[PATCH 1/4] ARM: dts: sun8i-a83t-tbs-a711: OOB WiFi interrupt doesn't work

12 views
Skip to first unread message

Ondrej Jirman

unread,
Feb 22, 2020, 5:31:59 PM2/22/20
to linux...@googlegroups.com, Maxime Ripard, Chen-Yu Tsai, Ondrej Jirman, Tomas Novotny, Rob Herring, Mark Rutland, open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS, moderated list:ARM/Allwinner sunXi SoC support, open list
It just causes a constant rate of 5000 interrupts per second for both
GPIO and MMC, even if nothing is happening. Rely on in-band interrupts
instead.

Fixes: 0e23372080def7bb ("arm: dts: sun8i: Add the TBS A711 tablet devicetree")
Signed-off-by: Ondrej Jirman <meg...@megous.com>
---
arch/arm/boot/dts/sun8i-a83t-tbs-a711.dts | 3 ---
1 file changed, 3 deletions(-)

diff --git a/arch/arm/boot/dts/sun8i-a83t-tbs-a711.dts b/arch/arm/boot/dts/sun8i-a83t-tbs-a711.dts
index 2fd31a0a0b344..ee5ce3556b2ad 100644
--- a/arch/arm/boot/dts/sun8i-a83t-tbs-a711.dts
+++ b/arch/arm/boot/dts/sun8i-a83t-tbs-a711.dts
@@ -214,9 +214,6 @@ &mmc1 {
brcmf: wifi@1 {
reg = <1>;
compatible = "brcm,bcm4329-fmac";
- interrupt-parent = <&r_pio>;
- interrupts = <0 3 IRQ_TYPE_LEVEL_LOW>; /* PL3 WL_WAKE_UP */
- interrupt-names = "host-wake";
};
};

--
2.25.1

Chen-Yu Tsai

unread,
Feb 22, 2020, 10:26:39 PM2/22/20
to Ondřej Jirman, linux-sunxi, Maxime Ripard, Tomas Novotny, Rob Herring, Mark Rutland, open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS, moderated list:ARM/Allwinner sunXi SoC support, open list, Arend van Spriel
Hi,


On Sun, Feb 23, 2020 at 6:32 AM Ondrej Jirman <meg...@megous.com> wrote:
>
> It just causes a constant rate of 5000 interrupts per second for both
> GPIO and MMC, even if nothing is happening. Rely on in-band interrupts
> instead.
>
> Fixes: 0e23372080def7bb ("arm: dts: sun8i: Add the TBS A711 tablet devicetree")
> Signed-off-by: Ondrej Jirman <meg...@megous.com>

What WiFi chip/module does this use? It might be worth asking Broadcom
people to help with this and fix the driver.

ChenYu

> ---
> arch/arm/boot/dts/sun8i-a83t-tbs-a711.dts | 3 ---
> 1 file changed, 3 deletions(-)
>
> diff --git a/arch/arm/boot/dts/sun8i-a83t-tbs-a711.dts b/arch/arm/boot/dts/sun8i-a83t-tbs-a711.dts
> index 2fd31a0a0b344..ee5ce3556b2ad 100644
> --- a/arch/arm/boot/dts/sun8i-a83t-tbs-a711.dts
> +++ b/arch/arm/boot/dts/sun8i-a83t-tbs-a711.dts
> @@ -214,9 +214,6 @@ &mmc1 {
> brcmf: wifi@1 {
> reg = <1>;
> compatible = "brcm,bcm4329-fmac";
> - interrupt-parent = <&r_pio>;
> - interrupts = <0 3 IRQ_TYPE_LEVEL_LOW>; /* PL3 WL_WAKE_UP */
> - interrupt-names = "host-wake";
> };
> };
>
> --
> 2.25.1
>
> --
> You received this message because you are subscribed to the Google Groups "linux-sunxi" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to linux-sunxi...@googlegroups.com.
> To view this discussion on the web, visit https://groups.google.com/d/msgid/linux-sunxi/20200222223154.221632-2-megous%40megous.com.

Chen-Yu Tsai

unread,
Feb 22, 2020, 11:04:01 PM2/22/20
to Ondřej Jirman, linux-sunxi, Maxime Ripard, Tomas Novotny, Rob Herring, Mark Rutland, open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS, moderated list:ARM/Allwinner sunXi SoC support, open list, Arend van Spriel
On Sun, Feb 23, 2020 at 11:26 AM Chen-Yu Tsai <we...@csie.org> wrote:
>
> Hi,
>
>
> On Sun, Feb 23, 2020 at 6:32 AM Ondrej Jirman <meg...@megous.com> wrote:
> >
> > It just causes a constant rate of 5000 interrupts per second for both
> > GPIO and MMC, even if nothing is happening. Rely on in-band interrupts
> > instead.
> >
> > Fixes: 0e23372080def7bb ("arm: dts: sun8i: Add the TBS A711 tablet devicetree")
> > Signed-off-by: Ondrej Jirman <meg...@megous.com>
>
> What WiFi chip/module does this use? It might be worth asking Broadcom
> people to help with this and fix the driver.

Based on the comments in the device tree file, it uses an AP6210, which
is a BCM43362 inside for SDIO-based WiFi. There is a recent fix in 5.6-rc1
for this,

8c8e60fb86a9 brcmfmac: sdio: Fix OOB interrupt initialization on brcm43362

which seems to fix things for me. Could you try it on your end?

ChenYu

Ondřej Jirman

unread,
Feb 23, 2020, 5:02:29 AM2/23/20
to Chen-Yu Tsai, linux-sunxi, Maxime Ripard, Tomas Novotny, Rob Herring, Mark Rutland, open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS, moderated list:ARM/Allwinner sunXi SoC support, open list, Arend van Spriel
On Sun, Feb 23, 2020 at 12:03:46PM +0800, Chen-Yu Tsai wrote:
> On Sun, Feb 23, 2020 at 11:26 AM Chen-Yu Tsai <we...@csie.org> wrote:
> >
> > Hi,
> >
> >
> > On Sun, Feb 23, 2020 at 6:32 AM Ondrej Jirman <meg...@megous.com> wrote:
> > >
> > > It just causes a constant rate of 5000 interrupts per second for both
> > > GPIO and MMC, even if nothing is happening. Rely on in-band interrupts
> > > instead.
> > >
> > > Fixes: 0e23372080def7bb ("arm: dts: sun8i: Add the TBS A711 tablet devicetree")
> > > Signed-off-by: Ondrej Jirman <meg...@megous.com>
> >
> > What WiFi chip/module does this use? It might be worth asking Broadcom
> > people to help with this and fix the driver.
>
> Based on the comments in the device tree file, it uses an AP6210, which
> is a BCM43362 inside for SDIO-based WiFi. There is a recent fix in 5.6-rc1
> for this,
>
> 8c8e60fb86a9 brcmfmac: sdio: Fix OOB interrupt initialization on brcm43362
>
> which seems to fix things for me. Could you try it on your end?

Interesting, thanks for finding out! I'll test it.

I think it will work, since my tablet started having overheating issue recently,
and I tracked it down to this.

regards,
o.

Ondřej Jirman

unread,
Feb 23, 2020, 5:55:28 AM2/23/20
to Chen-Yu Tsai, linux-sunxi, Maxime Ripard, Tomas Novotny, Rob Herring, Mark Rutland, open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS, moderated list:ARM/Allwinner sunXi SoC support, open list, Arend van Spriel
Hello,

On Sun, Feb 23, 2020 at 12:03:46PM +0800, Chen-Yu Tsai wrote:
> On Sun, Feb 23, 2020 at 11:26 AM Chen-Yu Tsai <we...@csie.org> wrote:
> >
> > Hi,
> >
> >
> > On Sun, Feb 23, 2020 at 6:32 AM Ondrej Jirman <meg...@megous.com> wrote:
> > >
> > > It just causes a constant rate of 5000 interrupts per second for both
> > > GPIO and MMC, even if nothing is happening. Rely on in-band interrupts
> > > instead.
> > >
> > > Fixes: 0e23372080def7bb ("arm: dts: sun8i: Add the TBS A711 tablet devicetree")
> > > Signed-off-by: Ondrej Jirman <meg...@megous.com>
> >
> > What WiFi chip/module does this use? It might be worth asking Broadcom
> > people to help with this and fix the driver.
>
> Based on the comments in the device tree file, it uses an AP6210, which
> is a BCM43362 inside for SDIO-based WiFi. There is a recent fix in 5.6-rc1
> for this,
>
> 8c8e60fb86a9 brcmfmac: sdio: Fix OOB interrupt initialization on brcm43362
>
> which seems to fix things for me. Could you try it on your end?

I can confirm that it works as you say (on linus/master). 5.5 still doesn't have
the patch, so it's broken there, which confused me I guess.

Please ignore this patch.

thank you,
Ondrej

Arend Van Spriel

unread,
Feb 23, 2020, 6:29:28 AM2/23/20
to Ondřej Jirman, Chen-Yu Tsai, linux-sunxi, Maxime Ripard, Tomas Novotny, Rob Herring, Mark Rutland, open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS, moderated list:ARM/Allwinner sunXi SoC support, open list
On February 23, 2020 11:55:25 AM Ondřej Jirman <meg...@megous.com> wrote:

> Hello,
>
> On Sun, Feb 23, 2020 at 12:03:46PM +0800, Chen-Yu Tsai wrote:
>> On Sun, Feb 23, 2020 at 11:26 AM Chen-Yu Tsai <we...@csie.org> wrote:
>>>
>>> Hi,
>>>
>>>
>>> On Sun, Feb 23, 2020 at 6:32 AM Ondrej Jirman <meg...@megous.com> wrote:
>>>>
>>>> It just causes a constant rate of 5000 interrupts per second for both
>>>> GPIO and MMC, even if nothing is happening. Rely on in-band interrupts
>>>> instead.
>>>>
>>>> Fixes: 0e23372080def7bb ("arm: dts: sun8i: Add the TBS A711 tablet devicetree")
>>>> Signed-off-by: Ondrej Jirman <meg...@megous.com>
>>>
>>> What WiFi chip/module does this use? It might be worth asking Broadcom
>>> people to help with this and fix the driver.
>>
>> Based on the comments in the device tree file, it uses an AP6210, which
>> is a BCM43362 inside for SDIO-based WiFi. There is a recent fix in 5.6-rc1
>> for this,
>>
>> 8c8e60fb86a9 brcmfmac: sdio: Fix OOB interrupt initialization on brcm43362
>>
>> which seems to fix things for me. Could you try it on your end?
>
> I can confirm that it works as you say (on linus/master). 5.5 still doesn't
> have
> the patch, so it's broken there, which confused me I guess.

Hi Ondrej,

I have seen emails from GregKH including this patch in 5.5 stable so it
will also land there eventually.

Regards,
Arend
Reply all
Reply to author
Forward
0 new messages