[PATCH 1/2] ARM: imx6: remove config space from PCIe controller ranges property

[PATCH 1/2] ARM: imx6: remove config space from PCIe controller ranges property

[Lucas Stach]

This has been moved to the reg property where it belongs for quite some
time. The range has been unused by the kernel since then and with kernel
4.4 it's flagged as an unparsable range, as it does not comply to the
PCI ranges DT binding. Fix this by removing the superfluous range.

Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
---
 arch/arm/boot/dts/imx6qdl.dtsi | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/arch/arm/boot/dts/imx6qdl.dtsi b/arch/arm/boot/dts/imx6qdl.dtsi
index 2b6cc8bf3c5c..a5a731519d9e 100644
--- a/arch/arm/boot/dts/imx6qdl.dtsi
+++ b/arch/arm/boot/dts/imx6qdl.dtsi
@@ -173,8 +173,7 @@
 			#address-cells = <3>;
 			#size-cells = <2>;
 			device_type = "pci";
-			ranges = <0x00000800 0 0x01f00000 0x01f00000 0 0x00080000 /* configuration space */
-				  0x81000000 0 0          0x01f80000 0 0x00010000 /* downstream I/O */
+			ranges = <0x81000000 0 0          0x01f80000 0 0x00010000 /* downstream I/O */
 				  0x82000000 0 0x01000000 0x01000000 0 0x00f00000>; /* non-prefetchable memory */
 			num-lanes = <1>;
 			interrupts = <GIC_SPI 120 IRQ_TYPE_LEVEL_HIGH>;
-- 
2.6.2

[PATCH 2/2] ARM: imx_v6_v7_defconfig: enable PCI_MSI

[Lucas Stach]

This has been stable on i.MX6 for a good while now and there is no
reason to keep it disabled, as it allows to allocate non-shared
IRQs if several PCIe devices are connected to a single host.

Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
---
 arch/arm/configs/imx_v6_v7_defconfig | 1 +
 1 file changed, 1 insertion(+)

diff --git a/arch/arm/configs/imx_v6_v7_defconfig b/arch/arm/configs/imx_v6_v7_defconfig
index 4187f69f6630..e2ce60d00e8a 100644
--- a/arch/arm/configs/imx_v6_v7_defconfig
+++ b/arch/arm/configs/imx_v6_v7_defconfig
@@ -45,6 +45,7 @@ CONFIG_SOC_IMX7D=y
 CONFIG_SOC_LS1021A=y
 CONFIG_SOC_VF610=y
 CONFIG_PCI=y
+CONFIG_PCI_MSI=y
 CONFIG_PCI_IMX6=y
 CONFIG_SMP=y
 CONFIG_PREEMPT_VOLUNTARY=y
-- 
2.6.2

[PATCH 1/2] ARM: imx6: remove config space from PCIe controller ranges property

[Shawn Guo]

On Mon, Nov 30, 2015 at 06:00:10PM +0100, Lucas Stach wrote:
> This has been moved to the reg property where it belongs for quite some
> time. The range has been unused by the kernel since then and with kernel
> 4.4 it's flagged as an unparsable range, as it does not comply to the
> PCI ranges DT binding. Fix this by removing the superfluous range.
> 
> Signed-off-by: Lucas Stach <l.stach@pengutronix.de>

Changed patch prefix to 'ARM: dts: imx6: ...', and applied.

Shawn

> ---
>  arch/arm/boot/dts/imx6qdl.dtsi | 3 +--
>  1 file changed, 1 insertion(+), 2 deletions(-)
> 
> diff --git a/arch/arm/boot/dts/imx6qdl.dtsi b/arch/arm/boot/dts/imx6qdl.dtsi
> index 2b6cc8bf3c5c..a5a731519d9e 100644
> --- a/arch/arm/boot/dts/imx6qdl.dtsi
> +++ b/arch/arm/boot/dts/imx6qdl.dtsi
> @@ -173,8 +173,7 @@
>  			#address-cells = <3>;
>  			#size-cells = <2>;
>  			device_type = "pci";
> -			ranges = <0x00000800 0 0x01f00000 0x01f00000 0 0x00080000 /* configuration space */
> -				  0x81000000 0 0          0x01f80000 0 0x00010000 /* downstream I/O */
> +			ranges = <0x81000000 0 0          0x01f80000 0 0x00010000 /* downstream I/O */
>  				  0x82000000 0 0x01000000 0x01000000 0 0x00f00000>; /* non-prefetchable memory */
>  			num-lanes = <1>;
>  			interrupts = <GIC_SPI 120 IRQ_TYPE_LEVEL_HIGH>;
> -- 
> 2.6.2
> 
> 

[PATCH 2/2] ARM: imx_v6_v7_defconfig: enable PCI_MSI

[Shawn Guo]

On Mon, Nov 30, 2015 at 06:00:11PM +0100, Lucas Stach wrote:
> This has been stable on i.MX6 for a good while now and there is no
> reason to keep it disabled, as it allows to allocate non-shared
> IRQs if several PCIe devices are connected to a single host.
> 
> Signed-off-by: Lucas Stach <l.stach@pengutronix.de>

Applied, thanks.

[PATCH 2/2] ARM: imx_v6_v7_defconfig: enable PCI_MSI

[Tim Harvey]

On Mon, Nov 30, 2015 at 9:00 AM, Lucas Stach <l.stach@pengutronix.de> wrote:
> This has been stable on i.MX6 for a good while now and there is no
> reason to keep it disabled, as it allows to allocate non-shared
> IRQs if several PCIe devices are connected to a single host.
>
> Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
> ---
>  arch/arm/configs/imx_v6_v7_defconfig | 1 +
>  1 file changed, 1 insertion(+)
>
> diff --git a/arch/arm/configs/imx_v6_v7_defconfig b/arch/arm/configs/imx_v6_v7_defconfig
> index 4187f69f6630..e2ce60d00e8a 100644
> --- a/arch/arm/configs/imx_v6_v7_defconfig
> +++ b/arch/arm/configs/imx_v6_v7_defconfig
> @@ -45,6 +45,7 @@ CONFIG_SOC_IMX7D=y
>  CONFIG_SOC_LS1021A=y
>  CONFIG_SOC_VF610=y
>  CONFIG_PCI=y
> +CONFIG_PCI_MSI=y
>  CONFIG_PCI_IMX6=y
>  CONFIG_SMP=y
>  CONFIG_PREEMPT_VOLUNTARY=y
> --
> 2.6.2
>
>
> _______________________________________________
> linux-arm-kernel mailing list
> linux-arm-kernel at lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Lucas,

Any thoughts on why enabling MSI on IMX6 breaks legacy interrupts?

Again, many PCI devices/drivers don't support MSI and thus enabling
MSI as this patch does makes this config break many PCI devices on
IMX6.

I have found that MSI does indeed work on the IMX6 for cards/drivers
that use MSI, but for those that don't the legacy interrupts never
fire regardless of going through a bridge or directly hanging off the
IMX6. I assume you can confirm this on your boards/devices as well?

Regards,

Tim

[PATCH 2/2] ARM: imx_v6_v7_defconfig: enable PCI_MSI

[Lucas Stach]

Am Donnerstag, den 14.04.2016, 08:13 -0700 schrieb Tim Harvey:
> On Mon, Nov 30, 2015 at 9:00 AM, Lucas Stach <l.stach@pengutronix.de> wrote:
> > This has been stable on i.MX6 for a good while now and there is no
> > reason to keep it disabled, as it allows to allocate non-shared
> > IRQs if several PCIe devices are connected to a single host.
> >
> > Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
> > ---
> >  arch/arm/configs/imx_v6_v7_defconfig | 1 +
> >  1 file changed, 1 insertion(+)
> >
> > diff --git a/arch/arm/configs/imx_v6_v7_defconfig b/arch/arm/configs/imx_v6_v7_defconfig
> > index 4187f69f6630..e2ce60d00e8a 100644
> > --- a/arch/arm/configs/imx_v6_v7_defconfig
> > +++ b/arch/arm/configs/imx_v6_v7_defconfig
> > @@ -45,6 +45,7 @@ CONFIG_SOC_IMX7D=y
> >  CONFIG_SOC_LS1021A=y
> >  CONFIG_SOC_VF610=y
> >  CONFIG_PCI=y
> > +CONFIG_PCI_MSI=y
> >  CONFIG_PCI_IMX6=y
> >  CONFIG_SMP=y
> >  CONFIG_PREEMPT_VOLUNTARY=y
> > --
> > 2.6.2
> >
> >
> > _______________________________________________
> > linux-arm-kernel mailing list
> > linux-arm-kernel at lists.infradead.org
> > http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
> 
> Lucas,
> 
> Any thoughts on why enabling MSI on IMX6 breaks legacy interrupts?
> 
> Again, many PCI devices/drivers don't support MSI and thus enabling
> MSI as this patch does makes this config break many PCI devices on
> IMX6.
> 
> I have found that MSI does indeed work on the IMX6 for cards/drivers
> that use MSI, but for those that don't the legacy interrupts never
> fire regardless of going through a bridge or directly hanging off the
> IMX6. I assume you can confirm this on your boards/devices as well?
> 
No, this seems to be an errata in the DW PCIe core. The manual states
that legacy IRQs and MSIs can't be used together, but I thought this
only applies to the EP part of the core.

We can probably fix the simple configuration by not enabling MSI support
until someone actually requests to set up an MSI IRQ, so both IRQ types
work when only one EP is connected.
I don't have any idea yet how to fix setups with a PCIe switch where
some connected EPs might like to use MSIs while some others can't use
them.

Regards,
Lucas

[PATCH 2/2] ARM: imx_v6_v7_defconfig: enable PCI_MSI

[Tim Harvey]

On Fri, Apr 15, 2016 at 1:21 AM, Lucas Stach <l.stach@pengutronix.de> wrote:
> Am Donnerstag, den 14.04.2016, 08:13 -0700 schrieb Tim Harvey:
>> On Mon, Nov 30, 2015 at 9:00 AM, Lucas Stach <l.stach@pengutronix.de> wrote:
>> > This has been stable on i.MX6 for a good while now and there is no
>> > reason to keep it disabled, as it allows to allocate non-shared
>> > IRQs if several PCIe devices are connected to a single host.
>> >
>> > Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
>> > ---
>> >  arch/arm/configs/imx_v6_v7_defconfig | 1 +
>> >  1 file changed, 1 insertion(+)
>> >
>> > diff --git a/arch/arm/configs/imx_v6_v7_defconfig b/arch/arm/configs/imx_v6_v7_defconfig
>> > index 4187f69f6630..e2ce60d00e8a 100644
>> > --- a/arch/arm/configs/imx_v6_v7_defconfig
>> > +++ b/arch/arm/configs/imx_v6_v7_defconfig
>> > @@ -45,6 +45,7 @@ CONFIG_SOC_IMX7D=y
>> >  CONFIG_SOC_LS1021A=y
>> >  CONFIG_SOC_VF610=y
>> >  CONFIG_PCI=y
>> > +CONFIG_PCI_MSI=y
>> >  CONFIG_PCI_IMX6=y
>> >  CONFIG_SMP=y
>> >  CONFIG_PREEMPT_VOLUNTARY=y
>> > --
>> > 2.6.2
>> >
>> >
>> > _______________________________________________
>> > linux-arm-kernel mailing list
>> > linux-arm-kernel at lists.infradead.org
>> > http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
>>
>> Lucas,
>>
>> Any thoughts on why enabling MSI on IMX6 breaks legacy interrupts?
>>
>> Again, many PCI devices/drivers don't support MSI and thus enabling
>> MSI as this patch does makes this config break many PCI devices on
>> IMX6.
>>
>> I have found that MSI does indeed work on the IMX6 for cards/drivers
>> that use MSI, but for those that don't the legacy interrupts never
>> fire regardless of going through a bridge or directly hanging off the
>> IMX6. I assume you can confirm this on your boards/devices as well?
>>
> No, this seems to be an errata in the DW PCIe core. The manual states
> that legacy IRQs and MSIs can't be used together, but I thought this
> only applies to the EP part of the core.
>
> We can probably fix the simple configuration by not enabling MSI support
> until someone actually requests to set up an MSI IRQ, so both IRQ types
> work when only one EP is connected.
> I don't have any idea yet how to fix setups with a PCIe switch where
> some connected EPs might like to use MSIs while some others can't use
> them.
>
> Regards,
> Lucas
>

Lucas,

I feel a bit dumb here for forgetting about this issue and letting it
drop, but I've just spent a day debugging an issue with the ath9k
wireless driver only to realize it was that the ath9k card and/or
driver doesn't support MSI interrupts and thus no longer work on
mainline kernels that have MSI enabled. To be honest I'm not sure how
many PCIe devices out there can't support MSI because of hardware
limitations vs drivers that simply have not implemented it.

To make matters worse 3ee803641e76bea76ec730c80dcc64739a9919ff makes
it such that you can't disable MSI on the IMX6.

Is there some fix you can think of to make the IMX6 PCIe host
controller raise an interrupt for cards/drivers that use legacy irq's?
If not, then we need to allow MSI to be disabled for IMX6 and default
it to disabled for compatibility.

Regards,

Tim

Re: [PATCH 2/2] ARM: imx_v6_v7_defconfig: enable PCI_MSI

[Tim Harvey]

On Thu, Mar 30, 2017 at 4:42 PM, Tim Harvey <tharvey@gateworks.com> wrote:
> On Fri, Apr 15, 2016 at 1:21 AM, Lucas Stach <l.stach@pengutronix.de> wrote:
>> Am Donnerstag, den 14.04.2016, 08:13 -0700 schrieb Tim Harvey:
>>> On Mon, Nov 30, 2015 at 9:00 AM, Lucas Stach <l.stach@pengutronix.de> wrote:
>>> > This has been stable on i.MX6 for a good while now and there is no
>>> > reason to keep it disabled, as it allows to allocate non-shared
>>> > IRQs if several PCIe devices are connected to a single host.
>>> >
>>> > Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
>>> > ---
>>> >  arch/arm/configs/imx_v6_v7_defconfig | 1 +
>>> >  1 file changed, 1 insertion(+)
>>> >
>>> > diff --git a/arch/arm/configs/imx_v6_v7_defconfig b/arch/arm/configs/imx_v6_v7_defconfig
>>> > index 4187f69f6630..e2ce60d00e8a 100644
>>> > --- a/arch/arm/configs/imx_v6_v7_defconfig
>>> > +++ b/arch/arm/configs/imx_v6_v7_defconfig
>>> > @@ -45,6 +45,7 @@ CONFIG_SOC_IMX7D=y
>>> >  CONFIG_SOC_LS1021A=y
>>> >  CONFIG_SOC_VF610=y
>>> >  CONFIG_PCI=y
>>> > +CONFIG_PCI_MSI=y
>>> >  CONFIG_PCI_IMX6=y
>>> >  CONFIG_SMP=y
>>> >  CONFIG_PREEMPT_VOLUNTARY=y
>>> > --
>>> > 2.6.2
>>> >
>>> >
>>> > _______________________________________________
>>> > linux-arm-kernel mailing list
>>> > linux-arm-kernel@lists.infradead.org
>>> > http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
>>>
>>> Lucas,
>>>
>>> Any thoughts on why enabling MSI on IMX6 breaks legacy interrupts?
>>>
>>> Again, many PCI devices/drivers don't support MSI and thus enabling
>>> MSI as this patch does makes this config break many PCI devices on
>>> IMX6.
>>>
>>> I have found that MSI does indeed work on the IMX6 for cards/drivers
>>> that use MSI, but for those that don't the legacy interrupts never
>>> fire regardless of going through a bridge or directly hanging off the
>>> IMX6. I assume you can confirm this on your boards/devices as well?
>>>
>> No, this seems to be an errata in the DW PCIe core. The manual states
>> that legacy IRQs and MSIs can't be used together, but I thought this
>> only applies to the EP part of the core.
>>
>> We can probably fix the simple configuration by not enabling MSI support
>> until someone actually requests to set up an MSI IRQ, so both IRQ types
>> work when only one EP is connected.
>> I don't have any idea yet how to fix setups with a PCIe switch where
>> some connected EPs might like to use MSIs while some others can't use
>> them.
>>
>> Regards,
>> Lucas
>>
>
> Lucas,
>
> I feel a bit dumb here for forgetting about this issue and letting it
> drop, but I've just spent a day debugging an issue with the ath9k
> wireless driver only to realize it was that the ath9k card and/or
> driver doesn't support MSI interrupts and thus no longer work on
> mainline kernels that have MSI enabled. To be honest I'm not sure how
> many PCIe devices out there can't support MSI because of hardware
> limitations vs drivers that simply have not implemented it.
>
> To make matters worse 3ee803641e76bea76ec730c80dcc64739a9919ff makes
> it such that you can't disable MSI on the IMX6.
>
> Is there some fix you can think of to make the IMX6 PCIe host
> controller raise an interrupt for cards/drivers that use legacy irq's?
> If not, then we need to allow MSI to be disabled for IMX6 and default
> it to disabled for compatibility.
>

Joao / Jingoo,

(Adding linux-pci to the list to pick up maintainers of other DWC PCIe
core drivers)

Is Luca's statement that 'legacy IRQs and MSIs can't be used together'
true in general for the DW PCIe core? If any of the host controllers
using this core can't support both legacy and MSI irqs together I
still believe we shouldn't enable/require MSI as it breaks any
card/driver that only supports legacy interrupts (such as ath9k).

Regards,

Tim

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH 2/2] ARM: imx_v6_v7_defconfig: enable PCI_MSI

[Tim Harvey]

On Thu, Mar 30, 2017 at 4:42 PM, Tim Harvey <tharvey@gateworks.com> wrote:
> On Fri, Apr 15, 2016 at 1:21 AM, Lucas Stach <l.stach@pengutronix.de> wrote:
>> Am Donnerstag, den 14.04.2016, 08:13 -0700 schrieb Tim Harvey:
>>> On Mon, Nov 30, 2015 at 9:00 AM, Lucas Stach <l.stach@pengutronix.de> wrote:
>>> > This has been stable on i.MX6 for a good while now and there is no
>>> > reason to keep it disabled, as it allows to allocate non-shared
>>> > IRQs if several PCIe devices are connected to a single host.
>>> >
>>> > Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
>>> > ---
>>> >  arch/arm/configs/imx_v6_v7_defconfig | 1 +
>>> >  1 file changed, 1 insertion(+)
>>> >
>>> > diff --git a/arch/arm/configs/imx_v6_v7_defconfig b/arch/arm/configs/imx_v6_v7_defconfig
>>> > index 4187f69f6630..e2ce60d00e8a 100644
>>> > --- a/arch/arm/configs/imx_v6_v7_defconfig
>>> > +++ b/arch/arm/configs/imx_v6_v7_defconfig
>>> > @@ -45,6 +45,7 @@ CONFIG_SOC_IMX7D=y
>>> >  CONFIG_SOC_LS1021A=y
>>> >  CONFIG_SOC_VF610=y
>>> >  CONFIG_PCI=y
>>> > +CONFIG_PCI_MSI=y
>>> >  CONFIG_PCI_IMX6=y
>>> >  CONFIG_SMP=y
>>> >  CONFIG_PREEMPT_VOLUNTARY=y
>>> > --
>>> > 2.6.2
>>> >
>>> >
>>> > _______________________________________________
>>> > linux-arm-kernel mailing list
>>> > linux-arm-kernel at lists.infradead.org
>>> > http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
>>>
>>> Lucas,
>>>
>>> Any thoughts on why enabling MSI on IMX6 breaks legacy interrupts?
>>>
>>> Again, many PCI devices/drivers don't support MSI and thus enabling
>>> MSI as this patch does makes this config break many PCI devices on
>>> IMX6.
>>>
>>> I have found that MSI does indeed work on the IMX6 for cards/drivers
>>> that use MSI, but for those that don't the legacy interrupts never
>>> fire regardless of going through a bridge or directly hanging off the
>>> IMX6. I assume you can confirm this on your boards/devices as well?
>>>
>> No, this seems to be an errata in the DW PCIe core. The manual states
>> that legacy IRQs and MSIs can't be used together, but I thought this
>> only applies to the EP part of the core.
>>
>> We can probably fix the simple configuration by not enabling MSI support
>> until someone actually requests to set up an MSI IRQ, so both IRQ types
>> work when only one EP is connected.
>> I don't have any idea yet how to fix setups with a PCIe switch where
>> some connected EPs might like to use MSIs while some others can't use
>> them.
>>
>> Regards,
>> Lucas
>>
>
> Lucas,
>
> I feel a bit dumb here for forgetting about this issue and letting it
> drop, but I've just spent a day debugging an issue with the ath9k
> wireless driver only to realize it was that the ath9k card and/or
> driver doesn't support MSI interrupts and thus no longer work on
> mainline kernels that have MSI enabled. To be honest I'm not sure how
> many PCIe devices out there can't support MSI because of hardware
> limitations vs drivers that simply have not implemented it.
>
> To make matters worse 3ee803641e76bea76ec730c80dcc64739a9919ff makes
> it such that you can't disable MSI on the IMX6.
>
> Is there some fix you can think of to make the IMX6 PCIe host
> controller raise an interrupt for cards/drivers that use legacy irq's?
> If not, then we need to allow MSI to be disabled for IMX6 and default
> it to disabled for compatibility.
>

Joao / Jingoo,

(Adding linux-pci to the list to pick up maintainers of other DWC PCIe
core drivers)

Is Luca's statement that 'legacy IRQs and MSIs can't be used together'
true in general for the DW PCIe core? If any of the host controllers
using this core can't support both legacy and MSI irqs together I
still believe we shouldn't enable/require MSI as it breaks any
card/driver that only supports legacy interrupts (such as ath9k).

Regards,

Tim

Re: [PATCH 2/2] ARM: imx_v6_v7_defconfig: enable PCI_MSI

[Jingoo Han]

On Friday, March 31, 2017 12:05 PM, Tim Harvey wrote:
> On Thu, Mar 30, 2017 at 4:42 PM, Tim Harvey <tharvey@gateworks.com> wrote:
> > On Fri, Apr 15, 2016 at 1:21 AM, Lucas Stach <l.stach@pengutronix.de>
> wrote:
> >> Am Donnerstag, den 14.04.2016, 08:13 -0700 schrieb Tim Harvey:
> >>> On Mon, Nov 30, 2015 at 9:00 AM, Lucas Stach <l.stach@pengutronix.de>
> wrote:
> >>> > This has been stable on i.MX6 for a good while now and there is no
> >>> > reason to keep it disabled, as it allows to allocate non-shared
> >>> > IRQs if several PCIe devices are connected to a single host.
> >>> >
> >>> > Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
> >>> > ---
> >>> >  arch/arm/configs/imx_v6_v7_defconfig | 1 +
> >>> >  1 file changed, 1 insertion(+)
> >>> >
> >>> > diff --git a/arch/arm/configs/imx_v6_v7_defconfig
> b/arch/arm/configs/imx_v6_v7_defconfig
> >>> > index 4187f69f6630..e2ce60d00e8a 100644
> >>> > --- a/arch/arm/configs/imx_v6_v7_defconfig
> >>> > +++ b/arch/arm/configs/imx_v6_v7_defconfig
> >>> > @@ -45,6 +45,7 @@ CONFIG_SOC_IMX7D=y
> >>> >  CONFIG_SOC_LS1021A=y
> >>> >  CONFIG_SOC_VF610=y
> >>> >  CONFIG_PCI=y
> >>> > +CONFIG_PCI_MSI=y
> >>> >  CONFIG_PCI_IMX6=y
> >>> >  CONFIG_SMP=y
> >>> >  CONFIG_PREEMPT_VOLUNTARY=y
> >>> > --
> >>> > 2.6.2
> >>> >
> >>> >
> >>> > _______________________________________________
> >>> > linux-arm-kernel mailing list
> >>> > linux-arm-kernel@lists.infradead.org
> >>> > http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
> >>>
> >>> Lucas,
> >>>
> >>> Any thoughts on why enabling MSI on IMX6 breaks legacy interrupts?
> >>>
> >>> Again, many PCI devices/drivers don't support MSI and thus enabling
> >>> MSI as this patch does makes this config break many PCI devices on
> >>> IMX6.
> >>>
> >>> I have found that MSI does indeed work on the IMX6 for cards/drivers
> >>> that use MSI, but for those that don't the legacy interrupts never
> >>> fire regardless of going through a bridge or directly hanging off the
> >>> IMX6. I assume you can confirm this on your boards/devices as well?
> >>>
> >> No, this seems to be an errata in the DW PCIe core. The manual states
> >> that legacy IRQs and MSIs can't be used together, but I thought this
> >> only applies to the EP part of the core.
> >>
> >> We can probably fix the simple configuration by not enabling MSI
> support
> >> until someone actually requests to set up an MSI IRQ, so both IRQ types
> >> work when only one EP is connected.
> >> I don't have any idea yet how to fix setups with a PCIe switch where
> >> some connected EPs might like to use MSIs while some others can't use
> >> them.
> >>
> >> Regards,
> >> Lucas
> >>
> >
> > Lucas,
> >
> > I feel a bit dumb here for forgetting about this issue and letting it
> > drop, but I've just spent a day debugging an issue with the ath9k
> > wireless driver only to realize it was that the ath9k card and/or
> > driver doesn't support MSI interrupts and thus no longer work on
> > mainline kernels that have MSI enabled. To be honest I'm not sure how
> > many PCIe devices out there can't support MSI because of hardware
> > limitations vs drivers that simply have not implemented it.
> >
> > To make matters worse 3ee803641e76bea76ec730c80dcc64739a9919ff makes
> > it such that you can't disable MSI on the IMX6.
> >
> > Is there some fix you can think of to make the IMX6 PCIe host
> > controller raise an interrupt for cards/drivers that use legacy irq's?
> > If not, then we need to allow MSI to be disabled for IMX6 and default
> > it to disabled for compatibility.
> >
> 
> Joao / Jingoo,
> 
> (Adding linux-pci to the list to pick up maintainers of other DWC PCIe
> core drivers)
> 
> Is Luca's statement that 'legacy IRQs and MSIs can't be used together'
> true in general for the DW PCIe core? If any of the host controllers
> using this core can't support both legacy and MSI irqs together I
> still believe we shouldn't enable/require MSI as it breaks any
> card/driver that only supports legacy interrupts (such as ath9k).

If there are 2 DW PCIe controllers, one controller can be used for legacy
and another controller can be used for MSI.

But, I am not sure that one DW PCIe controller can support both MSI device
and legacy interrupt device at the same time.

To Joao Pinto,
Will you confirm this?

Or if there is anyone who knows this well, please share your knowledge. :-)
Thanks.

Best regards,
Jingoo Han

> 
> Regards,
> 
> Tim


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH 2/2] ARM: imx_v6_v7_defconfig: enable PCI_MSI

[Jingoo Han]

On Friday, March 31, 2017 12:05 PM, Tim Harvey wrote:
> On Thu, Mar 30, 2017 at 4:42 PM, Tim Harvey <tharvey@gateworks.com> wrote:
> > On Fri, Apr 15, 2016 at 1:21 AM, Lucas Stach <l.stach@pengutronix.de>
> wrote:
> >> Am Donnerstag, den 14.04.2016, 08:13 -0700 schrieb Tim Harvey:
> >>> On Mon, Nov 30, 2015 at 9:00 AM, Lucas Stach <l.stach@pengutronix.de>
> wrote:
> >>> > This has been stable on i.MX6 for a good while now and there is no
> >>> > reason to keep it disabled, as it allows to allocate non-shared
> >>> > IRQs if several PCIe devices are connected to a single host.
> >>> >
> >>> > Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
> >>> > ---
> >>> >  arch/arm/configs/imx_v6_v7_defconfig | 1 +
> >>> >  1 file changed, 1 insertion(+)
> >>> >
> >>> > diff --git a/arch/arm/configs/imx_v6_v7_defconfig
> b/arch/arm/configs/imx_v6_v7_defconfig
> >>> > index 4187f69f6630..e2ce60d00e8a 100644
> >>> > --- a/arch/arm/configs/imx_v6_v7_defconfig
> >>> > +++ b/arch/arm/configs/imx_v6_v7_defconfig
> >>> > @@ -45,6 +45,7 @@ CONFIG_SOC_IMX7D=y
> >>> >  CONFIG_SOC_LS1021A=y
> >>> >  CONFIG_SOC_VF610=y
> >>> >  CONFIG_PCI=y
> >>> > +CONFIG_PCI_MSI=y
> >>> >  CONFIG_PCI_IMX6=y
> >>> >  CONFIG_SMP=y
> >>> >  CONFIG_PREEMPT_VOLUNTARY=y
> >>> > --
> >>> > 2.6.2
> >>> >
> >>> >
> >>> > _______________________________________________
> >>> > linux-arm-kernel mailing list
> >>> > linux-arm-kernel at lists.infradead.org
> >>> > http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
> >>>
> >>> Lucas,
> >>>
> >>> Any thoughts on why enabling MSI on IMX6 breaks legacy interrupts?
> >>>
> >>> Again, many PCI devices/drivers don't support MSI and thus enabling
> >>> MSI as this patch does makes this config break many PCI devices on
> >>> IMX6.
> >>>
> >>> I have found that MSI does indeed work on the IMX6 for cards/drivers
> >>> that use MSI, but for those that don't the legacy interrupts never
> >>> fire regardless of going through a bridge or directly hanging off the
> >>> IMX6. I assume you can confirm this on your boards/devices as well?
> >>>
> >> No, this seems to be an errata in the DW PCIe core. The manual states
> >> that legacy IRQs and MSIs can't be used together, but I thought this
> >> only applies to the EP part of the core.
> >>
> >> We can probably fix the simple configuration by not enabling MSI
> support
> >> until someone actually requests to set up an MSI IRQ, so both IRQ types
> >> work when only one EP is connected.
> >> I don't have any idea yet how to fix setups with a PCIe switch where
> >> some connected EPs might like to use MSIs while some others can't use
> >> them.
> >>
> >> Regards,
> >> Lucas
> >>
> >
> > Lucas,
> >
> > I feel a bit dumb here for forgetting about this issue and letting it
> > drop, but I've just spent a day debugging an issue with the ath9k
> > wireless driver only to realize it was that the ath9k card and/or
> > driver doesn't support MSI interrupts and thus no longer work on
> > mainline kernels that have MSI enabled. To be honest I'm not sure how
> > many PCIe devices out there can't support MSI because of hardware
> > limitations vs drivers that simply have not implemented it.
> >
> > To make matters worse 3ee803641e76bea76ec730c80dcc64739a9919ff makes
> > it such that you can't disable MSI on the IMX6.
> >
> > Is there some fix you can think of to make the IMX6 PCIe host
> > controller raise an interrupt for cards/drivers that use legacy irq's?
> > If not, then we need to allow MSI to be disabled for IMX6 and default
> > it to disabled for compatibility.
> >
> 
> Joao / Jingoo,
> 
> (Adding linux-pci to the list to pick up maintainers of other DWC PCIe
> core drivers)
> 
> Is Luca's statement that 'legacy IRQs and MSIs can't be used together'
> true in general for the DW PCIe core? If any of the host controllers
> using this core can't support both legacy and MSI irqs together I
> still believe we shouldn't enable/require MSI as it breaks any
> card/driver that only supports legacy interrupts (such as ath9k).

If there are 2 DW PCIe controllers, one controller can be used for legacy
and another controller can be used for MSI.

But, I am not sure that one DW PCIe controller can support both MSI device
and legacy interrupt device at the same time.

To Joao Pinto,
Will you confirm this?

Or if there is anyone who knows this well, please share your knowledge. :-)
Thanks.

Best regards,
Jingoo Han

> 
> Regards,
> 
> Tim

Re: [PATCH 2/2] ARM: imx_v6_v7_defconfig: enable PCI_MSI

[Joao Pinto]
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[PATCH 2/2] ARM: imx_v6_v7_defconfig: enable PCI_MSI

[Joao Pinto]

?s 9:54 PM de 4/4/2017, Jingoo Han escreveu:
> On Friday, March 31, 2017 12:05 PM, Tim Harvey wrote:
>> On Thu, Mar 30, 2017 at 4:42 PM, Tim Harvey <tharvey@gateworks.com> wrote:
>>> On Fri, Apr 15, 2016 at 1:21 AM, Lucas Stach <l.stach@pengutronix.de>
>> wrote:
>>>> Am Donnerstag, den 14.04.2016, 08:13 -0700 schrieb Tim Harvey:
>>>>> On Mon, Nov 30, 2015 at 9:00 AM, Lucas Stach <l.stach@pengutronix.de>
>> wrote:
>>>>>> This has been stable on i.MX6 for a good while now and there is no
>>>>>> reason to keep it disabled, as it allows to allocate non-shared
>>>>>> IRQs if several PCIe devices are connected to a single host.
>>>>>>
>>>>>> Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
>>>>>> ---
>>>>>>  arch/arm/configs/imx_v6_v7_defconfig | 1 +
>>>>>>  1 file changed, 1 insertion(+)
>>>>>>
>>>>>> diff --git a/arch/arm/configs/imx_v6_v7_defconfig
>> b/arch/arm/configs/imx_v6_v7_defconfig
>>>>>> index 4187f69f6630..e2ce60d00e8a 100644
>>>>>> --- a/arch/arm/configs/imx_v6_v7_defconfig
>>>>>> +++ b/arch/arm/configs/imx_v6_v7_defconfig
>>>>>> @@ -45,6 +45,7 @@ CONFIG_SOC_IMX7D=y
>>>>>>  CONFIG_SOC_LS1021A=y
>>>>>>  CONFIG_SOC_VF610=y
>>>>>>  CONFIG_PCI=y
>>>>>> +CONFIG_PCI_MSI=y
>>>>>>  CONFIG_PCI_IMX6=y
>>>>>>  CONFIG_SMP=y
>>>>>>  CONFIG_PREEMPT_VOLUNTARY=y
>>>>>> --
>>>>>> 2.6.2
>>>>>>
>>>>>>
>>>>>> _______________________________________________
>>>>>> linux-arm-kernel mailing list
>>>>>> linux-arm-kernel at lists.infradead.org
>>>>>> https://urldefense.proofpoint.com/v2/url?u=http-3A__lists.infradead.org_mailman_listinfo_linux-2Darm-2Dkernel&d=DwICaQ&c=DPL6_X_6JkXFx7AXWqB0tg&r=s2fO0hii0OGNOv9qQy_HRXy-xAJUD1NNoEcc3io_kx0&m=yQx7ybu-JGcPPh6ZlQyxkC6_Uw62cuBKLb7o3zpK7hQ&s=o-tR0mZ5yKNTTXKX1MR7NuDLpHfvSHIjzi47YLue0aA&e= 
>>>>>
>>>>> Lucas,
>>>>>
>>>>> Any thoughts on why enabling MSI on IMX6 breaks legacy interrupts?
>>>>>
>>>>> Again, many PCI devices/drivers don't support MSI and thus enabling
>>>>> MSI as this patch does makes this config break many PCI devices on
>>>>> IMX6.
>>>>>
>>>>> I have found that MSI does indeed work on the IMX6 for cards/drivers
>>>>> that use MSI, but for those that don't the legacy interrupts never
>>>>> fire regardless of going through a bridge or directly hanging off the
>>>>> IMX6. I assume you can confirm this on your boards/devices as well?
>>>>>
>>>> No, this seems to be an errata in the DW PCIe core. The manual states
>>>> that legacy IRQs and MSIs can't be used together, but I thought this
>>>> only applies to the EP part of the core.
>>>>
>>>> We can probably fix the simple configuration by not enabling MSI
>> support
>>>> until someone actually requests to set up an MSI IRQ, so both IRQ types
>>>> work when only one EP is connected.
>>>> I don't have any idea yet how to fix setups with a PCIe switch where
>>>> some connected EPs might like to use MSIs while some others can't use
>>>> them.
>>>>
>>>> Regards,
>>>> Lucas
>>>>
>>>
>>> Lucas,
>>>
>>> I feel a bit dumb here for forgetting about this issue and letting it
>>> drop, but I've just spent a day debugging an issue with the ath9k
>>> wireless driver only to realize it was that the ath9k card and/or
>>> driver doesn't support MSI interrupts and thus no longer work on
>>> mainline kernels that have MSI enabled. To be honest I'm not sure how
>>> many PCIe devices out there can't support MSI because of hardware
>>> limitations vs drivers that simply have not implemented it.
>>>
>>> To make matters worse 3ee803641e76bea76ec730c80dcc64739a9919ff makes
>>> it such that you can't disable MSI on the IMX6.
>>>
>>> Is there some fix you can think of to make the IMX6 PCIe host
>>> controller raise an interrupt for cards/drivers that use legacy irq's?
>>> If not, then we need to allow MSI to be disabled for IMX6 and default
>>> it to disabled for compatibility.
>>>
>>
>> Joao / Jingoo,
>>
>> (Adding linux-pci to the list to pick up maintainers of other DWC PCIe
>> core drivers)
>>
>> Is Luca's statement that 'legacy IRQs and MSIs can't be used together'
>> true in general for the DW PCIe core? If any of the host controllers
>> using this core can't support both legacy and MSI irqs together I
>> still believe we shouldn't enable/require MSI as it breaks any
>> card/driver that only supports legacy interrupts (such as ath9k).
> 
> If there are 2 DW PCIe controllers, one controller can be used for legacy
> and another controller can be used for MSI.
> 
> But, I am not sure that one DW PCIe controller can support both MSI device
> and legacy interrupt device at the same time.
> 
> To Joao Pinto,
> Will you confirm this?
> 

Hi Jingoo,
I confirm that if a RC has MSI enable, it won't support legacy interrupts.

Regards,
Joao

> Or if there is anyone who knows this well, please share your knowledge. :-)
> Thanks.
> 
> Best regards,
> Jingoo Han
> 
>>
>> Regards,
>>
>> Tim
> 

Re: [PATCH 2/2] ARM: imx_v6_v7_defconfig: enable PCI_MSI

[Lucas Stach]

Am Mittwoch, den 05.04.2017, 13:07 +0100 schrieb Joao Pinto:
[...]
> >>> Lucas,
> >>>
> >>> I feel a bit dumb here for forgetting about this issue and letting it
> >>> drop, but I've just spent a day debugging an issue with the ath9k
> >>> wireless driver only to realize it was that the ath9k card and/or
> >>> driver doesn't support MSI interrupts and thus no longer work on
> >>> mainline kernels that have MSI enabled. To be honest I'm not sure how
> >>> many PCIe devices out there can't support MSI because of hardware
> >>> limitations vs drivers that simply have not implemented it.
> >>>
> >>> To make matters worse 3ee803641e76bea76ec730c80dcc64739a9919ff makes
> >>> it such that you can't disable MSI on the IMX6.
> >>>
> >>> Is there some fix you can think of to make the IMX6 PCIe host
> >>> controller raise an interrupt for cards/drivers that use legacy irq's?
> >>> If not, then we need to allow MSI to be disabled for IMX6 and default
> >>> it to disabled for compatibility.
> >>>
> >>
> >> Joao / Jingoo,
> >>
> >> (Adding linux-pci to the list to pick up maintainers of other DWC PCIe
> >> core drivers)
> >>
> >> Is Luca's statement that 'legacy IRQs and MSIs can't be used together'
> >> true in general for the DW PCIe core? If any of the host controllers
> >> using this core can't support both legacy and MSI irqs together I
> >> still believe we shouldn't enable/require MSI as it breaks any
> >> card/driver that only supports legacy interrupts (such as ath9k).
> > 
> > If there are 2 DW PCIe controllers, one controller can be used for legacy
> > and another controller can be used for MSI.
> > 
> > But, I am not sure that one DW PCIe controller can support both MSI device
> > and legacy interrupt device at the same time.
> > 
> > To Joao Pinto,
> > Will you confirm this?
> > 
> 
> Hi Jingoo,
> I confirm that if a RC has MSI enable, it won't support legacy interrupts.

Okay, so the only way to solve this in a generic way, that I can see at
the moment is to defer MSI enable until the downstream device actually
requests an MSI irq.

Also we need to disable the MSI capability if any bridge device turns up
during the bus topology scan, as we can have devices with conflicting
requirements connected to the RC in that case.

I'll cook up some patches to implement this.

Regards,
Lucas


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH 2/2] ARM: imx_v6_v7_defconfig: enable PCI_MSI

[Lucas Stach]

Am Mittwoch, den 05.04.2017, 13:07 +0100 schrieb Joao Pinto:
[...]
> >>> Lucas,
> >>>
> >>> I feel a bit dumb here for forgetting about this issue and letting it
> >>> drop, but I've just spent a day debugging an issue with the ath9k
> >>> wireless driver only to realize it was that the ath9k card and/or
> >>> driver doesn't support MSI interrupts and thus no longer work on
> >>> mainline kernels that have MSI enabled. To be honest I'm not sure how
> >>> many PCIe devices out there can't support MSI because of hardware
> >>> limitations vs drivers that simply have not implemented it.
> >>>
> >>> To make matters worse 3ee803641e76bea76ec730c80dcc64739a9919ff makes
> >>> it such that you can't disable MSI on the IMX6.
> >>>
> >>> Is there some fix you can think of to make the IMX6 PCIe host
> >>> controller raise an interrupt for cards/drivers that use legacy irq's?
> >>> If not, then we need to allow MSI to be disabled for IMX6 and default
> >>> it to disabled for compatibility.
> >>>
> >>
> >> Joao / Jingoo,
> >>
> >> (Adding linux-pci to the list to pick up maintainers of other DWC PCIe
> >> core drivers)
> >>
> >> Is Luca's statement that 'legacy IRQs and MSIs can't be used together'
> >> true in general for the DW PCIe core? If any of the host controllers
> >> using this core can't support both legacy and MSI irqs together I
> >> still believe we shouldn't enable/require MSI as it breaks any
> >> card/driver that only supports legacy interrupts (such as ath9k).
> > 
> > If there are 2 DW PCIe controllers, one controller can be used for legacy
> > and another controller can be used for MSI.
> > 
> > But, I am not sure that one DW PCIe controller can support both MSI device
> > and legacy interrupt device at the same time.
> > 
> > To Joao Pinto,
> > Will you confirm this?
> > 
> 
> Hi Jingoo,
> I confirm that if a RC has MSI enable, it won't support legacy interrupts.

Okay, so the only way to solve this in a generic way, that I can see at
the moment is to defer MSI enable until the downstream device actually
requests an MSI irq.

Also we need to disable the MSI capability if any bridge device turns up
during the bus topology scan, as we can have devices with conflicting
requirements connected to the RC in that case.

I'll cook up some patches to implement this.

Regards,
Lucas

Re: [PATCH 2/2] ARM: imx_v6_v7_defconfig: enable PCI_MSI

[Tim Harvey]

On Wed, Apr 5, 2017 at 5:16 AM, Lucas Stach <l.stach@pengutronix.de> wrote:
> Am Mittwoch, den 05.04.2017, 13:07 +0100 schrieb Joao Pinto:
> [...]
>> >>> Lucas,
>> >>>
>> >>> I feel a bit dumb here for forgetting about this issue and letting it
>> >>> drop, but I've just spent a day debugging an issue with the ath9k
>> >>> wireless driver only to realize it was that the ath9k card and/or
>> >>> driver doesn't support MSI interrupts and thus no longer work on
>> >>> mainline kernels that have MSI enabled. To be honest I'm not sure how
>> >>> many PCIe devices out there can't support MSI because of hardware
>> >>> limitations vs drivers that simply have not implemented it.
>> >>>
>> >>> To make matters worse 3ee803641e76bea76ec730c80dcc64739a9919ff makes
>> >>> it such that you can't disable MSI on the IMX6.
>> >>>
>> >>> Is there some fix you can think of to make the IMX6 PCIe host
>> >>> controller raise an interrupt for cards/drivers that use legacy irq's?
>> >>> If not, then we need to allow MSI to be disabled for IMX6 and default
>> >>> it to disabled for compatibility.
>> >>>
>> >>
>> >> Joao / Jingoo,
>> >>
>> >> (Adding linux-pci to the list to pick up maintainers of other DWC PCIe
>> >> core drivers)
>> >>
>> >> Is Luca's statement that 'legacy IRQs and MSIs can't be used together'
>> >> true in general for the DW PCIe core? If any of the host controllers
>> >> using this core can't support both legacy and MSI irqs together I
>> >> still believe we shouldn't enable/require MSI as it breaks any
>> >> card/driver that only supports legacy interrupts (such as ath9k).
>> >
>> > If there are 2 DW PCIe controllers, one controller can be used for legacy
>> > and another controller can be used for MSI.
>> >
>> > But, I am not sure that one DW PCIe controller can support both MSI device
>> > and legacy interrupt device at the same time.
>> >
>> > To Joao Pinto,
>> > Will you confirm this?
>> >
>>
>> Hi Jingoo,
>> I confirm that if a RC has MSI enable, it won't support legacy interrupts.
>
> Okay, so the only way to solve this in a generic way, that I can see at
> the moment is to defer MSI enable until the downstream device actually
> requests an MSI irq.
>
> Also we need to disable the MSI capability if any bridge device turns up
> during the bus topology scan, as we can have devices with conflicting
> requirements connected to the RC in that case.
>
> I'll cook up some patches to implement this.
>

Lucas,

Any thoughts yet? I'll likely want to backport this to stable although
the restructuring to drivers/pci/dwc and subsequent changes since 4.10
will likely require a re-written patch.

Best regards,

Tim

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH 2/2] ARM: imx_v6_v7_defconfig: enable PCI_MSI

[Tim Harvey]

On Wed, Apr 5, 2017 at 5:16 AM, Lucas Stach <l.stach@pengutronix.de> wrote:
> Am Mittwoch, den 05.04.2017, 13:07 +0100 schrieb Joao Pinto:
> [...]
>> >>> Lucas,
>> >>>
>> >>> I feel a bit dumb here for forgetting about this issue and letting it
>> >>> drop, but I've just spent a day debugging an issue with the ath9k
>> >>> wireless driver only to realize it was that the ath9k card and/or
>> >>> driver doesn't support MSI interrupts and thus no longer work on
>> >>> mainline kernels that have MSI enabled. To be honest I'm not sure how
>> >>> many PCIe devices out there can't support MSI because of hardware
>> >>> limitations vs drivers that simply have not implemented it.
>> >>>
>> >>> To make matters worse 3ee803641e76bea76ec730c80dcc64739a9919ff makes
>> >>> it such that you can't disable MSI on the IMX6.
>> >>>
>> >>> Is there some fix you can think of to make the IMX6 PCIe host
>> >>> controller raise an interrupt for cards/drivers that use legacy irq's?
>> >>> If not, then we need to allow MSI to be disabled for IMX6 and default
>> >>> it to disabled for compatibility.
>> >>>
>> >>
>> >> Joao / Jingoo,
>> >>
>> >> (Adding linux-pci to the list to pick up maintainers of other DWC PCIe
>> >> core drivers)
>> >>
>> >> Is Luca's statement that 'legacy IRQs and MSIs can't be used together'
>> >> true in general for the DW PCIe core? If any of the host controllers
>> >> using this core can't support both legacy and MSI irqs together I
>> >> still believe we shouldn't enable/require MSI as it breaks any
>> >> card/driver that only supports legacy interrupts (such as ath9k).
>> >
>> > If there are 2 DW PCIe controllers, one controller can be used for legacy
>> > and another controller can be used for MSI.
>> >
>> > But, I am not sure that one DW PCIe controller can support both MSI device
>> > and legacy interrupt device at the same time.
>> >
>> > To Joao Pinto,
>> > Will you confirm this?
>> >
>>
>> Hi Jingoo,
>> I confirm that if a RC has MSI enable, it won't support legacy interrupts.
>
> Okay, so the only way to solve this in a generic way, that I can see at
> the moment is to defer MSI enable until the downstream device actually
> requests an MSI irq.
>
> Also we need to disable the MSI capability if any bridge device turns up
> during the bus topology scan, as we can have devices with conflicting
> requirements connected to the RC in that case.
>
> I'll cook up some patches to implement this.
>

Lucas,

Any thoughts yet? I'll likely want to backport this to stable although
the restructuring to drivers/pci/dwc and subsequent changes since 4.10
will likely require a re-written patch.

Best regards,

Tim

Re: [PATCH 2/2] ARM: imx_v6_v7_defconfig: enable PCI_MSI

[Lucas Stach]

Hi Tim,

Am Dienstag, den 11.04.2017, 09:05 -0700 schrieb Tim Harvey:
> On Wed, Apr 5, 2017 at 5:16 AM, Lucas Stach <l.stach@pengutronix.de> wrote:
> > Am Mittwoch, den 05.04.2017, 13:07 +0100 schrieb Joao Pinto:
> > [...]
> >> >>> Lucas,
> >> >>>
> >> >>> I feel a bit dumb here for forgetting about this issue and letting it
> >> >>> drop, but I've just spent a day debugging an issue with the ath9k
> >> >>> wireless driver only to realize it was that the ath9k card and/or
> >> >>> driver doesn't support MSI interrupts and thus no longer work on
> >> >>> mainline kernels that have MSI enabled. To be honest I'm not sure how
> >> >>> many PCIe devices out there can't support MSI because of hardware
> >> >>> limitations vs drivers that simply have not implemented it.
> >> >>>
> >> >>> To make matters worse 3ee803641e76bea76ec730c80dcc64739a9919ff makes
> >> >>> it such that you can't disable MSI on the IMX6.
> >> >>>
> >> >>> Is there some fix you can think of to make the IMX6 PCIe host
> >> >>> controller raise an interrupt for cards/drivers that use legacy irq's?
> >> >>> If not, then we need to allow MSI to be disabled for IMX6 and default
> >> >>> it to disabled for compatibility.
> >> >>>
> >> >>
> >> >> Joao / Jingoo,
> >> >>
> >> >> (Adding linux-pci to the list to pick up maintainers of other DWC PCIe
> >> >> core drivers)
> >> >>
> >> >> Is Luca's statement that 'legacy IRQs and MSIs can't be used together'
> >> >> true in general for the DW PCIe core? If any of the host controllers
> >> >> using this core can't support both legacy and MSI irqs together I
> >> >> still believe we shouldn't enable/require MSI as it breaks any
> >> >> card/driver that only supports legacy interrupts (such as ath9k).
> >> >
> >> > If there are 2 DW PCIe controllers, one controller can be used for legacy
> >> > and another controller can be used for MSI.
> >> >
> >> > But, I am not sure that one DW PCIe controller can support both MSI device
> >> > and legacy interrupt device at the same time.
> >> >
> >> > To Joao Pinto,
> >> > Will you confirm this?
> >> >
> >>
> >> Hi Jingoo,
> >> I confirm that if a RC has MSI enable, it won't support legacy interrupts.
> >
> > Okay, so the only way to solve this in a generic way, that I can see at
> > the moment is to defer MSI enable until the downstream device actually
> > requests an MSI irq.
> >
> > Also we need to disable the MSI capability if any bridge device turns up
> > during the bus topology scan, as we can have devices with conflicting
> > requirements connected to the RC in that case.
> >
> > I'll cook up some patches to implement this.
> >
> 
> Lucas,
> 
> Any thoughts yet? I'll likely want to backport this to stable although
> the restructuring to drivers/pci/dwc and subsequent changes since 4.10
> will likely require a re-written patch.

I have a fix for this, that I'm currently testing on different setups.
I'll probably send it out in a few hours.

Regards,
Lucas


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH 2/2] ARM: imx_v6_v7_defconfig: enable PCI_MSI

[Lucas Stach]

Hi Tim,

Am Dienstag, den 11.04.2017, 09:05 -0700 schrieb Tim Harvey:
> On Wed, Apr 5, 2017 at 5:16 AM, Lucas Stach <l.stach@pengutronix.de> wrote:
> > Am Mittwoch, den 05.04.2017, 13:07 +0100 schrieb Joao Pinto:
> > [...]
> >> >>> Lucas,
> >> >>>
> >> >>> I feel a bit dumb here for forgetting about this issue and letting it
> >> >>> drop, but I've just spent a day debugging an issue with the ath9k
> >> >>> wireless driver only to realize it was that the ath9k card and/or
> >> >>> driver doesn't support MSI interrupts and thus no longer work on
> >> >>> mainline kernels that have MSI enabled. To be honest I'm not sure how
> >> >>> many PCIe devices out there can't support MSI because of hardware
> >> >>> limitations vs drivers that simply have not implemented it.
> >> >>>
> >> >>> To make matters worse 3ee803641e76bea76ec730c80dcc64739a9919ff makes
> >> >>> it such that you can't disable MSI on the IMX6.
> >> >>>
> >> >>> Is there some fix you can think of to make the IMX6 PCIe host
> >> >>> controller raise an interrupt for cards/drivers that use legacy irq's?
> >> >>> If not, then we need to allow MSI to be disabled for IMX6 and default
> >> >>> it to disabled for compatibility.
> >> >>>
> >> >>
> >> >> Joao / Jingoo,
> >> >>
> >> >> (Adding linux-pci to the list to pick up maintainers of other DWC PCIe
> >> >> core drivers)
> >> >>
> >> >> Is Luca's statement that 'legacy IRQs and MSIs can't be used together'
> >> >> true in general for the DW PCIe core? If any of the host controllers
> >> >> using this core can't support both legacy and MSI irqs together I
> >> >> still believe we shouldn't enable/require MSI as it breaks any
> >> >> card/driver that only supports legacy interrupts (such as ath9k).
> >> >
> >> > If there are 2 DW PCIe controllers, one controller can be used for legacy
> >> > and another controller can be used for MSI.
> >> >
> >> > But, I am not sure that one DW PCIe controller can support both MSI device
> >> > and legacy interrupt device at the same time.
> >> >
> >> > To Joao Pinto,
> >> > Will you confirm this?
> >> >
> >>
> >> Hi Jingoo,
> >> I confirm that if a RC has MSI enable, it won't support legacy interrupts.
> >
> > Okay, so the only way to solve this in a generic way, that I can see at
> > the moment is to defer MSI enable until the downstream device actually
> > requests an MSI irq.
> >
> > Also we need to disable the MSI capability if any bridge device turns up
> > during the bus topology scan, as we can have devices with conflicting
> > requirements connected to the RC in that case.
> >
> > I'll cook up some patches to implement this.
> >
> 
> Lucas,
> 
> Any thoughts yet? I'll likely want to backport this to stable although
> the restructuring to drivers/pci/dwc and subsequent changes since 4.10
> will likely require a re-written patch.

I have a fix for this, that I'm currently testing on different setups.
I'll probably send it out in a few hours.

Regards,
Lucas