03:03 <junari> MasterR3C0RD: I made a working patch for lpddr3 a long time ago https://gist.github.com/iuncuim/a7b01ecd919dfd82d90f0f0fe81b50fd

03:04 <junari> LPDDR4 does not need additional changes to the memory controller, only checking and / 4 *3

03:08 <MasterR3C0RD> junari: Any idea regarding the logic behind the address mapping changes done here?

03:11 <MasterR3C0RD> Looks like it's moving the rank 1 bit lower in the HIF address and moving the upper 2-3 row bits up 1 bit

03:12 <junari> no, it's just the values that work on my device, and that are obtained through vendor u-boot dump

03:12 <junari> I didn't go into RE

03:12 <MasterR3C0RD> Ahh okay

03:14 <MasterR3C0RD> Yeah, these changes seem eerily similar to changes needed when 4GB of RAM is used on A133 with 16 row 10 column; seems to be a hardware quirk but was curious if this difference was related

06:57 <tokyovigilante> the bad news is that I can never get back the last day of my life debugging TFTP ;) but the good news with the new LwIP network stack and a quick "nft flush ruleset" later... TFTPBOOTing much faster

08:07 <hlauer> <apritzel> "hlix: just remove or comment the..." <- That did the job! Great, thanks a lot! So v1.1 ist running now 6.12 and detects 4 usb root hubs. Of course, still no power on the USB-A ports. So PH23 is copy&pasted and was never checked. Will try your suggestion during the week.

09:20 <tokyovigilante> [ 1.737555] input: H616 Audio Codec Line Out as /devices/platform/soc/5096000.codec/sound/card0/input1

09:20 <tokyovigilante> Progress...

11:43 <apritzel> hlauer: if I got this correctly, you also have the old v1.0 version? Did you get this in the "old days" - so a few years back, when the board was released?

11:44 <apritzel> If yes, and you can confirm that USB still works without PH23, we should remove that GPIO requirement from there, in the mainline DT and U-Boot's defconfig

11:44 <apritzel> I can help you create and send a patch, if you like

13:05 <loki666> tokyovigilante: how is your DE33 patch going ?

13:36 <apritzel> hlauer: so can you confirm that this is the situation:

13:37 <apritzel> 1) v1.0 never needed PH23. It doesn't hurt, but also doesn't do anything. VBUS on the two USB-A ports is always-on

13:37 <apritzel> 2) v1.1 uses PH23 for something, enabling it somehow hangs the system

13:38 <apritzel> 3) USB VBUS is not always-on for v1.1, so there is some GPIO that would need to be set

17:11 <apritzel> MasterR3C0RD: my version of the PDF, after adding the clock and GPIO bookmarks, is a v1.7 PDF, and is about 4% smaller. Any particular reason for v1.5, or is it just what your tools generated?

18:11 <MasterR3C0RD> apritzel: Wasn't something I had set; just what got spat out at the end of everything

19:35 <tokyovigilante> loki666: had a small amount of feedback on v5, so will hopefully get round to addressing that and post a v6 after the merge window