Linux-ARM-Kernel Archive on lore.kernel.org
 help / color / mirror / Atom feed
* Re:[PATCH v7 08/17] ACPI: pci_link: Clear the dependencies after probe
       [not found] <20240729142241.733357-9-sunilvl@ventanamicro.com>
@ 2025-12-01 13:07 ` huyuye
  2025-12-01 13:33   ` [PATCH " Sunil V L
  2025-12-01 14:12 ` huyuye
  1 sibling, 1 reply; 4+ messages in thread
From: huyuye @ 2025-12-01 13:07 UTC (permalink / raw)
  To: sunilvl
  Cc: ajones, anup, aou, atishp, bhelgaas, catalin.marinas,
	conor.dooley, dfustini, haibo1.xu, lenb, linux-arm-kernel,
	linux-kernel, linux-pci, linux-riscv, palmer, rafael,
	robert.moore, samuel.holland, tglx, will, dai.hualiang,
	deng.weixian, guo.chang2, liu.qingtao2, wu.jiabao, lin.yongchun,
	hu.yuye, zhang.longxiang, huyuye

Hi, sunilvl

Based on the above patch, I understand that you previously resolved dependencies between Link devices and PCI Host Bridges by calling acpi_dev_clear_dependencies(device). I would like to ask: on RISC‑V platforms, if we need to manage dependencies between multiple PCI Host Bridges, could this be addressed by adding a call to acpi_dev_clear_dependencies(device) at the end of the acpi_pci_root_add enumeration function?

Initialization order dependencies can be defined via the ACPI _DEP method in the DSDT. For example, if host bridge B depends on host bridge A, bridge B should not be enumerated until bridge A is fully initialized.

Best regards,
Yuye



^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH v7 08/17] ACPI: pci_link: Clear the dependencies after probe
  2025-12-01 13:07 ` Re:[PATCH v7 08/17] ACPI: pci_link: Clear the dependencies after probe huyuye
@ 2025-12-01 13:33   ` Sunil V L
  0 siblings, 0 replies; 4+ messages in thread
From: Sunil V L @ 2025-12-01 13:33 UTC (permalink / raw)
  To: huyuye
  Cc: ajones, anup, aou, atishp, bhelgaas, catalin.marinas,
	conor.dooley, dfustini, haibo1.xu, lenb, linux-arm-kernel,
	linux-kernel, linux-pci, linux-riscv, palmer, rafael,
	robert.moore, samuel.holland, tglx, will, dai.hualiang,
	deng.weixian, guo.chang2, liu.qingtao2, wu.jiabao, lin.yongchun,
	hu.yuye, zhang.longxiang

On Mon, Dec 01, 2025 at 09:07:55PM +0800, huyuye wrote:
> Hi, sunilvl
> 
> Based on the above patch, I understand that you previously resolved dependencies between Link devices and PCI Host Bridges by calling acpi_dev_clear_dependencies(device). I would like to ask: on RISC‑V platforms, if we need to manage dependencies between multiple PCI Host Bridges, could this be addressed by adding a call to acpi_dev_clear_dependencies(device) at the end of the acpi_pci_root_add enumeration function?
> 
> Initialization order dependencies can be defined via the ACPI _DEP method in the DSDT. For example, if host bridge B depends on host bridge A, bridge B should not be enumerated until bridge A is fully initialized.
> 
Yes, that should work.

Regards,
Sunil


^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re:[PATCH v7 08/17] ACPI: pci_link: Clear the dependencies after probe
       [not found] <20240729142241.733357-9-sunilvl@ventanamicro.com>
  2025-12-01 13:07 ` Re:[PATCH v7 08/17] ACPI: pci_link: Clear the dependencies after probe huyuye
@ 2025-12-01 14:12 ` huyuye
  2025-12-02  4:28   ` [PATCH " Sunil V L
  1 sibling, 1 reply; 4+ messages in thread
From: huyuye @ 2025-12-01 14:12 UTC (permalink / raw)
  To: sunilvl
  Cc: ajones, anup, aou, atishp, bhelgaas, catalin.marinas,
	conor.dooley, dfustini, haibo1.xu, lenb, linux-arm-kernel,
	linux-kernel, linux-pci, linux-riscv, palmer, rafael,
	robert.moore, samuel.holland, tglx, will, dai.hualiang,
	deng.weixian, guo.chang2, liu.qingtao2, wu.jiabao, lin.yongchun,
	hu.yuye, zhang.longxiang, huyuye

> Hi, sunilvl

> Based on the above patch, I understand that you previously >resolved dependencies between Link devices and PCI Host Bridges by >calling acpi_dev_clear_dependencies(device). I would like to ask: >on RISC‑V platforms, if we need to manage dependencies between >multiple PCI Host Bridges, could this be addressed by adding a >call to acpi_dev_clear_dependencies(device) at the end of the >acpi_pci_root_add enumeration function?

> Initialization order dependencies can be defined via the ACPI >_DEP method in the DSDT. For example, if host bridge B depends on >host bridge A, bridge B should not be enumerated until bridge A is >fully initialized.

> Yes, that should work.
> Regards,
> Sunil

Hi,Sunil

I'm truly honored by your affirmation of my answer.Do you think this solution has a chance of being accepted into the upstream kernel? Are there any unintended side effects we may have overlooked?

Regards,
Yuye



^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH v7 08/17] ACPI: pci_link: Clear the dependencies after probe
  2025-12-01 14:12 ` huyuye
@ 2025-12-02  4:28   ` Sunil V L
  0 siblings, 0 replies; 4+ messages in thread
From: Sunil V L @ 2025-12-02  4:28 UTC (permalink / raw)
  To: huyuye
  Cc: ajones, anup, aou, atishp, bhelgaas, catalin.marinas,
	conor.dooley, dfustini, haibo1.xu, lenb, linux-arm-kernel,
	linux-kernel, linux-pci, linux-riscv, palmer, rafael,
	robert.moore, samuel.holland, tglx, will, dai.hualiang,
	deng.weixian, guo.chang2, liu.qingtao2, wu.jiabao, lin.yongchun,
	hu.yuye, zhang.longxiang

On Mon, Dec 01, 2025 at 10:12:29PM +0800, huyuye wrote:
> > Hi, sunilvl
> 
> > Based on the above patch, I understand that you previously >resolved dependencies between Link devices and PCI Host Bridges by >calling acpi_dev_clear_dependencies(device). I would like to ask: >on RISC‑V platforms, if we need to manage dependencies between >multiple PCI Host Bridges, could this be addressed by adding a >call to acpi_dev_clear_dependencies(device) at the end of the >acpi_pci_root_add enumeration function?
> 
> > Initialization order dependencies can be defined via the ACPI >_DEP method in the DSDT. For example, if host bridge B depends on >host bridge A, bridge B should not be enumerated until bridge A is >fully initialized.
> 
> > Yes, that should work.
> > Regards,
> > Sunil
> 
> Hi,Sunil
> 
> I'm truly honored by your affirmation of my answer.Do you think this solution has a chance of being accepted into the upstream kernel? Are there any unintended side effects we may have overlooked?
> 
Acceptance of this solution depends on the PCI/ACPI maintainers. To help
them evaluate it, you should clearly describe why your platform requires
this dependency in the patch.

Regards
Sunil


^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2025-12-02  4:28 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
     [not found] <20240729142241.733357-9-sunilvl@ventanamicro.com>
2025-12-01 13:07 ` Re:[PATCH v7 08/17] ACPI: pci_link: Clear the dependencies after probe huyuye
2025-12-01 13:33   ` [PATCH " Sunil V L
2025-12-01 14:12 ` huyuye
2025-12-02  4:28   ` [PATCH " Sunil V L

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox