05:18 <exkcmoeAdmin[m]> Which U-Boot guides

09:24 <f_> exkcmoeAdmin[m]: ...

09:24 <f_> The U-Boot guides?

09:24 <f_> https://u-boot.readthedocs.io/

11:05 <exkcmoeAdmin[m]> Idk

11:05 <exkcmoeAdmin[m]> Why does the U-Boot guides tell to run dd twice?...

11:06 <f_> Why are you posting the same message I posted though..is it appservice-IRC again?

11:48 <exkcmoeAdmin[m]> I mean

11:48 <exkcmoeAdmin[m]> I dont know why does the U-Boot guides tell to run dd twice

11:49 <exkcmoeAdmin[m]> https://u-boot.readthedocs.io/en/latest/board/amlogic/beelink-gt1-ultimate.html

11:50 <exkcmoeAdmin[m]> what going worg with it?

11:50 <exkcmoeAdmin[m]> narmstrong

11:55 <f_> *wrong??

11:55 <f_> I think I should upstream my KII Pro patches.

11:55 <f_> It seems to run quite well.

11:55 <exkcmoeAdmin[m]> Yeah wrong

11:55 <exkcmoeAdmin[m]> formating issue

11:56 <f_> Hmm

11:58 <f_> Ok

11:58 <f_> I can confirm amlbootsig from meson-tools works perfectly.

11:58 <f_> =)

11:59 <f_> narmstrong: Let's make fiptool work!

12:04 <narmstrong> It’s documented why, to not break the msdos partitioning

12:08 <f_> narmstrong: Ok. Thanks.

12:09 <f_> One question however. In the amlogic-boot-fip repo the makefile runs aml_chksum after running aml_encrypt_gxb. Is that required? aml_encrypt_gxb seems to do checksumming already, right?

12:58 <f_> fiptool with afaerber's patches ....doesn't work.

13:17 <narmstrong> honestly, I have no idea, I haven't used gxbb for a while now

13:21 <f_> Sure, but your assumption about an FIP with custom UUIDs was correct.

13:22 <f_> Well. There's only one custom FIP UUID -- BL301.

13:25 <f_> Looks like a "fake" UUID and I think we can safely ignore that.

13:35 <f_> narmstrong: I've discovered that I was dumb. aml_chksum is libre...https://github.com/LibreELEC/aml_chksum/blob/main/aml_chksum.c

13:52 <f_> Sorry. Disconnected.

14:17 <f_> So yeah, anyone experienced with gxbb FIPs?

14:17 <f_> (trying to get fiptool to work correctly)

14:26 <f_> https://bin.vitali64.duckdns.org/644d28d5

14:26 <f_> Hmm...

15:23 <f_> What's the difference between bl31.bin and bl31.img?

15:46 <narmstrong> there's a header on the .img version

15:50 <f_> And that's about it, right?

15:51 <f_> So, theoretically, I could compile TF-A's bl31 and include a header, right?

15:51 <f_> And it should boot, rrrrright?

15:58 <narmstrong> hmm yep, isn't it documented in upstream tf-a ?

15:59 <f_> No idea.

15:59 <f_> IIRC it isn't.

16:01 <f_> narmstrong: https://trustedfirmware-a.readthedocs.io/en/latest/plat/meson-gxbb.html It isn't.

16:01 <f_> Oh wait a second

16:01 <f_> Amlogic's fip_create seems to remove that header!

16:08 <exkcmoeAdmin[m]> so u dont need the header?

16:09 <narmstrong> yep i think it's unused when you're not in secure boot mode

16:10 <exkcmoeAdmin[m]> i think it is unseed bc Amlogic's fip_create removed it

16:10 <exkcmoeAdmin[m]> *unused

16:12 <f_> So.......magic `dd` command to get rid of the header before inclusion in the FIP by using `fiptool`?

16:17 <exkcmoeAdmin[m]> oh

16:18 <exkcmoeAdmin[m]> So u mean the blx_fix thingly?

16:21 <f_> No

16:21 <f_> ?

16:25 <f_> narmstrong: Is the header located between offset 0 and 1f0?

16:25 <exkcmoeAdmin[m]> I mean this script in libercecl fip repo

16:25 <exkcmoeAdmin[m]> https://github.com/LibreELEC/amlogic-boot-fip/blob/master/p201/blx_fix.sh

16:25 <f_> exkcmoeAdmin[m]: I don't understand what you meant by "u mean the blx_fix thingly"

16:26 <f_> I know what blx_fix.sh is.

16:26 <exkcmoeAdmin[m]> i mean this script https://github.com/LibreELEC/amlogic-boot-fip/blob/master/p201/blx_fix.sh

16:26 <exkcmoeAdmin[m]> Yes blx_fix.sh

16:26 <exkcmoeAdmin[m]> Is it for getting rid for the header

16:27 <f_> blx_fix.sh has nothing to do with BL31.

16:27 <exkcmoeAdmin[m]> ok

16:30 <f_> It seems like fip_create is getting rid of the header itself, rather.

16:48 <f_> Here we go!

16:48 <f_> diff became smaller!

16:48 <f_> Will send it real quick.

16:49 <f_> https://bin.vitali64.duckdns.org/644d4a85

16:49 <f_> I only have this (awkward) stuff in fip_create's output

16:49 <f_> No idea if that's needed.

16:53 <f_> Wow

16:53 <f_> It boots up.

16:54 <f_> narmstrong, chewitt (if you read this from logs): All this has been tested by using the amlogic-boot-fip repo. I basically modified the makefile to try to use as much free/libre tools as possible instead of the Amlogic counterparts. At some point (if you want) I will open a PR with my changes.

16:55 <f_> That is, if GitHub will let me sign in.

16:55 <f_> Else it'll probably be an email.

17:00 <f_> I however didn't test TF-A's BL31..

17:15 <narmstrong> nice

17:16 <f_> Indeed!

17:16 <f_> Let's document all this!

17:16 <narmstrong> yup1

17:52 <f_> https://wiki.postmarketos.org/wiki/Videostrong_KII_Pro_(videostrong-kii-pro)/U-Boot_as_primary_bootloader#How_to_build

17:52 <f_> Hope I didn't miss anything.

18:06 <exkcmoeAdmin[m]> Nice

18:44 <f_> https://gitlab.com/postmarketOS/pmaports/-/merge_requests/4046 One step forward for running U-Boot as the primary bootloader on pmOS!

18:44 <f_> :confetti:

18:56 <f_> Anyone with a KII Pro/WeTek Play2, can you please test this?

19:14 <f_> narmstrong: TF-A doesn't work though.

19:15 <narmstrong> Not sure the last time it has been tested…

19:15 <narmstrong> You’ll probably need to test older versions first en find out when it broke

19:17 <f_> I know which version to test, thanks to hexdump0815 =)

19:17 <f_> But I'll do that tomorrow, probably.

19:18 <f_> For now, cya.