Hi Ondrej, On Tue, 4 Jun 2019 at 17:30, Ondřej Jirman <megous@xxxxxxxxxx> wrote: > > Hi Clément, > > On Tue, Jun 04, 2019 at 05:04:07PM +0200, Clément Péron wrote: > > Hi Ondrej, > > > > On Tue, 4 Jun 2019 at 16:47, Ondřej Jirman <megous@xxxxxxxxxx> wrote: > > > > > > Hi Clément, > > > > > > On Tue, Jun 04, 2019 at 02:33:55PM +0200, verejna wrote: > > > > Hi Clément, > > > > > > > > On Mon, Jun 03, 2019 at 09:58:23PM +0200, Clément Péron wrote: > > > > > Hi Ondrej, > > > > > > > > > > On Fri, 31 May 2019 at 14:46, Ondřej Jirman <megous@xxxxxxxxxx> wrote: > > > > > > > > > > > > Hello Clément, > > > > > > > > > > > > On Fri, May 31, 2019 at 12:25:32AM +0200, Clément Péron wrote: > > > > > > > Hi Ondrej, > > > > > > > > > > > > > > On Thu, 30 May 2019 at 16:55, Ondřej Jirman <megous@xxxxxxxxxx> wrote: > > > > > > > > > > > > > > > > Hello Clément, > > > > > > > > > > > > > > > > On Tue, May 28, 2019 at 06:14:38PM +0200, Clément Péron wrote: > > > > > > > > > Allwinner H6 IR is similar to A31 and can use same driver. > > > > > > > > > > > > > > > > > > Add support for it. > > > > > > > > > > > > > > > > > > Signed-off-by: Clément Péron <peron.clem@xxxxxxxxx> > > > > > > > > > --- > > > > > > > > > arch/arm64/boot/dts/allwinner/sun50i-h6.dtsi | 19 +++++++++++++++++++ > > > > > > > > > 1 file changed, 19 insertions(+) > > > > > > > > > > > > > > > > > > diff --git a/arch/arm64/boot/dts/allwinner/sun50i-h6.dtsi b/arch/arm64/boot/dts/allwinner/sun50i-h6.dtsi > > > > > > > > > index 16c5c3d0fd81..649cbdfe452e 100644 > > > > > > > > > --- a/arch/arm64/boot/dts/allwinner/sun50i-h6.dtsi > > > > > > > > > +++ b/arch/arm64/boot/dts/allwinner/sun50i-h6.dtsi > > > > > > > > > @@ -647,6 +647,25 @@ > > > > > > > > > pins = "PL0", "PL1"; > > > > > > > > > function = "s_i2c"; > > > > > > > > > }; > > > > > > > > > + > > > > > > > > > + r_ir_rx_pin: r-ir-rx-pin { > > > > > > > > > + pins = "PL9"; > > > > > > > > > + function = "s_cir_rx"; > > > > > > > > > + }; > > > > > > > > > + }; > > > > > > > > > + > > > > > > > > > + r_ir: ir@7040000 { > > > > > > > > > + compatible = "allwinner,sun50i-h6-ir", > > > > > > > > > + "allwinner,sun6i-a31-ir"; > > > > > > > > > + reg = <0x07040000 0x400>; > > > > > > > > > + interrupts = <GIC_SPI 109 IRQ_TYPE_LEVEL_HIGH>; > > > > > > > > > + clocks = <&r_ccu CLK_R_APB1_IR>, > > > > > > > > > + <&r_ccu CLK_IR>; > > > > > > > > > + clock-names = "apb", "ir"; > > > > > > > > > + resets = <&r_ccu RST_R_APB1_IR>; > > > > > > > > > + pinctrl-names = "default"; > > > > > > > > > + pinctrl-0 = <&r_ir_rx_pin>; > > > > > > > > > + status = "disabled"; > > > > > > > > > }; > > > > > > > > > > > > > > > > Please make a comment here, that this is known broken on some boards and may > > > > > > > > result IRQ flood if enabled. Otherwise noone will know. > > > > > > > > > > > > > > I'm planning to send a v4 next week with the IRQ_NONE return as Maxime > > > > > > > suggested it. > > > > > > > https://github.com/clementperon/linux/tree/h6_ir_v4 > > > > > > > > > > > > > > But maybe we could also use the bit 5 of the IRQ status. > > > > > > > > > > > > Thanks, that's nice, but that will not make the HW work. That will just disable > > > > > > it. The comment is still necessary. > > > > > I have pushed a new version on my github. > > > > > https://github.com/clementperon/linux/commits/h6_ir_v4 > > > > > > > > > > I will submit it, if you are ok with it. > > > > > > > > the changes make it worse. Console is flooded with "Temporarily disable IRQ" > > > > and other symptoms are the same as I described before. Interrupts are not > > > > disabled in a any reasonable time. (I've waited for more > 5mins already.) > > > > > > > > You probably need to disable interrupts right away, not wait for 100k failures. > > > > > > Hmm, this is what the registers look like post-probe: > > > > > > R_CIR: > > > 0x07040000 : 00000030 > > > 0x07040004 : 00000030 > > > 0x07040008 : 00000030 > > > 0x0704000c : 00000030 > > > 0x07040010 : 00000030 > > > 0x07040014 : 00000030 > > > 0x07040018 : 00000030 > > > 0x0704001c : 00000030 > > > 0x07040020 : 00000030 > > > 0x07040024 : 00000030 > > > 0x07040028 : 00000030 > > > 0x0704002c : 00000030 > > > 0x07040030 : 00000030 > > > 0x07040034 : 00000030 > > > 0x07040038 : 00000030 > > > 0x0704003c : 00000030 > > > 0x07040040 : 00000030 > > > 0x07040044 : 00000030 > > > 0x07040048 : 00000030 > > > 0x0704004c : 00000030 > > > 0x07040050 : 00000030 > > > 0x07040054 : 00000030 > > > 0x07040058 : 00000030 > > > 0x0704005c : 00000030 > > > 0x07040060 : 00000030 > > > 0x07040064 : 00000030 > > > 0x07040068 : 00000030 > > > 0x0704006c : 00000030 > > > 0x07040070 : 00000030 > > > 0x07040074 : 00000030 > > > 0x07040078 : 00000030 > > > 0x0704007c : 00000030 > > > 0x07040080 : 00000030 > > > 0x07040084 : 00000030 > > > 0x07040088 : 00000030 > > > 0x0704008c : 00000030 > > > 0x07040090 : 00000030 > > > 0x07040094 : 00000030 > > > 0x07040098 : 00000030 > > > 0x0704009c : 00000030 > > > 0x070400a0 : 00000030 > > > 0x070400a4 : 00000030 > > > 0x070400a8 : 00000030 > > > 0x070400ac : 00000030 > > > 0x070400b0 : 00000030 > > > 0x070400b4 : 00000030 > > > 0x070400b8 : 00000030 > > > 0x070400bc : 00000030 > > > 0x070400c0 : 00000030 > > > 0x070400c4 : 00000030 > > > 0x070400c8 : 00000030 > > > 0x070400cc : 00000030 > > > 0x070400d0 : 00000030 > > > 0x070400d4 : 00000030 > > > 0x070400d8 : 00000030 > > > 0x070400dc : 00000030 > > > 0x070400e0 : 00000030 > > > 0x070400e4 : 00000030 > > > 0x070400e8 : 00000030 > > > 0x070400ec : 00000030 > > > 0x070400f0 : 00000030 > > > 0x070400f4 : 00000030 > > > 0x070400f8 : 00000030 > > > 0x070400fc : 00000030 > > > > > > Clearly not right. It's just the R_CIR module, other modules have normal values. > > > > > > I've checked: > > > 0x070101c0 : 81000002 > > > (IR clock config register) > > > 0x070101cc : 00010000 > > > (IR reset/bus clk gate reg) > > > > > > static const char * const r_mod0_default_parents[] = { "osc32k", "osc24M" }; > > > static SUNXI_CCU_MP_WITH_MUX_GATE(ir_clk, "ir", > > > r_mod0_default_parents, 0x1c0, > > > 0, 5, /* M */ > > > 8, 2, /* P */ > > > 24, 1, /* mux */ > > > BIT(31), /* gate */ > > > 0); > > > > > > static SUNXI_CCU_GATE(r_apb1_ir_clk, "r-apb1-ir", "r-apb1", > > > 0x1cc, BIT(0), 0); > > > > > > [RST_R_APB1_IR] = { 0x1cc, BIT(16) }, > > > > > > So parent clock seems to be OK. But gate clock is not enabled, so the bus > > > is not working. > > > > > > And look at this!!: > > > > > > static SUNXI_CCU_GATE(r_apb1_ir_clk, "r-apb1-ir", "r-apb1", > > > 0x1cc, BIT(0), 0); > > > static SUNXI_CCU_GATE(r_apb1_w1_clk, "r-apb1-w1", "r-apb1", > > > 0x1cc, BIT(0), 0); > > > > > > So, it's wrong w1 gate config! > > > > > > You can drop your changes, because I've probbably found the root cause. > > > > Nice to see that you have found the issue, but I don't understand why > > It's working on my board on not on yours. > > Maybe you use modules? I have a builtin driver. Correct, > > That may change initialization order. It would disable unused gates first, and > when you load the module later from userpsace then it would enable the gate. > > If builtin, then it would enable IR gate first, and then disable the unused > gates (W1 in this case), later on when entering userspace. Thanks for the explanation it makes sense. > > Anyway, I can confirm that now, when I turn on the light in the room, I get > around 10 interrupts with empty FIFO and than it stops. Ok I will push a V4 with just the introduction of the RXSTA. Regards, Clément > > It doesn't cause the flood anymore. > > regards, > o. > > > Regards, > > Clément > > > > > > > > regards, > > > o. > > > > > > > thank you and regards, > > > > o. > > > > > > > > > Thanks, > > > > > Clément > > > > > > > > > > > > > > > > > thank you, > > > > > > o. > > > > > > > > > > > > > Regards, Clement > > > > > > > > > > > > > > > > > > > > > > > thanks, > > > > > > > > o. > > > > > > > > > > > > > > > > > r_i2c: i2c@7081400 { > > > > > > > > > -- > > > > > > > > > 2.20.1 > > > > > > > > > > > > > > > > > > > > > > > > > > > _______________________________________________ > > > > > > > > > linux-arm-kernel mailing list > > > > > > > > > linux-arm-kernel@xxxxxxxxxxxxxxxxxxx > > > > > > > > > http://lists.infradead.org/mailman/listinfo/linux-arm-kernel > > > > > > > > > > > > > > -- > > > > > > > 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+unsubscribe@xxxxxxxxxxxxxxxx. > > > > > > > To view this discussion on the web, visit https://groups.google.com/d/msgid/linux-sunxi/CAJiuCce7nHSktVsDKcR8GLRpD3WrN5yP3Nb_Hbu_Q9NjUQbSMw%40mail.gmail.com. > > > > > > > For more options, visit https://groups.google.com/d/optout. > > > > > > > > > > _______________________________________________ > > > > > linux-arm-kernel mailing list > > > > > linux-arm-kernel@xxxxxxxxxxxxxxxxxxx > > > > > http://lists.infradead.org/mailman/listinfo/linux-arm-kernel > > > > > > > > _______________________________________________ > > > > linux-arm-kernel mailing list > > > > linux-arm-kernel@xxxxxxxxxxxxxxxxxxx > > > > http://lists.infradead.org/mailman/listinfo/linux-arm-kernel > > > > _______________________________________________ > > linux-arm-kernel mailing list > > linux-arm-kernel@xxxxxxxxxxxxxxxxxxx > > http://lists.infradead.org/mailman/listinfo/linux-arm-kernel