diff mbox series

[1/5] arm64: dts: imx8-ss-hsio: Wire up DMA IRQ for PCIe

Message ID 20250107140110.982215-2-alexander.stein@ew.tq-group.com (mailing list archive)
State New, archived
Headers show
Series TQMa8Xx update | expand

Commit Message

Alexander Stein Jan. 7, 2025, 2:01 p.m. UTC 
IRQ mapping is already present. Add the missing DMA interrupt.

Signed-off-by: Alexander Stein <alexander.stein@ew.tq-group.com>
---
 arch/arm64/boot/dts/freescale/imx8-ss-hsio.dtsi | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

Comments

Frank Li Jan. 7, 2025, 3:39 p.m. UTC | #1 
On Tue, Jan 07, 2025 at 03:01:06PM +0100, Alexander Stein wrote:
> IRQ mapping is already present. Add the missing DMA interrupt.

PCI host side have not use bridge's DMA yet although hardware support it.

Frank

>
> Signed-off-by: Alexander Stein <alexander.stein@ew.tq-group.com>
> ---
>  arch/arm64/boot/dts/freescale/imx8-ss-hsio.dtsi | 5 +++--
>  1 file changed, 3 insertions(+), 2 deletions(-)
>
> diff --git a/arch/arm64/boot/dts/freescale/imx8-ss-hsio.dtsi b/arch/arm64/boot/dts/freescale/imx8-ss-hsio.dtsi
> index 70a8aa1a67911..4bdfc15487cbc 100644
> --- a/arch/arm64/boot/dts/freescale/imx8-ss-hsio.dtsi
> +++ b/arch/arm64/boot/dts/freescale/imx8-ss-hsio.dtsi
> @@ -57,8 +57,9 @@ pcieb: pcie@5f010000 {
>  		ranges = <0x81000000 0 0x00000000 0x8ff80000 0 0x00010000>,
>  			 <0x82000000 0 0x80000000 0x80000000 0 0x0ff00000>;
>  		#interrupt-cells = <1>;
> -		interrupts = <GIC_SPI 102 IRQ_TYPE_LEVEL_HIGH>;
> -		interrupt-names = "msi";
> +		interrupts = <GIC_SPI 102 IRQ_TYPE_LEVEL_HIGH>,
> +			     <GIC_SPI 104 IRQ_TYPE_LEVEL_HIGH>;
> +		interrupt-names = "msi", "dma";
>  		#address-cells = <3>;
>  		#size-cells = <2>;
>  		clocks = <&pcieb_lpcg IMX_LPCG_CLK_6>,
> --
> 2.34.1
>
Alexander Stein Jan. 8, 2025, 9:58 a.m. UTC | #2 
Hi Frank,

Am Dienstag, 7. Januar 2025, 16:39:25 CET schrieb Frank Li:
> On Tue, Jan 07, 2025 at 03:01:06PM +0100, Alexander Stein wrote:
> > IRQ mapping is already present. Add the missing DMA interrupt.
> 
> PCI host side have not use bridge's DMA yet although hardware support it.

So this is a driver limitation, right? So IMHO the device description is
independent from that and still correct, right?

Best regards,
Alexander

> >
> > Signed-off-by: Alexander Stein <alexander.stein@ew.tq-group.com>
> > ---
> >  arch/arm64/boot/dts/freescale/imx8-ss-hsio.dtsi | 5 +++--
> >  1 file changed, 3 insertions(+), 2 deletions(-)
> >
> > diff --git a/arch/arm64/boot/dts/freescale/imx8-ss-hsio.dtsi b/arch/arm64/boot/dts/freescale/imx8-ss-hsio.dtsi
> > index 70a8aa1a67911..4bdfc15487cbc 100644
> > --- a/arch/arm64/boot/dts/freescale/imx8-ss-hsio.dtsi
> > +++ b/arch/arm64/boot/dts/freescale/imx8-ss-hsio.dtsi
> > @@ -57,8 +57,9 @@ pcieb: pcie@5f010000 {
> >  		ranges = <0x81000000 0 0x00000000 0x8ff80000 0 0x00010000>,
> >  			 <0x82000000 0 0x80000000 0x80000000 0 0x0ff00000>;
> >  		#interrupt-cells = <1>;
> > -		interrupts = <GIC_SPI 102 IRQ_TYPE_LEVEL_HIGH>;
> > -		interrupt-names = "msi";
> > +		interrupts = <GIC_SPI 102 IRQ_TYPE_LEVEL_HIGH>,
> > +			     <GIC_SPI 104 IRQ_TYPE_LEVEL_HIGH>;
> > +		interrupt-names = "msi", "dma";
> >  		#address-cells = <3>;
> >  		#size-cells = <2>;
> >  		clocks = <&pcieb_lpcg IMX_LPCG_CLK_6>,
> > --
> > 2.34.1
> >
>
Frank Li Jan. 8, 2025, 4:33 p.m. UTC | #3 
On Wed, Jan 08, 2025 at 10:58:18AM +0100, Alexander Stein wrote:
> Hi Frank,
>
> Am Dienstag, 7. Januar 2025, 16:39:25 CET schrieb Frank Li:
> > On Tue, Jan 07, 2025 at 03:01:06PM +0100, Alexander Stein wrote:
> > > IRQ mapping is already present. Add the missing DMA interrupt.
> >
> > PCI host side have not use bridge's DMA yet although hardware support it.
>
> So this is a driver limitation, right? So IMHO the device description is
> independent from that and still correct, right?

Yes, but dma register space may missed also. I suggest add later after
EP side support merged, which can verify informaiton is correct.

Anyway, I don't block this change.

Frank

>
> Best regards,
> Alexander
>
> > >
> > > Signed-off-by: Alexander Stein <alexander.stein@ew.tq-group.com>
> > > ---
> > >  arch/arm64/boot/dts/freescale/imx8-ss-hsio.dtsi | 5 +++--
> > >  1 file changed, 3 insertions(+), 2 deletions(-)
> > >
> > > diff --git a/arch/arm64/boot/dts/freescale/imx8-ss-hsio.dtsi b/arch/arm64/boot/dts/freescale/imx8-ss-hsio.dtsi
> > > index 70a8aa1a67911..4bdfc15487cbc 100644
> > > --- a/arch/arm64/boot/dts/freescale/imx8-ss-hsio.dtsi
> > > +++ b/arch/arm64/boot/dts/freescale/imx8-ss-hsio.dtsi
> > > @@ -57,8 +57,9 @@ pcieb: pcie@5f010000 {
> > >  		ranges = <0x81000000 0 0x00000000 0x8ff80000 0 0x00010000>,
> > >  			 <0x82000000 0 0x80000000 0x80000000 0 0x0ff00000>;
> > >  		#interrupt-cells = <1>;
> > > -		interrupts = <GIC_SPI 102 IRQ_TYPE_LEVEL_HIGH>;
> > > -		interrupt-names = "msi";
> > > +		interrupts = <GIC_SPI 102 IRQ_TYPE_LEVEL_HIGH>,
> > > +			     <GIC_SPI 104 IRQ_TYPE_LEVEL_HIGH>;
> > > +		interrupt-names = "msi", "dma";
> > >  		#address-cells = <3>;
> > >  		#size-cells = <2>;
> > >  		clocks = <&pcieb_lpcg IMX_LPCG_CLK_6>,
> > > --
> > > 2.34.1
> > >
> >
>
>
> --
> TQ-Systems GmbH | Mühlstraße 2, Gut Delling | 82229 Seefeld, Germany
> Amtsgericht München, HRB 105018
> Geschäftsführer: Detlef Schneider, Rüdiger Stahl, Stefan Schneider
> http://www.tq-group.com/
>
>
Frank Li Feb. 24, 2025, 5:49 p.m. UTC | #4 
On Wed, Jan 08, 2025 at 11:33:35AM -0500, Frank Li wrote:
> On Wed, Jan 08, 2025 at 10:58:18AM +0100, Alexander Stein wrote:
> > Hi Frank,
> >
> > Am Dienstag, 7. Januar 2025, 16:39:25 CET schrieb Frank Li:
> > > On Tue, Jan 07, 2025 at 03:01:06PM +0100, Alexander Stein wrote:
> > > > IRQ mapping is already present. Add the missing DMA interrupt.
> > >
> > > PCI host side have not use bridge's DMA yet although hardware support it.
> >
> > So this is a driver limitation, right? So IMHO the device description is
> > independent from that and still correct, right?
>
> Yes, but dma register space may missed also. I suggest add later after
> EP side support merged, which can verify informaiton is correct.
>
> Anyway, I don't block this change.


Did you run DTB_CHECK? I found new warning was added

/home/lizhi/source/linux-upstream-dts/arch/arm64/boot/dts/freescale/imx8dx-colibri-aster.dtb: pcie@5f010000: interrupts: [[0, 102, 4], [0, 104, 4]] is too long
        from schema $id: http://devicetree.org/schemas/pci/fsl,imx6q-pcie.yaml#
/home/lizhi/source/linux-upstream-dts/arch/arm64/boot/dts/freescale/imx8dx-colibri-aster.dtb: pcie@5f010000: interrupt-names: ['msi', 'dma'] is too long
        from schema $id: http://devicetree.org/schemas/pci/fsl,imx6q-pcie.yaml#
/home/lizhi/source/linux-upstream-dts/arch/arm64/boot/dts/freescale/imx8dx-colibri-eval-v3.dtb: pcie@5f010000: interrupts: [[0, 102, 4], [0, 104, 4]] is too long
        from schema $id: http://devicetree.org/schemas/pci/fsl,imx6q-pcie.yaml#
/home/lizhi/source/linux-upstream-dts/arch/arm64/boot/dts/freescale/imx8dx-colibri-eval-v3.dtb: pcie@5f010000: interrupt-names: ['msi', 'dma'] is too long
        from schema $id: http://devicetree.org/schemas/pci/fsl,imx6q-pcie.yaml#
/home/lizhi/source/linux-upstream-dts/arch/arm64/boot/dts/freescale/imx8dx-colibri-iris.dtb: pcie@5f010000: interrupts: [[0, 102, 4], [0, 104, 4]] is too long
        from schema $id: http://devicetree.org/schemas/pci/fsl,imx6q-pcie.yaml#
/home/lizhi/source/linux-upstream-dts/arch/arm64/boot/dts/freescale/imx8dx-colibri-iris.dtb: pcie@5f010000: interrupt-names: ['msi', 'dma'] is too long

...

Frank

>
> Frank
>
> >
> > Best regards,
> > Alexander
> >
> > > >
> > > > Signed-off-by: Alexander Stein <alexander.stein@ew.tq-group.com>
> > > > ---
> > > >  arch/arm64/boot/dts/freescale/imx8-ss-hsio.dtsi | 5 +++--
> > > >  1 file changed, 3 insertions(+), 2 deletions(-)
> > > >
> > > > diff --git a/arch/arm64/boot/dts/freescale/imx8-ss-hsio.dtsi b/arch/arm64/boot/dts/freescale/imx8-ss-hsio.dtsi
> > > > index 70a8aa1a67911..4bdfc15487cbc 100644
> > > > --- a/arch/arm64/boot/dts/freescale/imx8-ss-hsio.dtsi
> > > > +++ b/arch/arm64/boot/dts/freescale/imx8-ss-hsio.dtsi
> > > > @@ -57,8 +57,9 @@ pcieb: pcie@5f010000 {
> > > >  		ranges = <0x81000000 0 0x00000000 0x8ff80000 0 0x00010000>,
> > > >  			 <0x82000000 0 0x80000000 0x80000000 0 0x0ff00000>;
> > > >  		#interrupt-cells = <1>;
> > > > -		interrupts = <GIC_SPI 102 IRQ_TYPE_LEVEL_HIGH>;
> > > > -		interrupt-names = "msi";
> > > > +		interrupts = <GIC_SPI 102 IRQ_TYPE_LEVEL_HIGH>,
> > > > +			     <GIC_SPI 104 IRQ_TYPE_LEVEL_HIGH>;
> > > > +		interrupt-names = "msi", "dma";
> > > >  		#address-cells = <3>;
> > > >  		#size-cells = <2>;
> > > >  		clocks = <&pcieb_lpcg IMX_LPCG_CLK_6>,
> > > > --
> > > > 2.34.1
> > > >
> > >
> >
> >
> > --
> > TQ-Systems GmbH | Mühlstraße 2, Gut Delling | 82229 Seefeld, Germany
> > Amtsgericht München, HRB 105018
> > Geschäftsführer: Detlef Schneider, Rüdiger Stahl, Stefan Schneider
> > http://www.tq-group.com/
> >
> >
Alexander Stein Feb. 25, 2025, 10:20 a.m. UTC | #5 
Hi Frank,

Am Montag, 24. Februar 2025, 18:49:01 CET schrieb Frank Li:
> On Wed, Jan 08, 2025 at 11:33:35AM -0500, Frank Li wrote:
> > On Wed, Jan 08, 2025 at 10:58:18AM +0100, Alexander Stein wrote:
> > > Hi Frank,
> > >
> > > Am Dienstag, 7. Januar 2025, 16:39:25 CET schrieb Frank Li:
> > > > On Tue, Jan 07, 2025 at 03:01:06PM +0100, Alexander Stein wrote:
> > > > > IRQ mapping is already present. Add the missing DMA interrupt.
> > > >
> > > > PCI host side have not use bridge's DMA yet although hardware support it.
> > >
> > > So this is a driver limitation, right? So IMHO the device description is
> > > independent from that and still correct, right?
> >
> > Yes, but dma register space may missed also. I suggest add later after
> > EP side support merged, which can verify informaiton is correct.
> >
> > Anyway, I don't block this change.
> 
> 
> Did you run DTB_CHECK? I found new warning was added
> 
> /home/lizhi/source/linux-upstream-dts/arch/arm64/boot/dts/freescale/imx8dx-colibri-aster.dtb: pcie@5f010000: interrupts: [[0, 102, 4], [0, 104, 4]] is too long
>         from schema $id: http://devicetree.org/schemas/pci/fsl,imx6q-pcie.yaml#
> /home/lizhi/source/linux-upstream-dts/arch/arm64/boot/dts/freescale/imx8dx-colibri-aster.dtb: pcie@5f010000: interrupt-names: ['msi', 'dma'] is too long
>         from schema $id: http://devicetree.org/schemas/pci/fsl,imx6q-pcie.yaml#
> /home/lizhi/source/linux-upstream-dts/arch/arm64/boot/dts/freescale/imx8dx-colibri-eval-v3.dtb: pcie@5f010000: interrupts: [[0, 102, 4], [0, 104, 4]] is too long
>         from schema $id: http://devicetree.org/schemas/pci/fsl,imx6q-pcie.yaml#
> /home/lizhi/source/linux-upstream-dts/arch/arm64/boot/dts/freescale/imx8dx-colibri-eval-v3.dtb: pcie@5f010000: interrupt-names: ['msi', 'dma'] is too long
>         from schema $id: http://devicetree.org/schemas/pci/fsl,imx6q-pcie.yaml#
> /home/lizhi/source/linux-upstream-dts/arch/arm64/boot/dts/freescale/imx8dx-colibri-iris.dtb: pcie@5f010000: interrupts: [[0, 102, 4], [0, 104, 4]] is too long
>         from schema $id: http://devicetree.org/schemas/pci/fsl,imx6q-pcie.yaml#
> /home/lizhi/source/linux-upstream-dts/arch/arm64/boot/dts/freescale/imx8dx-colibri-iris.dtb: pcie@5f010000: interrupt-names: ['msi', 'dma'] is too long

I think I did, but I see the same warnings.
I've prepared a series which addresses schema and imx8qm dtsi.

Best regards
Alexander
diff mbox series

Patch

diff --git a/arch/arm64/boot/dts/freescale/imx8-ss-hsio.dtsi b/arch/arm64/boot/dts/freescale/imx8-ss-hsio.dtsi
index 70a8aa1a67911..4bdfc15487cbc 100644
--- a/arch/arm64/boot/dts/freescale/imx8-ss-hsio.dtsi
+++ b/arch/arm64/boot/dts/freescale/imx8-ss-hsio.dtsi
@@ -57,8 +57,9 @@  pcieb: pcie@5f010000 {
 		ranges = <0x81000000 0 0x00000000 0x8ff80000 0 0x00010000>,
 			 <0x82000000 0 0x80000000 0x80000000 0 0x0ff00000>;
 		#interrupt-cells = <1>;
-		interrupts = <GIC_SPI 102 IRQ_TYPE_LEVEL_HIGH>;
-		interrupt-names = "msi";
+		interrupts = <GIC_SPI 102 IRQ_TYPE_LEVEL_HIGH>,
+			     <GIC_SPI 104 IRQ_TYPE_LEVEL_HIGH>;
+		interrupt-names = "msi", "dma";
 		#address-cells = <3>;
 		#size-cells = <2>;
 		clocks = <&pcieb_lpcg IMX_LPCG_CLK_6>,