narmstrong changed the topic of #linux-amlogic to: Amlogic mainline kernel development discussion - our wiki http://linux-meson.com/ - ml linux-amlogic@lists.infradead.org - official channel moved from Freenode - publicly logged on https://libera.irclog.whitequark.org/linux-amlogic
elastic_dog has quit [Ping timeout: 256 seconds]
elastic_dog has joined #linux-amlogic
BlueMatt has joined #linux-amlogic
f11f12 has joined #linux-amlogic
JohnnyonF has quit [Ping timeout: 240 seconds]
JohnnyonFlame has joined #linux-amlogic
ldevulder has joined #linux-amlogic
naoki has joined #linux-amlogic
naoki has quit [Client Quit]
f_ has joined #linux-amlogic
elastic_dog has quit [Remote host closed the connection]
elastic_dog has joined #linux-amlogic
f_ has quit [Quit: disconnect.]
Terry137322934 has quit [Ping timeout: 240 seconds]
Terry137322934 has joined #linux-amlogic
camus has quit [Ping timeout: 240 seconds]
camus has joined #linux-amlogic
f11f12 has quit [Quit: Leaving]
ldevulder has quit [Quit: Leaving]
vagrantc has joined #linux-amlogic
f_ has joined #linux-amlogic
<f_> narmstrong: Are you sure Meson GXBB is using a Synopsys DRAM controller?
<narmstrong> f_: I have no idea actually, it can be anything…
<f_> I'm looking up random registers online and results are quite inconvincing.
<f_> OHHH
<f_> Found something I think
<f_> Let me see..
<f_> I suspect the faulty ram init is caused by a bad timing configuration, too.
<f_> And I stumbled(typo??) upon a REing effort that seems to reverse-engineer acs.bin
<f_> Not sure
<f_> If I can, I might write a quickly-made utility to generate a timing.h out of the acs.bin, maybe.
<f_> chewitt (logs...): Is it possible that you ask WeTek if you can share at least timing.h? Or is it something you can't do? (if it is, no problem)
<minute> mnt reform laptop with a311d is quite incredible
<minute> just need to wire up SAI audio and solve some random dsi glitches (i will try to lower the pixel clock, currently the panel runs at 70Hz) and solve the ethernet issue (currently using usb ethernet)
<f_> Didn't know there were laptops using an Amlogic SoC!
<minute> f_: there weren't before today!
<minute> (it's the first prototype)
<f_> Aren't you that person from MNT tooting every now and then on mastodon?
<minute> it's funny to see what issues panfrost has vs etnaviv. they both struggle with legacy X/gtk2 gui rendering through glamor, but in different ways
<minute> f_: yep
<f_> Congrats!
<f_> Still a little sad that Amlogic's Trusted Firmware is proprietary.
<minute> thanks! yes, that's the only downside... but you're tinkering on that?
<f_> (fourtunately there's a port of TF-A's BL31 for most of Amlogic's SoCs)
<f_> minute: I'm trying to port TF-A's BL2 to my set-top box, yes.
<minute> also, there is a hard memory limit of 4GB, right?
<minute> f_: cool!
<f_> Not really, it depends.
<minute> oh?
<f_> It's all in acs.bin, which is libre.
<f_> And there's a switch to either hardcode the RAM size, or to auto-detect it.
<minute> sorry, i meant i thought there is a 4GB limit in the chip itself
<minute> ohh
<f_> No idea.
<minute> bananapi are working on a CM5 som with a311d2 that doesn't have this limit apparently
<f_> Nice.
<f_> ¯\_(ツ)_/¯ But I have absolutely no idea if there was a limit in the first place.
<f_> And honestly, I don't really care that much. If there was a limit my set-top box is probably far behind it (it has 2 GB of RAM)
<f_> But I could be wrong.
<f_> I'm only familiar with the S905. That's the only Amlogic SoC for which I have at least one board.
<f_> minute: Still, congratulations!
<minute> f_: thank you!!
<minute> i think personally i can manage with 4GB, it's just not an easy sell in 2023 for a laptop
<f_> Honestly I think (for lightweight stuff) that 4 GB is more than enough.
<minute> yeah
<f_> I myself have 8 GB of RAM on my 2011 laptop, but I'm only using ~2 GB.
<f_> (it's mostly my web browser taking RAM)
<f_> narmstrong: The Amlogic S905's datasheet pdf, page 220, contains some.....interesting stuff..
<f_> Those registers do take 4 bytes/32 bits, so mmio_{read,write}_32 seems like the right function to use.
<f_> But that does list a lot of registers for controlling the DDR controller (that's what DMC means, right?)
<f_> Some other registers were left undocumented though.
<f_> Ok let me just show what I'm getting -> https://search.projectsegfau.lt/search?q=DDR0_PUB_RIDR (replace 'projectsegfau.lt' with whatever SearXNG instance)
<f_> I do sometimes get results about i.MX.
<f_> Weird..it's listed in this header file which has to do with DVB (it seems)
<minute> interesting, i reenabled HDMI and tried to switch from mipi dsi over to hdmi and i get a kernel oops in drm_scdc_set_scrambling -> i2c_transfer https://gist.github.com/mntmn/5c5dd95b9fabc04059a3e00b16653c4d
<minute> cc narmstrong
<minute> EDID appeared to work though
<f_> DDR0_PUB's base is documented (as "DDR TOP"), but there's no other mention, or Amlogic removed mentions of it before HardKernel publied the datasheets.
<f_> I'm searching for all those undocumented registers, but there isn't really any mention of it anywhere.
JohnnyonFlame has quit [Ping timeout: 268 seconds]
<f_> These DDR0_PUB* registers are effectively undocumented.
<f_> Anyways, see you all tomorrow.
<f_> Will disconnect......now!
f_ has quit [Quit: ·_·]
<narmstrong> minute: yep i had the same error yesterday, I’ll debug on Monday
<minute> narmstrong: ah cool, thanks
<minute> i am able to run the panel now at much lower DSI speed (832.68mhz and 138.780mhz pixel clock), but can't get rid of glitches that happen like every 10 or 20 seconds
<minute> i.e. the display seems to jump horizontally for 1 frame
<minute> sometimes it doesn't happen for a longer time hmm
<minute> hmm, could be temperature related
JohnnyonFlame has joined #linux-amlogic
Danct12 has quit [Remote host closed the connection]
JohnnyonFlame has quit [Read error: Connection reset by peer]
JohnnyonFlame has joined #linux-amlogic
Danct12 has joined #linux-amlogic
anessen97 has quit [Quit: Ping timeout (120 seconds)]
anessen97 has joined #linux-amlogic