00:23 <apritzel> GrantM11235[m]: so by just adding CONFIG_USB_MUSB_GADGET=y to the defconfig, and doing a clean build, USB gadget now works nicely for me, with Ethernet, fastboot, UMS

00:23 <apritzel> (plus the compatible string in sunxi.c, of course)

00:25 <apritzel> GrantM11235[m]: did you manage to get U-Boot's mtd command working, with SPI NAND on the F1C200s?

00:29 <apritzel> "mtd list" results in: "spi_nand flash@0: unknown raw ID ff ff ff ff" for me

00:29 <apritzel> GrantM11235[m]: also many thanks for creating the Wiki page!

02:21 <GrantM11235[m]> apritzel: do you read the irc logs from when you are offline?

09:21 <junari> apritzel: you wanted to share some improvements for lpddr4 driver

09:23 <junari> In the coming days I plan to send an updated patch for the opizero3

10:44 <apritzel> GrantM11235[m]: yes, I always do (the read logs), so feel free to use the channel asynchronously ;-)

10:44 <apritzel> junari: ah, right, thanks for the heads up

10:52 <GrantM11235[m]> apritzel: I got the mtd command to work, mostly. I can read the stock uboot image at offset 0, but when I try to read other partitions, like the linux image, I get a different chunk of data than what I expect. I assume that the stock firmware is doing some sort of wear-leveling

11:06 <apritzel> GrantM11235[m]: is that just a matter for the stock firmware mapping? So is the view self-consistent with the mainline U-Boot driver? If you write something to offset x, you can always read it back there?

11:09 <GrantM11235[m]> I don't know, I'm afraid to write anything without a backup of the original firmware

11:15 <GrantM11235[m]> apritzel: This is my current defconfig with working mtd and ums https://paste.debian.net/1295042/

18:48 <junari> apritzel: got your patch and try to adopt for lpddr4

18:48 <junari> https://gist.github.com/iuncuim/b452341e373f46349220f7c0a12773d8

18:49 <junari> But what I see, regs 0x3cc 0x40c 0x44c should be 0x4, but it 0x0

18:51 <junari> If you have time, please see what I do wrong

18:54 <apritzel> val_hi should be 0x04040404, at least that's what I put in

18:55 <apritzel> (hoping that I didn't mess this up...)

18:57 <apritzel> junari: ^^^

18:58 <junari> val_hi = (para->type == SUNXI_DRAM_TYPE_LPDDR4) ? 4 : para->dx_dri;

18:58 <junari> If SUNXI_DRAM_TYPE_LPDDR4 it should be 0x4

19:01 <apritzel> 0x4 for each register, but 0x04040404 for val_hi, it will be split up into 0x4 for each register later

19:02 <junari> with your patch val_hi should be 0x4, and in first address 0x38c it write 0x4, but to the next 0x0

19:03 <junari> don't understand why

19:06 <junari> how it look https://disk.yandex.ru/i/sGBx-wKVW33eEQ

19:16 <junari> oh, i understand my mistake

19:21 <junari> with "val_hi = (para->type == SUNXI_DRAM_TYPE_LPDDR4) ? 0x04040404 : para->dx_dri;" all correct, thanks, Andre