:
> - pinctrl-single: split patch in two parts, a first patch to remove the
>   dead code, a second to move suspend()/resume() callbacks to noirq.
> - i2c-omap: add a comments above the suspend_noirq() callback.
> - mux: now mux_chip_resume() try to restores all muxes, then return 0 if
>   all is ok or the first failure.
> - mmio: fix commit message.
> - phy-j721e-wiz: add a patch to use dev_err_probe() instead of dev_err() in
>   the wiz_clock_init() function.
> - phy-j721e-wiz: remove probe boolean for the wiz_clock_init(), rename the
>   function to wiz_clock_probe(), extract hardware configuration part in a
>   new function wiz_clock_init().
> - phy-cadence-torrent: use dev_err_probe() and fix commit messages
> - pcie-cadence-host: remove probe boolean for the cdns_pcie_host_setup()
>   function and extract the link setup part in a new function
>   cdns_pcie_host_link_setup().
> - pcie-cadence-host: make cdns_pcie_host_init() global.
> - pci-j721e: use the cdns_pcie_host_link_setup() cdns_pcie_host_init()
>   functions in the resume_noirq() callback.
> - Link to v2: https://lore.kernel.org/r/20240102-j7200-pcie-s2r-v2-0-8e4f7d228ec2@bootlin.com
> 
> Changes in v2:
> - all: fix commits messages.
> - all: use DEFINE_NOIRQ_DEV_PM_OPS and pm_sleep_ptr macros.
> - all: remove useless #ifdef CONFIG_PM.
> - pinctrl-single: drop dead code
> - mux: add mux_chip_resume() function in mux core.
> - mmio: resume sequence is now a call to mux_chip_resume().
> - phy-cadence-torrent: fix typo in resume sequence (reset_control_assert()
>   instead of reset_control_put()).
> - phy-cadence-torrent: use PHY instead of phy.
> - pci-j721e: do not shadow cdns_pcie_host_setup return code in resume
>   sequence.
> - pci-j721e: drop dead code.
> - Link to v1: https://lore.kernel.org/r/20240102-j7200-pcie-s2r-v1-0-84e55da52400@bootlin.com
> 
> ---
> Thomas Richard (5):
>       PCI: cadence: Extract link setup sequence from cdns_pcie_host_setup()
>       PCI: cadence: Set cdns_pcie_host_init() global
>       PCI: j721e: Use dev_err_probe() in the probe() function
>       PCI: Add T_PERST_CLK_US macro
>       PCI: j721e: Use T_PERST_CLK_US macro
> 
> Théo Lebrun (2):
>       PCI: j721e: Add reset GPIO to struct j721e_pcie
>       PCI: j721e: Add suspend and resume support
> 
>  drivers/pci/controller/cadence/pci-j721e.c         | 121 ++++++++++++++++++---
>  drivers/pci/controller/cadence/pcie-cadence-host.c |  44 +++++---
>  drivers/pci/controller/cadence/pcie-cadence.h      |  12 ++
>  drivers/pci/pci.h                                  |   3 +
>  4 files changed, 146 insertions(+), 34 deletions(-)
> ---
> base-commit: 7510725e693fb5e4b4cd17086cc5a49a7a065b9c
> change-id: 20240102-j7200-pcie-s2r-ecb1a979e357
> 
> Best regards,
Krzysztof Wilczyński Sept. 3, 2024, 6:32 p.m. UTC | #3 
Hello,

> This series has no remaining comment to address.

No, nothing.  Everything looks good.

> Is there any chance to get this series merged ?

Yes.  Apologies for the delay.

Bjorn, should be able to pull it for 6.12, hopefully.

	Krzysztof