00:21
jacobk has quit [Ping timeout: 272 seconds]
00:22
montjoie has quit [Ping timeout: 255 seconds]
00:24
montjoie has joined #linux-amlogic
01:56
Daanct12 has joined #linux-amlogic
03:16
JohnnyonF has quit [Ping timeout: 255 seconds]
03:28
hexdump0815 has quit [Ping timeout: 240 seconds]
03:30
hexdump0815 has joined #linux-amlogic
03:55
buzzmarshall has quit [Quit: Konversation terminated!]
04:04
jacobk has joined #linux-amlogic
04:26
jacobk has quit [Ping timeout: 272 seconds]
04:42
jacobk has joined #linux-amlogic
06:29
jacobk has quit [Ping timeout: 255 seconds]
09:56
Daanct12 has quit [Read error: Connection reset by peer]
10:02
Daanct12 has joined #linux-amlogic
10:07
f_[xmpp] has joined #linux-amlogic
10:08
f_ has joined #linux-amlogic
10:09
<
f_ >
lvrp16: what have you said again about how inaccurate Amlogic datasheets are? :P
10:16
<
f_ >
pretty much what I have in my laptop
10:16
<
f_ >
all datasheets I was able to find. Some redundant though.
10:16
<
f_ >
but I organised it all.
10:26
<
f_ >
Surprisingly DFIODTCFG is somewhat documented
10:50
psydroid has joined #linux-amlogic
11:12
f_[xmpp] has quit [Ping timeout: 248 seconds]
11:31
gabes2 has joined #linux-amlogic
12:12
Daanct12 has quit [Ping timeout: 255 seconds]
12:17
<
f_ >
^ lvrp16 short fedi convo between me and orangecms
12:26
<
lvrp16 >
f_: there are errors but generally accurate.
12:27
<
f_ >
AFAICT e.g. bit 6 in DMC_DDR_CTRL is 'not used'
12:28
<
f_ >
but clearly by looking at p212 bl2 it's used.
12:28
<
f_ >
Always gotta verify.
12:30
<
f_ >
Saw the cottonwood patches on the mailing list
14:06
jacobk has joined #linux-amlogic
14:18
dliviu has quit [Quit: Going away]
14:20
dliviu has joined #linux-amlogic
14:39
JohnnyonFlame has joined #linux-amlogic
15:49
vagrantc has joined #linux-amlogic
15:57
mripard has quit [Quit: mripard]
16:44
jacobk has quit [Ping timeout: 272 seconds]
16:55
jacobk has joined #linux-amlogic
17:01
<
f_ >
I feel like dram-settings-gx.h should move to a board-specific file
17:02
<
f_ >
e.g. p200/spl_dram.h
17:03
<
f_ >
That could make more sense and probably should be upstreamed that way. thoughts narmstrong ?
17:12
<
narmstrong >
Probably yes
17:15
<
f_ >
After all, amlogic devices supported in U-Boot currently don't really hack acs to work with low bin ram chips, right?
17:17
<
f_ >
WeTek doesn't hack acs, same can be said of librecomputer
17:17
<
f_ >
Videostrong..no idea, but the WeTek acs works totally fine
17:18
<
f_ >
really, WeTek/Videostrong and librecomputer are all I care about
17:18
<
lvrp16 >
We take the reference component so it doesn't need customization.
17:18
<
f_ >
lvrp16: makes the job easier!
17:18
<
lvrp16 >
If we were to modify the reference design, changes are definitely needed.
17:19
<
lvrp16 >
Eg, if u need to make the area smaller.
17:19
<
f_ >
AFAICT most/all boards U-Boot currently supports are based on the reference design
17:20
<
f_ >
still have to compare acs of each single board
17:21
<
f_ >
because I think at the same time having to add a brand new 'real' board to U-Boot just for timings in SPL looks a little unnecessary
17:24
<
f_ >
and not really worth the effort in my opinion.
17:30
jacobk has quit [Ping timeout: 272 seconds]
17:33
jacobk has joined #linux-amlogic
17:39
<
f_ >
Don't know why, but I feel like I want to reimplement pyamlboot in C
17:50
<
f_ >
I mean, python is great and all
17:52
<
f_ >
USB mode in U-Boot SPL?
17:52
<
narmstrong >
Would be great
17:52
<
f_ >
Custom, simple protocol or worldcup?
17:53
<
narmstrong >
Perhaps a build config to select either usb or mmc
17:53
<
narmstrong >
Hard to know which protocol implement
17:56
<
f_ >
WorldCup is proprietary anyway
17:58
<
f_ >
so I'd go for a simple protocol
18:02
<
f_ >
For now U-Boot SPL panics when booted from USB.
18:39
<
narmstrong >
the amlogic G12 boot protocol quite ok, basically the bl2 send "AMLC" requests which contains a size and an offset of the fip, it could be adapted to get a size and an offset of the fit image
18:40
<
narmstrong >
the bl2 has a copy of the fip table when booted as usb
18:40
<
narmstrong >
perhaps it could have the fit dtb and request the fit data which is after the fit dtb blob
19:10
f_ has quit [Ping timeout: 260 seconds]
19:42
ungeskriptet has joined #linux-amlogic
20:56
jacobk has quit [Ping timeout: 255 seconds]
21:05
jacobk has joined #linux-amlogic
21:56
psydroid has quit [Remote host closed the connection]
22:05
jacobk has quit [Ping timeout: 255 seconds]
22:11
jacobk has joined #linux-amlogic