02:56 <naoki> mmm, it happens on g12b too

05:35 <narmstrong> naoki: can you send a bug report to linux-amlogic@lists.infradead.org ?

08:58 <rockosov> narmstrong: xdarklight Hello guys! I hope you're doing well. Previously, we discussed applying the A1 clock driver to linux-amlogic next some days ago. Am I understanding correctly that we can't apply anything in clock meson subsystem without jbrunet's Reviewed-by or Acked-by?

09:00 <narmstrong> rockosov: I could, let me sync with him before

09:01 <narmstrong> rockosov: the bindings are lacking an ack from the dt maintainers

09:03 <rockosov> narmstrong: For PLL bindings there is Rob's RvB - https://lore.kernel.org/all/20230523135351.19133-4-ddrokosov@sberdevices.ru/

09:04 <rockosov> narmstrong: For Peripherals bindings it was Rob's RvB in the previous series, but I've renamed driver and bindings from a1-clkc to a1-peripherals-clkc and remove Rob's RvB

09:04 <narmstrong> rockosov: yep, for patch 5 there isn't, can you ping conor (Conor Dooley <conor@kernel.org>) to see if he can add his ack ? just explain what just just said

09:06 <rockosov> narmstrong: v14 has RvB - https://lore.kernel.org/all/20230426095805.15338-6-ddrokosov@sberdevices.ru/

09:06 <rockosov> narmstrong: sure, I can ping him

09:08 <rockosov> xdarklight: Also I've removed all stuff from Peripherals clock driver which we discussed on the previous week in the v16 version

13:25 <LucasTanure> Vim4 boots Kernel 6.4.0-rc3

13:26 <LucasTanure> [ 0.000000] Linux version 6.4.0-rc3-00034-g4934d80de378-dirty (tanureal@ryzen) (aarch64-none-linux-gnu-gcc (GNU Toolchain for the A-profile Architecture 10.3-2021.07 (arm-10.29)) 10.3.1 2021063

13:26 <LucasTanure> [ 0.000000] Booting Linux on physical CPU 0x0000000000 [0x410fd092]

13:26 <LucasTanure> [ 0.000000] Machine model: Khadas VIM4

13:26 <LucasTanure> it hangs at rootfs as it cant find sd card

13:26 <LucasTanure> I ported clock and uart driver

13:45 <narmstrong> LucasTanure: nice :-)

13:54 <rockosov> LucasTanure: Good news! Does it have A311D2 SoC inside?

13:56 <rockosov> narmstrong: Conor provided RvB to Peripherals clock bindings - https://lore.kernel.org/linux-amlogic/20230525-connected-skipper-442c6d0b52c1@wendy/ Thank you for suggestion!

13:56 <narmstrong> rockosov: ack, I'll ping jerome to see how we solve that

13:57 <rockosov> narmstrong: Thanks a lot!

14:12 <LucasTanure> rockosov: yes , it has A311D2

17:43 <LucasTanure> Vim4 Mainline kernel : https://github.com/lucastanure/linux/commits/vim4

17:43 <LucasTanure> It hangs, not fully funtional!

19:10 <xdarklight> rockosov: I'll be looking into the latest patches on Sunday or Monday. Jerome previously asked me to continue with the review process. My understanding is that once all patches have my Reviewed-by he'll be applying them and create a PR for the clock maintainers

19:10 <minute> narmstrong: did the bpi wifi module just work for you?

19:11 <minute> (bpi cm4)

19:11 <narmstrong> minute: I just saw it probe from the armbian image, haven’t tried further tbh

19:14 <minute> narmstrong: ah, ok!

19:14 <minute> narmstrong: as diagnosed a bit with xdarklight, i'm getting RX corruption, so still have to squash some bugs there

19:15 <xdarklight> LucasTanure: nice to see someone working on the VIM4 (even though it's going to be a long road to get all peripherals supported). I suggest starting with a smaller .dts{,i} to avoid issues for example with DVFS. also please note that the Amlogic vendor SDK .dts files are typically not compatible with mainline drivers

19:23 <LucasTanure> xdarklight: I am still working out a patch that works

19:23 <LucasTanure> that dtsi was the bare minimum that got stuff working

19:24 <LucasTanure> from that to the actual upstreaming will be a long road

19:27 <xdarklight> LucasTanure: I understand that you're still in the process of making it work. just wanted to add some of my thoughts in case you didn't go through similar pain with other Amlogic SoCs before ;-)

19:28 <LucasTanure> xdarklight, what is DVFS?

19:29 <LucasTanure> Yeah, it's my first board I am upstreaming alone

19:29 <xdarklight> LucasTanure: DVFS = dynamic voltage and frequency scaling, it's used for changing the CPU and GPU clocks and associated voltage regulators

19:31 <xdarklight> LucasTanure: higher clocks typically require higher voltages, so if it hangs randomly I'd look at that. if it hangs "just" because it can't find rootfs then I'd start looking at the following components in the order: MMC controller, pinctrl, GPIO

19:31 <LucasTanure> I am having a look ant IRQ now

19:31 <LucasTanure> I did PINCTRL already

19:31 <xdarklight> ah, IRQ is also a good candidate

19:32 <LucasTanure> after that I will have a horrible patch that makes VIm4 boot

19:32 <LucasTanure> with that in hands I will start to trim down, separete logic units into patches

19:33 <LucasTanure> and send to upstream only after I have a nice patchset for VIM4

19:36 <xdarklight> the bare minimum for an upstream .dts would be UART output, see for example https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/arch/arm64/boot/dts/amlogic/meson-a1.dtsi?id=b255e1268b0b3dcc5a347d203e8f38ab52bc9a1d

19:37 <xdarklight> so if you feel like your patch queue is getting (too) big: feel free to send basic stuff early

19:48 <LucasTanure> ok, I was aiming to "boot to sd card", but if gets too complicated I trimm down to uart only

20:54 <rockosov> xdarklight: Got it! Thank you for clarification!