diff mbox series

[linu-next,v1] PCI: dw-rockchip: Enable async probe by default

Message ID 20240625155759.132878-1-linux.amoon@gmail.com
State New
Headers show
Series [linu-next,v1] PCI: dw-rockchip: Enable async probe by default | expand

Commit Message

Anand Moon June 25, 2024, 3:57 p.m. UTC
Rockchip PCIe driver lets waits for the combo PHY link like PCIe 3.0,
PCIe 2.0 and SATA 3.0 controller to be up during the probe this
consumes several milliseconds during boot.

Establishing a PCIe link can take a while; allow asynchronous probing so
that link establishment can happen in the background while other devices
are being probed.

Signed-off-by: Anand Moon <linux.amoon@gmail.com>
---
 drivers/pci/controller/dwc/pcie-dw-rockchip.c | 1 +
 1 file changed, 1 insertion(+)

Comments

Anand Moon Aug. 7, 2024, 4:39 a.m. UTC | #1
Hi All,

On Tue, 25 Jun 2024 at 21:28, Anand Moon <linux.amoon@gmail.com> wrote:
>
> Rockchip PCIe driver lets waits for the combo PHY link like PCIe 3.0,
> PCIe 2.0 and SATA 3.0 controller to be up during the probe this
> consumes several milliseconds during boot.
>
> Establishing a PCIe link can take a while; allow asynchronous probing so
> that link establishment can happen in the background while other devices
> are being probed.
>
> Signed-off-by: Anand Moon <linux.amoon@gmail.com>

Gentle ping.

Thanks
-Anand
Bjorn Helgaas Aug. 7, 2024, 4:31 p.m. UTC | #2
On Tue, Jun 25, 2024 at 09:27:57PM +0530, Anand Moon wrote:
> Rockchip PCIe driver lets waits for the combo PHY link like PCIe 3.0,
> PCIe 2.0 and SATA 3.0 controller to be up during the probe this
> consumes several milliseconds during boot.

This needs some wordsmithing.  "driver lets waits" ... I guess "lets"
is not supposed to be there?  I'm not sure what the relevance of "PCIe
3.0, PCIe 2.0, SATA 3.0" is.  I assume the host controller driver
doesn't know what downstream devices might be present, and the async
probing is desirable no matter what they might be?

> Establishing a PCIe link can take a while; allow asynchronous probing so
> that link establishment can happen in the background while other devices
> are being probed.
> 
> Signed-off-by: Anand Moon <linux.amoon@gmail.com>
> ---
>  drivers/pci/controller/dwc/pcie-dw-rockchip.c | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/drivers/pci/controller/dwc/pcie-dw-rockchip.c b/drivers/pci/controller/dwc/pcie-dw-rockchip.c
> index 61b1acba7182..74a3e9d172a0 100644
> --- a/drivers/pci/controller/dwc/pcie-dw-rockchip.c
> +++ b/drivers/pci/controller/dwc/pcie-dw-rockchip.c
> @@ -367,6 +367,7 @@ static struct platform_driver rockchip_pcie_driver = {
>  		.name	= "rockchip-dw-pcie",
>  		.of_match_table = rockchip_pcie_of_match,
>  		.suppress_bind_attrs = true,
> +		.probe_type = PROBE_PREFER_ASYNCHRONOUS,
>  	},
>  	.probe = rockchip_pcie_probe,
>  };
> -- 
> 2.44.0
>
Manivannan Sadhasivam Aug. 7, 2024, 5 p.m. UTC | #3
On Wed, Aug 07, 2024 at 11:31:06AM -0500, Bjorn Helgaas wrote:
> On Tue, Jun 25, 2024 at 09:27:57PM +0530, Anand Moon wrote:
> > Rockchip PCIe driver lets waits for the combo PHY link like PCIe 3.0,
> > PCIe 2.0 and SATA 3.0 controller to be up during the probe this
> > consumes several milliseconds during boot.
> 
> This needs some wordsmithing.  "driver lets waits" ... I guess "lets"
> is not supposed to be there?  I'm not sure what the relevance of "PCIe
> 3.0, PCIe 2.0, SATA 3.0" is.  I assume the host controller driver
> doesn't know what downstream devices might be present, and the async
> probing is desirable no matter what they might be?
> 

Since the DWC driver is enabling link training during boot, it also waits for
the link to be 'up'. But if the device is 'up', then the wait time would be
usually negligible (few ms). But if there is no device, then the wait time of 1s
would be evident.

But here the patch is trying to avoid the few ms delay itself (which is fine).
The type of endpoint might have some impact on the link training also. But async
probe is always preferred.

- Mani

> > Establishing a PCIe link can take a while; allow asynchronous probing so
> > that link establishment can happen in the background while other devices
> > are being probed.
> > 
> > Signed-off-by: Anand Moon <linux.amoon@gmail.com>
> > ---
> >  drivers/pci/controller/dwc/pcie-dw-rockchip.c | 1 +
> >  1 file changed, 1 insertion(+)
> > 
> > diff --git a/drivers/pci/controller/dwc/pcie-dw-rockchip.c b/drivers/pci/controller/dwc/pcie-dw-rockchip.c
> > index 61b1acba7182..74a3e9d172a0 100644
> > --- a/drivers/pci/controller/dwc/pcie-dw-rockchip.c
> > +++ b/drivers/pci/controller/dwc/pcie-dw-rockchip.c
> > @@ -367,6 +367,7 @@ static struct platform_driver rockchip_pcie_driver = {
> >  		.name	= "rockchip-dw-pcie",
> >  		.of_match_table = rockchip_pcie_of_match,
> >  		.suppress_bind_attrs = true,
> > +		.probe_type = PROBE_PREFER_ASYNCHRONOUS,
> >  	},
> >  	.probe = rockchip_pcie_probe,
> >  };
> > -- 
> > 2.44.0
> > 
>
Anand Moon Aug. 8, 2024, 3:13 a.m. UTC | #4
Hi Manivannan, Bjorn

On Wed, 7 Aug 2024 at 22:30, Manivannan Sadhasivam
<manivannan.sadhasivam@linaro.org> wrote:
>
> On Wed, Aug 07, 2024 at 11:31:06AM -0500, Bjorn Helgaas wrote:
> > On Tue, Jun 25, 2024 at 09:27:57PM +0530, Anand Moon wrote:
> > > Rockchip PCIe driver lets waits for the combo PHY link like PCIe 3.0,
> > > PCIe 2.0 and SATA 3.0 controller to be up during the probe this
> > > consumes several milliseconds during boot.
> >
> > This needs some wordsmithing.  "driver lets waits" ... I guess "lets"
> > is not supposed to be there?  I'm not sure what the relevance of "PCIe
> > 3.0, PCIe 2.0, SATA 3.0" is.  I assume the host controller driver
> > doesn't know what downstream devices might be present, and the async
> > probing is desirable no matter what they might be?
> >
Ok I will improve this commit message.

Rockchip DWC PCIe driver currently waits for the combo PHY link
(PCIe 3.0, PCIe 2.0, and SATA 3.0) to be established during the probe,
which could consume several milliseconds during boot.
To optimize boot time, this commit allows asynchronous probing.
This change enables the PCIe link establishment to occur in the
background while other devices are being probed.

>
> Since the DWC driver is enabling link training during boot, it also waits for
> the link to be 'up'. But if the device is 'up', then the wait time would be
> usually negligible (few ms). But if there is no device, then the wait time of 1s
> would be evident.
>
> But here the patch is trying to avoid the few ms delay itself (which is fine).
> The type of endpoint might have some impact on the link training also. But async
> probe is always preferred.
>
> - Mani
>
Ok,

Thanks
-Anand
diff mbox series

Patch

diff --git a/drivers/pci/controller/dwc/pcie-dw-rockchip.c b/drivers/pci/controller/dwc/pcie-dw-rockchip.c
index 61b1acba7182..74a3e9d172a0 100644
--- a/drivers/pci/controller/dwc/pcie-dw-rockchip.c
+++ b/drivers/pci/controller/dwc/pcie-dw-rockchip.c
@@ -367,6 +367,7 @@  static struct platform_driver rockchip_pcie_driver = {
 		.name	= "rockchip-dw-pcie",
 		.of_match_table = rockchip_pcie_of_match,
 		.suppress_bind_attrs = true,
+		.probe_type = PROBE_PREFER_ASYNCHRONOUS,
 	},
 	.probe = rockchip_pcie_probe,
 };